[
  {
    "key": "CNAVE9HR",
    "title": "Incremental cost-effectiveness of the second Xpert MTB/RIF assay to detect Mycobacterium tuberculosis",
    "abstract": "Background: Due to the non-homogeneity of specimens collected from tuberculosis (TB) suspects, repeated Xpert MTB/RIF (Xpert) may have potential clinical benefits. Incremental cost-effectiveness was analyzed for the second Xpert assay to detect Mycobacterium tuberculosis (Mtb) and rifampicin (RIF) resistance.\nMethods: Specimens were collected from 1,063 pulmonary TB (PTB) and 398 extrapulmonary TB (EPTB) suspects, who had two Xpert tests sequentially within one week. The specimens were subjected to smear, culture, Xpert and drug susceptibility testing. Incremental cost-effectiveness of the serial Xpert assays was evaluated.\nResults: Among 813 Xpert-positive TB patients, 755 (92.87%) were identified by the first assay whereas the additional 58 (7.13%) were identified by the second assay. The second Xpert assay had higher incremental yield for smear-negative than for smear-positive specimens (12.07% vs. 1.84%, P<0.001), and higher incremental yield for EPTB than for PTB (10.71% vs. 4.65%, P=0.003). About 94.48% (137/145) of the RIF-resistant patients were identified by the first Xpert assay and 5.52% (8/145) were identified by the second Xpert assay. After the first assay, the incremental cost of performing a second Xpert was huge: US$22.82 vs. US$467.72 (P<0.001) and US$35.02 vs. US$291.87 (P<0.001) for PTB and EPTB, respectively. The incremental cost of performing a second Xpert is lower in smear-negative than in smear-positive group in both PTB and EPTB.\nConclusions: One Xpert assay is sufficient for smear-positive cases, and a second Xpert assay is beneficial not only for Mtb detection but also for RIF-resistant diagnosis for smear-negative TB suspects, whereas the incremental cost for the second Xpert is huge.",
    "full_text": "Original Article\nIncremental cost-effectiveness of the second Xpert MTB/RIF assay to detect Mycobacterium tuberculosis\nGuirong Wang, Shuqi Wang, Guanglu Jiang, Yuhong Fu, Yuanyuan Shang, Hairong Huang\nNational Clinical Laboratory for Tuberculosis, Beijing Key laboratory for Drug-resistant Tuberculosis Research, Beijing Chest Hospital, Capital Medical University, Beijing Tuberculosis and Thoracic Tumor Institute, Beijing 101149, China Contributions: (I) Conception and design: G Wang, H Huang; (II) Administrative support: None; (III) Provision of study materials or patients: Y Fu, Y Shang; (IV) Collection and assembly of data: S Wang, G Jiang; (V) Data analysis and interpretation: S Wang; (VI) Manuscript writing: All authors; (VII) Final approval of manuscript: All authors. Correspondence to: Dr. Hairong Huang. Beimachang Rd 97, Beijing 101149, China. Email: huanghairong@tb123.org.\nBackground: Due to the non-homogeneity of specimens collected from tuberculosis (TB) suspects, repeated Xpert MTB/RIF (Xpert) may have potential clinical benefits. Incremental cost-effectiveness was analyzed for the second Xpert assay to detect Mycobacterium tuberculosis (Mtb) and rifampicin (RIF) resistance. Methods: Specimens were collected from 1,063 pulmonary TB (PTB) and 398 extrapulmonary TB (EPTB) suspects, who had two Xpert tests sequentially within one week. The specimens were subjected to smear, culture, Xpert and drug susceptibility testing. Incremental cost-effectiveness of the serial Xpert assays was evaluated. Results: Among 813 Xpert-positive TB patients, 755 (92.87%) were identified by the first assay whereas the additional 58 (7.13%) were identified by the second assay. The second Xpert assay had higher incremental yield for smear-negative than for smear-positive specimens (12.07% vs. 1.84%, P<0.001), and higher incremental yield for EPTB than for PTB (10.71% vs. 4.65%, P=0.003). About 94.48% (137/145) of the RIF-resistant patients were identified by the first Xpert assay and 5.52% (8/145) were identified by the second Xpert assay. After the first assay, the incremental cost of performing a second Xpert was huge: US$22.82 vs. US$467.72 (P<0.001) and US$35.02 vs. US$291.87 (P<0.001) for PTB and EPTB, respectively. The incremental cost of performing a second Xpert is lower in smear-negative than in smear-positive group in both PTB and EPTB. Conclusions: One Xpert assay is sufficient for smear-positive cases, and a second Xpert assay is beneficial not only for Mtb detection but also for RIF-resistant diagnosis for smear-negative TB suspects, whereas the incremental cost for the second Xpert is huge.\nKeywords: Tuberculosis (TB); diagnostic tests; rifampicin; resistance\nSubmitted Oct 10, 2017. Accepted for publication Jan 29, 2018. doi: 10.21037/jtd.2018.02.60 View this article at: http://dx.doi.org/10.21037/jtd.2018.02.60\n\nIntroduction\nTuberculosis (TB) is an ancient infectious disease caused by Mycobacterium tuberculosis (Mtb) that typically affects the lungs (pulmonary TB, PTB) but can affects other sites as well (extrapulmonary TB, EPTB) (1). Despite global efforts, TB is yet to be fully controlled and therefore requires improved methods for rapid identification and early\n\ntreatment of patients with active disease (2). Xpert MTB/ RIF (Xpert) assay (Cepheid, Sunnyvale, CA, USA) is an automated real-time nucleic acid amplification technology for rapid simultaneous detection of Mtb and rifampicin (RIF) resistance within 2 hours (3). The assay has been evaluated for PTB and EPTB diagnosis in abundant studies (4-7), and the results elucidated its excellent accuracy. In 2010 and 2013, the Xpert assay was endorsed by the World\n\n\u00a9 Journal of Thoracic Disease. All rights reserved.\n\njtd.amegroups.com\n\nJ Thorac Dis 2018;10(3):1689-1695\n\n1690\nHealth Organization (WHO) for PTB and EPTB diagnosis, respectively (8,9).\nExamination of multiple specimens from same TB patient improves the sensitivity of detection by microscopy and culture (10). Likewise, examination of multiple specimens would improve the detection sensitivity using Xpert assay, although increase the working load and cost as well. Economic evaluations of diagnostic strategies are needed to guide decisions on prioritizing the health care resources in TB control (11). However, little data on the utility of Xpert assay for additional specimen is available. Cost-effectiveness analysis is a measure to evaluate the efficiency of the repeated Xpert assay, and thus facilitate identifying the cost effective route for carrying out the test appropriately. Therefore, the current analysis was conducted to determine the incremental cost-effectiveness of the second Xpert test.\nMethods\nEthical approval\nThe ethical approvals for this study were obtained from the Beijing Chest Hospital Ethics Committee (ethical approval number: BJXK-2015-05). A written informed consent was acquired from each participant.\nPatients and sample collection\nThis prospective study was conducted from March 2015 to April 2017 in Beijing Chest Hospital (Beijing, China), which is the only national referral TB center in China. A total of 1,063 PTB suspects and 398 EPTB suspects who had two Xpert tests sequentially within one week were enrolled. Totally 2,922 Xpert tests were performed on 2,126 sputum, 288 pus, 176 cerebrospinal fluid, 208 pleural fluid, 24 urine and 100 bronchoalveolar lavage fluid specimens.\nSmear microbiology\nDirect smear was prepared and stained with auramine, and then examined by light-emitting diode (LED) microscopy. The smear was read and interpreted in accordance with WHO guidelines (12).\nXpert MTB/RIF\nThe assay was performed following the manufacturer\n\nWang et al. Incremental cost-effectiveness of 2nd Xpert\ninstructions. For pus specimens, 1 mL of pus was mixed with 2 mL of Xpert sample reagent, vortexed for at least 10 s and incubated at room temperature for 10 min. The mixture was again vortexed for 10 seconds and incubated at room temperature for 5 min. For other specimens, 1 mL was mixed with 2 mL of Xpert sample reagent. After vortexed for several seconds, the reaction mixture was incubated at room temperature for 15 min to inactivate the living bacteria. Then, a total of 2 mL processed mixture was transferred into the Xpert cartridge and loaded onto the GeneXpert instrument. The automatic detection procedure was run afterwards.\nMycobacterial culture and susceptibility testing by mycobacteria growth indicator tube (MGIT) 960 system\nThe MGIT 960 system is based on fluorescence detection of mycobacteria growth in a tube containing modified Middlebrook 7H9 medium together with fluorescence quenching-based oxygen sensor (13). The specimens were decontaminated with N-acetyl-L-cysteine-sodium hydroxide (BBL MycoPrep; Becton Dickinson, Sparks, MD) for 20 min, then neutralized with sterile saline phosphate buffer (PBS; pH 6.8) to a final volume of 45 mL, and the centrifuged at 3,000 \u00d7g for 15 min at 4 \u2103. The pellet was resuspended in 1.5 mL of PBS. Each MGIT960 tube was inoculated with 0.5 mL of the resulting specimen, incubated at 37 \u2103 in an automated MGIT960 apparatus (Becton Dickinson) for a maximum of 42 days. The MGIT960 outcomes were recorded according to the manufacturer\u2019s instruction. The standard drug susceptibility testing (DST) with RIF was carried out for the positive cultures using the MGIT960 IR kit (Becton Dickinson) according to the manufacturer\u2019s instructions.\nCost parameters\nSince its endorsement by WHO in 2010, over 23 million Xpert tests have been procured by 130 countries (14). Market price is around US$ 50,000 for the four-cartridge module plus computer extension and US$ 65 per cartridge. Nevertheless, it is provided at negotiated prices to some low- and middle-income countries with high TB burden at US$17,000 for the equipment and US$9.98 for each cartridge (15). The cost for one Xpert test was US$13.2 in this study, including equipment US$2.84 (at a price of US$17,000 per four-module instrument), building space US$0.02, maintenance US$0.18, staff US$0.11, cartridge\n\n\u00a9 Journal of Thoracic Disease. All rights reserved.\n\njtd.amegroups.com\n\nJ Thorac Dis 2018;10(3):1689-1695\n\nJournal of Thoracic Disease, Vol 10, No 3 March 2018\n\n1691\n\nTable 1 Mtb detection outcomes by Xpert MTB/RIF with two specimens examined\n\nCategory\n\nSuspects\n\nCases\n\nCases identified on 1st Xpert assay (%)\n\nPTB\n\nSmear-positive 464\n\n430\n\n422 (98.14)\n\nSmear-negative 599\n\n215\n\n193 (89.77)\n\nTotal\n\n1,063\n\n645\n\n615 (95.35)\n\nEPTB\n\nSmear-positive\n\n97\n\n60\n\n59 (98.33)\n\nSmear-negative 301\n\n108\n\n91 (84.26)\n\nTotal\n\n398\n\n168\n\n150 (89.29)\n\nTotal\n\n1,461\n\n813\n\n765 (94.10)\n\nPTB, pulmonary tuberculosis; EPTB, extrapulmonary tuberculosis.\n\nAdditional cases identified on 2nd Xpert assay (%)\n8 (1.86) 22 (10.23) 30 (4.65)\n1 (1.67) 17 (15.74) 18 (10.71) 48 (5.90)\n\nUS$9.98 (at negotiated price), and consumables US$0.07.\nCost-effectiveness analysis\nCost-effectiveness was calculated by determining the cost for each TB cases detected. An average cost-effectiveness ratio was estimated by dividing the total cost of an intervention by its measure of effectiveness. An incremental cost-effectiveness ratio considers change from the first to the second Xpert test; in this instance, the additional cost per additional TB case identified from examining progressively more specimens per patient.\n\nmore cases were identified by the second assay (Table 1). Overall, the second Xpert assay had higher incremental yield for EPTB samples than for PTB samples [10.71% (18/168) vs. 4.65% (30/645), \u03c72=8.820, P=0.003].\nThe suspects were further stratified into smear-positive and smear-negative groups, and the results showed that in smear-negative group, the second Xpert assay had apparent incremental yield for both PTB and EPTB patients (Table 1). Overall, the second Xpert assay had higher incremental yield for smear-negative specimens than for the smearpositive specimens [12.07% (39/323) vs. 1.84% (9/490), \u03c72=36.727, P<0.001].\n\nStatistical analysis\nThe incremental yields of the first and second Xpert assay were compared using \u03c72 test. The Student\u2019s t-test was performed to assess statistical significance between the costs of different groups. Statistical analyses were performed with SPSS (version 19.0). Differences were considered statistically significant at P<0.05.\nResults\nDetection of Mtb\nTotally 645 PTB suspects produced Xpert-positive outcomes. Among them, 615 (95.35%) could be identified by the first Xpert assay and 30 (4.65%) additional cases were identified by a second assay. Meanwhile, 168 EPTB suspects had Xpert-positive outcomes, and 150 (89.29%) of them were identified by the first Xpert assay, while 18 (10.71%)\n\nDetection of RIF resistance\nOne hundred and twenty-six PTB and 19 EPTB cases produced RIF resistance outcomes by 2 Xpert assays. Among the 145 RIF-resistant patients, 137 (94.48%) were identified by the first Xpert assay and 8 (5.52%) additional cases were identified by a second Xpert assay. Compared with the phenotypic DST, the first Xpert test detected 92.57% (137/148) of the RIF-resistant patients, and the second Xpert test gained 5.41% increment, so the total detection rate was 97.97% (145/148). Furthermore, 2 Xpert tests could correctly detect all of the RIF susceptible patients (Table 2).\nCost-effectiveness\nThere was a big difference regarding average cost (per TB case diagnosed) between 1st and 2nd Xpert assays: US$22.82\n\n\u00a9 Journal of Thoracic Disease. All rights reserved.\n\njtd.amegroups.com\n\nJ Thorac Dis 2018;10(3):1689-1695\n\n1692\n\nWang et al. Incremental cost-effectiveness of 2nd Xpert\n\nTable 2 Rifampicin resistance detection outcomes by Xpert MTB/RIF with two specimens examined\n\nCategory\n\nCases identified on 1st Xpert assay, n (%)\n\nCases identified on 2nd Xpert assay, n (%)\n\nRifampicin-resistant cases\n\n137 (94.48%)\n\n8 (5.52)\n\nRifampicin-susceptible cases\n\n390 (100%)\n\n0 (0)\n\nTable 3 Stratified analysis of Mtb detection and cost-effectiveness\n\nMtb detection\n\nCategory\n\nOrder of tests\n\nSpecimens examined\n\nCases detected\n\nPulmonary tuberculosis\n\nSmear-positive\n\n1st\n\n464\n\n422\n\n2nd\n\n464\n\n430\n\nSmear-negative\n\n1st\n\n599\n\n193\n\n2nd\n\n599\n\n215\n\nTotal\n\n1st\n\n1,063\n\n615\n\n2nd\n\n1,063\n\n645\n\nExtrapulmonary tuberculosis\n\nSmear-positive\n\n1st\n\n97\n\n59\n\n2nd\n\n97\n\n60\n\nSmear-negative\n\n1st\n\n301\n\n91\n\n2nd\n\n301\n\n108\n\nTotal\n\n1st\n\n398\n\n150\n\n2nd\n\n398\n\n168\n\nIncremental cases\n\u2013 8 \u2013 22 \u2013 30\n\u2013 1 \u2013 17 \u2013 18\n\nCosts (US $)\n\nAverage cost for Cost for per\n\nper case\n\nincremental case\n\n14.51 28.49 40.97 73.55 22.82 43.51\n\n\u2013 765.60\n\u2013 359.40\n\u2013 467.72\n\n21.70 42.68 43.66 73.58 35.02 62.54\n\n\u2013 1,280.40\n\u2013 233.72\n\u2013 291.87\n\nvs. US$43.51 (P<0.001) and US$35.02 vs. US$62.54 (P<0.001) for PTB and EPTB diagnosis, respectively. Compared with the first Xpert MTB/RIF assay, the incremental cost of performing a second test was huge: US$22.82 vs. US$467.72 (P<0.001) and US$35.02 vs. US$291.87 (P<0.001) for PTB and EPTB, respectively. Furthermore, the incremental cost of performing a second Xpert is lower in smear-negative group than in smearpositive group for both PTB and EPTB patients (Table 3, Figure 1).\nDiscussion\nAccurate, rapid detection of TB and TB drug resistance is critical for improving patient care and decreasing TB transmission. The development of Xpert technique was a landmark event. A Cochrane systematic review (6) found\n\nthat Xpert has a pooled sensitivity of 88% and pooled specificity of 98% compared with the gold standard of culture. It provides fast results, is easy to use and has a low biohazard risk which facilitates its implementation in rural settings. Cost-effectiveness evaluation for TB diagnostics often provides important information to the policymakers. Data on cost and cost-effectiveness of Xpert in diverse settings had been reported. Xpert was more sensitive, comparably specific, and more cost-effective than smear microscopy in intermediate and low burden areas (16,17). Xpert was superior over microscopic determination of drug susceptibility (MODS) in high TB/HIV prevalence setting (18). Whereas, Vassall et al. (19) reported that Xpert was cost-neutral and did not improve the cost-effectiveness of TB diagnosis in South Africa.\nBudgetary constraint is a major consideration influencing the choice of diagnostics in developing countries. This is\n\n\u00a9 Journal of Thoracic Disease. All rights reserved.\n\njtd.amegroups.com\n\nJ Thorac Dis 2018;10(3):1689-1695\n\nJournal of Thoracic Disease, Vol 10, No 3 March 2018\n\n1693\n\nTotal cost ($)\n\nA 16000\n12000 8000 4000\n\nSecond 765.60 First\n\n0 0 100 200 300 400 500\n\nB\n\n20000\n\n16000\n\nTotal cost ($)\n\n12000\n\n8000\n\n4000\n\n0 0\n\nSecond 359.40 First\n\n100\n\n200\n\n300\n\nTotal cost ($)\n\nC 30000\n25000 20000 15000 10000\n5000 0 0\n\nSecond 467.72 First\n200 400 600 800\n\nD\n\n3200\n\nTotal cost ($)\n\n2400\n\n1600\n\n800\n\nSecond 1280.40 First\n\n0\n\n0\n\n20\n\n40\n\n60\n\n80\n\nE 10000\n8000\n\nF\nSecond\n\nTotal cost ($)\n\n6000\n\n233.72\n\n4000\n\nFirst\n\n2000\n\n0\n\n0\n\n30\n\n60\n\n90\n\n120\n\nNumber of TB cases detected\n\nTotal cost ($)\n\n12000 10000\n\nSecond\n\n8000\n\n291.87\n\n6000 First\n4000\n\n2000\n\n0\n\n0\n\n40\n\n80 120 160 200\n\nNumber of TB cases detected\n\nFigure 1 Cost-effectiveness of sequential Xpert MTB/RIF assays. The dashed lines denote the incremental cost-effectiveness ratios of increasing the number of Xpert MTB/RIF assays that are examined per patient, the figures adjacent to these lines indicate the extra cost of detecting each additional TB case. The steeper the slope is, the less cost-effective the intervention is. (A) Smear-positive pulmonary tuberculosis; (B) smear-negative pulmonary tuberculosis; (C) total pulmonary tuberculosis; (D) smear-positive extrapulmonary tuberculosis; (E) smear-negative extrapulmonary tuberculosis; (F) total extrapulmonary tuberculosis.\n\nthe first study to assess the incremental cost-effectiveness of the second Xpert assay for detection of TB. Our results showed that the first Xpert assay detected 98.16% (481/490) of the Xpert-positive patients among the smear-positive TB patient group. Nevertheless, the cost-effectiveness analysis showed that the incremental cost of performing a second Xpert was very high for the smear-positive TB patients. As the incremental yield from a second Xpert was relatively small, so one Xpert assay was sufficient for smear-positive patients. Albeit the benefit of performing Xpert assays for those patients was to find RIF resistance, especially in high multidrug-resistant (MDR)-TB burden area.\nThe global priorities for TB care and control are to improve case detection and to detect cases earlier, including cases of smear-negative disease, and to enhance the capacity to diagnose MDR-TB. Our results showed that the second\n\nXpert assay had an incremental yield of 12.07% (39/323) for smear-negative TB patients. Boehme et al. (5) reported that among patients with smear-negative/culture-positive PTB, the addition of a second Xpert test increased sensitivity by 12.6%, which is similar to our results. Dorman et al. (20) showed that the new generation of Xpert-Xpert MTB/RIF Ultra (Ultra) could increase sensitivity by 17% compared with Xpert for TB detection with smear-negative-culturepositive sputum, while the cost of it will be similar with Xpert. Theoretically, Ultra would be more cost-effective comparing with 2 Xpert tests for smear-negative PTB diagnosis. Cowan et al. (21) reported that the two Xpert tests strategy was more expensive but still cost-effective compared with 3 smears. Due to the unavailable of Ultra in most of countries nowadays, our results suggested that two Xpert tests can benefit the poor diagnosis of smear-negative\n\n\u00a9 Journal of Thoracic Disease. All rights reserved.\n\njtd.amegroups.com\n\nJ Thorac Dis 2018;10(3):1689-1695\n\n1694\nTB. Notable, for smear-negative (and economically sustainable) TB suspects, a second Xpert assay is not only valuable for Mtb detection but also for RIF resistance diagnosis.\nMDR-TB is an increasing concern globally and directly threatens disease control efforts in many countries. Only 30,000 of nearly 500,000 new cases of multidrug-resistant TB are detected and reported every year, hence misdiagnosis causes thousands of deaths, nosocomial and community transmission, and amplification of drug resistance (22,23). In this study, 5.52% additional RIF resistant cases were identified by the second Xpert assays. Although Xpert had excellent repeatability for RIF resistance detection, our assay demonstrated a second Xpert assay has the benefit to detect more RIF resistant cases.\nThe incremental cost-effectiveness of the second Xpert is dependent on a number of different setting-specific factors. First, the cost for per TB case detection will decrease with the increase of TB prevalence in the setting (21). Second, higher proportion of TB cases among the suspect population improves the cost-effectiveness of Xpert. Third, the decision analytic modeling demonstrated that when transmission effects are excluded, the cost-effectiveness of Xpert increases as the MDR-TB prevalence increases (24). Fourth, the cost for per Xpert-positive case was higher in sites with lower volumes of testing. Previous study in South Africa suggested that low testing volume and a high number of sites involved could increase Xpert testing cost by 50% or more (25). Other factors are likely to influence the costeffectiveness as well, such as the proportion of those coinfected with HIV. These findings may help to inform the decision-makers about the appropriateness of a second Xpert deployment in different settings.\nThe main drawback of Xpert is its cost. As the goal of TB control is to correctly identify as many cases as possible for effective treatment, a cost-effective but simple, easy and rapid diagnostic method that could be readily and widely adopted is need. Our results showed that after the first Xpert assay, the incremental cost of performing a second test is huge. In lowincome countries, resourcing for TB services is extremely constrained. In order to end the global TB epidemic, it is the responsibility of the manufacturers, governments and nonprofit organizations to lower the price of Xpert assay to make it affordable in low-income countries.\nConclusions\nAccording to our assay, one Xpert assay is sufficient for\n\nWang et al. Incremental cost-effectiveness of 2nd Xpert\nsmear-positive cases, and a second Xpert assay is beneficial not only for Mtb detection but also for RIF-resistant diagnosis for smear-negative TB suspects, whereas the incremental cost for the second Xpert test is huge.\nAcknowledgements\nFunding: This work was supported by the research funding from National Science and Technology Major Project (2017ZX10201301-004-002, 2017ZX09304009004), Natural Science Fund of China (81703632), Beijing Natural Science Foundation (7172050), and Beijing Municipal Administration of Hospitals Clinical Medicine Development of Special Funding Support (ZYLX201824).\nFootnote\nConflicts of Interest: The authors have no conflicts of interest to declare.\nEthical Statement: The ethical approvals for this study were obtained from the Beijing Chest Hospital Ethics Committee (ethical approval number: BJXK-2015-05). A written informed consent was acquired from each participant.\nReferences\n1. World Health Organization. Global tuberculosis report 2017. Geneva: World Health Organization, 2017.\n2. Raviglione M, Marais B, Floyd K, et al. Scaling up interventions to achieve global tuberculosis control: progress and new developments. Lancet 2012;379:1902-13.\n3. Boehme CC, Nabeta P, Hillemann D, et al. Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med 2010;363:1005-15.\n4. Theron G, Zijenah L, Chanda D, et al. Feasibility, accuracy, and clinical effect of point-of-care Xpert MTB/ RIF testing for tuberculosis in primary-care settings in Africa: a multicentre, randomised, controlled trial. Lancet. 2014;383:424-35.\n5. Boehme CC, Nicol MP, Nabeta P, et al. Feasibility, diagnostic accuracy, and effectiveness of decentralised use of the Xpert MTB/RIF test for diagnosis of tuberculosis and multidrug resistance: a multicentre implementation study. Lancet 2011;377:1495-505.\n6. Steingart KR, Schiller I, Horne DJ, et al. Xpert \u00ae MTB/ RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev\n\n\u00a9 Journal of Thoracic Disease. All rights reserved.\n\njtd.amegroups.com\n\nJ Thorac Dis 2018;10(3):1689-1695\n\nJournal of Thoracic Disease, Vol 10, No 3 March 2018\n2014;1:CD009593. 7. Denkinger CM, Schumacher SG, Boehme CC, et al.\nXpert MTB/RIF assay for the diagnosis of extrapulmonary tuberculosis: a systematic review and meta-analysis. Eur Respir J 2014;44:435-46. 8. World Health Oragnization. WHO endorses new rapid tuberculosis test. Available online: http://www.who.int/ mediacentre/news/releases/2010/tb_test_20101208/en/. Geneva, Switzerland: WHO; 2010. 9. World Health Organization. Xpert MTB/RIF assay for the diagnosis of pulmonary and extrapulmonary TB in adults and children Policy update. Available online: http://www. who.int/tb/publications/xpert-mtb-rif-assay-diagnosispolicy-update/en/. Geneva, Switzerland: WHO, 2013. 10. Bonnet M, Ramsay A, Gagnidze L, et al. Reducing the number of sputum samples examined and thresholds for positivity: an opportunity to optimise smear microscopy. Int J Tuberc Lung Dis 2007;11:953-8. 11. van Hoorn R, Jaramillo E, Collins D, et al. The Effects of Psycho-Emotional and Socio-Economic Support for Tuberculosis Patients on Treatment Adherence and Treatment Outcomes - A Systematic Review and MetaAnalysis. PLoS One 2016;11:e0154095. 12. International Union against Tuberculosis and Lung Disease. Technical guide: sputum examination for tuberculosis by direct microscopy in low income countries, 5th ed. Paris, France: International Union against Tuberculosis and Lung Disease, 2000. 13. Harausz E, Lusiba JK, Nsereko M, et al. Comparison of MGIT and Myco/F lytic liquid-based blood culture systems for recovery of Mycobacterium tuberculosis from pleural fluid. J Clin Microbiol 2015;53:1391-4. 14. Albert H, Nathavitharana RR, Isaacs C, et al. Development, roll-out and impact of Xpert MTB/RIF for tuberculosis: what lessons have we learnt and how can we do better? Eur Respir J 2016;48:516-25. 15. Diagnostics FFI. Negotiated prices for Xpert MTB/ RIF and FIND country list. Available online: http:// www.finddiagnostics.org/about/what_we_do/successes/ findnegotiated-prices/xpert_mtb_rif.html. 2013. 16. You JH, Lui G, Kam KM, Lee NL. Cost-effectiveness\n\n1695\nanalysis of the Xpert MTB/RIF assay for rapid diagnosis of suspected tuberculosis in an intermediate burden area. J Infect 2015;70:409-14. 17. Cowan JF, Chandler AS, Kracen E, et al. Clinical Impact and Cost-effectiveness of Xpert MTB/RIF Testing in Hospitalized Patients With Presumptive Pulmonary Tuberculosis in the United States. Clin Infect Dis 2017;64:482-9. 18. Wikman-Jorgensen PE, Llenas-Garcia J, Perez-Porcuna TM, et al. Microscopic observation drug-susceptibility assay vs. Xpert((R)) MTB/RIF for the diagnosis of tuberculosis in a rural African setting: a cost-utility analysis. Trop Med Int Health 2017;22:734-43. 19. Vassall A, Siapka M, Foster N, et al. Cost-effectiveness of Xpert MTB/RIF for tuberculosis diagnosis in South Africa: a real-world cost analysis and economic evaluation. Lancet Glob Health 2017;5:e710-9. 20. Dorman SE, Schumacher SG, Alland D, et al. Xpert MTB/ RIF Ultra for detection of Mycobacterium tuberculosis and rifampicin resistance: a prospective multicentre diagnostic accuracy study. Lancet Infect Dis 2018;18:76-84. 21. Cowan JF, Chandler AS, Kracen E, et al. Clinical Impact and Cost-effectiveness of Xpert MTB/RIF Testing in Hospitalized Patients With Presumptive Pulmonary Tuberculosis in the United States. Clin Infect Dis 2017;64:482-9. 22. Chang KC, Yew WW. Management of difficult multidrugresistant tuberculosis and extensively drug-resistant tuberculosis: update 2012. Respirology 2013;18:8-21. 23. Shin SS, Keshavjee S, Gelmanova IY, et al. Development of extensively drug-resistant tuberculosis during multidrug-resistant tuberculosis treatment. Am J Respir Crit Care Med 2010;182:426-32. 24. Vassall A, van Kampen S, Sohn H, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: a cost-effectiveness analysis. PLoS Med 2011;8:e1001120. 25. Schnippel K, Meyer-Rath G, Long L, et al. Scaling up Xpert MTB/RIF technology: the costs of laboratory- vs. clinic-based roll-out in South Africa. Trop Med Int Health 2012;17:1142-51.\n\nCite this article as: Wang G, Wang S, Jiang G, Fu Y, Shang Y, Huang H. Incremental cost-effectiveness of the second Xpert MTB/RIF assay to detect Mycobacterium tuberculosis. J Thorac Dis 2018;10(3):1689-1695. doi: 10.21037/jtd.2018.02.60\n\n\u00a9 Journal of Thoracic Disease. All rights reserved.\n\njtd.amegroups.com\n\nJ Thorac Dis 2018;10(3):1689-1695\n\n\n",
    "authors": [
      "Guirong Wang",
      "Shuqi Wang",
      "Guanglu Jiang",
      "Yuhong Fu",
      "Yuanyuan Shang",
      "Hairong Huang"
    ],
    "doi": "10.21037/jtd.2018.02.60",
    "year": null,
    "item_type": "journalArticle",
    "url": "http://jtd.amegroups.com/article/view/19447/15515"
  },
  {
    "key": "X8HRQHSD",
    "title": "Screening for Tuberculosis Among Adults Newly Diagnosed With HIV in Sub-Saharan Africa: A Cost-Effectiveness Analysis",
    "abstract": "Objective\u2014New tools, including light emitting diode (LED) fluorescence microscopy and the molecular assay Xpert MTB/RIF\u00ae offer increased sensitivity for TB in persons with HIV but come with higher costs. Using operational data from rural Malawi we explored the potential costeffectiveness of on-demand screening for TB in low-income countries of sub-Saharan Africa. Design & Methods\u2014Costs were empirically collected in four clinics and one hospital using a micro-costing approach, through direct interview and observation from the national TB program perspective. Using decision analysis newly diagnosed persons with HIV were modeled as being screened by one of three strategies: Xpert, LED or standard of care (i.e., at the discretion of the treating physician).\nResults\u2014Cost-effectiveness of TB screening among persons newly diagnosed with HIV was largely determined by two factors: prevalence of active TB among patients newly diagnosed with HIV and volume of testing. In facilities screening at least 50 people with a 6.5% prevalence of TB, or at least 500 people with a 2.5% TB prevalence, screening with Xpert is likely to be costeffective. At lower prevalence \u2013 including that observed in Malawi \u2013 LED microscopy may be the preferred strategy, whereas in settings of lower TB prevalence or small numbers of eligible patients, no screening may be reasonable (such that resources can be deployed elsewhere).\nConclusions\u2014TB screening at the point of HIV diagnosis may be cost-effective in low-income countries of sub-Saharan Africa, but only if a relatively large population with high prevalence of TB can be identified for screening.",
    "full_text": "Author Manuscript\n\nAuthor Manuscript\n\nHHS Public Access\nAuthor manuscript\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01. Published in final edited form as:\nJ Acquir Immune Defic Syndr. 2015 September 1; 70(1): 83\u201390. doi:10.1097/QAI.0000000000000712.\nScreening for tuberculosis among adults newly diagnosed with HIV in sub-Saharan Africa: a cost-effectiveness analysis\nAlice A. Zwerling, PhD, MSc1,*, Maitreyi Sahu, MPH1,*, Lucky G. Ngwira, BSc2, McEwen Khundi, BSc2, Tina Harawa, BSc2, Elizabeth L. Corbett, FRCP, PhD2,3, Richard E. Chaisson, MD4, and David W. Dowdy, MD, PhD1 1Johns Hopkins Bloomberg School of Public Health Department of Epidemiology 2Malawi-Liverpool Wellcome Trust Research Programme 3London School of Hygiene & Tropical Medicine 4Johns Hopkins Center for Tuberculosis Research, Johns Hopkins University\nAbstract\nObjective\u2014New tools, including light emitting diode (LED) fluorescence microscopy and the molecular assay Xpert MTB/RIF\u00ae offer increased sensitivity for TB in persons with HIV but come with higher costs. Using operational data from rural Malawi we explored the potential costeffectiveness of on-demand screening for TB in low-income countries of sub-Saharan Africa.\nDesign & Methods\u2014Costs were empirically collected in four clinics and one hospital using a micro-costing approach, through direct interview and observation from the national TB program perspective. Using decision analysis newly diagnosed persons with HIV were modeled as being screened by one of three strategies: Xpert, LED or standard of care (i.e., at the discretion of the treating physician).\nResults\u2014Cost-effectiveness of TB screening among persons newly diagnosed with HIV was largely determined by two factors: prevalence of active TB among patients newly diagnosed with HIV and volume of testing. In facilities screening at least 50 people with a 6.5% prevalence of TB, or at least 500 people with a 2.5% TB prevalence, screening with Xpert is likely to be costeffective. At lower prevalence \u2013 including that observed in Malawi \u2013 LED microscopy may be the preferred strategy, whereas in settings of lower TB prevalence or small numbers of eligible patients, no screening may be reasonable (such that resources can be deployed elsewhere).\nConclusions\u2014TB screening at the point of HIV diagnosis may be cost-effective in low-income countries of sub-Saharan Africa, but only if a relatively large population with high prevalence of TB can be identified for screening.\nCorresponding author & requests for reprints: Dr. David Dowdy, Department of Epidemiology, E6531 Johns Hopkins Bloomberg School of Public Health, 615 N Wolfe Street, Baltimore, MD 21205 USA, ddowdy1@jhmi.edu. *These authors contributed equally to this work Presented at the 45th Union World Conference on Lung Health, Barcelona, Nov. 1, 2014 Conflicts of interest There are no conflicts of interest.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nZwerling et al.\n\nPage 2\n\nKeywords Costs and Benefits; Cost Effectiveness; Diagnostic tests; Tuberculosis; screening; Decision analysis\nIntroduction\nTuberculosis (TB) is the leading cause of death among people living with HIV and AIDS (PLWHA) worldwide1. Although antiretroviral therapy (ART) has greatly reduced the burden of TB mortality among PLWHA, the first six months after initiating ART remain a period of high risk for TB-associated mortality, likely due to prevalent subclinical TB at the time of ART initiation 2, 3. Intensified case finding (ICF) for TB is increasingly recommended for those newly diagnosed with HIV as a tool to reduce TB mortality 4, 5. Unfortunately, the test most widely used for TB ICF, namely sputum smear microscopy (SSM), has less than 40% sensitivity among PLWHA6, 7. New tools, including light emitting diode (LED) fluorescence microscopy and the molecular assay Xpert MTB/RIF\u00ae (\u201cXpert\u201d, Cepheid, Inc., Sunnyvale, CA, USA) offer increased sensitivity over traditional SSM and can be performed in under 2 hours. However, these diagnostic tests \u2013 especially if performed, at the point of care \u2013 are much more expensive than standard light microscopy, which is often performed in batches in centralized laboratories. We therefore use cost and operational data from a trial of LED microscopy and Xpert for point-of-treatment TB screening among people newly diagnosed with HIV in rural Malawi as a model to explore the potential cost-effectiveness of on-demand screening for TB in low-income countries of sub-Saharan Africa.\nMethods\nData Collection\nWe collected costs and operational data from a randomized trial of point-of-care screening for TB among people receiving a new HIV diagnosis in rural Malawi. The parent study is an ongoing cluster-randomized trial of Xpert versus LED microscopy for TB screening in 12 rural clinics, with both tests performed by nurses or trained assistants on the day of HIV diagnosis before the patient leaves the clinic. Consenting participants are initially evaluated by asking for any of four symptoms: cough of any duration, weight loss, fevers, and night sweats; those with any symptom are screened with LED microscopy or Xpert. Those diagnosed with TB are started immediately on treatment. Symptomatic patients who tested negative for TB were asked to return in one month, at which time they were screened for symptoms and tested a second time with either Xpert or LED microscopy if still symptomatic.\nHere, we use data from a cost and operational analysis performed at four study sites and the district hospital that serves the corresponding region, as a model evaluation for other similar sites in low-income countries of sub-Saharan Africa. These data were collected using a unitbased or \u201cingredients\u201d approach and included comprehensive budgetary reviews, interviews and logs of study staff (two staff members per clinic), direct observation of procedures, and\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nZwerling et al.\n\nPage 3\n\nprospective documentation of start-up costs. Other costs (e.g., TB treatment, ART) were obtained from Malawian notifications and published literature and were varied in sensitivity analysis. For this analysis, we also used study data on the prevalence of TB among eligible individuals. Diagnosis of TB in the study was made using SSM, LED microscopy or Xpert.\nWe coupled cost and operational data with data from the literature on diagnostic accuracy and likely patient outcomes to populate a decision analytic model of point-of-treatment TB screening among people obtaining a new diagnosis of HIV in different settings of TB prevalence and patient volume. The primary outcome was the incremental cost-effectiveness ratio (ICER), defined as the incremental cost per disability adjusted life year (DALY) averted comparing universal screening for TB (among people receiving a new HIV diagnosis) with Xpert or LED microscopy to a standard of care in which clinical judgment alone is used to refer patients for standard SSM.\n\nModel assumptions\nScreening for TB among persons newly diagnosed with HIV is not the current standard of care in many settings. Therefore, in addition to the Xpert and LED scenarios in which we assumed all patients newly diagnosed with HIV and at least one TB symptom would be screened, we also considered a standard of care scenario in which such patients are screened only at the discretion of the treating physician we assumed this was equivalent the probability of a future diagnosis of TB through the routine health system estimated at 66%8. Treatment regimens were indicated based on the diagnostic test result; patients diagnosed by smear, LED microscopy, or Xpert with a negative test for rifampin resistance were put on first-line therapy (two months of isoniazid, rifampin, ethambutol, and pyrazinamide, followed by four months of isoniazid and rifampin). Those diagnosed with rifampinresistant TB by Xpert would initiate second-line therapy. Patients with false negative Rif resistant results were assumed to start first-line therapy but have lower probability of success while false positive Rif resistant patients started second-line therapy, however the effects of unnecessary treatment are not explicitly modeled, apart from the costs. Patients with underlying multidrug-resistant (MDR) TB treated with first-line therapy were assumed to have worse outcomes. We assumed that 15% of patients testing positive would not complete a sufficient course of therapy to achieve cure9. Patients who were false negatives or lost to follow-up had the opportunity of a future TB diagnosis with sputum smear microscopy through the routine health system 8, 10, 11. Treatment failure and untreated TB were assumed to be universally fatal in this population of HIV-infected individuals.\nAccording to Malawian national guidelines, we assumed that patients with new HIV diagnoses received ART immediately if they had a CD4+ T-cell count \u2264350 cells/mm3 or a diagnosis of active TB. Those with CD4+ count >350 cells/mm3 were assumed to delay initiation of ART12. Model parameter values for tree probabilities, effectiveness measures and costs are included in Table 1. ART costs and DALYs averted were calculated over the individuals\u2019 lifetime assuming a life expectancy of 59.2 years for HIV-infected individuals on ART13.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nZwerling et al.\n\nPage 4\n\nEconomic Methods\nUnit costs for diagnosis and treatment included labor costs, material costs and overhead costs. Overhead costs were allocated based on discussions with experienced clinic staff and direct measurements of dimensions. Items such as building space, water utilization, housekeeping, and cleaning supplies were allocated based on proportional space required (approximately 1% of total clinic overhead costs allocated to Xpert or LED annually). Overhead staff costs and mobile airtime costs were allocated based on estimated staff time required to devote to Xpert or LED and were approximately 5% of overall overhead costs in these categories. Direct observations of supervisory staff were estimated based on interview, as actual allocation of study staff in this research setting was not reflective of typical operating scenarios. Staff time in reference to performing the symptom screen, LED and/or Xpert tests and follow-up was directly observed in time-motion studies. Equipment, supervisory costs and start-up costs were calculated for one year, and were allocated based on patient volume. Specific labour and consumable costs were estimated per individual test. Start up costs included recruitment and advertising costs for project and field managers, microscopy or Xpert training, petty cash and postage costs. We assumed these were costs incurred yearly due to high staff turnover, and were allocated based on patient volume. As patient volume increased, start-up costs per test decreased. Costs were measured from the healthcare perspective and inflated to 2010 US dollars (2010 selected as the year because of a devaluation of the Malawian currency in May 2012 during the period of data collection). DALYs were calculated without age weighting using standardized disability weights from the Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) 201014. To calculate the years of life lost (YLL) resulting from a TB death, we took the life expectancy of PLWHA in Malawi with a mean CD4 count at first presentation and the mean age at presentation of 31 years. All costs and DALYs were discounted at 3% per year, with sensitivity analysis for 0\u20137%15.\n\nSensitivity & Uncertainty analyses\nWe performed one-way sensitivity analysis on all model parameters by varying their values across broad yet plausible ranges. Upon finding that two parameters (testing volume and TB prevalence) were key drivers of cost-effectiveness, we conducted a two-way sensitivity analysis to include these parameters, reporting our primary results as a function of these two inputs. A probabilistic uncertainty analysis was performed using 10,000 Monte Carlo simulations of parameter values simultaneously drawn from beta distributions with upper and lower values as shown in Table 1 and a uniform alpha (shape) parameter of four. We report 95% uncertainty ranges as the 2.5th and 97.5th percentiles of those simulations. Scenario analyses were performed to specifically evaluate cost-effectiveness under conditions of high, medium, and low test volume; with and without ART; and across varying levels of symptom-driven diagnosis of TB in the standard of care. The decision analysis was performed using TreeAge Pro Version 2013 (TreeAge Software Inc., Williamstown, MA).\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nZwerling et al.\n\nPage 5\n\nResults\nResults under observed conditions\nUsing primary costing data from the parent trial we calculated the cost per test for LED microscopy and Xpert (including overhead costs, equipment, consumables and salaries). Cost per test was heavily influenced by test volume; therefore, three scenarios were separately considered: low volume of 50 tests/year, observed volume of 100 tests/year, and high volume of 1000 tests/year. Cost per test at observed volume was US$90.5 for Xpert and US$21.4 for LED; under high volume conditions these fell to US$24.8 for Xpert and US $4.05 for LED (Table 2). Volume-driven price reductions primarily reflect equipment costs for microscopes and Xpert systems (i.e., same equipment can be used to run more tests), but lower overhead costs also play an important role. These variations in test cost also strongly influenced initial estimates of cost-effectiveness, as shown in Table 3. Relative to the standard of care, both LED microscopy and Xpert \u2013 as performed at low volume \u2013 carried an incremental cost-effectiveness of over $1800 per DALY averted, above the per-capita gross domestic product of most low-income countries in sub-Saharan Africa. However, if higher patient volume could be achieved, incremental cost-effectiveness improved to $500\u2013$700 per DALY averted, and Xpert became more cost-effective than LED microscopy.\nDrivers of cost-effectiveness\nTo identify the factors most likely to drive cost-effectiveness estimates across different settings, we performed a one-way sensitivity analysis of all model parameters. Parameters that influenced model results by more than 10% are presented in the tornado diagram in Figure 1. Cost-effectiveness of intensified TB case finding among individuals newly diagnosed with HIV was largely determined by two factors: prevalence of active TB among patients newly diagnosed with HIV and volume of testing (which, as above, strongly influenced the unit cost of both Xpert and LED microscopy).\nWe present the impact of these two factors on cost-effectiveness in Figure 2, assuming a willingness to pay of $1417 USD per DALY averted, corresponding to the average per capita gross domestic product (GDP) of low-income countries in sub-Saharan Africa16. Scenarios in which LED microscopy was preferred to the standard of care include screening populations of 50 people per year with an expected TB prevalence of >3%, 100 people per year with an expected TB prevalence of >2%, or 500 people per year with an expected prevalence of >1.8%. Xpert was preferred to LED in screening populations of 50/year if prevalence were >6.5%, 100/year if prevalence were >3.5%, or 500/year if prevalence were >2.5%. More detailed results are given for \u201cfavorable,\u201d observed, and \u201cunfavorable\u201d scenarios in Table 3. The incremental cost-effectiveness of LED microscopy and Xpert relative to the standard of care varied by 12-fold and 20-fold across these scenarios, respectively \u2013 a range far greater than that induced by varying all other model parameters simultaneously (as seen in the corresponding uncertainty ranges).\nAs Xpert machines are rolled out across many settings for a wide range of patient groups, equipment and implementation costs may be reduced independent of HIV positive patient volume being screened for TB. To explore such scenarios we reduced equipment and\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nZwerling et al.\n\nPage 6\n\nmaintenance costs for the Xpert machine by half. In this scenario assuming patient volume in this population remains low (50 tests/year) ICERs for Xpert become almost identical to that of LED at $1816 per DALY averted, and at higher patient volumes of 100 or 1000 tests per year, the ICERs for Xpert becomes more cost-effective than LED at $1134 per DALY averted and $500 per DALY averted respectively.\nIn further sensitivity analyses, we assumed that the average number of DALYs averted from preventing a a death is only half the estimate that we obtained using our mean DALY estimate. In this sensitivity analysis, the effectiveness of the intervention is essentially reduced by half, leading to a doubling of the ICER in all scenarios.\n\nImpact of ART and symptom-driven TB diagnosis\nWe also considered the isolated costs and cost-effectiveness of LED microscopy and Xpert (i.e., excluding costs and impact of ART). Excluding ART costs only (while retaining the benefit of ART in terms of life expectancy) drove the cost per person screened, and thus incremental cost per DALY averted, down from $1216 to $917 (LED microscopy versus standard of care) and from $1615 to $1486 per DALY averted (Xpert versus standard of care).\nGiven the paucity of data on symptom-driven TB diagnosis in low-income countries of subSaharan Africa, we also varied the probability that TB cases missed by screening will get diagnosed and effectively treated before death from 36% to 80%. In a setting where the percentage of missed cases later become diagnosed through routine services decreases from 66% to 36% (i.e., screening becomes more important to avert eventual death), both LED and Xpert became more cost effective, with the incremental cost per DALY averted falling by about 40% (LED: $704, Xpert: $918). Conversely, if 80% of cases missed by initial screening are ultimately diagnosed through routine services, both LED and Xpert become less cost-effective, with an incremental cost per DALY averted of $1980 (LED) and $2651 (Xpert).\n\nDiscussion\nTuberculosis remains the leading cause of death among PLWHA in low-income countries of sub-Saharan Africa, but the cost-effectiveness of novel tests to screen for TB among adults receiving a new diagnosis of HIV remains uncertain. This economic evaluation, using data from a randomized trial in rural Malawi, suggests that the cost-effectiveness of TB screening at the point of HIV diagnosis in these settings depends critically on the volume of people being screened and the TB prevalence in the screened population. In facilities that can screen at least 50 people with a 6.5% prevalence of TB, or at least 500 people with a 2.5% TB prevalence, point-of-diagnosis TB screening with Xpert is likely to be cost-effective at a willingness to pay of per-capita GDP per DALY averted. At somewhat lower prevalence \u2013 including that observed in Malawi \u2013 LED microscopy may be the preferred screening strategy, whereas in settings of lower TB prevalence or small numbers of eligible patients, no screening may be reasonable such that resources can be deployed to other interventions for PLWHA. These results provide important guidance to low-income countries in sub-\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nZwerling et al.\n\nPage 7\nSaharan Africa as they contemplate the most appropriate approaches to implementing novel TB diagnostic test for screening among people newly diagnosed with HIV.\nAs a threshold of less than three times per-capita GDP per DALY averted is sometimes recommended as \u201ccost-effective\u201d, using the \u201chighly cost-effective\u201d threshold of one GDP per capita particularly in one of the lowest-income countries in Africa (per-capita GDP of $360 in 2010 16), represents a conservative approach.17 Even at the \u201ccost-effective\u201d threshold of three times GDP per capita for Malawi ($1080), the ICER is higher than the threshold; however, , screening for TB would be considered cost-effective in most countries (for example, neighboring Tanzania, per-capita GDP is $525 in 2010).16, 17\nLogistical feasibility is an important consideration in rolling out any TB screening strategy. In the rural Malawian setting, sputum smears are sent off-site and performed at centralized laboratories with high sample throughput; in other sub-Saharan African settings, LED microscopy (and even Xpert) at the point of HIV diagnosis may already be available, thereby markedly improving the cost-effectiveness of screening. In such settings, the incremental cost of LED microscopy (or Xpert, where already available) for screening may largely be limited to consumables and some staff costs (e.g., $1\u2013$2 per test for LED microscopy and $16\u2013$17 for Xpert), and incremental cost-effectiveness of TB screening \u2013 given that these tests are already available and being used for symptom-driven TB diagnosis \u2013 may approach that of the high-volume scenario shown here. Where implementation of TB screening would require new equipment, however, testing volume is a critical consideration. Indeed, in rural settings such as this one, technical and logistical problems (e.g., electrical outages, equipment breakdown, theft) may further reduce the effective testing volume and drive up costs. These findings emphasize that there is no \u201cone size fits all\u201d solution to TB screening in low-income sub-Saharan Africa; settings with high TB prevalence, high patient volumes, or existing capacity for Xpert or LED microscopy may find universal TB screening of symptomatic patients with new HIV diagnoses to be highly cost-effective, whereas small centers with lower TB prevalence and little existing capacity may be justified in deploying their resources elsewhere.\nIn a similar analysis in South Africa (where Xpert-based TB diagnosis has been scaled up throughout the country, and per-capita GDP is over 10 times that of the average used here), Andrews et al identified the prevalence of active TB as the most influential driver of costeffectiveness when considering TB screening for people with newly-diagnosed HIV 18. This analysis did not vary test volume as widely and concluded that TB screening was likely to be highly cost-effective in South Africa. Our results suggest where this may also be true in lower-income countries of sub-Saharan Africa.\nOther cost-effectiveness analyses of TB diagnosis in southern Africa have found ART costs to be very influential19, 20. In our analysis, by contrast, ART costs were less important because of the restriction of the population to those being newly diagnosed with ART and the corresponding assumption that all members of the population would eventually start ART. Furthermore, the prevalence of TB in this screening population was lower than among individuals seeking diagnosis for cough or other TB symptoms, making the cost of TB\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nZwerling et al.\n\nPage 8\n\nscreening relatively more expensive compared to the cost of ART (the cost of which continues to decline in the African setting).\nAs with any model-based economic evaluation, this analysis has certain important limitations. Data on symptom-driven TB diagnosis and outcomes of untreated TB in lowincome sub-Saharan Africa are very sparse. In the absence of convincing data, we assumed that 40% of people with TB would be captured at the time of HIV diagnosis, that 66% of people missed by initial screening would eventually be diagnosed before death, and that untreated TB was universally fatal among PLWHA. To the extent that these assumptions are not reflective of specific settings in low-income sub-Saharan Africa, our estimates of costeffectiveness may be incorrect. On wide variation of these parameters,, untreated TB mortality had little influence, but symptom-driven TB diagnosis patterns were important. Our results should therefore be interpreted with caution where symptom-driven TB diagnosis patterns are very different from those assumed here. Our cost and operational data from rural Malawi may not directly generalize to other countries or urban centers, and care should be taken when generalizing these findings to other countries in sub-Saharan Africa. However, sensitivity analyses suggest that test volume and TB prevalence are likely to be key considerations in most settings. In many settings (including Malawi), the current recommended algorithm suggests Xpert testing only after negative smear result. However, when implementing a test such as Xpert for TB screening (rather than diagnosis) in a rural clinic (as opposed to a centralized lab), it is unlikely to be realistic to have patients wait hours for their smear result, and another two hours for Xpert, particularly given the low probability of smear positivity in this screening patient population, where the prevalence of TB is less than 3% and therefore 98\u201399% of patients will be smear negative. Finally, our model does not account for potential reduction of TB transmission due to earlier TB diagnosis; thus, our estimates of TB screening cost-effectiveness may be somewhat conservative. Future studies could consider inclusion of transmission and collection of data to inform corresponding assumptions.\n\nConclusion\nThis analysis demonstrates that test volume and TB prevalence are key drivers of costeffectiveness when considering screening people newly diagnosed with HIV for TB using LED microscopy or Xpert in low-income sub-Saharan Africa. In settings of high patient volume and TB prevalence \u2013 or existing capacity and low logistical barriers \u2013 Xpert may be a highly cost-effective method to screen all people with new HIV diagnoses and any TB symptoms. In settings of moderate volume and TB prevalence, LED microscopy may be the preferred option, and in low-volume peripheral centers with high logistical barriers, resources may be better allocated to other interventions (which could include transport of sputum specimens to other centers). Future studies \u2013 including primary results from the parent trial \u2013 could improve estimates of long-term effectiveness of such TB screening strategies. In assessing the cost-effectiveness of TB screening among PLWHA in lowincome countries of sub-Saharan Africa, evaluations should move away from a \u201cone size fits all\u201d approach and toward consideration of key drivers including patient volume, TB prevalence in the screened population, existing capacity, and logistical feasibility.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nZwerling et al.\n\nPage 9\n\nAcknowledgements\nAuthors would like to thank all those who made this work possible, including the staff who graciously provided their time and assistance in costing efforts.\nThis work is supported by grants from the National Institute of Allergy and Infectious Diseases, NIH: R01AI093316, AZ is supported by a fellowship from the Canadian Institutes of Health Research.\nDD, RC, EC and MS conceived and designed the experiments; MS, LN and TH collected the data. DD, MS, MK and AZ analyzed the data, AZ wrote the first draft of the manuscript and DD and AZ wrote the manuscript.\nReferences\n1. UNAIDS. 2013 UNAIDS Report on the global AIDS epidemic. Geneva, Switzerland: UNAIDS; 2013.\n2. Lawn SD, Harries AD, Anglaret X, Myer L, Wood R. Early mortality among adults accessing antiretroviral treatment programmes in sub-Saharan Africa. Aids. 2008 Oct 1; 22(15):1897\u20131908. [PubMed: 18784453]\n3. Lawn SD, Kranzer K, Edwards DJ, McNally M, Bekker LG, Wood R. Tuberculosis during the first year of antiretroviral therapy in a South African cohort using an intensive pretreatment screening strategy. Aids. 2010 Jun 1; 24(9):1323\u20131328. [PubMed: 20386425]\n4. De Cock KM, Chaisson RE. Will DOTS do it? A reappraisal of tuberculosis control in countries with high rates of HIV infection. The international journal of tuberculosis and lung disease : the official journal of the International Union against Tuberculosis and Lung Disease. 1999 Jun; 3(6): 457\u2013465.\n5. Havlir DV, Getahun H, Sanne I, Nunn P. Opportunities and challenges for HIV care in overlapping HIV and TB epidemics. JAMA : the journal of the American Medical Association. 2008 Jul 23; 300(4):423\u2013430. [PubMed: 18647985]\n6. Getahun H, Harrington M, O'Brien R, Nunn P. Diagnosis of smear-negative pulmonary tuberculosis in people with HIV infection or AIDS in resource-constrained settings: informing urgent policy changes. Lancet. 2007 Jun 16; 369(9578):2042\u20132049. [PubMed: 17574096]\n7. Long R, Scalcini M, Manfreda J, Jean-Baptiste M, Hershfield E. The impact of HIV on the usefulness of sputum smears for the diagnosis of tuberculosis. American journal of public health. 1991 Oct; 81(10):1326\u20131328. [PubMed: 1928536]\n8. Organization, WH. Malawi: Tuberculosis profile 2011. Geneva, Switzerland: WHO; 2011. 9. MacPherson P, Corbett EL, Makombe SD, et al. Determinants and consequences of failure of\nlinkage to antiretroviral therapy at primary care level in Blantyre, Malawi: a prospective cohort study. PloS one. 2012; 7(9):e44794. [PubMed: 22984560] 10. Espinal MA, Kim SJ, Suarez PG, et al. Standard short-course chemotherapy for drug-resistant\ntuberculosis: treatment outcomes in 6 countries. JAMA : the journal of the American Medical Association. 2000 May 17; 283(19):2537\u20132545. [PubMed: 10815117] 11. Yew WW, Leung CC. Management of multidrug-resistant tuberculosis: Update 2007. Respirology. 2008 Jan; 13(1):21\u201346. [PubMed: 18197909] 12. Mellors JW, Munoz A, Giorgi JV, et al. Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection. Annals of internal medicine. 1997 Jun 15; 126(12):946\u2013954. [PubMed: 9182471] 13. Mills EJ, Bakanda C, Birungi J, et al. Life expectancy of persons receiving combination antiretroviral therapy in low-income countries: a cohort analysis from Uganda. Annals of internal medicine. 2011 Aug 16; 155(4):209\u2013216. [PubMed: 21768555] 14. Salomon JA, Vos T, Hogan DR, et al. Common values in assessing health outcomes from disease and injury: disability weights measurement study for the Global Burden of Disease Study 2010. Lancet. 2012 Dec 15; 380(9859):2129\u20132143. [PubMed: 23245605] 15. WHO. Making choices in health: WHO guide to cost-effectiveness analysis. Geneva, Switzerland: World Health Organization; 2003. 16. World Bank Open Data. The World Bank. 2012.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nAuthor Manuscript\n\nZwerling et al.\n\nPage 10\n17. Marseille E, Larson B, Kazi DS, James KG, Rosen S. Thresholds for the cost\u2013effectiveness of interventions: alternative approaches. Bull World Health Organ. 2015; 93:118\u2013124. [PubMed: 25883405]\n18. Andrews JR, Lawn SD, Rusu C, et al. The cost-effectiveness of routine tuberculosis screening with Xpert MTB/RIF prior to initiation of antiretroviral therapy: a model-based analysis. Aids. 2012 May 15; 26(8):987\u2013995. [PubMed: 22333751]\n19. Andrews JR, Lawn SD, Dowdy DW, Walensky RP. Challenges in evaluating the cost-effectiveness of new diagnostic tests for HIV-associated tuberculosis. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2013 Oct; 57(7):1021\u20131026. [PubMed: 23788239]\n20. Menzies NA, Cohen T, Lin HH, Murray M, Salomon JA. Population health impact and costeffectiveness of tuberculosis diagnosis with Xpert MTB/RIF: a dynamic simulation and economic evaluation. PLoS medicine. 2012; 9(11):e1001347. [PubMed: 23185139]\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nAuthor Manuscript\n\nZwerling et al.\n\nPage 11\n\nFigure 1. Tornado diagram depicting one-way sensitivity analysis of model parameters ICER calculated for screening arm with Xpert in refernce to standard of care. Bars in blue represent parameter values that increase, while bars in red reflect parameter values that decrease. Spread of the bars reflects the variability in ICER found when parameter was varied across range of interest. Bars on the left indicate ICERs were smaller, and bars on the right represent ICERS that were larger compared with default estimates.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nAuthor Manuscript\n\nZwerling et al.\n\nPage 12\n\nFigure 2. Impact of test volume and TB prevalence on cost-effectiveness of TB screening with Xpert and LED The shaded area corresponds to the combination of values for test volume and TB prevalence where one screening strategy is more cost-effective compared with the others, using a threshold of $1417 USD corresponding to the average GDP per capita of developing countries in Sub-Saharan Africa. For example the values of TB prevalence and test volume that fall in the red shaded area indicate a setting where the standard of care is more costeffectives compared with LED or Xpert, values in the yellow zone denote settings where LED is more cost-effective, and values in the teal zone correspond to settings where Xpert is preferred.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nZwerling et al.\n\nPage 13\n\nTable 1 Cohort model inputs including costs, effectiveness measures and tree probabilities\n\nModel Parameter Inputs\nCost Parameters Cost of 1st Line TB Treatment\n\nParameter Value\n\nRange (Min, Max)\n\n$185\n\n(154, 236)\n\nData Source\n20,21\n\nCost of 2nd Line TB Treatment\n\n$1759\n\n(1353, 2351)\n\n20,21\n\nLifetime cost of ART started immediately\n\n$2563\n\n(0, 4000)\n\nCHAI Treatment costs for HIV (MATCH study) 22\n\nCost of a symptom screen\n\n$0.20\n\n(0, 1)\n\nChepetsa costing study\n\nCost of standard smear (1000 tests/year at peripheral lab)\n\n$4.06\n\n(1, 10)\n\nChepetsa costing study\n\nEffectiveness parameters (Including effectiveness of ART) DALY \u2013 1st line treatment success\n\n\u22121.53\n\n(\u22122.53, \u22120.5)\n\n14,23\n\nDALY \u2013 2nd line treatment success\n\n\u22121.9878\n\n(\u22122.99, \u22120.99)\n\n14,23\n\nDALY \u2013 Death\n\n\u221223.8967\n\n(\u221227.8967, \u221229.90)\n\n9,13,14\n\nDALY \u2013 No TB, ART initiation (delayed and immediate)\n\n\u22121.2710\n\n(\u22122.2710, \u22120.27)\n\n9,13,14\n\nCohort proportions\n\nProbability of active TB among patients newly diagnosed with\n\n.024\n\nHIV in Malawi\n\n(0.01, 0.06)\n\nChepetsa, facility report\n\nProbability that symptomatic patients would receive smear\n\n.40\n\n(0,1)\n\nresults without screening\n\nProbability that CD4+ <350 at time of screening\n\n.60\n\n(0.5, 0.75)\n\n24\n\nProbability that missed TB case is later diagnosed with TB\n\n0.66\n\n(0.61, 0.71)\n\n8\n\noutside of screening\n\nProbability of loss to follow up during TB treatment\n\n.15\n\n(0.1, 0.21)\n\n23\n\nProbability of Rifampicin resistance among patients with TB\n\n.004\n\n(0.0014, 0.01)\n\n8\n\nProbability of death among TB patients with HIV given treatment failure, missed diagnosis, or loss to follow-up\n\n1\n\n(0.63, 1)\n\nAssumption 25\n\nProbability of treatment success, MDR-TB treated with 1st line\n\n.47\n\n(0.42, 0.52)\n\n10\n\ndrugs\n\nProbability of Treatment success, smear-negative TB treated\n\n.8\n\n(0.72, 0.88)\n\n8\n\nwith 1st line therapy\n\nProbability of Treatment success, smear positive TB treated with\n\n.87\n\n(0.78, 0.96)\n\n8\n\n1st line therapy\n\nProbability of treatment success, 2nd line\n\n.80\n\n(0.7, 0.9)\n\n11\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nAuthor Manuscript\n\nZwerling et al.\n\nPage 14\n\nModel Parameter Inputs\nDiagnostic Parameters Sensitivity of Xpert for RIF resistance\n\nParameter Value\n\nRange (Min, Max)\n\n.976\n\n(0.94, 0.99)\n\nData Source\n26,27\n\nSensitivity of Xpert for LED-negative TB\n\n.718\n\n(0.29, 0.79)\n\n26\u201328\n\nSensitivity of Xpert for LED-positive TB\n\n.977\n\n(0.92, 0.99)\n\n26,27\n\nSensitivity of LED for TB among Smear negatives\n\n0.095\n\n(0.09, 0.2)\n\n8,25,29\n\nSensitivity of LED for TB among Smear positives\n\n1\n\n(0, 1)\n\n29\n\nSpecificity of GXP\n\n.992\n\n(0.98, 0.996)\n\n26,27\n\nSpecificity of GXP for RIF resistance\n\n1\n\n(0.9, 1)\n\n26,27\n\nSpecificity of LED\n\n0.944\n\n(0.92, 0.96)\n\n30\n\nSensitivity of standard smear among LED positive people with\n\n.37\n\n(0.36, 0.7)\n\n8,25\n\nHIV\n\nSpecificity of smear among people with HIV\n\n.8\n\n(0, 1)\n\nAbbreviations: TB: active tuberculosis disease, GXP: Gene Xpert, LED: light emitting diode fluorescence microscopy, MDR: multi drug resistant tuberculosis, ART: antiretroviral therap\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nUnit cost of Xpert and LED by input type and annual test volume\n\nTable 2\n\nZwerling et al.\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nInput Type\n\nCosts per Test (US$ 2010)\n\nLED\n\nXpert\n\n50 /yr 100/yr 1000/yr 50/yr 100/yr 1000/yr\n\nOverhead*\n\n$21.99 $10.99 $1.10 $22.8 $11.4 $1.14\n\nEquipment**\n\n$13.77 $6.89 $0.68 $120.63 $60.32 $6.04\n\nStaff\n\n$2.67 $1.89 $1.19 $2.57 $1.79 $1.09\n\nConsumables & Reagents *** $1.02 $1.02 $1.02 $16.44 $16.44 $16.44\n\nTraining costs\n\n$1.22 $0.61 $0.06 $1.07 $0.57 $0.06\n\nTotal\n\n$40.7 $21.4 $4.05 $163.5 $90.5 $24.8\n\n*Approximately 75% of overhead costs are supervisory staff costs assuming 1 day/week at clinic to provide monitoring and ongoing training, 18% of overhead costs for Xpert and 14% of overhead costs for LED comprises building space, remaining 7% and 11% includes water, mobile airtime, other overhead staff costs, housekeeping and cleaning supplies.\n**Approximately 77% of equipment costs are the Xpert machine and computer, 20% include the solar panel. Approximately 87% of LED equipment costs are due to the microscope, and remaining equipment costs are for the battery.\n*** The Xpert cartridge accounts for 58% of Xpert consumables.\n\nPage 15\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nInfluence of test volume on cost-effectiveness estimates\n\nTable 3\n\nZwerling et al.\n\nJ Acquir Immune Defic Syndr. Author manuscript; available in PMC 2016 September 01.\n\nScreening with LED\n\nScreening with Xpert\n\nTest Volume (tests/year)\n\n50\n\n100\n\n1000\n\n50\n\n100\n\n1000\n\nObserved screening conditions: (prevalence of TB: 2.4%)\n\n\u00a0\u00a0\u00a0\u2022Cost per person screenedincremental to std of care\n\n$116 ($43 to $298)\n\n$78 (CS to $261)\n\n$45 (CS to $226)\n\n$343 ($302 to $516)\n\n$199 ($123\u2013$371)\n\n$69 ($16 to $240)\n\n\u00a0\u00a0\u00a0\u2022DALYs averted per person screened *\n\n0.064 (0.05 to 0.075)\n\n0.122 (0.10 to 0.138)\n\n\u00a0\u00a0\u00a0ICER** ($ per DALY averted)\n\n$1808\n\n$1216\n\n$699\n\n$2809\n\n$1615\n\n$564\n\n($567 to $$6023) (CS to $5313) (CS to $4782) ($2191 to $5039) ($898 to $3644) ($113 to $2386)\n\n\u00a0\u00a0\u00a0ICER** (Reference standard is LED)\n\n$2,205 ($507 to $2379)\n\nUnfavorable to screening: (low prevalence of TB: 1%)\n\n\u00a0\u00a0\u00a0\u2022Cost per person screened incremental to std of care\n\n$112 ($29 to $287)\n\n$336 ($290 to $501)\n\n\u00a0\u00a0\u00a0\u2022DALYs averted per person screened *\n\n0.027 (0.022 to 0.028)\n\n0.051 (0.045 to 0.054)\n\n\u00a0\u00a0\u00a0\u2022ICER** ($ per DALY averted)\n\n$4190 ($1036 to $12909)\n\n$6606 ($5398 to $11105)\n\nFavorable to screening: (high prevalence of TB: 6%)\n\n\u00a0\u00a0\u00a0\u2022Cost per person screened incremental to std of care\n\n$55 (CS to $229)\n\n$91 ($31 to $261)\n\n\u00a0\u00a0\u00a0\u2022DALYs averted per person screened incremental to std of care\n\n0.16 (0.095 to 0.226)\n\n0.305 (0.21 to 0.399)\n\n\u00a0\u00a0\u00a0\u2022ICER** ($ per DALY averted)\n\n$347 (CS to $2416)\n\n$298 ($77 to $1241)\n\nUncertainty ranges calculated using probabilistic sensitivity analysis, monte-carlo simulations over 10,000 runs. * Incremenal to standard of care; **ICER: Incremental cost-effectiveness ratio: Cost ($) per DALY averted, calculated incremental to cost/DALY averted with standard of care unless otherwise noted. CS: Cost saving compared with reference arm\n\nPage 16\n\n\n",
    "authors": [
      "Alice A. Zwerling",
      "Maitreyi Sahu",
      "Lucky G. Ngwira",
      "McEwen Khundi",
      "Tina Harawa",
      "Elizabeth L. Corbett",
      "Richard E. Chaisson",
      "David W. Dowdy"
    ],
    "doi": "10.1097/QAI.0000000000000712",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://journals.lww.com/00126334-201509010-00011"
  },
  {
    "key": "AFEZ8TW8",
    "title": "Scaling up Xpert MTB/RIF technology: the costs of laboratory\u2010 <i>vs.</i> clinic\u2010based roll\u2010out in South Africa",
    "abstract": "Abstract\n            \n              Objective\u2002\n              The World Health Organization recommends using Xpert MTB/RIF for diagnosis of pulmonary tuberculosis (PTB), but there is little evidence on the optimal placement of Xpert instruments in public health systems. We used recent South African data to compare the cost of placing Xpert at points of TB treatment (all primary clinics and hospitals) with the cost of placement at sub\u2010district laboratories.\n            \n            \n              Methods\u2002\n              We estimated Xpert\u2019s cost/test in a primary clinic pilot and in the pilot phase of the national Xpert roll\u2010out to smear microscopy laboratories; the expected future volumes for each of 223 laboratories or 3799 points of treatment; the number and cost of Xpert instruments required and the national cost of using Xpert for PTB diagnosis for each placement scenario in 2014.\n            \n            \n              Results\u2002\n              In 2014, South Africa will test 2.6 million TB suspects. Laboratory placement requires 274 Xpert instruments, while point\u2010of\u2010treatment placement requires 4020 instruments. With an Xpert cartridge price of $14.00, the cost/test is $26.54 for laboratory placement and $38.91 for point\u2010of\u2010treatment placement. Low test volumes and a high number of sites are the major contributors to higher point\u2010of\u2010treatment costs. National placement of Xpert at laboratories would cost $71 million/year; point\u2010of\u2010treatment placement would cost $107 million/year, 51% more.\n            \n            \n              Conclusion\u2002\n              Placing Xpert technology at points of treatment is substantially more expensive than placing the instruments in smear microscopy laboratories. The incremental benefits of point\u2010of\u2010treatment placement, in terms of better patient outcomes, will have to be equally substantial to justify the additional cost to the national health budget.\n            \n          , \n            \n              \n              \n                Objectif:\u2002\n                L\u2019Organisation mondiale de la Sant\u00e9 recommande l\u2019utilisation du test XpertMTB/RIF pour le diagnostic de la tuberculose pulmonaire (TBP), mais il y a peu de donn\u00e9es sur l\u2019impl\u00e9mentation optimale des outils Xpert dans les syst\u00e8mes de sant\u00e9 publique. Nous avons utilis\u00e9 des donn\u00e9es r\u00e9centes de l\u2019Afrique du Sud pour comparer le co\u00fbt de l\u2019impl\u00e9mentation de Xpert dans les points de traitement de la TB (toutes les cliniques de soins primaires et les h\u00f4pitaux) au co\u00fbt de l\u2019impl\u00e9mentation dans les laboratoires de sous\u2010district.\n              \n              \n                M\u00e9thodes:\u2002\n                Nous avons estim\u00e9 le co\u00fbt par test de Xpert dans un projet pilote de clinique primaire et dans la phase pilote du d\u00e9ploiement national de Xpert dans les laboratoires de microscopie des frottis, les futurs volumes attendus de tests pour chacun des 223 laboratoires ou des 3.799 points de traitement, le nombre et le co\u00fbt des tests Xpert n\u00e9cessaires et le co\u00fbt national de l\u2019utilisation de Xpert pour le diagnostic de la TBP pour chaque sc\u00e9nario d\u2019impl\u00e9mentation en 2014.\n              \n              \n                R\u00e9sultats:\u2002\n                En 2014, l\u2019Afrique du Sud testera 2,6 millions de cas suspects TB. L\u2019impl\u00e9mentation en laboratoire n\u00e9cessite 274 outils Xpert tandis que l\u2019impl\u00e9mentation dans les points de traitement exige 4.020 outils. Avec un prix par cartouche Xpert de 14,00 $, le co\u00fbt par test est de 26,54 $ pour l\u2019impl\u00e9mentation en laboratoire et 38,91 $ pour celle dans les points de traitement. Des volumes de tests faibles et un nombre \u00e9lev\u00e9 de sites sont les principaux contributeurs \u00e0 la hausse des co\u00fbts pour les points de traitement. L\u2019impl\u00e9mentation nationale de Xpert dans les laboratoires co\u00fbterait 71 millions de dollars par an et 107 millions de dollars par an pour celle dans les points de traitement, i.e. 51% de plus.\n              \n              \n                Conclusion:\u2002\n                L\u2019impl\u00e9mentation de la technologie Xpert dans les points de traitement est nettement plus ch\u00e8re que celle dans les laboratoires de microscopie des frottis. Les avantages suppl\u00e9mentaires de l\u2019impl\u00e9mentation dans les points de traitement, en termes de r\u00e9sultats pour les patients, devraient \u00eatre aussi importants pour justifier le co\u00fbt suppl\u00e9mentaire pour le budget national de la sant\u00e9.\n              \n            \n          , \n            \n              \n              \n                Objetivo:\u2002\n                La Organizaci\u00f3n Mundial de la Salud recomienda utilizar Xpert MTB/RIF para el diagn\u00f3stico de la tuberculosis pulmonar (TBP), pero existe poca evidencia sobre la colocaci\u00f3n \u00f3ptima de los instrumentos Xpert dentro de los sistemas de salud p\u00fablicos. Hemos utilizado datos recientes de Sud\u00e1frica para comparar los costes de colocar Xpert en puntos de tratamiento de la TB (todos centros de atenci\u00f3n primaria y hospitales) con el coste de colocarlos en laboratorios sub\u2010distritales.\n              \n              \n                M\u00e9todos:\u2002\n                Hemos estimado el coste por prueba Xpert en un centro primario piloto y en la fase piloto del despliegue nacional del Xpert en los laboratorios con microscop\u00eda; los vol\u00famenes esperados en el futuro para cada uno de los 223 laboratorios o 3,799 puntos de tratamiento; el n\u00famero y coste de los instrumentos Xpert requeridos; y el coste nacional de utilizar Xpert para el diagn\u00f3stico de la TBP en cada escenario en el 2014.\n              \n              \n                Resultados:\u2002\n                En el 2014, Sud\u00e1frica realizar\u00e1 pruebas a 2.6 millones de pacientes con sospecha de TB. La colocaci\u00f3n de los equipos en los laboratorios requerir\u00eda de 274 instrumentos Xpert, mientras que para hacerlo en los puntos de tratamiento se necesitar\u00edan 4,020 instrumentos. Con un precio por cartucho Xpert de $14.00, el ratio coste/prueba es de $26.54 con los equipos colocados en los laboratorios y de $38.91 si los equipos est\u00e1n colocados en los puntos de tratamiento. Unos bajo volumen de pruebas y un gran n\u00famero de emplazamientos son los principales contribuyentes al mayor coste para los puntos de tratamiento. Colocar el Xpert en los laboratorios a nivel nacional costar\u00eda $71 millones/a\u00f1o; hacerlo en los puntos de tratamiento costar\u00eda $107 millones/a\u00f1o, un 51% m\u00e1s.\n              \n              \n                Conclusi\u00f3n:\u2002\n                Colocar la tecnolog\u00eda Xpert en los puntos de tratamiento es sustancialmente m\u00e1s caro que colocarlos en laboratorios con microscop\u00eda. Los beneficios incrementales de colocar los equipos en puntos de tratamiento, en t\u00e9rminos de mejores resultados para los pacientes, tendr\u00eda que ser sustancial para justificar el coste adicional para el presupuesto nacional de sanidad.",
    "full_text": "Tropical Medicine and International Health volume 17 no 9 pp 1142\u20131151 september 2012\n\ndoi:10.1111/j.1365-3156.2012.03028.x\n\nScaling up Xpert MTB/RIF technology: the costs of laboratoryvs. clinic-based roll-out in South Africa\nKathryn Schnippel1, Gesine Meyer-Rath1,2, Lawrence Long1, William MacLeod1,2, Ian Sanne1,2, Wendy S. Stevens3,4 and Sydney Rosen1,21\n1 Health Economics and Epidemiology Research Of\ufb01ce, Department of Medicine, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa\n2 Center for Global Health and Development, Boston University, Boston, MA, USA 3 Department of Molecular Medicine and Haematology, University of the Witwatersrand, Johannesburg, South Africa 4 National Health Laboratory Service, Johannesburg, South Africa\n\nAbstract\n\nobjective The World Health Organization recommends using Xpert MTB \u2044 RIF for diagnosis of pulmonary tuberculosis (PTB), but there is little evidence on the optimal placement of Xpert instruments in public health systems. We used recent South African data to compare the cost of placing Xpert at points of TB treatment (all primary clinics and hospitals) with the cost of placement at sub-district laboratories. methods We estimated Xpert\u2019s cost \u2044 test in a primary clinic pilot and in the pilot phase of the national Xpert roll-out to smear microscopy laboratories; the expected future volumes for each of 223 laboratories or 3799 points of treatment; the number and cost of Xpert instruments required and the national cost of using Xpert for PTB diagnosis for each placement scenario in 2014. results In 2014, South Africa will test 2.6 million TB suspects. Laboratory placement requires 274 Xpert instruments, while point-of-treatment placement requires 4020 instruments. With an Xpert cartridge price of $14.00, the cost \u2044 test is $26.54 for laboratory placement and $38.91 for point-oftreatment placement. Low test volumes and a high number of sites are the major contributors to higher point-of-treatment costs. National placement of Xpert at laboratories would cost $71 million \u2044 year; point-of-treatment placement would cost $107 million \u2044 year, 51% more. conclusion Placing Xpert technology at points of treatment is substantially more expensive than placing the instruments in smear microscopy laboratories. The incremental bene\ufb01ts of pointof-treatment placement, in terms of better patient outcomes, will have to be equally substantial to justify the additional cost to the national health budget.\n\nkeywords tuberculosis, diagnostics, economics, scale-up, middle income\n\nIntroduction\nIn 2010, national tuberculosis (TB) control programmes diagnosed and noti\ufb01ed approximately 65% of 8.8 million estimated TB cases globally and only 18% of an estimated 290 000 multidrug-resistant TB (MDR-TB) cases (World Health Organization 2011a). Earlier and improved detection of TB and MDR-TB, especially in high-burden countries, is thus an international health priority. The Xpert MTB \u2044 RIF (Xpert) assay for the GeneXpert (GX) platform (Cepheid, Sunnyvale, CA, USA), which provides both rapid and speci\ufb01c detection of Mycobacterium tuberculosis (MTB) and identi\ufb01cation of rifampicin (RIF)\nRe-use of this article is permitted in accordance with the Terms and Conditions set out at http://wileyonlinelibrary.com/online open#OnlineOpen_Terms.\n\ndrug resistance, is seen as one promising solution to this problem (Boehme et al. 2010; Helb et al. 2010; Small & Pai 2010; Van Rie et al. 2010). In late 2010, WHO recommended that Xpert be used as the initial diagnostic for persons suspected of MDR-TB or TB \u2044 HIV co-infection (Stop TB Department 2010).\nSouth Africa, one of the countries most heavily burdened by both TB \u2044 HIV co-infection and MDR-TB, was among the \ufb01rst to begin large-scale roll-out of Xpert (World Health Organization 2011a). South Africa has extensive laboratory capacity for the detection of TB, with more than 200 active smear microscopy laboratories and an average of 1.5 liquid culture laboratories and 1.4 line probe assay (LPA) laboratories per 5 million population, exceeding WHO targets for TB diagnostic capacity (World Health Organization 2011a). Because of the high prevalence of smear-negative TB, however, diagnosis can still take weeks\n\n1142\n\n\u00aa 2012 Blackwell Publishing Ltd\n\n13653156, 2012, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-3156.2012.03028.x by Readcube (Labtiva Inc.), Wiley Online Library on [20/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License\n\nTropical Medicine and International Health K. Schnippel et al. Costs of Xpert MTB/RIF placement\n\nvolume 17 no 9 pp 1142\u20131151 september 2012\n\n(Chihota et al. 2010), and many patients are lost to care while waiting for culture results (Boehme et al. 2011). In early 2011, in response to both the WHO recommendation and locally generated evidence about the potential bene\ufb01ts of Xpert, especially for the detection of smear-negative TB (Boehme et al. 2011; Scott et al. 2011), the National Health Laboratory Service (NHLS) of South Africa, working with the National Department of Health, developed a national roll-out plan for Xpert within South Africa (Smart 2011). Implementation of the second phase of this plan is now underway, following a successful pilot phase launched in March 2011 (Erasmus et al. 2011).\nWhile the capital expense and operational complexity of the liquid culture and LPA technologies have limited placement of these services to centralized TB reference laboratories, the WHO recommendation for the placement of Xpert technology is at \u2018health facility level (ideally district or sub-district level)\u2019 (World Health Organization 2011b). The South African roll-out plan calls for the placement of Xpert instruments in more than 200 existing sub-district laboratories that currently provide smear microscopy to multiple healthcare facilities within their catchment areas, which include regional, provincial and district hospitals, 24h community health centres and primary healthcare (PHC) clinics. Sputum samples to be tested with Xpert will be transported from the healthcare facilities to the smear microscopy laboratories, some of which are located within hospitals. Results will be returned on average within 1\u20133 working days but usually not on the same day as the sputum sample was collected. Thus, although the plan will greatly accelerate diagnosis of TB once samples have reached the laboratories, as well as improve the accuracy of diagnosis, it will not allow most patients to receive test results and initiate TB treatment during the visit at which they provide the \ufb01rst sputum sample. It may thus not reduce loss of patients to TB treatment initiation as much as placement of the instruments at point of treatment (i.e. health centres and clinics) would (Lawn et al. 2012).\nEarly reports on the Xpert MTB \u2044 RIF technology suggest that it was designed to be placed at patient-contact sites where nurses could provide treatment immediately upon diagnosis (Helb et al. 2010; Morris 2010; Boehme et al. 2011). Among the reasons for South Africa\u2019s decision to place the Xpert technology at laboratories rather than at point of treatment were concerns about both the cost of the test and of the required improvements to the peripheral health facilities, such as air conditioning and stable electricity supply (Tre\u00b4bucq et al. 2011; World Health Organization 2011b). To assist the South African government to evaluate the potential costs of laboratory vs. pointof-treatment placement, we developed a cost model that uses recent data from a pilot study of Xpert at the PHC level\n\n(Page-Shipp et al. 2011) and the pilot phase of the South African national roll-out of Xpert (Meyer-Rath et al. 2012) to estimate the average cost per test performed and the total cost of rolling out Xpert for each placement scenario.\nMethods\nTaking the perspective of the public sector provider, we conducted a bottom-up cost analysis of the use of Xpert for laboratory diagnosis of pulmonary TB (PTB) using data collected by the NHLS during the pilot phase (March\u2013May 2011) of the national roll-out. We also conducted a bottom-up costing of Xpert during a pilot implementation study at an urban PHC. Further results of both cost analyses are reported elsewhere (Bistline et al. 2011; Meyer-Rath et al. 2012). We used these costs to parameterize a cost model that estimated the cost per test and the total annual cost of the laboratory vs. clinic placement scenarios based on 2010 test volumes at smear microscopy laboratories across South Africa. Capital costs were annualized over an estimated useful life of \ufb01ve years and discounted using the South African Reserve Bank average 2011 repo rate of 5.5% (South African Reserve Bank 2011). All costs were converted to USD using the 2011 average exchange rate (January\u2013October) of 1 USD = ZAR 7.05 (Oanda.com 2011) and are reported in 2011 USD.\nXpert test volumes, instrument placement and operations\nFor both scenarios, the number of Xpert tests required is based on NHLS data indicating the total volume of smear microscopy tests performed for public sector patients in 2010. The number was adjusted downward to exclude smears for extrapulmonary TB (EPTB) and TB treatment monitoring. Volume and operational parameters are summarized in Table 1. For the laboratory scenario, Xpert volumes required to replace smear microscopy for diagnosis by 2014 were calculated for each of the 223 smear microscopy laboratories. The end of 2014 was selected to allow for future growth in the number of suspects needing testing. These laboratories are spread across most districts of South Africa and are located in both urban and rural areas. The estimated Xpert volumes varied among the laboratories from six to 309 tests per day in 2014, re\ufb02ecting differences in current smear microscopy volumes for each laboratory. For the point-of-treatment scenario, the 2010 smear microscopy volume was summed for each district. The estimated volume of Xpert tests required in each district were then allocated across all healthcare facilities in the district by the types of facilities present in the district, assuming that 10% of tests would be performed at provincial hospital level, 15% at district\n\n\u00aa 2012 Blackwell Publishing Ltd\n\n1143\n\n13653156, 2012, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-3156.2012.03028.x by Readcube (Labtiva Inc.), Wiley Online Library on [20/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License\n\nTropical Medicine and International Health K. Schnippel et al. Costs of Xpert MTB/RIF placement\n\nvolume 17 no 9 pp 1142\u20131151 september 2012\n\nTable 1 Assumptions and sources for Xpert volumes and operational model parameters\n\nParameter\n\nValue (range)\n\nSource\n\nComments\n\nExtrapulmonary TB as proportion of all TB Annual growth in TB suspect numbers Proportion of all smears used for treatment monitoring Number of diagnostic smears per suspect Xpert error rate\nTesting days per year\nClinic down time requiring Xpert tests to be done in laboratories\n\n16%\n10% (0%, 6.5%)\n8\u201330% (varies by district)\n2\nLab: 3.4% (1\u20133.4%); Clinic: 6.8% (2\u20136.8%) Lab: 264; Clinic: 246\n2 months \u2044 year (0\u20136 months)\n\nWorld Health Organization 2011a Meyer-Rath et al. 2012\nHealth Systems Trust 2010\nNational Department of Health 2009 Lab: NHLS pilot phase; Clinic: assumption\nLab: standard NHLS working days; Clinic: working days within a year Lab: Not applicable (delays incorporated in the 1\u20133 day average processing time) Clinic: assumption\n\nBased on NDOH targets for intensi\ufb01ed HIV \u2044 TB case \ufb01nding Calculated from TB case load for each district\nBased on South African TB diagnostic algorithm\nErrors caused by operators and environment (high temperatures, excess dust, etc) assumed to be more frequent in clinic setting\nDays when clinic is unable to use Xpert because of cartridge or supply stock out, temperatures in excess of 30 degrees and air-conditioning not functioning due to poor maintenance or electrical outages, electrical and generator outages, staff leave or shortages, or other operational down time\n\nNHLS, National Health Laboratory Service; PHC, primary healthcare.\n\nhospital level, 10% at community health centre level and the remainder at PHC level. South Africa has 3799 hospitals, primary health clinics and community health clinics that should diagnose TB and could potentially initiate TB treatment (Health Systems Trust 2010).\nThe current Xpert MTB \u2044 RIF cartridge has a 2-h processing time, and GX instruments in use in South Africa have between 1 and 48 modules, which translates to a capacity of 3\u2013256 tests per day. In the model, GX instruments were allocated to each laboratory or treatment facility according to the number of Xpert tests estimated to be required per day by the end of 2014 and the maximum number of tests that could be run per 8-h day for laboratories and 6-h day for clinics. Clinics were assigned a 6-h working day because clinic opening hours are shorter than those for laboratories, and additional time is required at the beginning of each day for the \ufb01rst patients to progress through clinic reception and triage before being asked to produce a sputum sample for testing. Laboratories, in contrast, would typically begin each morning testing samples which were stored in the refrigerator overnight while awaiting pickups from the clinics.\nCost data and costing methods\nThe cost of the cartridge and of international freight, importation and local delivery of the cartridge were\n\nassumed to be the same regardless of the placement. Other costs were varied both across the placement scenarios and according to the estimated volume of the tests at the site. Estimation methods and sources of the cost components are summarized in Table 2. Costs per site, instrument or day were translated into costs per test according to the national average volume of tests for those instruments.\nCapital costs, as incurred by the NHLS in the pilot rollout phase, include the GX instruments (case and modules), desktop computers, printers, uninterrupted power supply systems, barcode readers, air conditioners, renovations, project management time for the roll-out and installation, data management systems, backup generators and refrigerators for sample storage. Bio-safety equipment, such as a biohazard hood, was excluded for point-of-treatment placement because of the low risk of aerosols in using Xpert (Banada et al. 2010) and for laboratory placement because of existing capacity.\nThe recurrent cost per Xpert test includes the Xpert MTB \u2044 RIF cartridge, cartridge procurement, module calibration, test consumables, labour, external quality assessment and operator training, transport of samples and consumables, and operating overheads. Required operator time was based on the bottom-up costing of the NHLS pilot roll-out and the PHC pilot for the laboratory and clinic placement scenarios, respectively. A 2-day on-site training for GX operators was included biannually for\n\n1144\n\n\u00aa 2012 Blackwell Publishing Ltd\n\n13653156, 2012, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-3156.2012.03028.x by Readcube (Labtiva Inc.), Wiley Online Library on [20/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License\n\nTropical Medicine and International Health K. Schnippel et al. Costs of Xpert MTB/RIF placement\n\nvolume 17 no 9 pp 1142\u20131151 september 2012\n\nTable 2 Model cost assumptions and sources\n\nCost\n\nValue (range)\n\nSource\n\nComments\n\nRecurrent costs Xpert MTB \u2044 RIF cartridge\nCartridge procurement\nModule calibration\nSample consumables\nSalaries\nOperator staff time per test\nManagement staff salaries\nTransport of supplies and \u2044 or samples\nExternal quality assessment\nTraining (2 days on-site)\nOverhead cost\nCapital costs GX instruments\nRenovations\nData management system\n\n$14.00 ($10.72\u2013$16.86)\n$2.68 ($2.05\u2013$3.23)\n$496 \u2044 module, exclusive of labor and travel See Table 4 for per test costs\nLab: Technician at $24 454 \u2044 year; Clinic: Staff nurse at $28 450 \u2044 year Lab: 0.2 h \u2044 test Clinic: 0.25 h \u2044 test Lab: Laboratory manager at $52 817 \u2044 year; Clinic: $55 516 \u2044 year Lab: 8% of cartridge + con sumables cost; Clinic: 3% cartridge + consumables cost See Table 4 for per test costs See Table 4 for per test costs\nLab: 12% of other direct test costs; Clinic: see Table 4 for per test costs\nGX-IV with 4 modules and desktop computer at $17 000\nSee Table 4 (Other equipment)\nSee Table 4 (Other equipment)\n\nPublished prices from manufacturer (Cepheid)\nQuotation from local supplier\nQuotation from local supplier\nLab: NHLS pilot phase expenditure; Clinic: PHC pilot study Lab: NHLS pilot expenditure; Clinic: NDOH salary scales Lab: NHLS pilot phase; Clinic: PHC pilot study Lab: NHLS pilot expenditure; Clinic: NDOH salary scales Lab: NHLS pilot phase Clinic: Quotation NHLS pilot phase expenditure NHLS pilot phase expenditure\nLab: NHLS pilot phase; Clinic: PHC pilot study\nGX-IV: Published prices from manufacturer (Cepheid); Other GX costs: quotation from local supplier Lab: NHLS pilot phase expenditure; Clinic: PHC pilot study expenditure\nNHLS pilot phase expenditure\n\nPrices dependent on cumulative global volume of cartridges procured, Stop TB Department 2010. Price assumed at $14.00 in 2014, based on expected procurement volumes by South Africa 2012\u20132013 Inclusive of air freight, customs and importation, insurance, and local delivery charges. Varies with cartridge cost Module calibration required after every 2000 tests or after 1 year, whichever occurs \ufb01rst. \u2018Swap pack\u2018calibration method used Includes gloves, disinfectant, and N-95 masks (per day) and sputum collection bottles, request forms, and specimen bags (per test) Lab: laboratory technician (1 year laboratory training) Clinic: staff nurse (2 years nursing school) Allocated at \u2018hands-on\u2019 time per test for GX1-GX12 100% effort for GX16 instruments and above 2% level of effort\nThree times per year for each module, following calibration Includes trainer, travel, meals, accommodation, training materials. Lab: every other year Clinic: every year (due to higher staff rotation) Clinic: Includes electricity, water, medical waste disposal, security services, cleaning services, and space (rent). Expenses allocated according to the proportion of total space required for each type of instrument (Cepheid n.d.).\nIncludes international freight, customs and importation, insurance, uninterrupted power supply unit, desktop computer, printer, barcode reader, installation and delivery. Includes minor renovations for shelves or security, air-conditioning, network points, and generator installation. Lab: additional extensive renovations for the GX48 because of large footprint and excess weight Included for any instrument using a GX16 case as well as GX48\n\n\u00aa 2012 Blackwell Publishing Ltd\n\n1145\n\n13653156, 2012, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-3156.2012.03028.x by Readcube (Labtiva Inc.), Wiley Online Library on [20/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License\n\nTropical Medicine and International Health K. Schnippel et al. Costs of Xpert MTB/RIF placement\n\nvolume 17 no 9 pp 1142\u20131151 september 2012\n\nTable 2 (Continued)\n\nCost\n\nValue (range)\n\nSource\n\nComments\n\nGenerator\nRefrigerator for sample storage Useful life of equipment\n\nLab: 85% existing coverage Clinic: 0% existing coverage\nLab: 85% existing coverage Clinic: Not included\n5 years (3\u20138 years)\n\nPublished local prices for generators; Lab: NHLS pilot phase for coverage Clinic: Assumption for coverage Lab: NHLS pilot for coverage; Published local prices for refrigerators Assumption\n\nBased on Cepheid Xpert speci\ufb01cations (Cepheid n.d.). Generator back-up capacity not calculated to power air conditioning\nClinics would not need to store samples as providing the service while patient waits\n\nNHLS, National Health Laboratory Service; PHC, primary healthcare.\n\nlaboratory technicians and annually for clinic nurses because of turnover and rotation of staff within clinics. Per site training costs incurred by the NHLS in the pilot rollout phase were applied to both scenarios. Annual quality assurance site visits by NHLS staff was also assumed for both scenarios, mirroring systems established and costs incurred in the NHLS pilot phase. Sample transport from clinics to the laboratories was included in the laboratory placement at the standard NHLS markup of 8% per test. For clinic placement, a charge of 3% for the transport of cartridges and sputum collection supplies from a district depot to the peripheral facilities was included. Overhead costs were included at the standard NHLS per test 12% markup for the laboratory placement. For the clinic placement scenario, overhead costs included electricity, generator fuel, water, medical waste disposal, security and cleaning services, and required clinic space.\nThe Xpert error rate from the NHLS pilot roll-out phase was used as the baseline laboratory error rate. Clinic placement was assumed to have twice as many errors as the laboratory scenario because of less experienced operators, off-site (and therefore less frequent) quality assurance and management, and greater environmental instability especially in terms of maintaining temperatures within the manufacturers\u2019 recommended operating range of less than 30 degrees Celsius (World Health Organization 2011b). We assumed that the same factors would also lead to a need for clinic staff to access laboratory-based Xpert for a proportion of tests throughout the year.\nSensitivity analysis\nA number of input parameters were varied in sensitivity analysis. The base case for both scenarios assumes a cost per Xpert cartridge of $14.00 in 2014, annual growth in TB suspects of 10%, a useful life of equipment of 5 years, an Xpert error rate of 3.4% for laboratory placement and 6.8% for clinic placement and the need to access a\n\nlaboratory-based Xpert instrument for on average 2 months each year when clinics would be unable to provide the service. For the cartridges, the current price of $16.86 available to high-burden countries and the future discounted price of $10.72 were both considered as alternatives, as the price is dependent on the cumulative international volume of cartridges procured (Stop TB Department 2010). The number of PTB suspects was varied to consider a scenario of no growth and one where the suspects increase by 6.5% annually, according to the assumption used in the Planning and Budgeting for TB Control Model for South Africa (Stop TB Department 2006). The useful life of the equipment was varied from 3 to 8 years, but was kept the same for both placement options. Scenarios were considered in which the error rate of 3.4% was the same for both placements and in which both rates were decreased, to 1% for laboratories and 2% for point of treatment, because of expected improvements to the Xpert MTB \u2044 RIF cartridge. Finally, the proportion of tests that would have to be performed in laboratories under point-of-treatment placement was varied from 0% (i.e. all tests performed in clinics, no \u2018down\u2019 time) to 50%.\nResults\nNational scale-up to existing smear microscopy laboratories at sub-district level will require 274 Xpert instruments ranging in size from GX1 to GX48, with a total of 2739 modules. These will cost $16 million to procure. Scale-up to points of treatment will require 4020 instruments (GX1\u2013 GX16, a total of 5056 modules) which will cost $41 million to procure, 2.5 times more than the procurement of instruments for laboratory placement. Table 3 details the estimated need for GX instruments and the capital costs of placement.\nIn 2014, at full national scale, the total cost per test performed is $26.54 in the laboratory scenario and $38.91\n\n1146\n\n\u00aa 2012 Blackwell Publishing Ltd\n\n13653156, 2012, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-3156.2012.03028.x by Readcube (Labtiva Inc.), Wiley Online Library on [20/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License\n\nTropical Medicine and International Health K. Schnippel et al. Costs of Xpert MTB/RIF placement\n\nvolume 17 no 9 pp 1142\u20131151 september 2012\n\nTable 3 Capital costs* of placement, by instrument type (2011 USD)\n\nLaboratory placement scenario\n\nPoint-of-treatment placement scenario\n\nGX instrument\n\nNumber of instruments\n\nTotal capital cost (2011 USD)\n\nNumber of instruments\n\nTotal capital cost (2011 USD)\n\nGX1\u00e0 (GX-IV case) GX2 (GX-IV case) GX3 (GX-IV case) GX4 (GX-IV case) GX8 (GX-XVI case) GX12 (GX-XVI case) GX16 (GX-XVI case) GX48 (GX-In\ufb01nity case) Total Total number of GX modules\n\n4 12 29 17 85 47 79\n1 274 2739\n\n$40 142 $174 035 $550 143 $366 887 $5 112 351 $3 590 829 $7 313 096 $399 812 $17 547 295\n\n3533 294 89 56 36 9 3 0\n4 020 5056\n\n$42 796 050 $4 801 101 $1 817 305 $1 303 122 $2 260 103 $714 712 $286 748\n\u2013 $53 979 142\n\n*Capital costs inclusive of GX instrument, desktop computer, uninterrupted power supply, desktop printer, generator, refrigerator, data management information system, air conditioning, renovations, and delivery, installation and roll-out of the above. GX instrument and module costs are from quotations from local supplier, August 2011. GX-IV case at \u2018compassionate\u2019 pricing level. Other cases at price negotiated between NHLS and local supplier. \u00e0GX-I case not eligible for discounted \u2018compassionate\u2019 pricing; therefore costs given here are for a GX-IV case equipped with one module, which as per local supplier quotation is less expensive than the GX-I.\n\nin the point-of-treatment scenario, an additional $12.37 or 47%. The per-test cost for both scenarios is driven by the cost of the cartridge, assumed to be $14.00 in 2014 and comprising 53% and 36% of total per-test costs in the laboratory and point-of-treatment placement scenarios, respectively. Three items make major contributions to the difference in the cost per test between the scenarios, as indicated in Table 4: instrument procurement, external quality assessment and training, and labour. The cost of the Xpert instruments comprises just 6% and 11% of the total cost per test in the laboratory and point-of-treatment scenarios, respectively, but contributes 20% of the difference between scenarios. On-site training and quality assurance delivered to more than 4000 sites comprises 10% of point-of-treatment test costs while remaining a negligible component of laboratory placement, thus contributing 30% of the difference between the scenarios. Ef\ufb01ciencies in sample preparation and operation for the larger scale (GX16 and GX48) instruments lead to lower labour cost per test in the laboratory scenario and account for 20% of the difference in the per-test cost. The breakdown of the cost per Xpert test in each scenario is shown in Table 4.\nIn 2014, South Africa will use Xpert as the \ufb01rst-line diagnostic for testing 2.6 million PTB suspects. For this volume, laboratory placement would cost $71 million per year. Point-of-treatment placement would cost $107 million per year, 51% more than laboratory placement.\n\nSensitivity analysis\nIn sensitivity analysis, presented in Table 5, we found that varying core assumptions leads to an annual cost for the point-of-treatment scenario that is 43\u201365% higher than laboratory placement. Apart from the cost per cartridge, cost per test in the laboratory placement scenario was less sensitive to changes in the core assumptions, varying by )2% to +5%. Cost per test in the point-of-treatment scenario varied by )5% to +8% as factors affecting the utilization of the instruments changed. Because the capital per-test cost is 3 times higher in the point-of-treatment scenario, differences in per-test cost are sensitive to assumptions about the estimated annual growth in TB suspects and the expected useful life of equipment.\nDiscussion\nThe literature on the use of Xpert MTB \u2044 RIF and other new diagnostic technologies typically referred to as \u2018point of care\u2019 has only recently begun to consider exactly where the technologies should be placed (Tre\u00b4bucq et al. 2011). In this analysis of the cost of the national rollout of Xpert for \ufb01rst-line PTB diagnosis in South Africa, based on locally generated data on test volumes and costs, we estimated that truly decentralized placement at the point of TB treatment (clinics and hospitals) is approximately 51% more expensive than placement at sub-district laboratories. This additional cost of $36 million would represent a 17%\n\n\u00aa 2012 Blackwell Publishing Ltd\n\n1147\n\n13653156, 2012, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-3156.2012.03028.x by Readcube (Labtiva Inc.), Wiley Online Library on [20/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License\n\nTropical Medicine and International Health K. Schnippel et al. Costs of Xpert MTB/RIF placement\n\nvolume 17 no 9 pp 1142\u20131151 september 2012\n\nTable 4 2014 Cost per successful Xpert MTB \u2044 RIF test, by scenario (2011 USD)\n\nCost component\nRecurrent costs Xpert MTB \u2044 RIF cartridge Cartridge procurement Labor Overhead operating costs Sample, supplies transport Module calibration Consumables Quality assessment and training\n\nCost per test (% of total) Laboratory placement Point-of-treatment placement\n\nAdditional cost for point of treatment (% of total difference)\n\n$14.00 (53) $2.68 (10) $2.90 (11) $2.68 (10) $1.36 (5) $0.60 (2) $0.36 (1) $0.15 (1)\n\n$14.00 (36) $2.68 (7) $5.35 (14) $4.25 (11) $0.68 (2) $1.47 (4) $1.20 (3) $3.87 (10)\n\n$0.00 $0.00 $2.45 (20) $1.57 (13) )$0.68 ()5) $0.87 (7) $0.84 (7) $3.72 (30)\n\nCapital costs GX instruments Other equipment, renovations\n\n$1.66 (6) $0.15 (1)\n\n$4.16 (11) $1.25 (3)\n\n$2.50 (20) $1.10 (9)\n\nTotals (% additional) Total cost per test Total cartridges procured to test 2.6 million TB suspects in 2014 Total annual cost (2011 USD) in 2014\n\n$26.54 2.7 million\n$71 million\n\n$38.91 2.8 million\n$107 million\n\n$12.37 (+47) 0.1 million (+3)\n$36 million (+51)\n\nTable 5 Results of sensitivity analysis: 2014 Cost per test and annual costs, by scenario (2011 USD)\n\nBase case*\n\nLaboratory placement\n\nCost per test Annual cost\n\n$26.54\n\n$71 045 331\n\nPoint of treatment placement\n\nCost per test Annual cost\n\n$38.91\n\n$107 282 983\n\nAdditional annual cost for point-of-treatment (%)\n$36 237 652 (51)\n\nGrowth in suspect population 0% annual growth 6.5 % annual growth\n\n$26.40 $26.41\n\n$48 716 554 $62 268 732\n\n$42.18 $39.97\n\n$80 171 695 $97 047 560\n\n$31 455 141 (65) $34 778 828 (56)\n\nClinic service gaps requiring laboratory back-up\n\nNo outages\n\n$26.54\n\nAverage 6 months \u2044 year\n\n$26.54\n\n$71 045 331 $71 045 331\n\n$38.01 $40.12\n\n$105 356 422 $109 287 682\n\n$34 311 091 (48) $38 242 351 (54)\n\nError rate Clinic same as lab (3.4%) Both reduced (1%, 2%)\n\n$26.54 $26.52\n\n$71 045 331 $69 316 411\n\n$39.44 $39.66\n\n$105 784 390 $104 770 496\n\n$34 739 059 (49) $35 454 085 (51)\n\nUseful life of GX and other equipment\n\n3 years\n\n$27.74\n\n8 years\n\n$25.89\n\n$74 229 660 $69 279 232\n\n$42.10 $37.14\n\n$116 018 843 $102 350 115\n\n$41 789 183 (56) $33 070 883 (48)\n\nFuture discount on Xpert MTB \u2044 RIF cartridges (current international price $16.86)\n\nNo discount ($16.86)\n\n$30.64\n\n$81 989 790 $42.52\n\nDiscounted price of $10.72 $21.87\n\n$58 522 086 $34.78\n\n$117 176 276 $95 846 446\n\n$35 186 486 (43) $37 324 360 (64)\n\n*Base case: 2.6 million suspects, 10% growth in suspects per year; 2 months service outage in clinics per year; 3.4% error in laboratories and 6.8% error in clinics; 5 years useful life of GX instruments; $14.00 cartridge cost.\n\nincrease in the overall estimated $218 million TB control budget for South Africa for 2011 (World Health Organization 2011a). The additional cost of point-of-treatment\n\nplacement is attributable to two main factors. First, many more sites must be initially capacitated with GX instruments, equipment and trained staff. Second, the lower\n\n1148\n\n\u00aa 2012 Blackwell Publishing Ltd\n\n13653156, 2012, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-3156.2012.03028.x by Readcube (Labtiva Inc.), Wiley Online Library on [20/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License\n\nTropical Medicine and International Health K. Schnippel et al. Costs of Xpert MTB/RIF placement\n\nvolume 17 no 9 pp 1142\u20131151 september 2012\n\nvolumes of tests conducted per day in clinics diminish the technical and economic ef\ufb01ciency with which each instrument can be operated.\nDespite its higher cost, placing Xpert in at point of treatment offers the potential to reduce the loss of patients before initiation of TB treatment. Healthcare facilities with access to both TB treatment and on-site Xpert diagnostic capability could potentially have TB suspects provide a sputum sample, receive test results and initiate TB treatment on the same day. Xpert also rapidly diagnoses RIF resistance, an important marker for MDR-TB. Current South African MDR-TB guidelines indicate that only a limited number of capacitated hospitals should initiate MDR-TB treatment. Thus, while the placement of Xpert technology at health facilities may reduce delays in MDR-TB treatment compared with the laboratory placement, it is unlikely that patients with Xpert-detected RIF resistance will be able to initiate MDR-TB treatment on the same day in either scenario. Further economic analysis, including research that incorporates treatment outcomes and analysis that takes into account the costs and bene\ufb01ts of the scenarios to patients, the health system and society for both drug-sensitive and drug-resistant TB, is needed to fully appreciate the differences between the scenarios.\nThe per-test cost of Xpert at both laboratory and point of treatment reported here are higher than previously reported estimates from South Africa (Theron et al. 2011; Vassall et al. 2011). The local prices for procurement of the cartridges and instruments used in our analysis were higher than those used in other estimates. Also, this analysis does not assume the exclusive use of GX4 instruments or average volumes, as was the case for the previous estimates, but rather uses the actual range of GX instruments anticipated to be in use and current daily TB test volumes from South Africa. Finally, this analysis was designed to include the costs of the overall roll-out of the Xpert technology at a national scale, which we show can be a signi\ufb01cant component of the total cost per test for a national TB control programme.\nAlthough the analysis reported here is based on primary data from South Africa, it has several limitations. First, unit cost estimates are based on small samples and early pilot studies which may not re\ufb02ect costs at scale. Second, current smear microscopy volumes may not accurately estimate the need for Xpert, even if allowing for future growth in the number of suspects. Adjustments made to current volumes to exclude smears for diagnosing EPTB may underestimate the volume of Xpert tests that will be required if Xpert becomes the diagnostic of choice for EPTB (Vadwai et al. 2011) and \u2044 or paediatric TB (Nicol et al. 2011) as well. Third, the analysis does not take into account the time required to implement either scenario, but\n\nrather assumes that Xpert provision will reach full scale immediately. Placing instruments and supporting their use at more than 4000 facilities for the point-of-treatment scenario, most of which lack existing laboratory infrastructure, will be a far more complicated undertaking than placing them at just over 200 laboratories and may limit access to this rapid diagnostic and its bene\ufb01ts for a far longer period than the roll-out to laboratories. We attempted to capture this in our cost estimates by including additional training and supervision time as well as laboratory backup of clinic testing capacity, but this might not capture the full difference in operational complexity. Fourth, the analysis is based on the cost of Xpert within the current diagnostic algorithm for South Africa. Alternative diagnostic algorithms, such as using Xpert only for smearnegative, HIV-infected TB suspects (Page-Shipp et al. 2011; Theron et al. 2011) would have to be analysed according to their impact on test volumes. Finally, the comparison presented here assumes an \u2018either \u2044 or\u2019 decision with regard to Xpert placement \u2013 either in sub-district laboratories or at point of treatment, but not both. A combination of laboratory and point-of-treatment placement may be preferable and is likely to be the strategy ultimately adopted by South Africa.\nThe results of the analysis pertain to South Africa and may not be readily generalizable to other high-TB-burden countries. Unlike many other low- and middle-income countries, South Africa has a strong infrastructure of existing sub-district laboratories and an excellent transport network that allows for ef\ufb01cient collection and processing of samples. The potential bene\ufb01ts of point-of-treatment diagnosis may be greater in countries where there are no or very few laboratories and where patient travel distances to clinics are longer; though, the challenges and costs of clinic placement in such settings may also be greater. Although the speci\ufb01c results of this analysis may not be readily transferrable to other countries, the issues it examines, such as the relationship between the cost per Xpert test and the volume of tests performed, will be of relevance to all countries that are considering its use.\nDespite the limitations described above, we conclude from this analysis that point-of-treatment placement of Xpert technology is ultimately more expensive per test because of the inadequacy of existing clinic infrastructure and low test volumes in each health facility. A substantial increase in treatment uptake, large improvement in treatment outcomes and \u2044 or signi\ufb01cant cost savings to patients would be needed to justify the higher costs of this placement. While access to Xpert may indeed facilitate achieving these goals, other health system investments may also be needed to secure them. Given the resource constraints faced by most high-TB-burden countries,\n\n\u00aa 2012 Blackwell Publishing Ltd\n\n1149\n\n13653156, 2012, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-3156.2012.03028.x by Readcube (Labtiva Inc.), Wiley Online Library on [20/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License\n\nTropical Medicine and International Health K. Schnippel et al. Costs of Xpert MTB/RIF placement\n\nvolume 17 no 9 pp 1142\u20131151 september 2012\n\nlower-cost interventions to reduce loss to TB treatment initiation, such as reducing other causes of delay in diagnosis and treatment (Sreeramareddy et al. 2009) and continued efforts to develop point-of-treatment tests (Dorman et al. 2012), combined with laboratory-based access to Xpert technology, may be more cost-effective investments than point-of-treatment placement of Xpert instruments. It is important that this question be investigated in a range of settings throughout high-TB-burden countries.\nAcknowledgements\nWe are grateful to Linda Erasmus, Naseem Cassim, Floyd Olsen and Lesley Scott for sharing cost and operational information from the NHLS pilot implementation and to Witkoppen Health and Welfare Centre for cost information for the PHC pilot. Funding for this study was provided by the South Africa Mission of the US Agency for International Development (USAID) under the terms of Cooperative Agreement GHSA-00-00020-00, Country Research Activity (G \u2044 PHN \u2044 HN \u2044 CS). KS received additional training support from Fogarty International Center \u2044 National Institutes of Health ICOHRTA AIDS \u2044 TB Grant No. U2RTW007373. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.\nReferences\nBanada PP, Sivasubramani SK, Blakemore R et al. (2010) Containment of bioaerosol infection risk by the Xpert MTB \u2044 RIF assay and its applicability to point-of-care settings. Journal of Clinical Microbiology 48, 3551\u20133557.\nBistline K, Van Rie A, Page-Shipp L, Basset J & Sanne I (2011) Cost of Xpert MTB \u2044 RIF for smear-negative TB suspects at primary care clinic in Johannesburg. International Journal of Tuberculosis and Lung Disease 15, S61. Lille, France.\nBoehme CC, Nabeta P, Hillemann D et al. (2010) Rapid molecular detection of tuberculosis and rifampin resistance. New England Journal of Medicine 363, 1005\u20131015.\nBoehme CC, Nicol MP, Nabeta P et al. (2011) Feasibility, diagnostic accuracy, and effectiveness of decentralised use of the Xpert MTB \u2044 RIF test for diagnosis of tuberculosis and multidrug resistance: a multicentre implementation study. Lancet 6736, 1\u201311.\nCepheid (n.d.). GeneXpert System. from http://www.genexpert.com/pdfs/GeneXpert Brochure_0112-04.pdf (accepted 20 May 2011).\nChihota VN, Grant AD, Fielding K et al. (2010) Liquid vs. solid culture for tuberculosis: performance and cost in a resourceconstrained setting. International Journal of Tuberculosis and Lung Disease 14, 1024\u20131031.\n\nDorman S, Manabe Y, Nicol M, Nakiyingi L, Moodley M & Zemanay W et al. (2012). Accuracy of Determine TB-LAM Lateral Flow Test for Diagnosis of TB in HIV+ Adults: interim results from a Multicenter Study. Conference on Retroviruses and Opportunistic Infections, Oral abstract, paper #149aLB. Seattle, USA.\nErasmus LK, Coetzee GJ & Stevens WS (2011) Scale up of Xpert MTB \u2044 RIF from the national laboratory perspective: issues and challenges. International Journal of Tuberculosis and Lung Disease 15, S61. Lille, France.\nHealth Systems Trust (2010). The District Health Barometer 2008 \u2044 09. Health Systems Trust. http://www.hst.org.za/ publications/district-health-barometer-200809.\nHelb D, Jones M, Story E et al. (2010) Rapid detection of Mycobacterium tuberculosis and Rifampin resistance by use of on-demand, near-patient technology. Journal of Clinical Microbiology 48, 229\u2013237.\nLawn SD, Kerkhoff AD & Wood R (2012) Location of Xpert\u00d2 MTB \u2044 RIF in centralized laboratories in South Africa undermines potential impact. International Journal of Tuberculosis and Lung Disease 16, 701\u2013710.\nMeyer-Rath G, Schnippel K, Long L et al. (2012) The impact and cost of scaling up GeneXpert MTB \u2044 RIF in South Africa. PLoS One 7(5), e36966. doi:10.1371/journal.pone.0036966.\nMorris K (2010) Xpert TB diagnostic highlights gap in point-ofcare pipeline. The Lancet Infectious Diseases 10, 742\u2013743. Elsevier Ltd.\nNational Department of Health (2009). South African National Tuberculosis Guidelines. National Department of Health, Pretoria.\nNicol MP, Workman L, Isaacs W et al. (2011) Accuracy of the Xpert MTB \u2044 RIF test for the diagnosis of pulmonary tuberculosis in children admitted to hospital in Cape Town, South Africa: a descriptive study. The Lancet Infectious Diseases 11, 819\u2013824. Elsevier Ltd\nOanda.com (2011). Oanda average exchange rates. from http:// www.oanda.com/currency/average (accepted 28 September 2011).\nPage-Shipp L, Dansey H, Basset J et al. (2011) Point of care Xpert MTB \u2044 RIF for smear-negative TB diagnosis at a primary care clinic in South Africa. International Journal of Tuberculosis and Lung Disease 15, S61. Lille, France.\nScott LE, Mccarthy K, Gous N et al. (2011) Comparison of Xpert MTB \u2044 RIF with other nucleic acid technologies for diagnosing pulmonary tuberculosis in a high HIV prevalence setting: a prospective study. PLoS Medicine 8, 1\u201311.\nSmall PM & Pai M (2010) Tuberculosis diagnosis \u2014 time for a game change. New England Journal of Medicine 363, 1070\u20131071.\nSmart T (2011). GeneXpert to be rolled out as \ufb01rst-line diagnostic for TB in South Africa. HIV & AIDS treatment in practice. from http://www.aidsmap.com/GeneXpert-to-be-rolled-out-as-\ufb01rstline-diagnostic-for-TB-in-South-Africa/page/1746803/ (accepted 10 November 2011).\nSouth African Reserve Bank (2011). Selected historical exchange rates and other interest rates. from http://www.resbank.co.za/\n\n1150\n\n\u00aa 2012 Blackwell Publishing Ltd\n\n13653156, 2012, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-3156.2012.03028.x by Readcube (Labtiva Inc.), Wiley Online Library on [20/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License\n\nTropical Medicine and International Health K. Schnippel et al. Costs of Xpert MTB/RIF placement\n\nvolume 17 no 9 pp 1142\u20131151 september 2012\n\nResearch/Rates/Pages/SelectedHistoricalExchangeAndInterest Rates.aspx (accepted 14 November 2011). Sreeramareddy CT, Panduru KV, Menten J & Van den Ende J (2009) Time delays in diagnosis of pulmonary tuberculosis: a systematic review of literature. BMC Infectious Diseases 9, 91 doi:10.1186/1471-2334-9-91. Stop TB Department (2006). Planning and budgeting for TB control model. World Health Organization. from http:// www.who.int/tb/dots/planning_budgeting_tool/en/ (accepted 28 July 2011). Stop TB Department (2010). Roadmap for Rolling Out Xpert MTB \u2044 RIF for Rapid Diagnosis of TB and MDR-TB (pp. 1\u201312). World Health Organization, Geneva. Theron G, Pooran A, Peter J et al. (2011) Adjunct TB tests, when combined with Xpert MTB \u2044 RIF, improve accuracy and the cost of diagnosis in a resource-poor setting? ERJ. doi: 10.1183/09031936.00145511. Tre\u00b4bucq A, Enarson DA, Chiang CY et al. (2011) Xpert\u00d2 MTB \u2044 RIF for national tuberculosis programmes in lowincome countries: when, where and how? International\n\nJournal of Tuberculosis and Lung Disease 15, 1567\u2013 1571. Vadwai V, Boehme CC, Nabeta P, Shetty A, Alland D & Rodrigues C (2011) Xpert MTB \u2044 RIF: a new pillar in diagnosis of extrapulmonary tuberculosis? Journal of Clinical Microbiology, 49 2540\u20132545. Van Rie A, Page-Shipp L, Scott L, Sanne I & Stevens W (2010) Xpert \u00d2 MTB \u2044 RIF for point-of- care diagnosis of TB in highHIV burden, resource-limited countries: hype or hope? Expert Review of Molecular Diagnosis 10, 937\u2013946. Vassall A, van Kampen S, Sohn H et al. (2011) Rapid diagnosis of tuberculosis with the Xpert MTB \u2044 RIF assay in high burden countries: a cost-effectiveness analysis. PLoS Medicine 8, e1001120. World Health Organization (2011a) Global Tuberculosis Control. World Health Organization, Geneva. World Health Organization (2011b) Rapid Implementation of the Xpert MTB \u2044 RIF Diagnostic test: Technical and Operational \u2018\u2018How-to\u2019\u2019; Practical Considerations. World Health (WHO \u2044 HTM \u2044 TB.). World Health Organization, Geneva.\n\nCorresponding Author Kathryn Schnippel, Health Economics and Epidemiology Research Of\ufb01ce, Department of Medicine, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa. E-mail: kschnippel@heroza.org\n\n\u00aa 2012 Blackwell Publishing Ltd\n\n1151\n\n\n",
    "authors": [
      "Kathryn Schnippel",
      "Gesine Meyer\u2010Rath",
      "Lawrence Long",
      "William MacLeod",
      "Ian Sanne",
      "Wendy S. Stevens",
      "Sydney Rosen"
    ],
    "doi": "10.1111/j.1365-3156.2012.03028.x",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://onlinelibrary.wiley.com/doi/10.1111/j.1365-3156.2012.03028.x"
  },
  {
    "key": "3ME3CJ2L",
    "title": "Do adjunct tuberculosis tests, when combined with Xpert MTB/RIF, improve accuracy and the cost of diagnosis in a resource-poor setting?",
    "abstract": "Information regarding the utility of adjunct diagnostic tests in combination with Xpert MTB/RIF (Cepheid, Sunnyvale, CA, USA) is limited. We hypothesised adjunct tests could enhance accuracy and/or reduce the cost of tuberculosis (TB) diagnosis prior to MTB/RIF testing, and rulein or rule-out TB in MTB/RIF-negative individuals.",
    "full_text": "Eur Respir J 2012; 40: 161\u2013168 DOI: 10.1183/09031936.00145511 Copyright\u00dfERS 2012\nDo adjunct tuberculosis tests, when combined with Xpert MTB/RIF, improve accuracy and the cost of diagnosis in a resource-poor setting?\nGrant Theron*, Anil Pooran*, Jonny Peter*, Richard van Zyl-Smit*, Hridesh Kumar Mishra#, Richard Meldau*, Greg Calligaro*, Brian Allwood*, Surendra Kumar Sharma#, Rod Dawson* and Keertan Dheda*,\"\n\nABSTRACT: Information regarding the utility of adjunct diagnostic tests in combination with Xpert MTB/RIF (Cepheid, Sunnyvale, CA, USA) is limited. We hypothesised adjunct tests could enhance accuracy and/or reduce the cost of tuberculosis (TB) diagnosis prior to MTB/RIF testing, and rulein or rule-out TB in MTB/RIF-negative individuals.\nWe assessed the accuracy and/or laboratory-associated cost of diagnosis of smear microscopy, chest radiography (CXR) and interferon-c release assays (IGRAs; T-SPOT-TB (Oxford Immunotec, Oxford, UK) and QuantiFERON-TB Gold In-Tube (Cellestis, Chadstone, Australia)) combined with MTB/RIF for TB in 480 patients in South Africa.\nWhen conducted prior to MTB/RIF: 1) smear microscopy followed by MTB/RIF (if smear negative) had the lowest cost of diagnosis of any strategy investigated; 2) a combination of smear\n\nAFFILIATIONS *Lung Infection and Immunity Unit, Division of Pulmonology & UCT Lung Institute, Dept of Medicine, University of Cape Town, and \"Institute of Infectious Diseases and Molecular Medicine, University of Cape Town, Cape Town, South Africa. #Dept of Medicine, All India Institute of Medical Sciences, New Delhi, India.\n\nmicroscopy, CXR (if smear negative) and MTB/RIF (if imaging compatible with active TB) did not further reduce the cost per TB case diagnosed; and 3) a normal CXR ruled out TB in 18% of patients (57 out of 324; negative predictive value (NPV) 100%). When downstream adjunct tests were applied to MTB/RIF-negative individuals, radiology ruled out TB in 24% (56 out of 234; NPV 100%), smear microscopy ruled in TB in 21% (seven out of 24) of culture-positive individuals and IGRAs were not useful in either context.\n\nCORRESPONDENCE K. Dheda University of Cape Town H47 Old Main Building Groote Schuur Hospital Observatory 7925\n\nIn resource-poor settings, smear microscopy combined with MTB/RIF had the highest accuracy and lowest cost of diagnosis compared to either technique alone. In MTB/RIF-negative individuals, CXR has poor rule-in value but can reliably rule out TB in approximately one in four\n\nCape Town South Africa E-mail: keertan.dheda@uct.ac.za\n\ncases. These data inform upon the programmatic utility of MTB/RIF in high-burden settings.\n\nReceived:\n\nAug 23 2011\n\nKEYWORDS: Adjunct diagnostics, chest radiography, tuberculosis, Xpert MTB/RIF\n\nAccepted after revision: Oct 17 2011\n\nFirst published online:\n\nX pert MTB/RIF (Cepheid, Sunnyvale, CA, USA) is an accurate molecular test for the diagnosis of tuberculosis (TB). The World Health Organization (WHO) recently endorsed MTB/RIF for rapid implementation as a frontline test in individuals suspected of HIV/TB coinfection [1, 2]. The performance of a single MTB/ RIF has been prospectively assessed in over 7,500 patients from a variety of settings [3, 4], where its\n\nvalue (PPV) ,97%, negative predictive value (NPV) ,96%).\nAlthough repeated MTB/RIF testing offers small improvements in diagnostic accuracy [3, 4], the test remains expensive and is likely to be performed in routine care only once per patient [1, 2]. Given resource constraints in high-burden settings, national TB programmes need to make important policy decisions on the positioning of\n\nNov 10 2011\n\nsensitivity in smear-positive and smear-negative MTB/RIF within existing diagnostic algorithms.\n\nindividuals was found to be ,98% and ,75%, res- Few data about the performance and cost of MTB/\n\npectively, with overall excellent specificity (,99%) RIF in combination with adjunct tests are available\n\nand good predictive value (positive predictive to guide policymakers. Additionally, given the\n\nThis article has supplementary material available from www.erj.ersjournals.com\n\nEuropean Respiratory Journal Print ISSN 0903-1936\n\nc\n\nOnline ISSN 1399-3003\n\nEUROPEAN RESPIRATORY JOURNAL\n\nVOLUME 40 NUMBER 1\n\n161\n\nTUBERCULOSIS\n\nG. THERON ET AL.\n\ndiminished NPV of a single MTB/RIF in HIV-infected individuals (,91%; therefore, approximately one in 10 HIV-infected, MTB/RIF-negative individuals have TB) [4\u20136], combined with the unavailability of culture facilities in many HIV-prevalent settings, there is a need to assess the performance of adjunct tests for ruling-out TB in HIV-infected individuals who are MTB/RIF negative.\nThere are limited studies evaluating the cost of diagnosis of MTB/RIF alone or in conjunction with adjunct tests for the diagnosis of TB. Recent articles [7, 8], including a WHO policy document pertaining to the roll-out of MTB/RIF [2], have highlighted the urgent need for research in this area. In addition to assessing the diagnostic accuracy of certain adjunct TB tests, we performed a cost analysis to determine the laboratorybased cost of diagnosis of these tests when combined with MTB/RIF alone.\nMETHODS Study sites and population We recently assessed the performance of a single MTB/RIF assay in archived spot sputum samples collected from 496 selfreporting patients with suspected TB. Patients were consecutively recruited from primary care clinics in a high HIV prevalence setting in Cape Town, South Africa [5, 9]. Liquid culture positivity for Mycobacterium tuberculosis from a single sample served as a reference standard. Informed consent was obtained from all participants and the study was approved by the University of Cape Town Faculty of Health Sciences Research Ethics Committee (Cape Town). Detailed demographic and clinical information, as well as MTB/RIF performance data, have been published previously for this cohort [5].\nCase definitions Each patient was allocated to one of three diagnostic categories. 1) Definite TB: a clinical presentation compatible with TB with at least one sputum sample smear positive or culture positive for M. tuberculosis. 2) Probable TB: a clinicalradiological picture highly suggestive of TB and/or anti-TB treatment was initiated by an attending clinician based on clinical suspicion but the patient did not meet the criteria for definite TB (smear negative and no culture-based evidence of M. tuberculosis). 3) Non-TB: no evidence of TB based on smear microscopy and culture, no anti-TB treatment initiated with response to alternative treatment where appropriate and, when available, no radiological evidence to support the diagnosis of TB.\nDiagnostic tests Each patient gave two spot sputum samples and one earlymorning sputum sample (the latter provided no longer than 1 week after the initial visit) (fig. 1). An arbitrarily selected spot sputum sample was stored at -20uC for later MTB/RIF analysis. The MTB/RIF procedure has been detailed previously [3, 10]. The remaining samples were used for concentrated fluorescent smear microscopy and cultured using the BACTEC MGIT 960 system (BD Diagnostics, Franklin Lakes, NJ, USA). Where possible, a chest radiograph (CXR), read by two investigators using the validated CXR scoring and recording system (CRRS) [11, 12], and standardised interferon-c release assays (IGRAs), T-SPOT.TB (Oxford Immunotec, Oxford, UK) and QuantiFERONTB Gold In-Tube (QFT-GIT; Cellestis, Chadstone, Australia) were performed.\n\nTest performance assessment and statistical analysis We assessed the diagnostic accuracy of adjunct tests (alone and in combination with MTB/RIF) when used to pre-screen individuals for MTB/RIF testing (table 1). Differences in diagnostic accuracy between strategies are outlined in table S3. We also assessed the diagnostic utility of these adjunct tests for the detection of TB in MTB/RIF-negative individuals (table 2). For the analysis of test sensitivity, culture positivity from any sample (spot or morning) served as a reference standard. Individuals who were culture negative for all samples (from the probable or non-TB group) were used in specificity calculations. Comparative performance data were obtained when the analysis was restricted to individuals classified as non-TB (i.e. probable TB excluded from the culture-negative group) (tables S1 and S2). Test performance assessment and Chi-squared analyses were performed using OpenEpi version 2.3.1 [13].\nCost analysis The cost analysis was performed from a healthcare provider perspective and limited to the laboratory-associated costs of diagnosis (table 3). Results of the cost analysis stratified by HIV status are shown in tables S6 and S7 of the supplementary material. A decision tree model was used to determine the outcomes and costs of using MTB/RIF, either on its own or in combination with other pre-screening tests (smear microscopy and/or CXR), for the diagnosis of TB (fig. 2) [7]. IGRAs were not included due to their poor clinical utility for ruling-in or ruling-out patients for MTB/RIF testing. The model was run on a hypothetical cohort of 1,000 TB patients normalised to our clinical performance data (table 1). Additional costs associated with clinic visits, TB treatment or drug sensitivity testing were not considered in this analysis. Actual smear microscopy and CXR costs were taken from reference laboratory sources (table S4) and correspond to those reported elsewhere [14]. The actual cost of performing an MTB/RIF test was calculated using WHO estimates [2] and data specific to South Africa (table S5). Costs are presented in US$ according to the currency conversion rate of 2011. The number of cases (TB and non-TB) and the cost per TB case detected compared to the baseline of screening with smear microscopy were reported. A univariate and a multivariate sensitivity analysis were also performed. The costing and sensitivity analyses methodologies are described in the supplementary material.\nRESULTS\nDoes pre-screening with adjunct diagnostic tests improve MTB/RIF performance and/or the cost of TB diagnosis?\nSmear microscopy\nSmear microscopy had a good rule-in value for TB (sensitivity, specificity and PPV were 69% (102 out of 149), 99% (328 out of 331) and 97% (102 out of 105), respectively) (table 1). A combination of smear microscopy and MTB/RIF (performed if smear negative) had better overall sensitivity than smearmicroscopy alone (82% (95% CI 75\u201387%), 122 out of 149 versus 69% (61\u201375%), 102 out of 149; p,0.01) but did not outperform MTB/RIF alone (77% (115 out of 149); p50.32). Using smear microscopy to pre-screen individuals with suspected TB prior to MTB/RIF testing reduced the cost to detect a TB case by $115 compared to MTB/RIF alone ($401 versus $516) (table 3).\n\n162\n\nVOLUME 40 NUMBER 1\n\nEUROPEAN RESPIRATORY JOURNAL\n\nG. THERON ET AL.\n\nTUBERCULOSIS\n\n496 patients with suspected pulmonary TB\n2 spot sputa collected at first visit\n\nCXR performed at first visit using CRRS (n=324)#\n\nSputum 1\n\nSmear microscopy (n=496)\nMGIT culture (n=496)\n\nSputum 2\nArchived\nXpert MTB/RIF (n=496)\n\nBlood collected at first visit\nT-SPOT.TB (n=350)\u00b6 QFT-GIT (n=399)+\n\nExcluded: 15 culture contaminated 1 indeterminate Xpert MTB/RIF result\n480 included in main analysis\n\nDefinite TB\u00a7 Culture positive (n=149)\n102 smear positive 47 smear negative\n\nCulture negative (n=331)\n\nProbable TB (n=182)\u0192\n\nNo TB (n=149)##\n\n115 Xpert MTB/RIF positive 34 Xpert MTB/RIF negative\n\n7 Xpert MTB/RIF positive 175 Xpert MTB/RIF negative\n\n8 Xpert MTB/RIF positive 141 Xpert MTB/RIF negative\n\nFIGURE 1. Patient flow diagram and diagnostic outcomes stratified by final diagnostic category. TB: tuberculosis; CXR: chest radiograph; CRRS: CXR scoring and\nrecording system; QFT-GIT: QuantiFERON-TB Gold In-Tube (Cellestis, Chadstone, Australia); MGIT: mycobacteria growth indicator tube. T-SPOT-TB is manufactured by Oxford Immunotec Ltd, Oxford, UK. #: CXR data missing in 156 patients. \": T-SPOT.TB data missing in 130 patients and an indeterminate result in nine patients. +: QFT-GIT data missing in 91 patients and an indeterminate result in 44 patients. 1: in this group: all 106 individuals with CXR data had a CXR compatible with active TB; 91 (85%) out of 107 individuals with T-SPOT.TB had a positive result; 49 (46%) out of 106 individuals with QFT-GIT had a positive result. e: in this group: 161 (97%) out of 166 individuals with CXR data had a CXR compatible with active TB; 72 (57%) out of 127 individuals with T-SPOT.TB data had a positive result; 75 (61%) out of 123 individuals with QFT-GIT data had a positive result. ##: in this group: none of the 78 individuals with CXR data had a CXR compatible with active TB; 60 (57%) out of 106 individuals with T-SPOT.TB data had a positive result; 66 (58%) out of 113 individuals with QFT-GIT data had a positive result.\n\nChest radiography The sensitivity and NPV of CXR in our cohort were both 100% (n5106 and 57, respectively) (table 1). Thus, of the 324 individuals who had CXR data, 18% (n557) could have been ruled out as non-TB prior to MTB/RIF testing. The predictive value of CXR combined with MTB/RIF (performed only in individuals with a CXR compatible with active TB) did not significantly differ compared to MTB/RIF alone (PPV 92% (82 out of 89) versus 89% (115 out of 130) for MTB/RIF alone (p50.38); NPV 90.0% (211 out of 235) versus 90.3% (316 out of 350) for MTB/RIF alone (p50.84)). Consequently, this strategy (pre-screening with CXR prior to MTB/RIF testing) had a higher cost per detected TB case than MTB/RIF alone ($698 versus $516 per TB case detected) (table 3).\nSmear microscopy and CXR combined Smear microscopy followed by CXR and MTB/RIF (i.e. MTB/ RIF performed only on smear-negative individuals with a CXR compatible with active TB) had similar diagnostic accuracy to other strategies involving a combination of tests: smear\n\nmicroscopy followed by MTB/RIF, CXR followed by MTB/ RIF, CXR followed by smear microscopy, and MTB/RIF (sensitivity, specificity, PPV and NPV for the smear microscopy/CXR/MTB/RIF strategy of 83% (88 out of 106), 96% (210 out of 218), 92% (88 out of 96) and 92% (210 out of 228), respectively) (table 1).\nBoth combined strategies (i.e. smear, CXR and MTB/RIF, or CXR, smear and MTB/RIF) detected an equivalent number of TB cases (59.05 TB cases per 1,000 screened) (table 3). However, the cost per case detected of the latter strategy was higher ($531 versus $401 per TB case detected). This is because more individuals required an upfront CXR, which is more expensive per test than smear microscopy. Both strategies were still more costly than smear microscopy followed by MTB/RIF.\nInterferon-c release assays As outlined in table 1, IGRAs (T-SPOT.TB and QFT-GIT) had\nc sub-optimal sensitivities (,85% each; 91 out of 107 and 90 out\nof 106, respectively) and sub-optimal NPV (,85% each; 102 out of 118 and 96 out of 112, respectively).\n\nEUROPEAN RESPIRATORY JOURNAL\n\nVOLUME 40 NUMBER 1\n\n163\n\nTUBERCULOSIS\n\n164\n\nTABLE 1 The performance of different diagnostic tests, stratified by HIV status, alone or in combination for the detection of tuberculosis (TB)\nFrontline performance in all TB suspects#\n\nAll patients\"\n\nHIV uninfected+\n\nHIV infected1\n\nAll patients\"\n\nHIV uninfected+\n\nHIV infected1\n\nSensitivity %e Specificity %## Sensitivity %e Specificity %## Sensitivity %e Specificity %##\n\nPPV\n\nNPV\n\nPPV\n\nNPV\n\nPPV\n\nNPV\n\n(95% CI) n/N (95% CI) n/N (95% CI) n/N (95% CI) n/N (95% CI) n/N (95% CI) n/N (95% CI) n/N (95% CI) n/N (95% CI) n/N (95% CI) n/N (95% CI) n/N (95% CI) n/N\n\nVOLUME 40 NUMBER 1\n\nXpert MTB/RIF\nSmear microscopy\nSmear microscopy followed by Xpert MTB/RIF\"\"\nCXR for active TB\nCXR followed by Xpert MTB/RIF++\nSmear microscopy followed by CXR\"\" followed by Xpert MTB/RIF++\nCXR followed by smear microscopy++ followed by Xpert MTB/RIF\"\"\nSmear microscopy followed by Xpert MTB/RIF\"\" and CXR11\nT-SPOT.TB\nQFT-GIT\n\n77.2 (69.8\u201383.2)\n115/149 68.5\n(60.6\u201375.4) 102/149 81.9\n(74.9\u201387.2) 122/149 100\n(96.5\u2013100) 106/106 77.4\n(68.5\u201384.3) 82/106 83.0\n(74.8\u201389.0) 88/106\n83.0 (74.8\u201389.0)\n88/106\n100 (96.5\u2013100)\n106/106 85.0\n(77.1\u201390.6) 91/107 84.9\n(76.9\u201390.5) 90/106\n\n95.5 (92.7\u201397.2)\n316/331 99.1\n(97.4\u201399.7) 328/331 94.9\n(91.9\u201396.8) 314/331 26.2\n(20.8\u201332.4) 57/218 96.8\n(93.5\u201398.4) 211/218 96.3\n(93.9\u201398.1) 210/218\n96.3 (93.9\u201398.1)\n210/218\n25.7 (20.3\u201331.9)\n56/218 43.6\n(37.4\u201350.0) 102/234 40.2\n(34.2\u201346.5) 96/239\n\n81.0 (71.3\u201387.9)\n68/84 77.4 (67.4\u201385.0) 65/84 85.7 (76.7\u201391.6) 72/84 100 (93.9\u2013100) 59/59 83.1 (71.5\u201390.5) 49/59 89.8 (79.5\u201395.3) 53/59\n89.8 (79.5\u201395.3)\n53/59\n100 (93.9\u2013100)\n59/59 85.9 (75.4\u201392.4) 55/64 87.7 (77.6\u201393.6) 57/65\n\n95.1 (91.2\u201397.3)\n193/203 99.0\n(96.5\u201399.7) 200/202 95.1\n(91.1\u201397.3) 192/202 28.7\n(21.6\u201337.0) 37/129 96.1\n(91.3\u201398.3) 124/129 96.1\n(91.3\u201398.3) 124/129\n96.1 (91.3\u201398.3)\n124/129\n27.9 (20.9\u201336.2)\n36/129 32.6\n(25.5\u201340.7) 47/144 34.6\n(37.6\u201342.3) 55/159\n\n70.0 (56.3\u201380.9)\n35/50 52.0 (38.5\u201365.2) 26/50 74.0 (60.5\u201384.1) 37/50 100 (90.6\u2013100) 37/37 64.9 (48.8\u201378.2) 24/37 67.6 (51.5\u201380.4) 25/37\n67.6 (51.5\u201380.4)\n25/37\n100 (90.6\u2013100)\n37/37 82.4 (66.5\u201391.7) 28/34 80.6 (63.7\u201390.8) 25/31\n\n95.0 (87.8\u201398.0)\n76/80 98.8 (93.3\u201399.8) 79/80 93.8 (86.2\u201397.3) 75/80 27.3 (17.3\u201340.2) 15/55 98.2 (90.4\u201399.7) 54/55 96.4 (87.7\u201399.0) 53/55\n96.4 (87.7\u201399.0)\n53/55\n28.3 (18.0\u201341.6)\n15/53 64.9 (51.9\u201376.0) 37/57 57.8 (43.3\u201371.0) 26/45\n\n88.5 (81.8\u201393.2)\n115/130 97.1\n(91.9\u201399.0) 102/105 87.8\n(81.3\u201392.2) 122/139 39.7\n(34.0\u201346.7) 106/267 92.1\n(84.6\u201396.1) 82/89 91.7\n(84.4\u201395.7) 88/96\n\n90.3 (86.7\u201393.0)\n316/350 87.5\n(83.7\u201390.4) 328/375 92.1\n(88.7\u201394.5) 314/341 100\n(93.7\u2013100) 57/57 89.8\n(85.3\u201393.0) 211/235 92.1\n(87.9\u201395.0) 210/228\n\n87.2 (78.0\u201392.9)\n68/79 97.0 (89.8\u201399.2) 65/67 87.8 (79.0\u201393.2) 72/82 39.1 (31.7\u201347.0) 59/151 90.7 (80.1\u201396.0) 49/54 91.4 (81.4\u201396.3) 53/58\n\n91.7 (84.4\u201395.7)\n88/96\n\n92.1 (87.9\u201395.0)\n210/228\n\n91.4 (81.4\u201396.3)\n53/58\n\n39.6 (33.9\u201345.5)\n106/268 40.8\n(34.6\u201347.4) 91/223 38.6\n(32.6\u201345.0) 90/233\n\n100 (93.6\u2013100)\n56/56 86.4 (79.1\u201391.5) 102/118 85.7 (78.1\u201391.0) 96/112\n\n38.8 (31.4\u201346.8)\n59/152 36.2\n(29.0\u201344.1) 55/152 35.4\n(28.4\u201343.1) 57/161\n\n92.3 (87.9\u201395.2)\n193/209 91.3\n(86.9\u201394.4) 200/219 94.1\n(90.0\u201396.6) 192/204 100\n(90.6\u2013100) 37/37 92.5\n(86.8\u201395.9) 124/134 95.3\n(90.3\u201397.9) 124/130\n95.3 (90.3\u201397.9)\n124/130\n100 (90.4\u2013100)\n36/36 83.9 (72.2\u201391.3) 47/56 87.3 (76.9\u201393.4) 55/63\n\n89.7 (76.4\u201395.4)\n35/39 96.3 (81.7\u201399.3) 26/27 88.1 (75.0\u201394.8) 37/42 48.1 (37.3\u201359.0) 37/77 96.0 (80.5\u201399.3) 24/25 92.6 (76.6\u201397.9) 25/27\n\n83.5 (74.6\u201389.8)\n76/91 76.7 (67.7\u201383.8) 79/103 85.2 (76.4\u201391.2) 75/88 100 (79.6\u2013100) 15/15 80.6 (69.6\u201388.3) 54/67 81.5 (70.5\u201389.1) 53/65\n\n92.6 (76.6\u201397.9)\n25/27\n\n81.5 (70.5\u201389.1)\n53/65\n\n49.3 (38.3\u201360.4)\n37/75 58.3 (44.3\u201371.2) 28/48 56.8 (42.2\u201371.0) 25/44\n\n100 (79.6\u2013100)\n15/15 86.1 (72.7\u201393.4) 37/43 81.3 (64.7\u201391.1) 26/32\n\nG. THERON ET AL.\n\nEUROPEAN RESPIRATORY JOURNAL\n\nPPV: positive predictive value; NPV: negative predictive value; CXR: chest radiograph; QFT-GIT: QuantiFERON-TB Gold In-Tube (Cellestis, Chadstone, Australia). T-SPOT-TB is manufactured by Oxford Immunotec Ltd, Oxford, UK. #: liquid culture positivity for Mycobacterium tuberculosis served as a reference standard. \": n5480. +: n5286, 64 were of unknown HIV status (test refused or data missing). 1: n5130, 64 were of unknown HIV\nstatus (test refused or data missing). e: calculated by dividing the number of individuals positive for the adjunct test(s) by the number of culture-positive individuals who had received the same adjunct test(s); not all individuals received the same combination of tests. ##: calculated by dividing the number of culture-negative individuals negative for the adjunct test(s) by the number of culture-negative individuals who received the same adjunct test(s) (i.e. both probable and non-TB groups were included); not all individuals received the same combination of tests. \"\": performed if smear negative. ++: performed if CXR compatible with active TB. 11: performed if MTB/RIF-negative.\n\nTABLE 2 Performance of different diagnostic tests stratified by HIV status for the detection of tuberculosis (TB) in individuals negative for a single Xpert MTB/RIF test\n\nPerformance in Xpert MTB/RIF negative individuals#\n\nAll patients\"\n\nHIV uninfected+\n\nHIV infected1\n\nAll patients\"\n\nHIV uninfected+\n\nHIV infected1\n\nG. THERON ET AL.\nEUROPEAN RESPIRATORY JOURNAL\n\nSensitivity %e (95% CI) n/N\n\nSpecificity %## (95% CI) n/N\n\nSensitivity %e (95% CI) n/N\n\nSpecificity %## (95% CI) n/N\n\nSensitivity %e (95% CI) n/N\n\nSpecificity %## (95% CI) n/N\n\nPPV (95% CI) n/N\n\nNPV (95% CI) n/N\n\nPPV (95% CI) n/N\n\nNPV (95% CI) n/N\n\nPPV (95% CI) n/N\n\nNPV (95% CI) n/N\n\nSmear microscopy\nCXR for active TB\nSmear microscopy combined with CXR\"\"\nT-SPOT.TB QFT-GIT\n\n20.6\n\n99.4\n\n25\n\n99.5\n\n(10.4\u201336.8) 7/34 (97.7\u201399.8) 314/316 (10.2\u201349.5) 4/16 (97.1\u201399.9) 192/193\n\n100\n\n26.7\n\n100\n\n29.3\n\n(86.2\u2013100) 24/24 (21.1\u201333.0) 56/210 (72.3\u2013100) 10/10 (22.0\u201337.8) 36/123\n\n100\n\n26.7\n\n100\n\n29.3\n\n(86.2\u2013100) 24/24 (21.1\u201333.0) 56/210 (72.3\u2013100) 10/10 (22.0\u201337.8) 36/123\n\n13.3 (3.7\u201337.9) 2/15\n100 (77.2\u2013100) 13/13\n100 (77.2\u2013100) 13/13\n77.8 (45.3\u201393.7) 7/9\n77.8 (45.3\u201393.7) 7/9\n\n98.7 (92.9\u201399.8) 75/76\n27.8 (17.6\u201340.9) 15/54\n27.8 (17.6\u201340.9) 15/54\n63.6 (50.4\u201375.1) 35/55\n60.5 (45.6\u201373.6) 26/43\n\n77.8\n\n92.1\n\n80.0\n\n94.1\n\n66.7\n\n85.2\n\n(45.3\u201393.7) 7/9 (88.7\u201394.5) 314/341 (37.6\u201396.4) 4/5 (90.0\u201396.6) 192/204 (20.8\u201393.9) 2/3 (76.4\u201391.2) 75/88\n\n13.5\n\n100\n\n10.3\n\n100\n\n25\n\n100\n\n(9.2\u201319.3) 24/178 (93.6\u2013100) 56/56 (5.7\u201318.0) 10/97 (90.4\u2013100) 26/26 (15.2\u201338.2) 13/52 (79.6\u2013100) 15/15\n\n13.5\n\n100\n\n10.3\n\n100\n\n25\n\n100\n\n(9.2\u201319.3) 24/178 (93.6\u2013100) 56/56 (5.7\u201318.0) 10/97 (90.4\u2013100) 26/26 (15.2\u201338.2) 13/52 (79.6\u2013100) 15/15\n\n74.1\n\n43.4\n\n73.3\n\n33.1\n\n13.5\n\n93.3\n\n10.6\n\n92\n\n(55.3\u201386.8) 20/27 (37.1\u201349.9) 98/226 (48.1\u201389.1) 11/15 (25.8\u201341.3) 46/139\n\n(8.9\u201320.0) 20/148 (86.9\u201396.7) 98/105 (6.0\u201318.0) 11/104 (81.2\u201396.9) 46/50\n\n25.9 (13.2\u201344.7) 7/27\n29.2 (14.9\u201349.2) 7/24\n\n94.6 (82.3\u201398.5) 35/37\n92.9 (77.4\u201398.0) 26/28\n\n80.0\n\n41.0\n\n84.6\n\n34.9\n\n12.9\n\n94.9\n\n10\n\n96.4\n\n(60.9\u201391.1) 20/25 (34.9\u201347.5) 94/229 (57.8\u201395.7) 11/13 (27.8\u201342.7) 53/152\n\n(8.5\u201319.1) 20/155 (88.7\u201397.8) 94/99 (5.7\u201317.0) 11/110 (87.7\u201399.0) 53/55\n\nPPV: positive predictive value; NPV: negative predictive value; CXR: chest radiograph; QFT-GIT: QuantiFERON-TB Gold In-Tube (Cellestis, Chadstone, Australia). T-SPOT-TB is manufactured by Oxford Immunotec Ltd, Oxford, UK. #: liquid culture positivity for Mycobacterium tuberculosis served as a reference standard. \": n5350. +: n5209, 50 individuals were of unknown HIV status (test refused or data missing). 1: n591, 50 individuals were of unknown HIV status (test refused or data missing). e: calculated by dividing the number of MTB/RIF-negative individuals positive for the adjunct test(s) by the number of MTB/RIF-negative, culture-positive individuals who had received the same adjunct test(s); not all individuals received the same combination of tests. ##: calculated by dividing the number of MTB/RIF-negative, culture-negative individuals negative for the adjunct test(s) by the number of MTB/RIF-negative, culture-negative individuals (i.e. both probable and non-TB groups were included) who had received the same adjunct test(s); not all individuals received the same combination of tests. \"\": performed if smear negative.\n\nTUBERCULOSIS\n\nDiagnostic costs stratified by HIV status\nThe cost per TB case detected in the HIV-infected cohort for each strategy was much lower than in the HIV-uninfected cohort (tables S6 and S7). This is primarily due to the poorer performance of smear microscopy in HIV-infected patients, which consequently increases the number of cases detected by downstream adjunct tests. As a result, screening with MTB/ RIF alone had the second lowest cost per TB case detected (after performing MTB/RIF in smear-negative individuals) in the HIV-infected cohort ($202 per TB case detected) but the highest in the HIV-uninfected cohort ($1,446 per TB case detected). The strategies combining all three techniques (smear microscopy, CXR and MTB/RIF) had the lowest costs per TB case detected in the HIV-uninfected group ($427 per TB case detected when pre-screening with smear microscopy followed by CXR versus $566 when pre-screening with CXR followed by smear microscopy).\nSensitivity analysis for costing strategies involving smear microscopy and/or CXR\nA univariate sensitivity analysis revealed that the cost rankings, in terms of the cost per TB case detected, were most sensitive to changes in MTB/RIF test cost and the sensitivity of the baseline strategy (smear microscopy alone), as well as that of MTB/RIF (either on its own or in combination with other pre-screening tests) (table S8).\nIn most cases, increasing the sensitivity of a particular strategy made that strategy the least costly in terms of TB cases detected. For example, when MTB/RIF sensitivity in the MTB/ RIF alone strategy was increased to 90%, this strategy detected more TB cases and consequently had the lowest cost ($209 per TB case detected). Conversely, lowering MTB/RIF sensitivity to 65% results in this strategy detecting fewer TB cases than with smear microscopy alone. Increasing TB prevalence decreased the cost per TB case detected for each strategy due to an increase in TB cases detected, but did not significantly change the cost ranking of the strategies. However, in many cases, performing smear microscopy followed by MTB/RIF remained the least costly strategy.\nWhen a multivariate sensitivity analysis was performed (table S9) using smear microscopy and MTB/RIF diagnostic accuracy data from a larger study by BOEHME et al. [3] (MTB/ RIF sensitivity of 90% (933 out of 1,033) versus 77% (115 out of 149) for our study), screening with MTB/RIF alone ($191 per TB case detected) was less costly than a combination of smear microscopy and MTB/RIF ($222 per TB case detected) and smear microscopy followed by CXR and MTB/RIF ($270 per TB case detected). This was primarily due to the improved diagnostic accuracy of MTB/RIF in the larger study, which resulted in a lower cost per TB case detected. The cost ranking of other strategies remained unchanged.\n\nDo adjunct tests possess diagnostic utility for ruling out TB\n\nin MTB/RIF-negative individuals?\n\nOf 73% (350 out of 480) of patients with a negative MTB/RIF\n\nresult, ,10% (n534) had culture-confirmed TB and approxi-\n\nmately half of these (15 out of 31) were HIV infected (three had no HIV data). Detailed performance data for each test in MTB/\n\nc\n\nRIF-negative patients is shown in table 2.\n\nVOLUME 40 NUMBER 1\n\n165\n\nTUBERCULOSIS\n\nG. THERON ET AL.\n\nTABLE 3 Cost of diagnosis, per 1,000 cases screened, of different strategies involving pre-screening with an alternative diagnostic test prior to Xpert MTB/RIF assay#\n\nSmear microscopy\nalone\n\nXpert MTB/RIF alone\n\nSmear microscopy followed by:\n\nCXR followed by:\n\nXpert MTB/RIF (if smear negative)\n\nCXR (if smear negative), followed by Xpert MTB/RIF\n(if CXR suggestive of active TB)\n\nXpert MTB/RIF (if CXR suggestive\nof active TB)\n\nSmear microscopy (if CXR suggestive of active TB), followed by Xpert MTB/RIF\n(if smear negative)\n\nTotal test cost $ Incremental cost versus\nsmear microscopy alone $ Number of correctly diagnosed non-TB cases Number of correctly diagnosed TB cases Additional number of TB cases correctly diagnosed versus smear microscopy alone Cost per TB case detected (compared to smear microscopy) $\n\n7420.00 683.33 212.50\n\n21389.75 13969.75\n658.33 239.58 27.08\n515.81\n\n24130.74 16710.74\n654.17 254.17 41.67\n401.06\n\n32471.71 25051.71\n648.15 271.60 59.10\n423.85\n\n35746.74 28326.74\n651.23 253.09 40.59\n697.94\n\n36778.00 29358.00\n648.15 271.60 59.10\n496.71\n\nCXR: chest radiograph; TB: tuberculosis. #: similar tables stratified by HIV status can be found in the supplementary material (tables S6 and S7).\n\nSmear microscopy 21% (seven out of 34) of the MTB/RIF-negative, culturepositive group were smear positive. Thus, 5% (seven out of 149) of culture-positive samples were MTB/RIF-negative but smear positive.\nChest radiography The sensitivity and NPV performance of CXR for ruling-out TB in MTB/RIF-negative individuals were 100% (n524) and 100% (n556), respectively. Thus, 24% of the MTB/RIF-negative individuals who received a CXR (56 out of 234) could be ruled out as non-TB by the presence of a normal CXR. The proportion of individuals who had a normal CXR did not differ by HIV status (21% (26 out of 123) of HIV-uninfected versus 22% (14 out of 67) of HIV-infected individuals; p50.97). All smear-positive individuals had a CXR compatible with active TB and, thus, CXR performance did not improve when combined with smear microscopy (table 1).\nInterferon-c release assays IGRAs had moderate performance for ruling out TB in MTB/ RIF-negative individuals (NPV: 93% (98 out of 105) and 94% (94 out of 99) for T-SPOT.TB and QFT-GIT, respectively) but suboptimal clinical utility (sensitivity ,75% in each).\nDISCUSSION A single frontline MTB/RIF test has been endorsed by the WHO and is currently in the process of being rolled out in resource scarce settings. However, a recent WHO global consultation on the rapid implementation of MTB/RIF has highlighted the need\n\nfor research investigating the utility of adjunct tests in the diagnostic pathway [2]. Adjunct tests may have cost saving utility when used to pre-screen TB suspects for MTB/RIF testing without compromising overall diagnostic accuracy. They may also be useful in MTB/RIF-negative individuals to guide further patient management. There are currently no published data on how adjunct tests may be combined with MTB/RIF or their impact on the overall cost of diagnosis.\nThe key findings of our study were as follows. 1) Smear microscopy followed by MTB/RIF (performed if smear negative) had a lower cost of diagnosis than MTB/RIF alone. 2) Approximately one in four MTB/RIF-negative individuals has a normal CXR (using the validated CRRS system) [11, 12] and all of these individuals are culture negative; thus, CXR can be used to reliably rule-out TB in MTB/RIF-negative individuals. 3) CXR can reliably rule out TB in approximately one in five TB suspects prior to MTB/RIF testing but is still more costly than performing MTB/RIF alone or the combination of smear microscopy and MTB/RIF. 4) Adjunctive diagnostic strategies can be less costly per TB case detected than a single upfront MTB/RIF. In order of increasing cost: smear microscopy followed by MTB/RIF (performed if smear negative); smear microscopy followed by CXR (performed if smear negative), followed by MTB/RIF (performed if the CXR is compatible with active TB); MTB/RIF alone; CXR followed by smear microscopy (performed if CXR is suggestive of active TB) and MTB/RIF testing (performed if smear negative); and, finally, CXR followed by MTB/RIF (performed if CXR is suggestive of active TB). However, this ranking is sensitive to\n\n166\n\nVOLUME 40 NUMBER 1\n\nEUROPEAN RESPIRATORY JOURNAL\n\nG. THERON ET AL.\n\nTUBERCULOSIS\n\nSmear Xpert Smear TB suspect CXR\nSmear\nCXR\n\nTrue + Smear +\n\nTB\n\nFalse + No TB\n\nTrue - No TB Smear -\nFalse - TB\n\nXpert + Xpert -\n\nTrue + TB False + No TB True - No TB False - TB\n\nTrue +\n\nSmear +\n\nTB\n\nFalse + No TB\n\nSmear -\n\nXpert\n\nXpert + Xpert -\n\nCXR -\n\nTrue False -\n\nNo TB TB\n\nCXR +\n\nXpert\n\nXpert + Xpert -\n\nTrue + Smear +\n\nTB\n\nFalse + No TB\n\nCXR -\n\nSmear - CXR\n\nTrue + TB\nFalse + No TB\nTrue - No TB False -\nTB\n\nTrue + TB False + No TB\nTrue - No TB False - TB\n\nTrue False -\n\nNo TB TB\n\nXpert +\n\nCXR + Xpert\n\nXpert -\n\nCXR CXR +\n\nTrue False -\nSmear\n\nNo TB\n\nTB\n\nTrue + Smear +\n\nNo TB\n\nFalse + TB\n\nXpert +\n\nSmear - Xpert\n\nXpert -\n\nTrue + False + True False -\n\nTB No TB No TB TB\n\nTrue + TB False + No TB True - No TB False - TB\n\nFIGURE 2. A decision tree for the diagnosis of tuberculosis (TB) using six\ndifferent screening strategies. 1) Smear microscopy alone. 2) Xpert MTB/RIF (Xpert) alone. 3) Smear microscopy followed by Xpert (performed if smear negative). 4) Chest radiograph (CXR) followed by Xpert (performed if the CXR is suggestive of TB). 5) Smear microscopy followed by CXR (performed if smear negative) and Xpert (performed if CXR is suggestive of active TB). 6) CXR followed by smear microscopy (performed if CXR is suggestive of active TB) and Xpert (performed if smear negative). TB and non-TB outcomes refer to an initial diagnosis based on a single or combination test result. Square node: decision branches; circular nodes: chance branches; triangular nodes: terminal branches. Data modified from [7].\nbaseline MTB/RIF performance. 5) IGRAs have little utility as a pre-screening tool for MTB/RIF testing.\nPre-screening individuals with smear microscopy followed by MTB/RIF (performed if smear negative) was more cost-effective than MTB/RIF alone. A key advantage of this approach is that the potential of same-day diagnosis may be retained, although the feasibility will be setting specific. Thus, given that smear\n\nmicroscopy facilities already exist or are immediately accessible to many primary care clinics in resource-scarce settings, it might be more suitable to target MTB/RIF at smear-negative individuals; however, this requires prospective validation. A major disadvantage of this approach is the lack of drug resistance testing for smear-positive individuals, thus, perhaps limiting this strategy to areas with a low prevalence of drug-resistant TB or to individuals not suspected of drug-resistant TB. A normal CXR, by contrast, could reliably exclude TB in ,20% of cases prior to MTB/RIF testing and, although the caveat of drugresistance testing would not apply, the applicability of this finding is restricted by the limited nature of radiology facilities and trained readers in resource-scarce settings.\nIn MTB/RIF-negative individuals, smear microscopy retained some utility and was able to detect ,20% of TB cases. CXR was able to reliably rule out TB in approximately one in four MTB/ RIF-negative individuals and, thus, is useful for the downstream investigation of MTB/RIF-negative individuals. Although the NPVs of IGRAs were relatively high in this study, the NPV of IGRAs in unselected persons with suspected TB [15] and in HIV-infected persons with smear-negative TB [16] in high-burden settings is sub-optimal and, in line with the recent WHO guidelines about IGRAs in TB-endemic countries [17], we would not recommend their use in this context. Furthermore, their cost and incompatibility with same-day diagnosis are major caveats, especially given that patient nonreturn rates for follow-up test results are significant [18].\nOur cost analysis used a simple decision tree model to compare the short-term laboratory-associated costs of correctly diagnosing TB using different MTB/RIF-based diagnostic strategies, making it of interest to policy makers who prefer to examine how implementation of a new clinical intervention affects their annual budget, rather than long-term costs. We did not include further downstream costs related to TB treatment and transmission. Costs associated with misdiagnosing patients, such as increased transmission and morbidity, together with the unnecessary use of TB treatment can be significant. Thus, the lack of morbidity and mortality data for our cohort is a major limitation of our cost analysis.\nOur model did not account for potential cost benefits arising from the drug susceptibility testing capability of MTB/RIF. Given that our data was generated using the first-generation cartridge and an intermediate version of the software (which is being modified to improve the PPV of the rifampicin resistance result), as well as the limited number of rifampicin resistant cases, this would have been of limited conclusiveness. Furthermore, many of the cost benefits associated with drug resistance detection will probably be incurred in terms of reduced TB transmission, which is outside the scope of this model. Additionally, there is very limited capacity to diagnose or treat multidrug-resistant TB in Africa (WHO currently recommends a confirmatory phenotypic test when MTB/RIF indicates drug resistance) and, thus, cost analyses for this variable is likely to vary considerably on a setting-by-setting basis. Our economic analysis, whilst not a true cost-effectiveness analysis, aims to\nc inform future detailed cost-effectiveness studies on MTB/RIF\nfocusing on inclusion of the long-term costs associated with these strategies, including TB treatment, drug susceptibility\n\nEUROPEAN RESPIRATORY JOURNAL\n\nVOLUME 40 NUMBER 1\n\n167\n\nTUBERCULOSIS\n\nG. THERON ET AL.\n\ntesting, wider TB transmission to the community and those from a patient perspective.\nThere are several additional limitations of our study. We had limited data for all tests, including CXR, on each patient (either the test was not performed or the data were missing), thereby restricting our test-specific sample size and the conclusiveness of our findings. This was also limited by the use of archived specimens for MTB/RIF. Additionally, the relevance of CXR when used in combination with smear microscopy and/or MTB/RIF is limited by the lack of radiology facilities being available in resource-scarce settings. The high NPVs of CXR will be modulated by the degree of immunosuppression and findings may differ in settings where HIV-infected patients present with lower CD4 counts. Thus, findings presented here may not be applicable to low HIV prevalence, resource-poor settings, or those in which patients have more advanced immunosuppression at presentation. Our use of concentrated fluorescence smear microscopy, which is not available in many resource-poor settings, also affects the generalisability. Our findings now require prospective validation in diagnostic trials.\nIn summary, we have shown that smear microscopy combined with MTB/RIF (performed if smear negative) had the lowest cost per TB case detected. Furthermore, we have shown that radiology can be a useful tool for ruling-out TB in MTB/ RIF-negative individuals. Further prospective studies and costeffectiveness analyses are now required to assess the performance and cost benefits of these strategies.\nSUPPORT STATEMENT This work was supported by an EU-FP7 award (TBsusgent). The Xpert MTB/RIF cartridges were a gift from the Foundation for Innovative New Diagnostics (FIND; Geneva, Switzerland). The FIND had no input into the design of the study, analysis of the data or preparation of the manuscript. G. Theron was supported by the European and Developing Countries Clinical Trials Partnership (EDCTP) TB-NEAT, the Claude Leon Foundation and the South African National Research Foundation. A. Pooran and J. Peter were supported by a South African TB/AIDS Research Training fellowship. R. van Zyl-Smit and J. Peter were supported by a Discovery Foundation Fellowship, the Fogarty International Clinical Research Scholars/Fellows Support Centre National Institutes of Health grant (R24TW007988) and the EDCTP (TB-NEAT). K. Dheda was supported by the EDCTP (TB-NEAT/TESA), the South African Department of Science and Technology and the National Research Foundation (South African Research Chairs Initiative).\nSTATEMENT OF INTEREST None declared.\nREFERENCES 1 World Health Organization. Automated real-time nucleic acid amplification technology for rapid and simultaneous detection of tuberculosis and rifampicin resitance: Xpert MTB/RIF System.\n\nWHO/HTM/TB/2011.4. Geneva, World Health Organization, 2011. 2 World Health Organization. Rapid Implementation of the Xpert MTB/RIF diagnostic test. Technical and Operational \u2018\u2018How to\u2019\u2019 Practicle Considerations. WHO/HTM/TB/2011.2. Geneva, World Health Organization, 2011. 3 Boehme CC, Nabeta P, Hillemann D, et al. Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med 2010; 363: 1005\u20131015. 4 Boehme CC, Nicol MP, Nabeta P, et al. Feasibility, diagnostic accuracy, and effectiveness of decentralised use of the Xpert MTB/ RIF test for diagnosis of tuberculosis and multidrug resistance: a multicentre implementation study. Lancet 2011; 377: 1495\u20131505. 5 Theron G, Peter J, van Zyl-Smit R, et al. Evaluation of the Xpert MTB/RIF assay for the diagnosis of pulmonary tuberculosis in a high HIV prevalence setting. Am J Respir Crit Care Med 2011; 184: 132\u2013140. 6 Theron G, Peter J, Dheda K. Xpert MTB/RIF test for tuberculosis. Lancet 2011; 378: 481. 7 Dowdy DW, Cattamanchi A, Steingart KR, et al. Is scale-up worth it? Challenges in economic analysis of diagnostic tests for tuberculosis. PLoS Medicine 2011; 8: e1001063. 8 Evans CA. GeneXpert \u2013 a game-changer for tuberculosis control? PLoS Med 2011; 8: e1001064. 9 Theron G, Pinto L, Peter J, et al. The use of an automated quantitative polymerase chain reaction (Xpert MTB/RIF) to predict the sputum smear status of tuberculosis patients. Clin Infect Dis 2012; 54: 384\u2013388. 10 Helb D, Jones M, Story E, et al. Rapid detection of Mycobacterium tuberculosis and rifampin resistance by use of on-demand, nearpatient technology. J Clin Microbiol 2010; 48: 229\u2013237. 11 Dawson R, Masuka P, Edwards D, et al. Chest radiograph reading and recording system: evaluation for tuberculosis screening in patients with advanced HIV. Int J Tuberc Lung Dis 2010; 14: 52\u201358. 12 Den Boon S, Bateman E, Enarson D, et al. Development and evaluation of a new chest radiograph reading and recording system for epidemiological surveys of tuberculosis and lung disease. Int J Tuberc Lung Dis 2005; 9: 1088\u20131096. 13 Dean AG, Sullivan KM, Soe MM. OpenEpi: Open Source Epidemiologic Statistics for Public Health, Version 2.3.1. www. OpenEpi.com Date last updated: September 19, 2010. Date last accessed: November 30, 2010. 14 Sinanovic E, Floyd K, Dudley L, et al. Cost and cost-effectiveness of community-based care for tuberculosis in Cape Town, South Africa. Int J Tuberc Lung Dis 2003; 7: Suppl. 1, S56\u2013S62. 15 Ling DI, Pai M, Davids V, et al. Are interferon-c release assays useful for diagnosing active tuberculosis in a high-burden setting? Eur Respir J 2011; 38: 649\u2013656. 16 Rangaka MX, Gideon HP, Wilkinson KA, et al. Interferon release does not add discrimiatory value to smear negative HIVtuberculosis algorithms. Eur Respir J 2012; 39: 163\u2013171. 17 World Health Organization. Commercial Serodiagnostic Tests for Diagnosis of Tuberculosis. WHO/HTM/TB/2011.5. Geneva, World Health Organization, 2011. 18 Millen SJ, Uys PW, Hargrove J, et al. The effect of diagnostic delays on the drop-out rate and the total delay to diagnosis of tuberculosis. PloS One 2008; 3: e1933.\n\n168\n\nVOLUME 40 NUMBER 1\n\nEUROPEAN RESPIRATORY JOURNAL\n\n\n",
    "authors": [
      "Grant Theron",
      "Anil Pooran",
      "Jonny Peter",
      "Richard Van Zyl-Smit",
      "Hridesh Kumar Mishra",
      "Richard Meldau",
      "Greg Calligaro",
      "Brian Allwood",
      "Surendra Kumar Sharma",
      "Rod Dawson",
      "Keertan Dheda"
    ],
    "doi": "10.1183/09031936.00145511",
    "year": null,
    "item_type": "journalArticle",
    "url": "http://erj.ersjournals.com/lookup/doi/10.1183/09031936.00145511"
  },
  {
    "key": "ZLUWHFTL",
    "title": "Cost-effectiveness analysis of microscopic observation drug susceptibility test versus Xpert MTB/Rif test for diagnosis of pulmonary tuberculosis in HIV patients in Uganda",
    "abstract": "Background: Microscopic Observation Drug Susceptibility (MODS) and Xpert MTB/Rif (Xpert) are highly sensitive tests for diagnosis of pulmonary tuberculosis (PTB). This study evaluated the cost effectiveness of utilizing MODS versus Xpert for diagnosis of active pulmonary TB in HIV infected patients in Uganda.\nMethods: A decision analysis model comparing MODS versus Xpert for TB diagnosis was used. Costs were estimated by measuring and valuing relevant resources required to perform the MODS and Xpert tests. Diagnostic accuracy data of the tests were obtained from systematic reviews involving HIV infected patients. We calculated base values for unit costs and varied several assumptions to obtain the range estimates. Cost effectiveness was expressed as costs per TB patient diagnosed for each of the two diagnostic strategies. Base case analysis was performed using the base estimates for unit cost and diagnostic accuracy of the tests. Sensitivity analysis was performed using a range of value estimates for resources, prevalence, number of tests and diagnostic accuracy.\nResults: The unit cost of MODS was US$ 6.53 versus US$ 12.41 of Xpert. Consumables accounted for 59 % (US$ 3.84 of 6.53) of the unit cost for MODS and 84 % (US$10.37 of 12.41) of the unit cost for Xpert. The cost effectiveness ratio of the algorithm using MODS was US$ 34 per TB patient diagnosed compared to US$ 71 of the algorithm using Xpert. The algorithm using MODS was more cost-effective compared to the algorithm using Xpert for a wide range of different values of accuracy, cost and TB prevalence. The cost (threshold value), where the algorithm using Xpert was optimal over the algorithm using MODS was US$ 5.92.\nConclusions: MODS versus Xpert was more cost-effective for the diagnosis of PTB among HIV patients in our setting. Efforts to scale-up MODS therefore need to be explored. However, since other non-economic factors may still favour the use of Xpert, the current cost of the Xpert cartridge still needs to be reduced further by more than half, in order to make it economically competitive with MODS.",
    "full_text": "Walusimbi et al. BMC Health Services Research (2016) 16:563 DOI 10.1186/s12913-016-1804-9\n\nRESEARCH ARTICLE\n\nOpen Access\n\nCost-effectiveness analysis of microscopic observation drug susceptibility test versus Xpert MTB/Rif test for diagnosis of pulmonary tuberculosis in HIV patients in Uganda\nSimon Walusimbi1,2, Brendan Kwesiga3, Rashmi Rodrigues2, Melles Haile4, Ayesha de Costa2, Lennart Bogg2,6 and Achilles Katamba5*\n\nAbstract\nBackground: Microscopic Observation Drug Susceptibility (MODS) and Xpert MTB/Rif (Xpert) are highly sensitive tests for diagnosis of pulmonary tuberculosis (PTB). This study evaluated the cost effectiveness of utilizing MODS versus Xpert for diagnosis of active pulmonary TB in HIV infected patients in Uganda.\nMethods: A decision analysis model comparing MODS versus Xpert for TB diagnosis was used. Costs were estimated by measuring and valuing relevant resources required to perform the MODS and Xpert tests. Diagnostic accuracy data of the tests were obtained from systematic reviews involving HIV infected patients. We calculated base values for unit costs and varied several assumptions to obtain the range estimates. Cost effectiveness was expressed as costs per TB patient diagnosed for each of the two diagnostic strategies. Base case analysis was performed using the base estimates for unit cost and diagnostic accuracy of the tests. Sensitivity analysis was performed using a range of value estimates for resources, prevalence, number of tests and diagnostic accuracy.\nResults: The unit cost of MODS was US$ 6.53 versus US$ 12.41 of Xpert. Consumables accounted for 59 % (US$ 3.84 of 6.53) of the unit cost for MODS and 84 % (US$10.37 of 12.41) of the unit cost for Xpert. The cost effectiveness ratio of the algorithm using MODS was US$ 34 per TB patient diagnosed compared to US$ 71 of the algorithm using Xpert. The algorithm using MODS was more cost-effective compared to the algorithm using Xpert for a wide range of different values of accuracy, cost and TB prevalence. The cost (threshold value), where the algorithm using Xpert was optimal over the algorithm using MODS was US$ 5.92.\nConclusions: MODS versus Xpert was more cost-effective for the diagnosis of PTB among HIV patients in our setting. Efforts to scale-up MODS therefore need to be explored. However, since other non-economic factors may still favour the use of Xpert, the current cost of the Xpert cartridge still needs to be reduced further by more than half, in order to make it economically competitive with MODS.\nKeywords: Cost-effectiveness, MODS, Xpert MTB/Rif, Diagnosis, Tuberculosis, HIV\n\n* Correspondence: akatamba@chs.mak.ac.ug; akatamba@yahoo.com 5Department of Medicine, Clinical Epidemiology Unit, Makerere University, College of Health Sciences, Kampala, Uganda Full list of author information is available at the end of the article\n\u00a9 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.\n\nWalusimbi et al. BMC Health Services Research (2016) 16:563\n\nPage 2 of 10\n\nBackground In most low resource settings where Tuberculosis (TB) is huge problem, diagnosis conventionally relies on microscopy. However, TB microscopy has a sensitivity of only 40\u201360 % under field conditions, falling to as low as 20 % in the presence of HIV co-infection [1]. Two-thirds of HIV infected people live in sub-Saharan Africa, and 75 % of the global burden of HIV-associated TB is found in the region with limited health care resources [2]. With the launch of the Global Laboratory Initiative, strengthening and modernization of TB laboratories in low resource settings became a priority for global TB control, particularly in high HIV prevalence settings [3]. Consequently, since 2007, the array of diagnostics for TB has expanded tremendously and several of them have been endorsed by the World Health Organization (WHO) in such settings [4, 5].\nThe Xpert MTB/Rif test (Xpert) is an automated rapid molecular test with high sensitivity for simultaneous detection of pulmonary TB (PTB) and rifampicin resistance in a one off-test [6].\nXpert relies on real time polymerase chain reaction (PCR) to amplify a portion of Mycobacteria DNA. The steps involved in processing the sample, amplification and detection of the Mycobacterial DNA are automated. This enables reporting of test results in two-three hours [7].\nA number of modelling studies in settings with high prevalence of TB-HIV co-infection, found Xpert was cost effective for diagnosis of PTB and reducing mortality in comparison to microscopy or conventional mycobacterial culture [8\u201311]. Thus, the WHO currently recommends Xpert as the primary diagnostic for HIV-associated TB as a replacement for TB microscopy [12]. Through the support of international donors and multilateral development assistance partners, the Xpert test has been rolled out on a large scale in several sub-Sahara African countries where TB and HIV co-infection is prevalent [13, 14]. However, the rollout of Xpert is faced with affordability and implementation challenges [15, 16]. There is also emerging evidence currently, that using Xpert in resource limited health-care settings may not be cost-effective because of its limited impact on patient mortality [17, 18].\nThe microscopic observation drug susceptibility (MODS) assay is an inexpensive test with high sensitivity for diagnosis of PTB in HIV infected patients [19], targeted for resource-limited settings [20, 21]. MODS is a liquid culture test, for simultaneous detection of TB and resistance to both rifampicin and isoniazid. MODS relies on two wellknown properties of Mycobacterium tuberculosis (MTB): First, the rate of growth of TB bacilli in liquid medium is considerably higher than that on solid medium. Second, the morphology in liquid culture is characteristic and recognizable, consisting of so called \u201ccord\u201d like structures. By using an inverted light microscope to examine culture\n\nplates inoculated with sputum from patients with presumptive TB, MTB growth can be detected within 7\u201310 days, for both smear positive and negative samples, compared to conventional solid culture that takes 3\u20138 weeks [22, 23]. The MODS test has received increased attention in recent years and has been improved and standardized further for more widespread use [24, 25]. However, there is inadequate information about the full cost of the MODS procedure, including costs of materials, labour, laboratory equipment and overhead, which need to be properly evaluated.\nThe comparable diagnostic performance of the Xpert test with MODS and the urgent need of affordable tests for diagnosis of TB in HIV-infected patients, led us to perform this study in our setting in Uganda where HIV and TB are a high burden with an estimated incidence of 0.51 per 100 person year and 161 per 100,000 population respectively [26, 27]. The aim was to compare the cost-effectiveness of the utilizing the MODS test versus Xpert as primary tests for diagnosis of pulmonary tuberculosis (PTB) among patients infected with HIV. Our results could be useful for low income settings where implementation of the tests is planned or is already established.\nMethods\nStudy population The study population comprised adult HIV-infected patients older than 18 years, with presumptive active pulmonary TB. An HIV-infected patient could present with presumptive PTB regardless of whether they were on antiretroviral treatment or not, CD4 count, HIV clinical stage or history of previous treatment for TB. An HIV patient was presumed to have active PTB if they had cough for two or more weeks with or without fever, night sweats, loss of weight, or blood stained sputum [18].\nStudy setting The diagnostic procedures were conducted in a TB research laboratory located within the campus of Mulago National referral Hospital in Kampala, Uganda. Sputum specimens were obtained from consecutively presenting patients to the Mulago Hospital HIV outpatient clinic and from patients admitted to the medical department of the hospital. HIV-infected adults presenting with symptoms and signs of PTB were enrolled on the basis of the WHO TB screening criteria. Symptomatic patients provided a spot and morning sputum in a universal sterile sputum container. At the laboratory, the two samples were pooled and examined using MODS and Xpert.\nDiagnostic procedures All tests were performed by trained technicians. For the Xpert MTB/RIF assay, a sample reagent was added to the pooled sputum sample in a 2:1 ratio. The mixture\n\nWalusimbi et al. BMC Health Services Research (2016) 16:563\n\nPage 3 of 10\n\nwas incubated at room temperature for 15\u201330 min and agitated manually. A total of 1 ml of the mixture was introduced into an Xpert MTB/RIF cartridge, which was then loaded into a GeneXpert instrument, where the subsequent steps of sample lysis, nucleic acid extraction, and amplification occurred automatically. The instrument generates the test report automatically within 3 h which is printed and signed by the technician.\nThe MODS test was performed in a 24-well tissue plate. The patient sputum was processed (digestion and decontamination) with NALC/NAOH 2 % method for 15 min, followed by homogenization. The homogenized sample was then centrifuged at 3000 X g (Allegra\u00ae X-12 series) for 15 min to prepare a sediment. The sediment was re-suspended with phosphate buffer (pH 6.8) to make 1\u20132 ml. The media for the MODS was prepared with 4.7 g Middlebrook 7H9 broth (Difco, Sparks, MD) and 2 ml glycerol in 900 ml of distilled water. This media was autoclaved at 121 \u00b0C for 10 min, cooled to 45 \u00b0C and enriched with 100 ml of Oleic, Albumin, Dextrose, catalase (OADC). A portion of the processed sample (100 \u03bcl) and Middlebrook 7H9 broth (800 \u03bcl) and of antibiotic mixture (100 \u03bcl) of polymyxin B, Amphotericin B, Nalidixic acid, Trimethoprim and Azlocillin (PANTA) were then transferred into wells giving a final volume of 1 ml/well. Two wells were used for each processed and quality control sample. For positive control, 100 \u03bcl of a suspension of H37Rv isolate 0.5 McFarland standard, was used. For negative control, 800 \u03bcl of Middlebrook 7H9 broth, 100 \u03bcl PANTA without sample was used. The tissue plates were sealed with tape and ziplock bags and incubated at 37 oC. They were examined under an inverted light microscope at magnifications of X10 and X40 for cord formation.\n\nModel structure A decision-analysis model was constructed using TreeAge software (version 3.5) to compare the cost effectiveness of the MODS algorithm to the Xpert algorithm for diagnosis of TB (Fig. 1). The model involved 10,000 HIV patients with presumptive PTB. A positive MODS or Xpert test was either a true positive or a false negative based on the sensitivity of MODS or Xpert. A negative test was either a true negative or a false positive based on the specificity of MODS or Xpert.\nModel parameters The data for diagnostic accuracy in the model were sourced from systematic reviews of studies on diagnostic accuracy of Xpert [28] and MODS [19, 20] among HIV infected patients. We used the pooled values from the systematic reviews as the base estimates for sensitivity and specificity of the tests, and the 95 % confidence interval values as the outer limits for diagnostic accuracy of the tests (Table 1).\nCost data Estimates of the costs were made from the provider\u2019s perspective (Table 2). The costs for the diagnostic procedure of each test were collected by identifying all the reagents required to perform the test and their quantities. These were assessed by reviewing the standard operating procedures (SOP) of the tests and observation of laboratory technicians during performance of the tests in the research laboratory in Mulago Hospital. We then computed the cost per test by applying a price per quantity of the resource used for the respective test. Prices of the reagents, equipment, calibration and training costs were obtained from laboratory invoices between 2010\n\nFig. 1 Decision analysis model for diagnosis of pulmonary tuberculosis in HIV patients using MODS or Xpert strategy\n\nWalusimbi et al. BMC Health Services Research (2016) 16:563\n\nPage 4 of 10\n\nTable 1 Model assumptions for TB diagnosis using MODS or Xpert\n\nModel input\n\nBase value\n\nMin-Max\n\nReference\n\n# Test sensitivity Xpert\n\n0.79\n\n0.70\u20130.86\n\n[6]\n\n# Test specificity Xpert\n\n0.98\n\n0.97\u20130.99\n\n[6]\n\n# Test sensitivity MODS\n\n0.88\n\n0.86\u20130.9\n\n[19]\n\n# Test specificity MODS\n\n0.98\n\n0.97\u20130.99\n\n[19]\n\n# TB prevalence\n\n0.20\n\n0.10\u20130.30\n\n[49\u201352]\n\nand 2014. All costs were estimated in US$ based 2014 prices. The local costs were converted using the average exchange rate for 2014 of 2700 Uganda shillings for one US dollar. In the model we included some sunk costs, e g water, space and overheads, but since these costs were the same for the two alternatives, they were automatically cancelled out in the calculation. We also reviewed previous studies and guidelines to direct our data collection [8, 29\u201332]. We used standard tables of annualization\n\nand a discount rate of 3 % for the capital equipment. We made the following general assumptions to arrive at the unit costs for each test: With regard to equipment, the useful life of the centrifuge, incubator, autoclave, biosafety hood, inverted microscope were assumed to be 10 years and 100 % of their use was allocated to the MODS test. The useful life of a laboratory fridge was assumed to be 10 years and 20 % its use was allocated to the MODS test. The useful life of the GeneXpert machine was assumed to be 5 years and 100 % of its use was allocated to the Xpert test [33]. We used the concessional price of the GeneXpert machine and cartridges which is provided to resource poor settings. The useful life of digital pipettes was assumed to be 3 years and 100 % of their use was allocated to the MODS test. We assumed that both tests would require 25 M2 of work space and allocated the cost for space equally between the two tests. We allocated staff salary based on the time required to process the MODS test (2 h) and Xpert test (30 min). We assumed a maximum of\n\nTable 2 Provider costs involved in the MODS and Xpert diagnostic procedure (US $)\n\nMethod\n\nComponent\n\nMODS\n\nEquipment Xpert test\n\n# Xpert MTB/Rif Machine 4 module and accessories\n\n\u2013\n\nEquipment MODS test\n\n# Centrifuge (Beckmann Coulter x12R)\n\n18000\n\n# Incubator (CO2)\n\n19500\n\n# Autoclave\n\n24500\n\n# Bio-safety cabinet (class 2)\n\n11500\n\n# Inverted microscope\n\n2700\n\n# Fridge\n\n1300\n\n# Vortex\n\n524\n\n# Pipettes (200ul-1 ml pipette)\n\n218\n\n# Pipettes (50ul-200ul pipette)\n\n218\n\nConsumables\n\n# Xpert cartridge & reagent kit\n\n\u2013\n\n# MODS culture media per year\n\n245\n\n# MODS Culture plate\n\n5\n\n# MODS digestion & decontamination reagents per year\n\n416\n\nStaff\n\n# Annual salary for a laboratory technician\n\n4000\n\n# Training costs (5 days for Xpert, 22 for MODS)\n\n990\n\nOverheads: Utilities, space\n\n# Utilities (water, power, stationary) per year\n\n540\n\n# Space cost per M2 (25 M2 for either Xpert or MODS)\n\n463\n\nQuality control\n\n# Xpert calibration cartridge per 2000 tests\n\n\u2013\n\n# MODS proficiency panels per year\n\n940\n\nXpert\n19900\n\u2013 \u2013 \u2013 \u2013 \u2013 \u2013 \u2013 \u2013 \u2013\n10 \u2013 \u2013 \u2013\n4000 225\n540 463\n450 \u2013\n\nSource\nInvoice\nInvoice Invoice Invoice Invoice Invoice Invoice Invoice Invoice Invoice\nInvoice Invoice Invoice Invoice\nInvoice Invoice\nInvoice Invoice\nInvoice Invoice\n\nWalusimbi et al. BMC Health Services Research (2016) 16:563\n\nPage 5 of 10\n\n20 tests per day for MODS and 16 tests per day for Xpert and a total of 264 working days per year. We assumed MODS training required 22 days and Xpert 5 days and a refresher training for both tests by the laboratory staff every 3 years.\nModel outcomes The model\u2019s outcome measure were cost per TB patient diagnosed when the MODS test or the Xpert test were used for TB diagnosis in HIV infected individuals. We also derived incremental cost-effectiveness ratios (ICERs), expressed as US $ per TB patient diagnosed.\nSensitivity analysis We performed sensitivity analysis of our model based on adjustments of: (1) the diagnostic accuracy of the MODS and Xpert tests using the minimum and maximum values from the systematic reviews, (2) the useful life of the capital equipment of both tests between five and ten years, (3) the average number of tests performed per day between five and twenty, (4) different prices of the reagents for MODS and the Xpert cartridge, (5) the percentage allocated for shared equipment or staff time, (6) TB prevalence (10\u201330 %), corresponding to the most common values in this patient group from the studies in the systematic reviews.\nData analysis Cost data was entered and analyzed in Excel. The cost of equipment for each test was obtained by dividing the annualized cost of the equipment over the number of tests performed each year. The cost of consumables for each test was obtained by the dividing the gross cost of a given measure of each reagent over the average number of tests that can performed using that amount. The cost of quality control (QC) for a MODS test was obtained by dividing the total costs incurred for QC per year over the average number of MODS tests that can be performed each year. The cost of QC for Xpert was obtained by dividing the cost of the Xpert calibration cartridge over 2000 tests, which is the number recommended by the manufacturer when QC should be performed. Cost-effectiveness analysis was performed by in-putting the test probabilities and unit costs into the TreeAge software. A base case analysis was performed using the pooled estimates for diagnostic accuracy of the tests. Sensitivity analysis was performed by modifying the parameters in the model.\nResults\nCost The average cost for the MODS test was US$ 6.53 compared to US$ 12.41 for the Xpert test. Consumables (reagents and chemicals) accounted for 59 % of the cost for the MODS test while the Xpert cartridge with the reagent\n\nkit accounted for 84 % of the cost for the Xpert test (Table 3).\nThe effect of changes in the base-case assumptions on the unit cost of MODS and Xpert are summarized in Table 4. In the case of MODS, reducing the useful life of the capital equipment from ten to five years, increased the cost of the test moderately to US$ 8.04. Reducing the number of tests performed each day to five from twenty increased the cost of the test substantially to US$ 11.8. Increasing the price of reagents and chemicals by two-fold increased cost of the test minimally to US$ 7.8. Allocating 100 % of all shared equipment and staff time to MODS increased the cost for the test moderately to US$ 8.9.\nIn the case of Xpert, increasing the useful life of the GeneXpert machine from five to ten years lowered the cost of the test substantially to US$ 6.5. Reducing the number of tests performed each day to five from sixteen lowered the cost of the test minimally to US$ 11.8. Reducing the cartridge price and reagent kit by two-fold lowered the cost of the test substantially to US$ 7.1. Allocating 100 % of staff time to Xpert raised the cost of the test minimally to US$13.2.\nOutcomes and Cost effectiveness The MODS test generally detected more PTB patients by 11 % (range 5\u201323 %) compared to the Xpert test. In the base-case analysis, involving a cohort of 10,000 HIV patients with a PTB prevalence of 20 %, the algorithm using MODS would diagnose 1920 patients compared to 1740 patients by the algorithm using Xpert. The costeffectiveness ratio of using MODS was US$ 34 per TB patient diagnosed compared to US$ 71 when using Xpert. The algorithm using MODS therefore detected more patients at lower costs, making it dominant over the algorithm using Xpert (Table 5).\nSensitivity analysis In one-way sensitivity analyses, TB prevalence, followed by the cost of the MODS test had the most influence on results. The accuracy of the tests had the least influence (Fig. 2). However, the ratio of the total costs for diagnosis\n\nTable 3 Costs of MODS and Xpert by type of input, (2014$)\n\nMethod, (% of total)\n\nType of input\n\nMODS\n\nXpert\n\nTotal (US$)\n\n6.53\n\n12.41\n\n# Equipment\n\n1.76, (27)\n\n1.37, (11)\n\n# Consumables\n\n3.84, (59)\n\n10.37, (84)\n\n# Staff (salary and training)\n\n0.46, (07)\n\n0.15, (01)\n\n# Quality control\n\n0.18, (03)\n\n0.23, (02)\n\n# Overheads (utilities and space)\n\n0.29, (04)\n\n0.29, (02)\n\nWalusimbi et al. BMC Health Services Research (2016) 16:563\n\nPage 6 of 10\n\nTable 4 Effect of changes in the base-case assumptions on the unit cost of MODS and Xpert, (US $)\n\nType of test/parameter\n\nEffect on cost Increase/ decrease (%)\n\nMODS (base estimate US$ 6.53)\n\n-Reduce useful life of capital equipment 8.04\n\n+23\n\nfrom 10 to 5 years\n\n-Reduce number of tests to 5 each day 11.8\n\n+ 81\n\n-Double price of consumables\n\n7.8\n\n+ 19\n\n-Allocate 100 % of shared equipment & 8.9\n\n+ 36\n\nstaff time to MODS\n\nXpert (base estimate US$ 12.41)\n\n-Increasing useful life of Xpert from 5 6.5\n\n- 48\n\nto 10 years\n\n-Reduce number of tests to 5 each day 11.8\n\n-5\n\n-Reduce price of cartridge by half\n\n7.1\n\n- 43\n\n- Allocate 100 % of staff time to Xpert 13.2\n\n+6\n\nof TB patients using either the MODS or Xpert algorithm remained constant across variable prevalence situations (Table 6). The dominance of the algorithm using MODS was persistent across various values of PTB prevalence and for all the cost values of Xpert between 6.5\u201313.2 US$. The threshold value for cost, where using the Xpert algorithm would be optimal over the MODS algorithm was US $ 5.92.\nDiscussion We evaluated the cost-effectiveness of using an algorithm based on the MODS test versus Xpert test for diagnosis of Pulmonary Tuberculosis (PTB) in HIV patients from the perspective of the provider. The MODS algorithm was dominant over the Xpert algorithm despite adjustments in test accuracy, cost and TB prevalence.\nPrevalence of PTB and the costs of the tests were the most influential parameters in our findings. However, while the prevalence of PTB had high influence on the cost per TB patient detected, it did not change the ratio of the total costs for the cases detected using either MODS or Xpert. Thus, in settings where prevalence of PTB in HIV patients is for example 30 % or more, implementation of the Xpert algorithm could be worthwhile despite the higher total costs incurred in comparison to MODS algorithm.\nFor both tests, the consumables had the most influence on the unit costs, although variation in the useful\n\nlife of equipment and the average number of tests done each day also had substantial influence on the unit costs. The threshold for cost in order for Xpert to be optimal over MODS for diagnosis of PTB in our study population was about US$ 6. This value lies within the recommended US$ 4\u20136 for any new diagnostic to be placed at the microscopy-center level of the health care system [34]. This therefore would require further reduction of the currently subsidized price of the Xpert cartridge by more than half.\nWe did not consider data and costs of X-ray in our study. This is because we focused on new tests (MODS and Xpert) that provide a definitive (bacteriological) diagnosis of PTB which is important to ensure correct TB treatment. Although radiological tests such as X-ray have an important role in evaluating presumed PTB patients, their use often results in over diagnosis of PTB [35, 36]. They are therefore utilized primarily for diagnosis of extra-pulmonary TB and to assess presumed PTB patients for other etiologies of respiratory illness. In regard to cost-effectiveness analysis the costs of X-ray would cancel out one another since they would be the same in either the MODS or Xpert diagnostic algorithm.\nA recurrent concern limiting the use of MODS is the total cost of the test arising from infrastructure requirements. While it is argued that roll out of Xpert requires minimal laboratory modifications, the costs involved in modifying available space to make it suitable for operating the Xpert test, and the costs for installation of some accessories like power inverters, air conditioners, have limited its placement at the microscopy-center level [37, 38]. Moreover, we found in our study that equipment and space accounted just over 30 % of the total cost of the MODS test. The MODS could therefore be a promising method for decentralizing sputum culture services up-to the microscopy-center level of the health care system. Besides, the test provides a platform for extended drugsusceptibility testing for drug resistant TB and can be assembled on site. Therefore, despite the low incidence of drug resistant TB among newly diagnosed patients in Uganda [39], MODS would offer rapid diagnosis of drug-resistant TB in a one-off test. On the other hand, presumed drug-resistant patients identified using the Xpert test require confirmation with culture. Thus additional costs are incurred in such situations. Our study therefore, could have under-estimated the benefits of implementing MODS in our setting.\n\nTable 5 Cost-effectiveness of TB diagnosis using MODS or Xpert in a base-case analysis for a cohort of 10,000 HIV presumptive PTB patients\n\nStrategy Mean cost per test ($) incremental cost per test ($) Cases detected Incremental cases detected Cost effectiveness ICER\n\nMODS\n\n6.53\n\n1920\n\n34\n\nMore cost effective\n\nXpert\n\n12.41\n\n(5.88)\n\n1740\n\n180\n\n71\n\nDominated\n\nWalusimbi et al. BMC Health Services Research (2016) 16:563\n\nPage 7 of 10\n\nFig. 2 One-way sensitivity analysis comparing the influence of model parameters on cost effectiveness of MODS versus Xpert strategy for diagnosis of pulmonary tuberculosis in HIV patients. The x-axis is the cost per patient diagnosed. Each horizontal bar represents a parameter varied over the range indicated. Wider bars indicate larger differences in the cost per patient diagnosed by varying the parameter\n\nEven though the MODS was the preferred alternative in our cost-effectiveness analysis, the Xpert has several advantages over the MODS test in regard to time to detection, biosafety, level of skills required to operate the test, labour intensity during performance of the test, and minimal variation in the test performance and quality assurance.\nThe MODS requires a median of seven days to detect growth in comparison to Xpert that provides results in 3 h allowing for same day detection and treatment. Xpert therefore has more potential than MODS to avert patient loss during the process of TB diagnosis. Xpert also has more potential to avert transmission of disease arising from early treatment upon detection of TB [40]. Unfortunately, although the turn-around-time of Xpert is short, in real life settings there is significant delay in\n\ngetting the Xpert results and initiating TB treatment [41, 42] which counters these potential benefits.\nFurther, current Treatment algorithms suggest that all patients with positive Xpert results should immediately start anti-tuberculosis treatment. However, Xpert can detect DNA from both viable and nonviable TB bacilli. This presents a challenge that needs to be addressed as the Xpert test is rolled-out. This is important because despite the low likelihood of false Xpert positivity among new TB patients, false Xpert positivity among previously treated PTB patients may be common. [43, 44]. In such situations, clinicians may consider awaiting confirmatory testing using culture tests-which is a major advantage the MODS offers\nA major advantage of utilization of Xpert is the limited concerns about biosafety during its use. On the other\n\nTable 6 Expected diagnostic yield and costs of MODS versus Xpert strategy across variable prevalence situations for a cohort of 10,000 HIV presumptive PTB patients\n\nPrevalence/Strategy\n\nTotal TB cases detected\n\nCost per case detected\n\nTotal cost for cases detected\n\nRatio total cost\n\nTB prevalence (10 %)\n\nMODS\n\n1060\n\n62\n\n65720\n\nXpert\n\n970\n\n128\n\n124160\n\n1.89\n\nTB prevalence (20 %)\n\nMODS\n\n1920\n\n34\n\n65280\n\nXpert\n\n1740\n\n71\n\n123540\n\n1.89\n\nTB prevalence (30 %)\n\nMODS\n\n2780\n\n23\n\n63940\n\nXpert\n\n2510\n\n49\n\n122990\n\n1.92\n\nWalusimbi et al. BMC Health Services Research (2016) 16:563\n\nPage 8 of 10\n\nhand, biosafety is an important concern with the MODS and the test has so far therefore, been limited to referral or research TB culture laboratories. The risks about utilization of the MODS could however, be addressed by undertaking the procedures to perform the test inside a biosafety (Class 2) cabinet and having personal respiratory protection for laboratory staff such as N-95 masks. Further, since the MODS simply involves the inoculation of a sputum sample into a plate, after which the plate is sealed within a plastic bag and never again opened, the biosafety risks of the MODS test could be comparable to sputum smear microscopy as one study has shown previously [45].\nMODS is more labour intensive and requires more skilled training to perform in comparison to Xpert. Recent innovations enabling automated interpretation of the test could make the labour and skills required to perform the test comparable to Xpert [46]. This could enable deployment of the tests to peripheral laboratories even more feasible. Still, there would be need to standardize the procedures for the test and set up quality assurance systems. Currently, the probability of invalid results from the MODS test requiring repeat testing, is comparable to Xpert but could be reduced further through these measures [38].\nCost-effectiveness analysis is not an evaluation of affordability. Thus the affordability of deploying the MODS or Xpert in relation to the current and future economic developments in several of the resource poor settings was not answered by our study. One study that has evaluated the cost and affordability of Xpert found that targeted use of the test would be affordable in the majority of high burden TB countries [47].\nThe study also did not compare the epidemiological and health system effects of using either MODS or Xpert for diagnosis of HIV associated TB. One study that evaluated the population effects of Xpert found that the test could substantially reduce the TB burden in a resource limited and HIV prevalent setting [11]. A similar study involving MODS is required given the dominance of MODS over Xpert. Based on modelling, MODS could have similar population effects with Xpert [48]. Our study assumed only a single diagnostic attempt during the patient\u2019s disease course with no repeat diagnostic attempts. We also did not explore diagnostic attempts for multidrug-resistant tuberculosis using either MODS or Xpert.\nConclusions The algorithm using MODS was more cost-effective compared to the algorithm using Xpert for the diagnosis of TB among HIV patients in our setting. Efforts to scale-up MODS therefore need to be explored. However, other non-economic factors may still favour the use of\n\nXpert in our setting or other similar settings. But the current cost of the Xpert test, with subsidies, needs to be reduced further by more than half to make it economically competitive with MODS.\nAbbreviations HIV: Human immune deficiency virus; M.TB: Mycobacterium tuberculosis; MODS: Microscopic observation drug susceptibility test; PTB: Pulmonary tuberculosis; QC: Quality control; SOP: Standard operating procedures; TB: Tuberculosis; Xpert/XPT: Xpert MTB/Rif test\nAcknowledgements We sincerely thank the management and clinical staff of Makerere University Joint AIDS Program (MJAP) at Mulago National Referral Hospital, for their efforts in patient care. MJAP is implemented with funding from PEPFAR and technical support from CDC Uganda. SW also received a research scholarship from the World Federation of Scientists (WFS).\nFunding This study was supported by a grant from the Swedish International Development Agency (SIDA). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\nAvailability of data and material Data is not available for online access. Readers who wish to gain access to the data can write to the corresponding author Dr. Achilles Katamba at akatamba@chs.mak.ac.ug with their requests.\nAuthors' contributions SW, ADC, LB, AK: Conceived and designed the study. SW, BK. RR: Collected and analyzed the data. SW drafted the manuscript. ADC, MH, LB, AK: Critically reviewed the manuscript: All authors read and approved the final manuscript.\nAuthors' information Not applicable.\nCompeting interests The authors declare that they have no competing interests.\nConsent for publication Not applicable.\nEthics approval and consent to participate The study was approved by the Uganda National Council for Science and Technology (reference number: HS 1214). The diagnostic procedures in the laboratory were performed on material provided by patients with written informed consent.\nAuthor details 1Department of Microbiology, Makerere University College of Health Sciences, Kampala, Uganda. 2Department of Public Health Sciences, Karolinska Institute, Solna, Sweden. 3HealthNet Consult, Kampala, Uganda. 4Department of Microbiology, Public Health Agency of Sweden, Solna, Sweden. 5Department of Medicine, Clinical Epidemiology Unit, Makerere University, College of Health Sciences, Kampala, Uganda. 6School of Health, Care and social Welfare, Malardalen University, Vasteras, Sweden.\nReceived: 4 September 2015 Accepted: 28 September 2016\nReferences 1. Steingart KR, Ng V, Henry M, Hopewell PC, Ramsay A, Cunningham J,\nUrbanczik R, Perkins MD, Aziz MA, Pai M. Sputum processing methods to improve the sensitivity of smear microscopy for tuberculosis: a systematic review. Lancet Infect Dis. 2006;6(10):664\u201374. 2. World Health Organization. Global tuberculosis report. 2014. http://apps. who.int/iris/bitstream/10665/137094/1/9789241564809_eng.pdf. Accessed 04 Apr 2015. 3. World Health Organization. The End TB Strategy. http://www.who.int/tb/ post2015_TBstrategy.pdf. Accessed 04 Apr 2015.\n\nWalusimbi et al. BMC Health Services Research (2016) 16:563\n\nPage 9 of 10\n\n4. Pai NP, Pai M. Point-of-care diagnostics for HIV and tuberculosis: landscape, pipeline, and unmet needs. Discov Med. 2012;13(68):35\u201345.\n5. McNerney R, Maeurer M, Abubakar I, Marais B, McHugh TD, Ford N, Weyer K, Lawn S, Grobusch MP, Memish Z, et al. Tuberculosis diagnostics and biomarkers: needs, challenges, recent advances, and opportunities. J Infect Dis. 2012;205 Suppl 2:S147\u2013158.\n6. Steingart KR, Schiller I, Horne DJ, Pai M, Boehme CC, Dendukuri N. Xpert(R) MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev. 2014;1:CD009593.\n7. Boehme CC, Nabeta P, Hillemann D, Nicol MP, Shenai S, Krapp F, Allen J, Tahirli R, Blakemore R, Rustomjee R, et al. Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med. 2010;363(11):1005\u201315.\n8. Vassall A, van Kampen S, Sohn H, Michael JS, John KR, den Boon S, Davis JL, Whitelaw A, Nicol MP, Gler MT, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: a cost-effectiveness analysis. PLoS Med. 2011;8(11):e1001120.\n9. Abimbola TO, Marston BJ, Date AA, Blandford JM, Sangrujee N, Wiktor SZ. Cost-effectiveness of tuberculosis diagnostic strategies to reduce early mortality among persons with advanced HIV infection initiating antiretroviral therapy. J Acquir Immune Defic Syndr. 2012;60(1):e1\u20137.\n10. Menzies NA, Cohen T, Lin HH, Murray M, Salomon JA. Population health impact and cost-effectiveness of tuberculosis diagnosis with Xpert MTB/RIF: a dynamic simulation and economic evaluation. PLoS Med. 2012;9(11):e1001347.\n11. Langley I, Lin HH, Egwaga S, Doulla B, Ku CC, Murray M, Cohen T, Squire SB. Assessment of the patient, health system, and population effects of Xpert MTB/RIF and alternative diagnostics for tuberculosis in Tanzania: an integrated modelling approach. Lancet Global Health. 2014;2(10):e581\u2013591.\n12. Murray CJ, Ortblad KF, Guinovart C, Lim SS, Wolock TM, Roberts DA, Dansereau EA, Graetz N, Barber RM, Brown JC, et al. Global, regional, and national incidence and mortality for HIV, tuberculosis, and malaria during 1990-2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet. 2014;384(9947):1005\u201370.\n13. Pinto LM, Pai M, Dheda K, Schwartzman K, Menzies D, Steingart KR. Scoring systems using chest radiographic features for the diagnosis of pulmonary tuberculosis in adults: a systematic review. Eur Respir J. 2013;42:480\u201394.\n14. Corbett EL, Watt CJ, Walker N, Maher D, Williams BG, Raviglione MC, Dye C. The growing burden of tuberculosis: global trends and interactions with the HIV epidemic. Arch Intern Med. 2003;163(9):1009\u201321.\n15. Schnippel K, Meyer-Rath G, Long L, MacLeod W, Sanne I, Stevens WS, Rosen S. Scaling up Xpert MTB/RIF technology: the costs of laboratoryvs. clinic-based roll-out in South Africa. Trop Med Int Health. 2012;17(9): 1142\u201351.\n16. Clouse K, Page-Shipp L, Dansey H, Moatlhodi B, Scott L, Bassett J, Stevens W, Sanne I, Van Rie A. Implementation of Xpert MTB/RIF for routine pointof-care diagnosis of tuberculosis at the primary care level. S Afr Med J. 2012;102(10):805\u20137.\n17. Mupfumi L, Makamure B, Chirehwa M, Sagonda T, Zinyowera S, Mason P, Metcalfe JZ, Mutetwa R. Impact of Xpert MTB/RIF on Antiretroviral TherapyAssociated Tuberculosis and Mortality: A Pragmatic Randomized Controlled Trial. Open forum infectious diseases. 2014;1(1):ofu038.\n18. Theron G, Peter J, Dowdy D, Langley I, Squire SB, Dheda K. Do high rates of empirical treatment undermine the potential effect of new diagnostic tests for tuberculosis in high-burden settings? Lancet Infect Dis. 2014;14(6):527\u201332.\n19. Philip Wikman-Jorgensen, Jara Llenas-Garc\u0131a, Michael Hobbins, Jochen Ehmer, Rosa Abellana, Alessandra Queiroga Goncalves, Toma`s Maria Pe\u00b4rezPorcuna, Ascaso C: Microscopic observation drug susceptibility assay for the diagnosis of TB and MDR-TB in HIV-infected patients: a systematic review and meta-analysis. Eur Respir J 2014; in press | DOI: 10.1183/ 0903193600079614\n20. Leung E, Minion J, Benedetti A, Pai M, D M. Microcolony culture techniques for tuberculosis diagnosis: a systematic review. Int J Tuberc Lung Dis. 2011;16(1):16\u201323.\n21. Uganda National Population and Housing Census 2014. Provisional results. http://www.ubos.org/onlinefiles/uploads/ubos/NPHC/2014%20National% 20Census%20Main%20Report.pdf. Accessed 07 Oct 2016.\n22. Caviedes L, Lee TS, Gilman RH, Sheen P, Spellman E, Lee EH, Berg DE, Montenegro-James S. Rapid, efficient detection and drug susceptibility testing of Mycobacterium tuberculosis in sputum by microscopic observation of broth cultures. The Tuberculosis Working Group in Peru. J Clin Microbiol. 2000;38(3):1203\u20138.\n\n23. Moore DA, Evans CA, Gilman RH, Caviedes L, Coronel J, Vivar A, Sanchez E, Pinedo Y, Saravia JC, Salazar C, et al. Microscopic-observation drug-susceptibility assay for the diagnosis of TB. N Engl J Med. 2006;355(15):1539\u201350.\n24. Comina G, Mendoza D, Velazco A, Coronel J, Sheen P, Gilman RH, Moore DAJ, M Z. Development of an automated MODS plate reader to detect early growth of Mycobacterium tuberculosis. J Microsc. 2011;242:325\u201330.\n25. Chaiyasirinroje B, Aung MN, Moolphate S, Kasetjaroen Y, Rienthong S, Rienthong D, Nampaisan O, Nedsuwan S, Sangchun W, Suriyon N, et al. Prospective evaluation of simply modified MODS assay: an effective tool for TB diagnosis and detection of MDR-TB. Inf Drug Resist. 2012;5:79\u201386.\n26. UNAIDS 2016 estimates: HIV ESTIMATES WITH UNCERTAINTY BOUNDS 19902015. Available at http://www.unaids.org/en/resources/documents/2016/HIV_ estimates_with_uncertainty_bounds_1990-2015. Accessed 28 June 2016.\n27. World Health Organization: Global Tuberculosis Report 2015. Country profiles FOR 22 HIGH-BURDEN COUNTRIES. Available at http://www.who.int/tb/publications/ global_report/gtbr15_annex02.pdf?ua=1. Accessed 20 June 2016.\n28. Steingart KR, Schiller I, Horne DJ, Pai M, Boehme CC, Dendukuri N. Xpert\u00ae MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev. 2014, Issue 1. Art. No.: CD009593.doi:10. 1002/14651858.CD009593.pub3.\n29. Bwanga F, Haile M, Joloba ML, Ochom E, Hoffner S. Direct nitrate reductase assay versus microscopic observation drug susceptibility test for rapid detection of MDR-TB in Uganda. PLoS One. 2011;6(5):e19565.\n30. Russell S. The economic burden of illness for households in developing countries: a review of studies focusing on malaria, tuberculosis, and human immunodeficiency virus/acquired immunodeficiency syndrome. Am J Trop Med Hyg. 2004;71(2 Suppl):147\u201355.\n31. Langley I, Adams E, Doulla B, Squire SB. Operational modelling to guide implementation and scale-up of diagnostic tests within the health system: exploring opportunities for parasitic disease diagnostics based on example application for tuberculosis. Parasitology. 2014;141(14):1795\u2013802.\n32. Sohn H, Minion J, Albert H, Dheda K, Pai M. TB diagnostic tests: how do we figure out their costs? Expert Rev Anti-Infect Ther. 2009;7(6):723\u201333.\n33. Vincent Zishiri, Salome Charalambous, Maunank R. Shah, Violet Chihota, Liesl Page-Shipp, Gavin J. Churchyard, Christopher J. Hoffmann: Implementing a Large-Scale Systematic Tuberculosis Screening Program in Correctional Facilities in South Africa. Open Forum Infect Dis 2014. Doi: 10.1093/ofid/ofu121.\n34. Tameris MD, Hatherill M, Landry BS, Scriba TJ, Snowden MA, Lockhart S, Shea JE, McClain JB, Hussey GD, Hanekom WA, et al. Safety and efficacy of MVA85A, a new tuberculosis vaccine, in infants previously vaccinated with BCG: a randomised, placebo-controlled phase 2b trial. Lancet. 2013;381(9871):1021\u20138.\n35. Wekesa C, Kirenga BJ, Joloba ML, Bwanga F, Katamba A, Kamya MR. Chest X-ray vs. Xpert(R) MTB/RIF assay for the diagnosis of sputum smear-negative tuberculosis in Uganda. Int J Tuberc Lung Dis. 2014;18(2):216\u20139.\n36. Theron G, Pooran A, Peter J, van Zyl-Smit R, Kumar Mishra H, Meldau R, Calligaro G, Allwood B, Sharma SK, Dawson R, et al. Do adjunct tuberculosis tests, when combined with Xpert MTB/RIF, improve accuracy and the cost of diagnosis in a resource-poor setting? Eur Respir J. 2012;40(1):161\u20138.\n37. Abdurrahman ST, Emenyonu N, Obasanya OJ, Lawson L, Dacombe R, Muhammad M, Oladimeji O, Cuevas LE. The hidden costs of installing xpert machines in a tuberculosis high-burden country: experiences from Nigeria. Pan Afr Med J. 2014;18:277. doi:10.11604/pamj.2014.18.277.3906.\n38. Creswell J, Codlin AJ, Andre E, Micek MA, Bedru A, Carter EJ, Yadav RP, Mosneaga A, Rai B, Banu S, et al. Results from early programmatic implementation of Xpert MTB/RIF testing in nine countries. BMC Infect Dis. 2014;14:2.\n39. Lukoye D, Adatu F, Musisi K, Kasule GW, Were W, Odeke R, Kalamya JN, Awor A, Date A, Joloba ML. Anti-tuberculosis drug resistance among new and previously treated sputum smear-positive tuberculosis patients in Uganda: results of the first national survey. PLoS One. 2013;8(8):e70763.\n40. Cox HS, Daniels JF, Muller O, Nicol MP, Cox V, van Cutsem G, Moyo S, De Azevedo V, Hughes J. Impact of Decentralized Care and the Xpert MTB/RIF Test on Rifampicin-Resistant Tuberculosis Treatment Initiation in Khayelitsha, South Africa. Open Forum Infect Dis. 2015;2(1):ofv014.\n41. Cohen GM, Drain PK, Noubary F, Cloete C, Bassett IV. Diagnostic delays and clinical decision making with centralized Xpert MTB/RIF testing in Durban, South Africa. J Acquir Immune Defic Syndr. 2014;67(3):e88\u201393.\n42. Hanrahan CF, Haguma P, Ochom E, Kinera I, Cobelens F, Cattamanchi A, Davis L, Katamba A, Dowdy D. Implementation of Xpert MTB/RIF in Uganda: Missed Opportunities to Improve Diagnosis of Tuberculosis. Open Forum Infect Dis. 2016;3(2):ofw068.\n\nWalusimbi et al. BMC Health Services Research (2016) 16:563\n43. Theron G, Venter R, Calligaro G, Smith L, Limberis J, Meldau R, Chanda D, Esmail A, Peter J, Dheda K. Xpert MTB/RIF Results in Patients With Previous Tuberculosis: Can We Distinguish True From False Positive Results? Clin Infect Dis. 2016;62(8):995\u20131001.\n44. Boyles TH, Hughes J, Cox V, Burton R, Meintjes G, Mendelson M. Falsepositive Xpert(R) MTB/RIF assays in previously treated patients: need for caution in interpreting results. Int J Tuberc Lung Dis. 2014;18(7):876\u20138.\n45. Kim SJ, Lee SH, Kim IS, Kim HJ, Kim SK, Rieder HL. Risk of occupational tuberculosis in National Tuberculosis Programme laboratories in Korea. Int J Tuberc Lung Dis. 2007;11(2):138\u201342.\n46. Alva A, Aquino F, Gilman RH, Olivares C, Requena D, Gutierrez AH, Caviedes L, Coronel J, Larson S, Sheen P, et al. Morphological characterization of Mycobacterium tuberculosis in a MODS culture for an automatic diagnostics through pattern recognition. PLoS One. 2013;8(12):e82809.\n47. Pantoja A, Fitzpatrick C, Vassall A, Weyer K, Floyd K. Xpert MTB/Rif for diagnosis of tuberculosis and drug-resistant tuberculosis: a cost and affordability analysis. Eur Respir J. 2013;42:708\u201320. 2012.\n48. Lin HH, Dowdy D, Dye C, Murray M, Cohen T. The impact of new tuberculosis diagnostics on transmission: why context matters. Bull World Health Organ. 2012;90(10):739\u2013747A.\n49. Abebe G, Deribew A, Apers L, Abdissa A, Kiflie Y, Koole O, Colebunders R. Evaluation of the 2007 WHO guideline to diagnose smear negative tuberculosis in an urban hospital in Ethiopia. BMC Infect Dis. 2013;13:427.\n50. Huerga H, Varaine F, Okwaro E, Bastard M, Ardizzoni E, Sitienei J, Chakaya J, M B. Performance of the 2007 WHO algorithm to diagnose smear-negative pulmonary tuberculosis in a HIV prevalent setting. PLoS One. 2012;7(12): e51336.\n51. Mupfumi L, Makamure B, Chirehwa M, Sagonda T, Zinyowera S, Mason P, Metcalfe JZ, Mutetwa R. Impact of Xpert MTB/RIF on Antiretroviral TherapyAssociated Tuberculosis and Mortality: A Pragmatic Randomized Controlled Trial. Open Forum Infect Dis. 2014;1(1):ofu038.\n52. Theron G, Zijenah L, Chanda D, Clowes P, Rachow A, Lesosky M, Bara W, Mungofa S, Pai M, Hoelscher M, et al. Feasibility, accuracy, and clinical effect of point-of-care Xpert MTB/RIF testing for tuberculosis in primary-care settings in Africa: a multicentre, randomised, controlled trial. Lancet. 2014;383(9915):424\u201335.\n\nPage 10 of 10\n\nSubmit your next manuscript to BioMed Central and we will help you at every step:\n\u2022 We accept pre-submission inquiries \u2022 Our selector tool helps you to \ufb01nd the most relevant journal \u2022 We provide round the clock customer support \u2022 Convenient online submission \u2022 Thorough peer review \u2022 Inclusion in PubMed and all major indexing services \u2022 Maximum visibility for your research\nSubmit your manuscript at www.biomedcentral.com/submit\n\n\n",
    "authors": [
      "Simon Walusimbi",
      "Brendan Kwesiga",
      "Rashmi Rodrigues",
      "Melles Haile",
      "Ayesha De Costa",
      "Lennart Bogg",
      "Achilles Katamba"
    ],
    "doi": "10.1186/s12913-016-1804-9",
    "year": null,
    "item_type": "journalArticle",
    "url": "http://bmchealthservres.biomedcentral.com/articles/10.1186/s12913-016-1804-9"
  },
  {
    "key": "5RSZT2IK",
    "title": "Bottom-up or top-down: unit cost estimation of tuberculosis diagnostic tests in India",
    "abstract": "S E T T I N G : Of 18 sites that participated in an implementation study of the Xpertw MTB/RIF assay in India, we selected five microscopy centres and two reference laboratories.",
    "full_text": "INT J TUBERC LUNG DIS 21(4):375\u2013380 Q 2017 The Union http://dx.doi.org/10.5588/ijtld.16.0496\n\nBottom-up or top-down: unit cost estimation of tuberculosis diagnostic tests in India\n\nS. Rupert,* A. Vassall,*\u2020 N. Raizada,\u2021 S. D. Khaparde,\u00a7 C. Boehme,\u2021 V. S. Salhotra,\u00b6 K. S. Sachdeva,\u00a7 S. A. Nair,# A. H. van\u2019t Hoog*,**\n*Amsterdam Institute for Global Health and Development, Amsterdam, The Netherlands; \u2020Social and Mathematical Epidemiology Group, Department of Global Health and Development, London School of Hygiene & Tropical Medicine, London, UK; \u2021Foundation for Innovative New Diagnostics, New Delhi, \u00a7Central TB Division, Government of India, New Delhi, \u00b6Central TB Division, Ministry of Health and Family Welfare, New Delhi, #World Health Organization, Country Office for India, New Delhi, India; **Academic Medical Centre, University of Amsterdam, Department of Global Health, Amsterdam, The Netherlands\n\nSUMMARY\n\nS E T T I N G : Of 18 sites that participated in an implementation study of the Xpertw MTB/RIF assay in India, we selected five microscopy centres and two reference laboratories. O B J E C T I V E : To obtain unit costs of diagnostic tests for tuberculosis (TB) and drug-resistant TB. D E S I G N : Laboratories were purposely selected to capture regional variations and different laboratory types. Both bottom-up and the top-down methods were used to estimate unit costs. R E S U LT S : At the microscopy centres, mean bottom-up unit costs were respectively US$0.83 (range US$0.60\u2013 US$1.10) and US$12.29 (US$11.61\u2013US$12.89) for sputum smear microscopy and Xpert. At the reference laboratories, mean unit costs were US$1.69 for the\n\ndecontamination procedure, US$9.83 for a solid culture, US$11.06 for a liquid culture, US$29.88 for a drug susceptibility test, and US$18.18 for a line-probe assay. Top-down mean unit cost estimates were higher for all tests, and for sputum smear microscopy and Xpert these increased to respectively US$1.51 and US$13.58. The difference between bottom-up and top-down estimates was greatest for tests performed at the reference laboratories. C O N C L U S I O N : These unit costs for TB diagnostics can be used to estimate resource requirements and costeffectiveness in India, taking into account geographical location, laboratory type and capacity utilisation. K E Y W O R D S : tuberculosis; costs; bottom-up; topdown; diagnostic tests\n\nIMPROVED DIAGNOSTIC TESTING for tuberculosis (TB) and drug-resistant TB is one of the pillars of the global strategy to combat TB.1 New tests have started to become available and are currently being implemented.2 The Xpertw MTB/RIF assay (Cepheid, Sunnyvale, CA, USA) was endorsed by the World Health Organization as a cost-effective test for the detection of TB and rifampicin (RMP) resistance.1\nThe scale-up of new diagnostic tests and strategies for TB and drug-resistant TB have important economic implications.3 Economic analyses to determine the cost-effectiveness, affordability or budgetimpact implications of scaling up new diagnostic tests require reliable cost estimates of diagnostic tests and procedures.\nTo assess the feasibility of the Xpert assay roll-out in India, an implementation study was conducted to investigate the effect on case detection under programmatic conditions and assess the cost of the roll-out.4,5 For an economic analysis of the poten-\n\ntial scale-up of Xpert testing in India, unit cost estimates of Xpert, sputum smear microscopy (SSM), as well as of other commonly used tests to diagnose drug-resistant TB in India, were required. Earlier TB diagnostic costing studies in India were from limited settings and conducted during the early stages of Xpert demonstration studies.6 Costs may differ at scale and in routine settings.7\nWe present the unit costs of the Xpert assay, SSM and other commonly used drug resistance tests in India across a range of geographic and laboratory settings. For scale-up purposes, costs estimated by bottom-up and top-down methods may be equally important.7 We therefore estimated costs with the bottom-up or ingredient approach and with the topdown method, and we present a full capacity costing of Xpert. These estimates, under different conditions of service volume, inform scale-up decisions of Xpert across different sites in India.\n\nCorrespondence to: Neeraj Raizada, Foundation for Innovative New Diagnostics (FIND), New Delhi, Flat No 6\u201314, 9th floor, Vijaya Building, 17 Barakhamba Road, New Delhi 110001, India. e-mail: Neeraj.raizada@finddx.org\nArticle submitted 8 July 2016. Final version accepted 28 December 2016.\n\n376 The International Journal of Tuberculosis and Lung Disease\n\nTable 1 Characteristics of the laboratories included in the study\n\nDMC #1\n\nDMC #2\n\nDMC #3\n\nDMC #4\n\nDMC #5\n\nIRL #1\n\nIRL #2\n\nState\nUrban or rural Type of laboratory Year of assessment Population covered\nby each laboratory Persons requiring\ndiagnostic testing for TB in 2013, n Diagnostic tests available\n\nDelhi\nUrban DMC 2013\u20132014\n4 500 000\n4 466 SSM, Xpert\n\nAssam\nUrban DMC 2013\u20132014\n7 804 100\n5 803 SSM, Xpert\n\nAssam\nRural DMC 2013\u20132014\n3 341 550\n1 160 SSM, Xpert\n\nAndhra Pradesh\nRural DMC 2013\u20132014\n5 956 170\n6 232 SSM, Xpert\n\nAndhra Pradesh\nRural DMC 2013\u20132014\n5 375 670\n4 289 SSM, Xpert\n\nAssam\nUrban IRL 2013\u20132014\nNA\n3 775 SSM, culture\n(solid/liquid), DST, LPA, MTB ID\n\nDelhi\nUrban IRL 2013\u20132014\nNA\n10 923 SSM, culture\n(solid/liquid), DST, LPA, MTB ID\n\nDMC \u00bc designated microscopy centre; IRL \u00bc intermediate reference laboratory; NA \u00bc not available; SSM \u00bc sputum smear microscopy; DST \u00bc drug susceptibility testing; LPA \u00bc line-probe assay; MTB ID \u00bc M. tuberculosis rapid speciation.\n\nSTUDY POPULATION AND METHODS\nSetting and selection\nWe selected laboratories among the 18 sites that were part of the Xpert implementation study in India, located in 13 of the 29 states.5 Of the 18 sites, 13 were part of a hospital and five were located at primary health care facilities; 6 were located in urban settings, 7 in rural settings and 5 in tribal and hilly areas. To obtain representative costing estimates for TB diagnostic tests, we purposely selected five designated microscopy centres, 4 of which were located in hospitals and 1 in a primary health care facility. Besides variations in the physical location, we considered variations in workload, urban vs. rural settings, geographic distribution and proximity to intermediate reference laboratories (IRLs). The five microscopy centres selected represented an average workload of between 89 and 476 patients requiring TB testing each month. We selected 1 designated microscopy centre (DMC) in the north of India, 2 DMCs in the south and 2 in the east. Two microscopy centres were in urban and three in rural settings. To observe drug resistance tests, we selected two reference laboratories in close proximity to three of the five microscopy centres, in the north and the east, respectively. Table 1 provides an overview of the laboratory characteristics.\nEthics approval was obtained from the Institution Ethics Committee of the National Tuberculosis Institute, Bangalore, India. Approval for the study was granted by the Central TB Division, Ministry of Health and Family Welfare, Government of India, New Delhi, India.\nAt the DMCs, we observed the Xpert assay and Ziehl-Neelsen and auramine-stained SSM. At the IRLs, we observed solid culture on Lo\u00a8 wensteinJensen (LJ) medium, liquid culture and drug susceptibility testing (DST) with the BACTECe MGITe (Mycobacterium Growth Indicator Tube; BD, Sparks, MD, USA) system, the GenoTypew\n\nMTBDRplus (Hain Lifescience, Nehren, Germany) line-probe assay (LPA) v. 2 and M. tuberculosis rapid speciation (MTB ID). The MGIT DST was observed for first-line drugs (i.e., RMP, ethambutol, isoniazid and streptomycin). Specimen decontamination was costed as a separate procedure. The number of observations as well as the specimen being observed was determined by the workload on the day of the observation and the days available at each laboratory. One investigator (SR) used a stopwatch to time all the activities required to perform each laboratory procedure. Observations were made in December 2013 and January 2014, when the implementation study had been running for at least 1.5 years at the selected sites.\nBottom-up vs. top-down\nOur primary estimate was the cost of performing a single test, i.e., the unit cost calculated using the ingredient approach, where the measured resource use of the input is multiplied by the cost of the input. This estimate only includes directly observed resources, which is different from a top-down estimate, which takes into account total laboratory expenditures and allocates them to specific tests. The top-down method better captures underutilised resources, as it allocates total cost among all activities. For example, while in the bottom-up method the cost of the GeneXpert machine would be estimated based on inputs used to conduct one test, in the top-down method the total cost of the machine is divided by the number of tests. If the GeneXpert machine is underutilised, top-down estimates would be higher than bottom-up estimates.\nBottom-up\nFor each test, we measured resources used for the preparation, inoculation and incubation of the specimen, if applicable, and resources used for\n\nTB diagnostic tests costs India 377\n\nreporting the results. We collected inputs and their resources for the following categories: building, staff, equipment, reagents and chemicals, consumables and overheads. The unit of observation used for building, staff and equipment was time, for reagents and chemicals it was the volume used and for consumables it was the number of units used. We collected general information about the site, including operational minutes and building surface, using interviews and measurements. Non-observable overhead costs, such as building costs, administrative expenses and expenses incurred for utilities, were first determined through an examination of the financial records, and then through staff interviews and laboratory records. As is common in costing exercises, overhead resources were allocated to tests based on the amount of staff time or building space being used to perform the test, as measured during the observation.8,9 The prices of the equipment, drugs and medical supplies were provided by the procurement department of the Foundation for Innovative New Diagnostics India Office (New Delhi, India). Where information was not available, we consulted the United Nations Children\u2019s Fund Supply Catalogue10 and the Loba Chemie price list.11 Costs related to buildings and equipment were annualised over their expected lifetime using a standard discount rate of 3%.12 For buildings, an expected lifetime of 30 years was used; for equipment it was set between 2 and 15 years and for furniture we used 5 years.\nAll unit costs were calculated in a purpose-built Excel spreadsheet (MicroSoft, Redmond, WA, USA) that was developed for TB diagnostic test costing. All locally collected costs were in Indian rupees (INR) and are converted to US dollars (USD) using the exchange rate for March 2014.13\nTop-down\nIn the top-down method, we used expenditure data and collected figures on the number of Xpert assays and SSM tests performed annually at each microscopy centre. The top-down estimate was calculated by dividing total annual expenditures on salaries and the annual cost of buildings and equipment allocated specifically to each test by the numbers of tests performed annually. We first divided the total cost of the building space, the testing equipment (on the basis of observation) and annual staff salaries by the annual workload. We then allocated the total overhead costs per laboratory based on the building space that was used for either SSM or Xpert testing, and then also divided by the annual workload. For reagents and chemicals, and consumables, as annual expenditure data were unavailable, we used the costs as estimated in the bottom-up method.\n\nScenario and sensitivity analyses\nTo assess the effect of capacity underutilisation on unit cost for all tests, we compared bottom-up and top-down estimates. As GeneXpert machines can simultaneously run 1\u20134 tests, we also assessed whether underutilisation of the machine affected the bottom-up unit costs. We changed all fixed resources (i.e., building space, equipment, staff and overheads) as if used at full machine capacity of 8 (two 4-slot machines), instead of the actual observed capacity.\nFor solid culture, we investigated whether or not the number of specimens included in an observation session affected the unit cost. We simulated an increase in the number of specimens included in the first observation from 4 (observed) to 26 (number of specimens included in the second observation). Resource use increased proportionally.\nWe validated the annual maintenance cost for Xpert by comparing the value obtained from standard formulae that were applied to all tests and assume an expected lifetime, with annual maintenance costs as observed in the programme.\nCost of the full diagnostic process for drug resistance\nFinally, we calculated the cost of the diagnosis of drug resistance, which requires a diagnostic algorithm, by summing the bottom-up estimates of all procedures and tests included in the algorithm.\nRESULTS\nA total of 32 observations were made to estimate resource use (Table 2). Using the bottom-up method, the mean unit cost for SSM was US$0.83, ranging across observations and sites from US$0.60 to US$1.10. Using the top-down method, the mean unit cost for SSM was higher, US$1.51 (range US$0.63\u2013 US$1.84). The mean unit cost for the Xpert assay was US$12.29 (range US$11.61\u2013US$12.89) using the bottom-up method and increased to US$13.58 (range US$11.98\u2013US$16.43) when applying the top-down method. The mean unit cost per test at reference laboratories was US$1.69 for the decontamination procedure, US$9.83 for a solid culture test, US$11.06 for a liquid culture test, US$29.88 for DST, US$18.18 for an LPA and US$3.95 for an MTB ID test with the bottom-up method; these also increased for all tests when using the top-down method.\nWhen the GeneXpert machines were used at full capacity, the bottom-up unit cost fell by 6% to US$11.58. The bottom-up cost estimate of a solid culture varied between US$2.29 and US$17.83, depending on the number of specimens included in the observation session. Simulating the inclusion of a larger number of specimens reduced the unit cost from US$17.83 to US$3.39. The decrease was\n\n378 The International Journal of Tuberculosis and Lung Disease\n\nTable 2 Number of annual tests, observations per test, specimen per test and unit costs per test*\n\nTest SSM: ZN SSM: auramine Xpert\nDecontamination Solid culture Liquid culture DST LPA MTB ID\u2021\n\nSite(s) All 5 DMCs All 5 DMCs\n\u00fe IRL #2 All 5 DMCs\nBoth IRLs Both IRLs Both IRLs IRL #2 Both IRLs Both IRLs\n\nMethod\nTop-down Bottom-up Top-down Bottom-up Top-down Bottom-up Full capacity Top-down Bottom-up Top-down Bottom-up Top-down Bottom-up Top-down Bottom-up Top-down Bottom-up Top-down Bottom-up\n\nMean annual number of tests\n(min\u2013max)\n8 780 (2 320 \u201312 464)\u2020 NA\n15 389 (4 811\u201321 846) NA\n4 336 (1 384\u20135 966) NA NA\n11 805 (6 330\u201317 280) NA\n4 453 (2 661\u20136 245) NA\n3 747 (1 618\u20135 875) NA 72 NA\n3 565 (2 042\u20135 088) NA\nUnobtainable3 NA\n\nNumber of observations\nNA2 7 NA 3 NA 6 6 NA 4 NA 3 NA 2 NA 2 NA 2 Unobtainable 2\n\nMean number of specimen per observations (min\u2013max)\nNA 11.43 (6\u201324)\nNA 14 (7\u201325)\nNA 4.83 (2\u20137)\nNA NA 15 (9\u201320) NA 12.67 (4\u201326) NA 17 (16\u201318) NA 3 (2\u20134) NA 17 (14\u201320) Unobtainable 1 (1\u20131)\n\nMean unit cost (min\u2013max)\n1.44 (0.63\u20131.84) 0.85 (0.6\u20131.1) 1.89 (1.75\u20132.13) 0.84 (0.67\u20131.01) 13.58 (11.97\u201316.43) 12.29 (11.61\u201312.89) 11.58 (10.7\u201312.72) 3.38 (2.97\u20134.45) 1.69 (0.98\u20132.3) 10.94 (7.35\u201313.16) 9.83 (2.29\u201317.83) 15.95 (10.11\u201321.79) 11.06 (9.01\u201313.12) 546.76 (546.22\u2013547.3) 29.88 (26.34\u201333.42) 33.75 (28.13\u201339.37) 18.18 (17.09\u201319.27)\n\u2014 3.95 (3.74\u20134.16)\n\n* Only data on total microscopy tests were recorded; no distinction was made between ZN and auramine staining. \u2020 Except for \u2018reagents and chemicals and consumables\u2019, for which we used the bottom-up method to calculate costs. \u2021 We were not able to estimate the unit cost with the top-down method as figures on total annual number of tests performed were not known at the time of data\ncollection.\nSSM \u00bc sputum smear microscopy; ZN \u00bc Ziehl-Neelsen; DMC \u00bc designated microscopy centre; NA \u00bc not applicable; IRL \u00bc intermediate reference laboratory; DST \u00bc\ndrug susceptibility testing; LPA \u00bc line-probe assay; MTB ID \u00bc M. tuberculosis rapid speciation.\n\nprimarily caused by a drop in building costs, of approximately US$10. In the top-down method, the same simulation reduced unit costs by 2%.\nSSM reagents and chemicals comprised on average 36% of the mean bottom-up unit cost, followed by consumables (30%) and overheads (21%) (Table 3). The composition of the unit cost changed when the top-down method was used, and\n\noverhead costs comprised on average 40% of the unit cost, while reagents and chemicals comprised 20%. Regardless of the method used, the cost per Xpert assay was primarily determined by the cost of the cartridge (US$9.98), included in the component reagents and chemicals, and comprised on average 85% of the total unit costs (Table 4). The cost categories for DST and other TB tests and proce-\n\nTable 3 Mean costs per sputum smear microscopy test in designated microscopy centres (in 2014 USD)\n\nNumber of specimens observed\n\nSite #1\n\nSite #2 Site #3\n\nSite #4\n\nSite #5\n\nProportion\n\nAuramine Auramine ZN ZN ZN\n\nZN\n\nZN ZN\n\nZN Mean cost across of total\n\n(n \u00bc 7) (n \u00bc 10) (n \u00bc 6) (n \u00bc 6) (n \u00bc 6) (n \u00bc 24) (n \u00bc 7) (n \u00bc 8) (n \u00bc 23) observations (range) %\n\nBottom-up method Overhead Building space Equipment Staff Reagents and chemicals Consumables\nTotal\nTop-down method Overhead Building space Equipment Staff Reagents and chemicals Consumables\nTotal\n\n0.288 0.098 0.002 0.053 0.086\n0.327 0.85\n0.951 0.138 0.027 0.261 0.086\n0.327 1.79\n\n0.211 0.041 0.002 0.042 0.086\n0.285\n0.67\n\n0.253 0.044 0.003 0.05 0.235\n\n0.408 0.025 0.011 0.036 0.135\n\n0.13 0.003 0.003 0.019 0.285\n\n0.088 0.035 0.004 0.108 0.496\n\n0.088 0.039 0.003 0.104 0.490\n\n0.083 0.034 0.003 0.101 0.571\n\n0.03 0.008 0.004 0.059 0.309\n\n0.18 (0.03\u20130.408) 0.04 (0.003\u20130.098) 0.004 (0.002\u20130.011) 0.06 (0.019\u20130.108)\n0.3 (0.086\u20130.571)\n\n0.215 0.088 0.327 0.192 0.321 0.303 0.185 0.25 (0.088\u20130.327)\n\n0.8 0.7 0.77 0.92 1.04 1.1 0.6\n\n0.83 (0.6\u20131.1)\n\n21.2 4.4 0.5 7.7\n36.1\n30.1\n100\n\n0.951 0.138 0.027 0.261 0.086\n0.285\n1.75\n\n0.951 0.138 0.027 0.261 0.235\n\n0.278 0.055 0.027 0.05 0.135\n\n0.83 0.015 0.128 0.251 0.285\n\n0.435 0.068 0.019 0.281 0.496\n\n0.435 0.068 0.019 0.281 0.49\n\n0.435 0.068 0.019 0.281 0.571\n\n0.153 0.067 0.035 0.238 0.309\n\n0.215 0.088 0.327 0.192 0.321 0.303 0.185 1.83 0.63 1.84 1.49 1.61 1.68 0.99\n\n0.6 (0.153\u20130.951) 0.08 (0.015\u20130.138) 0.04 (0.019\u20130.128) 0.24 (0.05\u20130.281)\n0.3 (0.086\u20130.571)\n0.25 (0.088\u20130.327)\n1.51 (0.63\u20131.84)\n\n39.8 5.5 2.4\n15.9 19.8\n16.5\n100\n\nZN \u00bc Ziehl-Neelsen.\n\nTB diagnostic tests costs India 379\n\nTable 4 Mean costs per XpertW MTB/RIF test in designated microscopy centres for all three methods (in 2014 USD)\n\nNumber of specimens observed\nIngredient method Overhead Building space Equipment Staff Reagents and chemicals Consumables\nTotal\nBottom-up method Overhead Building space Equipment Staff Reagents and chemicals Consumables\nTotal\nFull capacity method Overhead Building space Equipment Staff Reagents and chemicals Consumables\nTotal\n\nSite #1 (n \u00bc 7)\n0.508 0.036 1.029 0.085 9.980 0.273 11.911\n2.017 0.440 0.950 0.485 9.980 0.273 14.145\n0.254 0.018 0.515 0.042 9.980 0.273 11.082\n\nSite #2 (n \u00bc 10)\n0.540 0.021 1.896 0.045 9.980 0.096 12.578\n0.836 0.153 0.807 0.103 9.980 0.096 11.975\n0.135 0.005 0.474 0.011 9.980 0.096 10.701\n\nSite #3 (n \u00bc 6)\n0.172 0.002 1.254 0.019 9.980 0.464 11.891\n2.074 0.063 3.424 0.421 9.980 0.464 16.426\n0.064 0.001 0.470 0.007 9.980 0.464 10.987\n\nSite #4 (n \u00bc 6)\n\n0.162 0.071 1.060 0.042 11.368 0.169\n12.872\n\n0.160 0.073 1.077 0.040 11.368 0.169\n12.887\n\n0.379 0.181 0.766 0.391 11.368 0.169\n13.254\n\n0.379 0.181 0.766 0.391 11.368 0.169\n13.254\n\n0.142 0.062 0.927 0.037 11.368 0.169\n12.704\n\n0.140 0.064 0.942 0.035 11.368 0.169\n12.717\n\nSite #5 (n \u00bc 24)\n0.036 0.005 1.241 0.044 9.980 0.307 11.613\n0.270 0.118 1.179 0.602 9.980 0.307 12.456\n0.027 0.004 0.931 0.033 9.980 0.307 11.281\n\nMean cost across observations (range)\n0.26 (0.036\u20130.54) 0.03 (0.002\u20130.073) 1.26 (1.029\u20131.896) 0.05 (0.019\u20130.085) 10.44 (9.98\u201311.368) 0.25 (0.096\u20130.464) 12.29 (11.61\u201312.89)\n0.99 (0.27\u20132.074) 0.19 (0.063\u20130.44) 1.32 (0.766\u20133.424)\n0.4 (0.103\u20130.602) 10.44 (9.98\u201311.368)\n0.25 (0.096\u20130.464) 13.58 (11.97\u201316.43)\n0.13 (0.027\u20130.254) 0.03 (0.001\u20130.064) 0.71 (0.47\u20130.942) 0.03 (0.007\u20130.042) 10.44 (9.98\u201311.368) 0.25 (0.096\u20130.464) 11.58 (10.7\u201312.72)\n\nProportion of total %\n2.1 0.3 10.2 0.4 85 2 100\n7.3 1.4 9.7 2.9 76.9 1.8 100\n1.1 0.22 6.13 0.24 90.19 2.13 100\n\ndures performed at the IRLs are presented in the Appendix.1\nSubstituting the annual maintenance costs of the GeneXpert machine calculated using the formulae with observed annual maintenance costs minimally increased the average unit costs for the Xpert assay by respectively 1.7% and 1.6% in the bottom-up and top-down methods.\nThe average cost of obtaining a DST profile on a sample would be equal to US$42.65, if we add the unit cost of decontamination, liquid culture and DST. The cost of a full procedure to perform an LPA was US$20.89, if the costs of decontamination, direct SSM and LPA are combined.\nDISCUSSION\nOn comparing our findings for the SSM and Xpert assay with previous results from the demonstration study in India from 2010,6 we found that the cost of Xpert was lower, primarily due to the reduced price for Xpert reagents.14 In addition, cost estimates for equipment and staff cost components were lower in our study, possibly due to economies of scale, as we costed larger sites. Our unit costs for liquid culture, DST and Xpert were similar to those found in a South African study.15 Unit costs for SSM were considerably lower than found in South Africa, explained by the\n1 The appendix is available in the online version of this article, at h t t p : / / w w w. i n g e n t a c o n n e c t . c o m / c o n t e n t / i u a t l d / i j t l d / 2 0 1 7 / 00000021/00000004/art00005\n\nhigher labour costs there, which comprise a large percentage of SSM unit cost. We also found lower costs for Xpert than those reported in South Africa, due to differences in prices of non-tradable goods (such as salaries) and the fact that we were estimating costs at full implementation rather than during scale-up.7\nBottom-up unit cost estimates for SSM and Xpert varied between sites, mostly due to differences in the use of reagents, chemicals and consumables. Fixed costs, such as building costs, did not vary substantially between settings. Variations in overhead costs were primarily related to costs of management staff. Variations in overheads may also have been due to the extent of staff time, as this increases the amount of total overhead costs that are then allocated to SSM. Differences in the cost of consumables by site were primarily explained by some staff not wearing laboratory coats, which is not advised.\nFor solid and liquid culture, and DST, building costs comprised a large proportion of unit costs. The need for a long incubation period in laboratories with advanced biosafety requirements increased the costs of utilisation of laboratory space and time in reference laboratories compared to microscopy centres. Both bottom-up unit costs and top-down costs were sensitive to the number of specimens tested per batch. In the top-down method, a change in the number of specimens per batch affected the unit cost, as some of the costs were allocated by dividing the estimated annual building cost by the annual number of tests performed. High costs for buildings also explained the cost difference between direct SSM\n\n380 The International Journal of Tuberculosis and Lung Disease\n\n(US$1.01) at the microscopy centres and SSM performed on liquid culture isolates at the reference laboratories (US$7.70). For DST, top-down cost estimates were much higher than the bottom-up approach because the use of DST (n \u00bc 72) was low compared to the other tests, suggesting a high level of excess capacity in DST resourcing.\nNot surprisingly, we generally found top-down costs to be higher than bottom-up costs. There was one exception, likely due to measurement bias by the staff when using a bottom-up method. In this case, some staff may take longer to perform a test than usual. The top-down method increased the mean unit cost across sites for SSM by 80% and for the Xpert assay by 11% compared to the bottom-up method. This suggests underutilisation of current diagnostic capacity. The smaller proportional difference for the Xpert assay is because a higher proportion of the cost of Xpert is due to items such as consumables, where \u2018excess capacity\u2019 is unlikely to occur. This difference in top-down and bottom-up costs therefore suggests that further scale-up of interventions could lead to improvements in efficiency and cost reductions,7 as capacity in resources such as building and staff become better utilised. A scale-up may also enhance the knowledge of staff, which could lead to improved efficiency in processes, and possibly task shifting to more junior staff over time, which may also further reduce costs.\nIn conclusion, we obtained unit costs for TB diagnostics across different settings in India, taking into account geographical location, laboratory type and capacity utilisation. These costs suggest room for further efficiencies and provide an essential base for estimating resource requirements and the costeffectiveness of Xpert roll-out in India.\nAcknowledgements\nThe authors would like to thank the site coordinators of the designated microscopy centres and the persons in charge of the intermediate reference laboratories for their assistance and cooperativeness in collecting the data. We would also like to thank all the friendly and helpful laboratory staff for letting us observe their work.\nThe study was funded by the Foundation for Innovative New Diagnostics (FIND; Geneva, Switzerland) and by the Amsterdam Institute for Global Health and Development (Amsterdam, The Netherlands)\nCompeting interests: The authors declare that NR and CB are employed by FIND, a non-profit organisation that collaborates with industry partners, including Cepheid Inc, in the development and evaluation of new diagnostic tests. These partners with whom we have partnered have in no way contributed to the study and would not be benefited by the results of the study. As such, there is no conflict of interest to the publication of this article. No other conflicts declared.\nSpreadsheet: The Excel spreadsheet used to calculate the unit costs is available upon request from the corresponding author: s. rupert@aighd.org.\n\nReferences\n1 World Health Organization Global TB Programme. Automated real-time nucleic acid amplification technology for rapid and simultaneous detection of tuberculosis and rifampicin resistance: Xpert MTB/RIF assay for the diagnosis of pulmonary and extrapulmonary TB in adults and children. WHO/HTM/TB/2013.16. Geneva, Switzerland: WHO, 2014.\n2 UNITAID. Tuberculosis diagnostics technology and market landscape. Report No. 3. Geneva, Switzerland: UNITAID, 2014: p 97. http://www.unitaid.eu/images/marketdynamics/ publications/UNITAID_TB_Diagnostics_Landscape_3rdedition.pdf Accessed January 2017.\n3 Pantoja A, Fitzpatrick C, Vassall A, Weyer K, Floyd K. Xpert MTB/RIF for diagnosis of tuberculosis and drug-resistant tuberculosis: a cost and affordability analysis. Eur Respir J 2013; 42: 708\u2013720.\n4 Raizada N, Sachdeva K S, Sreenivas A, et al. Feasibility of decentralised deployment of Xpert MTB/RIF test at lower level of health system in India. PLOS ONE 2014; 9: e89301.\n5 Sachdeva K S, Raizada N, Sreenivas A, et al. Use of Xpert MTB/ RIF in decentralized public health settings and its effect on pulmonary TB and DR-TB case finding in India. PLOS ONE 2015; 10: e0126065.\n6 Vassall A, van Kampen S, Sohn H, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: a cost-effectiveness analysis. PLOS MED 2011; 8: e1001120.\n7 Cunnama L, Sinanovic E, Ramma L, et al. Using top-down and bottom-up costing approaches in LMICs: the case for using both to assess the incremental costs of new technologies at scale: top-down and bottom-up costing approaches in LMICs. Health Econ 2016; 25 (Suppl 1): 53\u201366.\n8 World Health Organization. Cost analysis in primary health care: a training manual for programme managers. Geneva, Switzerland: WHO, 1994.\n9 UNAIDS. Manual for costing HIV facilities and services. Geneva, Switzerland: UNAIDS, 2011. http: //www.unaids.org/ sites/default/files/en/media/unaids/contentassets/documents/ document/2011/20110523_manual_costing_HIV_facilities_en. pdf Accessed January 2017.\n10 United Nations Children\u2019s Fund. The UNICEF Supply Catalogue. New York, NY, USA: UNICEF, 2014. https:// supply.unicef.org Accessed January 2017.\n11 Loba Chemie. Loba Chemie Pricelist. Mumbai, India: Loba Chemie, 2014 http://www.lobachemie.com/contactus/requestcatalogue.aspx Accessed January 2017.\n12 World Health Organization. Making choices in health: WHO guide to cost-effectiveness analysis. Geneva, Switzerland: WHO, 2003.\n13 International Monetary Fund. Exchange rate archives by month. Washington DC, USA: IMF, 2014. https: //www.imf. org/external/np/fin/data/param_rms_mth.aspx Accessed January 2017.\n14 Foundation for Innovative New Diagnostics. Price for Xpertw MTB/RIF and FIND country list. Geneva, Switzerland: FIND, 2015. http://www.finddiagnostics.org/about/what_we_do/ successes/find-negotiated-prices/xpert_mtb_rif.html Accessed January 2017.\n15 Shah M, Chihota V, Coetzee G, Churchyard G, Dorman S E. Comparison of laboratory costs of rapid molecular tests and conventional diagnostics for detection of tuberculosis and drugresistant tuberculosis in South Africa. BMC Infect Dis 2013; 13: 352.\n\nAPPENDIX\n\nTB diagnostic tests costs India i\n\nTable A.1 Cost per test in IRL#1 (in 2014 USD)\n\nNumber of specimens observed\n\nDecontamination (n \u00bc 9)\n\nCulture: solid (n \u00bc 4)\n\nCulture: solid (n \u00bc 8)\n\nCulture: liquid (n \u00bc 16)\n\nBottom-up method Overhead Building space Equipment Staff Reagents and chemicals Consumables\nTotal\nTop-down method Overhead Building space Equipment Staff Reagents and chemicals Consumables\nTotal\n\n0.201 0.506 0.219 0.105 0.662 0.603\n2.296\n0.952 0.970 0.802 0.460 0.662 0.603\n4.449\n\n1.658 11.925\n3.371 0.095 0.446 0.337\n17.832\n4.858 3.654 1.907 1.962 0.446 0.337\n13.165\n\n0.890 6.004 1.701 0.109 0.456 0.225\n9.385\n4.858 3.654 1.156 1.962 0.456 0.225\n12.311\n\n0.478 9.064 0.457 0.067 0.433 2.616\n13.115\n4.858 3.796 6.859 3.227 0.433 2.616\n21.789\n\nIRL \u00bc intermediate reference laboratory; LPA \u00bc line-probe assay; MTB ID \u00bc M. tuberculosis rapid speciation; NA \u00bc not applicable.\n\nLPA (n \u00bc 14)\n0.853 0.042 1.274 1.150 13.382 2.567 19.268\n4.123 5.889 4.481 8.931 13.382 2.567 39.374\n\nMTB ID (n \u00bc 1)\n0.101 1.090 0.528 0.055 1.400 0.564 3.738\nNA NA NA NA \u2014 \u2014 NA\n\nTable A.2 Cost per test in IRL #2 (in 2014 USD)\n\nNumber of specimens observed\n\nDecontamination Observation 1 (n \u00bc 20)\n\nDecontamination Observation 2 (n \u00bc 18)\n\nDecontamination Observation 3 (n \u00bc 13)\n\nCulture: solid\n(n \u00bc 26)\n\nCulture: liquid\n(n \u00bc 18)\n\nMicroscopy (ZN) Liquid culture (n \u00bc 6)\n\nMicroscopy (auramine) Direct smear (n \u00bc 25)\n\nDST Observation 1\n(n \u00bc 4)\n\nBottom-up method Overhead Building space Equipment Staff Reagents and chemicals Consumables\nTotal\nTop-down method Overhead Building space Equipment Staff Reagents and chemicals Consumables\nTotal\n\n0.190 0.395 0.100 0.031 0.814\n0.249 1.779\n0.821 0.654 0.344 0.124 0.814\n0.249 3.006\n\n0.202 0.366 0.095 0.074 0.736\n0.255 1.728\n0.821 0.654 0.344 0.281 0.736\n0.255 3.091\n\n0.086 0.232 0.074 0.050 0.376\n0.158 0.976\n0.821 0.749 0.417 0.450 0.376\n0.158 2.970\n\n0.117 1.191 0.385 0.075 0.445\n0.072 2.285\n1.043 4.067 0.480 1.244 0.445\n0.072 7.351\n\n0.182 5.368 0.290 0.105 0.449\n2.620 9.014\n1.108 1.923 1.861 2.149 0.449\n2.620 10.112\n\n0.847 2.533 0.416 0.420 0.632\n2.855 7.703\nNA NA NA \u2014\n\u2014 NA\n\n0.203 0.361 0.001 0.074 0.141\n0.228 1.008\n0.715 0.581 0.011 0.454 0.141\n0.228 2.129\n\n1.148 5.480 0.477 0.667 1.910\n16.655 26.337\n108.526 156.948 154.244 107.933\n1.910\n16.655 546.216\n\nIRL \u00bc Intermediate Reference Laboratory; ZN \u00bc Ziehl-Neelsen; DST \u00bc drug susceptibility testing; LPA \u00bc line-probe assay; MTB ID \u00bc M. tuberculosis rapid speciation; NA \u00bc not applicable.\n\nDST Observation 2\n(n \u00bc 2)\n1.289 10.943\n0.792 0.741 2.145\n17.506 33.416\n108.526 156.948 154.244 107.933\n2.145\n17.506 547.301\n\nLPA (n \u00bc 20)\n0.859 0.061 0.291 0.759 13.259\n1.863 17.092\n0.768 6.050 1.797 4.391 13.259\n1.863 28.129\n\nMTB ID (n \u00bc 1)\n0.306 1.552 0.386 0.243 1.400\n0.275 4.162\nNA e NA NA NA \u2014\n\u2014 NA\n\nii The International Journal of Tuberculosis and Lung Disease\n\nTB diagnostic tests costs India iii\n\nC O N T E X T E : Parmi les 18 sites qui ont participe\u00b4 a` une e\u00b4tude de mise en \u0153uvre du test Xpertw MTB/RIF en Inde, nous avons se\u00b4lectionne\u00b4 cinq centres de microscopie et deux laboratoires de re\u00b4fe\u00b4rence. O B J E C T I F : Obtenir les cou\u02c6 ts unitaires des tests de diagnostic de la tuberculose (TB) et de la TB pharmacore\u00b4 sistante. S C H E\u00b4 M A : Les laboratoires ont e\u00b4te\u00b4 se\u00b4lectionne\u00b4s de fac\u00b8on a` e\u02c6tre repre\u00b4sentatifs des variations re\u00b4gionales et des diffe\u00b4rents types de laboratoire. Des me\u00b4thodes ascendantes et descendantes ont e\u00b4te\u00b4 utilise\u00b4es afin de calculer les cou\u02c6 ts par unite\u00b4. R E\u00b4 S U LTAT S : Dans les centres de microscopie, les cou\u02c6 ts unitaires moyens ascendants ont e\u00b4te\u00b4 de US$0,83 (fourchette US$0,60\u2013US$1,10) et de US$12,29 (US$11,61\u2013US$12,89) pour le frottis de crachats et l\u2019Xpert, respectivement. Dans les laboratoires de\n\nRESUME\nre\u00b4fe\u00b4rence, les cou\u02c6 ts unitaires moyens ont e\u00b4te\u00b4 de US$1,69 pour la proce\u00b4dure de contamination, de US$9,83 pour la culture en milieu solide, de US$11,06 pour la culture en milieu liquide, de US$29,88 pour un test de pharmacosensibilite\u00b4 et de US$18,18 pour un test de sonde en ligne. Les estimations des cou\u02c6 ts unitaires moyens descendants ont e\u00b4te\u00b4 plus e\u00b4leve\u00b4es pour tous les tests et pour le frottis de crachats et l\u2019Xpert ont augmente\u00b4 a` US$1,51 et a` US$13,58, respectivement. Les diffe\u00b4rences entre les estimations ascendantes et descendantes ont e\u00b4te\u00b4 les plus importantes pour les tests re\u00b4alise\u00b4s dans les laboratoires de re\u00b4fe\u00b4rence. C O N C L U S I O N : Ces cou\u02c6 ts unitaires du diagnostic de la TB peuvent e\u02c6tre utilise\u00b4s pour estimer les besoins en ressources et la rentabilite\u00b4 en Inde, en tenant compte de la localisation ge\u00b4ographique, de type de laboratoire et de sa capacite\u00b4 d\u2019utilisation.\n\nM A R C O D E R E F E R E N C I A: De los 18 centros que participaron en un estudio de aplicacio\u00b4 n de la prueba Xpertw MTB/RIF en la India, se escogieron cinco centros de microscopia y dos laboratorios de referencia. O B J E T I V O: Conocer los costos unitarios de las pruebas diagno\u00b4 sticas de la tuberculosis (TB) y la TB farmacorresistente. M E\u00b4 T O D O: Se practico\u00b4 un muestreo intencional con el fin de captar las diferencias en las regiones y en los diversos tipos de laboratorios. La estimacio\u00b4 n de los costos unitarios se obtuvo mediante me\u00b4todos de ca\u00b4 lculo descendente y ascendente. R E S U LTA D O S: En los centros de microscopia, las estimaciones ascendentes revelaron un costo unitario promedio de la baciloscopia del esputo de US$0,83 (entre US$0,60 y US$1,10) y de la prueba Xpert de US$12,29 (entre US$11,61 y US$12,89). En los laboratorios de referencia el promedio de los costos\n\nRESUMEN\nunitarios fueron US$1,69 por procedimiento de descontaminacio\u00b4 n US$1,69; US$9,83 por cultivo en medio so\u00b4 lido; US$11,06 por cultivo en medio l\u00b4\u0131quido; US$29,88 por prueba de sensibilidad a los medicamentos; y US$18,18 por cada prueba molecular con sondas en l\u00b4\u0131nea. Las estimaciones descendentes revelaron un costo unitario promedio ma\u00b4 s alto para todas las pruebas, pues el costo de la baciloscopia del esputo aumento\u00b4 a US$1,51 y el costo de la prueba Xpert a US$13,58. Las diferencias entre las estimaciones ascendentes y descendentes fueron ma\u00b4 s grandes con las pruebas realizadas en los laboratorios de referencia. C O N C L U S I O\u00b4 N: Estos costos unitarios de las pruebas diagno\u00b4 sticas de la TB se pueden utilizar con el fin de calcular la necesidad de recursos y la relacio\u00b4 n de costoefectividad de las intervenciones en la India, teniendo en cuenta la ubicacio\u00b4 n geogra\u00b4 fica, el tipo de laboratorio y la utilizacio\u00b4 n de las capacidades existentes.\n\n\n",
    "authors": [
      "S. Rupert",
      "A. Vassall",
      "N. Raizada",
      "S. D. Khaparde",
      "C. Boehme",
      "V. S. Salhotra",
      "K. S. Sachdeva",
      "S. A. Nair",
      "A. H. Van'T Hoog"
    ],
    "doi": "10.5588/ijtld.16.0496",
    "year": null,
    "item_type": "journalArticle",
    "url": "http://www.ingentaconnect.com/content/10.5588/ijtld.16.0496"
  },
  {
    "key": "54U7ARSL",
    "title": "Yield, Efficiency, and Costs of Mass Screening Algorithms for Tuberculosis in Brazilian Prisons",
    "abstract": "Background.\u2003 Tuberculosis (TB) is a major cause of morbidity and mortality among incarcerated populations globally. We performed mass TB screening in 3 prisons and assessed yield, efficiency, and costs associated with various screening algorithms.\nMethods.\u2003 Between 2017 and 2018, inmates from 3 prisons in Brazil were screened for TB by symptom assessment, chest radiography, sputum testing by Xpert MTB/RIF fourth-generation assay, and culture. Chest radiographs were scored by an automated interpretation algorithm (Computer-Aided Detection for Tuberculosis [CAD4TB]) that was locally calibrated to establish a positivity threshold. Four diagnostic algorithms were evaluated. We assessed the yield (percentage of total cases found) and efficiency (prevalence among those screened) for each algorithm. We performed unit costing to estimate the costs of each screening or diagnostic test and calculated the cost per case detected for each algorithm.\nResults.\u2003 We screened 5387 prisoners, of whom 214 (3.9%) were diagnosed with TB. Compared to other screening strategies initiated with chest radiography or symptoms, the trial of all participants with a single Xpert MTB/RIF sputum test detected 74% of all TB cases at a cost of US$249 per case diagnosed. Performing Xpert MTB/RIF screening tests only on those with symptoms had a similar cost per case diagnosed (US$255) but missed 35% more cases (73 vs 54) as screening all inmates.\nConclusions.\u2003 In this prospective study in 3 prisons in a high TB burden country, we found that testing all inmates with sputum Xpert MTB/RIF was a sensitive approach, while remaining cost-efficient. These results support use of Xpert MTB/RIF for mass screening in TB-endemic prisons.",
    "full_text": "Clinical Infectious Diseases MAJOR ARTICLE\n\nDownloaded from https://academic.oup.com/cid/article/72/5/771/5736588 by guest on 20 August 2024\n\nYield, Efficiency, and Costs of Mass Screening Algorithms for Tuberculosis in Brazilian Prisons\nAndrea da Silva Santos,1, Roberto Dias de Oliveira,2 Everton Ferreira Lemos,2 Fabiano Lima,2 Ted Cohen,3 Olivia Cords,4 Leonardo Martinez,4, Crhistinne Gonc\u0327alves,2 Albert I. Ko,3 Jason R. Andrews,4,a and Julio Croda2,3,5,a\n1Faculty of Health Sciences, Federal University of Grande Dourados, Dourados, Brazil, 2School of Medicine, Federal University of Mato Grosso do Sul, Campo Grande, Brazil, 3Department of Epidemiology of Microbial Diseases, Yale University School of Public Health, New Haven, Connecticut, USA, 4Division of Infectious Diseases and Geographic Medicine, Stanford University School of Medicine, Stanford, California, USA, and 5Oswaldo Cruz Foundation Mato Grosso do Sul, Campo Grande, Brazil\n(See the Editorial Commentary by Woodman and Grandjean on pages 778\u20139.)\nBackground.\u2003 Tuberculosis (TB) is a major cause of morbidity and mortality among incarcerated populations globally. We performed mass TB screening in 3 prisons and assessed yield, efficiency, and costs associated with various screening algorithms.\nMethods.\u2003 Between 2017 and 2018, inmates from 3 prisons in Brazil were screened for TB by symptom assessment, chest radiography, sputum testing by Xpert MTB/RIF fourth-generation assay, and culture. Chest radiographs were scored by an automated interpretation algorithm (Computer-Aided Detection for Tuberculosis [CAD4TB]) that was locally calibrated to establish a positivity threshold. Four diagnostic algorithms were evaluated. We assessed the yield (percentage of total cases found) and efficiency (prevalence among those screened) for each algorithm. We performed unit costing to estimate the costs of each screening or diagnostic test and calculated the cost per case detected for each algorithm.\nResults.\u2003 We screened 5387 prisoners, of whom 214 (3.9%) were diagnosed with TB. Compared to other screening strategies initiated with chest radiography or symptoms, the trial of all participants with a single Xpert MTB/RIF sputum test detected 74% of all TB cases at a cost of US$249 per case diagnosed. Performing Xpert MTB/RIF screening tests only on those with symptoms had a similar cost per case diagnosed (US$255) but missed 35% more cases (73 vs 54) as screening all inmates.\nConclusions.\u2003 In this prospective study in 3 prisons in a high TB burden country, we found that testing all inmates with sputum Xpert MTB/RIF was a sensitive approach, while remaining cost-efficient. These results support use of Xpert MTB/RIF for mass screening in TB-endemic prisons.\nKeywords.\u2003 mass screening; tuberculosis; algorithms; prisons; cost-effectiveness.\n\nTuberculosis (TB) is the leading cause of death by an infectious disease worldwide [1] and as a response, the World Health Assembly set a goal to reduce the global TB incidence by 90% by 2035 [2]. Despite an elevated focus on TB and increased funding, the TB burden is declining by only 1%\u20132% per year globally. To reach global targets, complementary interventions are needed to supplement current TB control. Recently, there has been a push to target interventions to populations with a high TB burden to reduce disease incidence and transmission to the broader community [3, 4].\nPrisons frequently have a very high burden of TB [5]. A metaanalysis of 19 studies found that the incidence of TB in prisons was 23 times greater than the surrounding population [5]. This high incidence leads to markedly elevated transmission rates.\n\u2002\nReceived 24 September 2019; editorial decision 3 January 2020; accepted 13 February 2020; published online February 17, 2020.\naJ. R. A. and J. C. contributed equally to this work.\nCorrespondence: J. Croda, Funda\u00e7\u00e3o Oswaldo Cruz, Mato Grosso do Sul, Campo Grande, Mato Grosso do Sul, 79074\u2013460, Brazil (julio.croda@fiocruz.br). Clinical Infectious Diseases\u00ae\u2003\u20032021;72(5):771\u20137 \u00a9 The Author(s) 2020. Published by Oxford University Press for the Infectious Diseases Society of America. All rights reserved. For permissions, e-mail: journals.permissions@oup.com. DOI: 10.1093/cid/ciaa135\n\nFor example, 3 prisoner cohorts from Brazil, Colombia, and Iran have shown annual tuberculin conversion rates between 15% and 25% [6\u20138]. Effective case detection for TB in prisons is necessary to reduce ongoing transmission.\nDespite the high rates of TB in prisoners and the potential importance of this population in the overall epidemic [3, 5, 9], few studies have assessed the efficiency and costs of different approaches for screening for TB among incarcerated populations [7]. Studies reporting the yield of TB detected in prisons rarely compare distinct screening modalities or report their costs. Additionally, the use of sensitive molecular diagnostic tests, such as the Xpert MTB/RIF assay, for TB screening among prisoners has not been widely explored. Studies using mathematical modeling suggest that annual mass screening can reduce the incidence of TB in prisons [3, 10]. However, there are no specific guidelines on how screening should be performed. Due to costs and a lack of evidence on effective screening approaches in this population, few prisons in lowand middle-income countries perform systematic screening for TB. Our objective was to identify effective and efficient approaches to TB screening in prisons that could be implemented in low- and middle-income countries.\n\nMass Screening for TB in Prisons\u2002 \u2022\u2002 cid\u20022021:72\u2002(1 March)\u2002\u2022\u2002771\n\nDownloaded from https://academic.oup.com/cid/article/72/5/771/5736588 by guest on 20 August 2024\n\nMETHODS\nStudy Population\nThis study was carried out in 3 prisons in Mato Grosso do Sul, Brazil. Brazil\u2019s national prison population is > 700 000 individuals, the third largest globally [11]. Three prisons were included in this study: Penitenci\u00e1ria Estadual de Dourados (PED), Estabelecimento Penal Jair Ferreira de Carvalho (EPJFC), and Instituto Penal de Campo Grande (IPCG). These prisons exclusively incarcerate males \u226518 years old and were selected because they are the largest in the state and had the highest TB infection and disease rates in preliminary studies [7, 12].\nStudy Procedures\nAll prisoners were invited to participate, and those who accepted provided written informed consent. The study was approved by the Federal University of Grande Dourados, the National Committee on Research Ethics (number 2.195.047), and the Institutional Review Board of Stanford University (number 40285). Each participant was then interviewed using a standardized questionnaire to collect demographic and clinical information. We asked each participant about TB-related symptoms according to World Health Organization (WHO) guidelines [13, 14]. All participants were instructed to produce a sputum sample with a target volume of at least 2 mL. On this primary sample, the Xpert MTB/RIF assay (Cepheid, Sunnyvale; fourth genetation; hereinafter Xpert) was performed; the remainder of the sample was transported to the support laboratory for culture. A second sputum sample was collected on the following day for a second culture. Participants who were unable to produce sputum were coached by nursing staff; however, sputum induction was not performed, and many participants were unable to produce a sample. Participants without sufficient samples were included in the study.\nAll participants underwent posterior-anterior chest radiography (CXR). Chest radiographs were then evaluated with Computer-Aided Detection for Tuberculosis (CAD4TB) software version 5 [15]. CAD4TB assigns a quality assessment to a CXR, produces a heat map indicating areas with possible abnormalities, and designates a score between 1 and 100 related to the likelihood of radiological abnormalities suggestive of a TB diagnosis. CAD4TB was calibrated with training data from radiographic images of participants with (n = 80) and without (n = 200) microbiologically confirmed TB. Training data demonstrated high accuracy (area under the curve = 0.88), with a sensitivity and specificity >\u200980% using a CAD4TB score \u2265\u200960. Participants with a CAD4TB score \u2265\u200960 were clinically reevaluated; those who had been unable to produce sputum on the first occasion were given another opportunity and additional coaching to produce a sample for testing by Xpert assay. Those participants with negative results or who were still unable to produce sputum were assessed by a physician. All TB cases identified during screening were provided free treatment according to national guidelines [16].\n\nDerivation of Mass Diagnostic Screening Algorithms\nAll participants were prospectively and systematically screened as outlined above; we then retrospectively evaluated 4 hypothetical, intensive screening algorithms (Figure 1) consisting of more limited sets of diagnostics:\n\u2022\u2002 Strategy 1: Sputum testing by Xpert for all participants who could produce sputum at the moment of questionnaire, regardless of presence of symptoms.\n\u2022\u2002 Strategy 2: Sputum testing by Xpert only for those who reported any TB-related symptom of any duration and who could produce sputum at the moment of questionnaire.\n\u2022\u2002 Strategy 3: CXR with CAD4TB scoring for all participants. Those with CAD4TB score \u2265\u200960 underwent sputum testing by Xpert.\n\u2022\u2002 Strategy 4: Symptom screening, followed by sputum Xpert testing for participants who reported a TB-related symptom. Those without any TB symptoms underwent CXR with CAD4TB scoring. Sputum collection and Xpert testing were then offered to participants with a CAD4TB score \u2265\u200960.\nOutcome Definitions\nWe followed national Brazilian guidelines and WHO definitions for TB diagnosis. We defined a TB case as any individual with a positive sputum Xpert, sputum culture, or with a physician diagnosis based on clinical-epidemiological data and radiographic abnormalities. All participants with TB were administered a rapid human immunodeficiency virus (HIV) test and evaluated through a nursing and medical examination.\nAnalytical Approach and Cost Evaluation\nWe calculated the cost of each screening procedure: symptom screening interviews, Xpert, culture, and radiographic and clinical evaluations. These costs include equipment, maintenance, consumables, and personnel time. The Supplementary Table S1 has description of the full cost for each diagnostic procedure (Supplementary Appendix).\nIn our calculations, we assumed the equipment used during mass screening would remain useful for a period of 10 years and amortized the cost over this period. Personnel time was calculated based on the salary of staff members involved in each screening component, time devoted to each component of screening, and the number of individuals who could be screened during that unit time. We calculated average unit cost by dividing total cost of each diagnostic procedure by the total number of procedures during the study period. The cost of each Xpert cartridge was US$9.90.\nTo calculate the cost per case detected, we assumed the definition of fixed and variable costs. Fixed costs are costs that apply to the entire cohort, regardless of how many people are screened, such as purchasing and maintaining equipment and software. Variable costs are costs related to use, such as human\n\n772\u2002\u2022\u2002cid\u20022021:72\u2002(1 March)\u2002\u2022\u2002Santos et al\n\nDownloaded from https://academic.oup.com/cid/article/72/5/771/5736588 by guest on 20 August 2024\n\nFigure 1.\u2003 Outline of tuberculosis screening strategies assessed among prisoners in the study. Abbreviations: CAD4TB, Computer-Aided Detection for Tuberculosis; MTB, Mycobacterium tuberculosis; RIF, rifampicin.\n\nresources, inputs, and evaluation of each radiographic image in CAD4TB. The cost per case detected was calculated for each strategy by multiplying the average unit cost by the number of procedures performed in each strategy and then dividing by the cases detected in the strategy. The values in Brazilian reais were converted to US dollars using the quotation of 28 November 2018 (R$3.87 = US$1.00).\nRESULTS\nStudy Population\nBetween November 2017 and July 2018, we screened 5387 of 6054 eligible study participants (88.9%). These three prisons can hold up to 1610 prisoners, but currently hold 6054 incarcerated individuals in total and have high turnover rates (Supplementary Appendix). Reasons for not participating included lack of interest, lack of clothing to leave the cell, and fear of meeting members of rival groups. Participating inmates had a median age of 30.5 years (Table 1). More than half of participants were smokers (58.3%) and used some type of illicit drug in the past year (58.8%), and 70.3% were previously incarcerated. More than 71.4% of participants reported knowing a person diagnosed with TB, and 8.2% reported having prior TB. During the study period, a total of 214 participants were diagnosed with pulmonary TB, equating to a\n\nprevalence of 3973 per 100 000 participants (95% confidence interval, 3483\u20134528). Disaggregating by prison, we identified TB prevalence of 5567/100 000 (101/1814) in EPJFC, 3607/100 000 (82/2273) in PED, and 2384/100 000 (31/1300) in IPCG.\nOf the diagnosed cases on the initial visit, 172 (80.3%) were diagnosed by the Xpert assay and sputum culture. At the initial visit, sputum was obtained from 1467 inmates. Among these, Xpert was performed on almost all participants (1452 [98.9%]) and detected 160 TB cases (Figure 2). Culture was performed on 1385 participants and 12 additional cases were identified by culture. Sputum smear tests were performed on 1386 participants; among the 214 TB cases who had smear microscopy performed, 49 (22.8%) had a positive smear. All TB cases that tested smearpositive were positive by Xpert. Among 1295 participants who were culture and/or Xpert negative, 261 had a CAD4TB score \u2265\u200960. According to the study protocol, these participants were reevaluated for TB, 114 did Xpert, and 22 were diagnosed (11 by Xpert and 11 by clinical evaluation). Among participants who did not produce a sputum sample at the initial visit (n = 3920), 523 (13.3%) had a CAD4TB score \u2265\u200960. A second attempt was made to collect sputum among these participants, of which 155 (29.6%) were successful and 9 were Xpert positive. A further\nMass Screening for TB in Prisons\u2002 \u2022\u2002 cid\u20022021:72\u2002(1 March)\u2002\u2022\u2002773\n\nTable 1.\u2003 Sociodemographic Characteristics and Risk Factors for Tuberculosis Among Screened Inmates\n\nVariables\n\nTotal (N = 5387)\n\nTB Cases (n = 214)\n\nNo TB (n = 5173)\n\nP Value\n\nPrison unit PED EPJFC IPCG\nMedian age, y (IQR) Ethnicity\nMixed White Black Indigenous Asian < 8 y of schooling Current smoker Illicit drug use over the last year BCG vaccinated Previous TB Know someone with TB Report any WHO TB symptoms Report cough Previously incarcerated\n\n2272 (42.2) 1814 (33.7) 1300 (24.1) 30.5 (25\u201337)\n3312 (61.5) 1306 (24.2)\n617 (11.5) 144 (2.6)\n8 (0.1) 3540 (65.7) 3139 (58.3) 3172 (58.8) 4736 (87.9)\n482 (8.2) 3849 (71.4) 2127 (39.4) 1527 (28.3) 3786 (70.3)\n\n82 (38.3) 101 (47.2) 31 (14.5) 30 (25\u201337)\n136 (63.6) 49 (22.9) 23 (10.7) 6 (2.8) 0 (0.0)\n154 (72.0) 161 (75.2) 169 (78.9) 185 (86.4)\n56 (26.2) 181 (84.6) 174 (81.3) 151 (70.6) 167 (78.0)\n\n2191 (42.4) 1713 (33.1) 1269 (24.5)\n31 (25\u201337)\n3176 (61.3) 1257 (24.3)\n593 (11.4) 138 (2.6)\n8 (0.2) 3386 (65.4) 2978 (57.5) 3003 (58.0) 4551 (87.9)\n426 (8.2) 3668 (70.9) 1953 (37.7) 1376 (26.6) 3619 (70.0)\n\n.24 <\u2009.01 <\u2009.01 <\u2009.01\n.52 .64 .76 >\u2009.99 >\u2009.99 .04 <\u2009.01 <\u2009.01 .49 <\u2009.01 <\u2009.01 <\u2009.01 <\u2009.01 <\u2009.01\n\nData are presented as no. (%) unless otherwise indicated. Abbreviations: BCG, Bacillus Calmette-Gu\u00e9rin; EPJFC, Estabelecimento Penal Jair Ferreira de Carvalho; IPCG, Instituto Penal de Campo Grande; IQR, interquartile range; PED, Penitenci\u00e1ria Estadual de Dourados; TB, tuberculosis; WHO, World Health Organization.\n\nDownloaded from https://academic.oup.com/cid/article/72/5/771/5736588 by guest on 20 August 2024\n\n11 cases were clinically diagnosed after physician reevaluation. Among cases of active TB, 4 (1.9%) were HIV positive.\nAccuracy and Yield of Symptom- and Radiograph-based Screening\nIn the initial screening interview, 2127/5387 patients (39.5%) reported at least 1 WHO-defined TB symptom; the most common of the symptoms was cough (71.8%). Symptom screening alone had a sensitivity and specificity of 81.3% and 61.6%, respectively. If screening was initiated based on cough alone rather than a comprehensive symptom screen, 151 (70.6%) cases would be detected (Table 2). The sensitivity of CXR with a CAD4TB score \u226560 was 77.1% and specificity was 85.6%.\nThe prevalence of TB was very low (0.2%) among participants with no symptoms and a CAD4TB score <\u200960, and this group comprised just over half (51.6%) of all participants. Prevalence among participants with no TB symptoms but CAD4TB score \u2265\u200960, comprising 8.7% of the cohort, was 7.0%. Among the 22.0% of participants with cough and CAD4TB score <\u200960, prevalence was 3.0%. The highest-risk group consisted of participants with both cough and CAD4TB score \u2265\u200960, in whom prevalence was 34.0%; although only 6.3% of participants met both criteria, this accounted for 53.8% of all TB cases detected.\nCosts and Efficiency of Screening Strategies\nAmong diagnostic modalities used during mass screening, the highest cost per participant was for Xpert (US$19.20), followed by CXR with CAD4TB scoring (US$6.28), clinical evaluation (US$2.60), and symptom screening interviews (US$1.90)\n\n(Table 3). The costliest component of Xpert was consumables (54.8%). For CAD4TB score, CXR, human resources, equipment, and CAD4TB score analysis contributed almost equally to main costs (29.6%, 29.0%, and 29.8%, respectively).\nThe cost per case detected for all strategies ranged from US$249 to US$395 (Table 4). Strategy 1 (Xpert assay for all participants) resulted in the second highest yield, detecting 74% of all cases, at lowest cost (US$249) per case detected. Strategy 2 (Xpert for individuals with any symptom) had a low cost per case detected (US$255) but resulted in lower yield (65%) compared with strategies 1 and 4. Strategies 2 and 3 had lower yield (65% and 64%, respectively) than strategies 1 and 4. And strategies 3 and 4 had a higher cost per case diagnosed (US$370 and US$395, respectively).\nDISCUSSION\nTuberculosis is a major infectious disease problem within prisons worldwide. However, there is a dearth of evidence concerning how to effectively detect TB while controlling costs in these environments. As a result, screening modalities in prisons globally remain variable, with few high-TB-burden countries enacting systematic screening policies in correctional facilities. In 3 prisons in Brazil, we found a very high prevalence (3973 per 100 000) of TB through systematic screening of inmates. This prevalence is higher to the identified in other studies in prisons in Brazil and other countries [17\u201321].\nThe strategies 1 and 4 had similar yields. We found that systematic Xpert testing among all individuals able to produce\n\n774\u2002\u2022\u2002cid\u20022021:72\u2002(1 March)\u2002\u2022\u2002Santos et al\n\nDownloaded from https://academic.oup.com/cid/article/72/5/771/5736588 by guest on 20 August 2024\n\nFigure 2.\u2003 Tuberculosis (TB) cases overlap of screening strategies. Strategy 1: Xpert MTB/RIF assay for all prisoners. Strategy 2: Xpert test only for those who reported any TB symptom. Strategy 3: Chest radiography, those with ComputerAided Detection for Tuberculosis (CAD4TB) score \u2265\u200960 undergo Xpert test. Strategy 4: Symptom screening, followed by Xpert test, those without any TB symptoms undergo chest radiography with CAD4TB scoring followed by Xpert test. Abbreviations: MTB, Mycobacterium tuberculosis; RIF, rifampicin.\nsputum was effective at a modest cost per case diagnosed of US$249 dollars, and detected only 3 fewer cases compared to strategy 4, which detected the most due to radiographically diagnosed cases. Implementing a symptom screen to identify individuals for sputum testing had a similar efficiency at US$255 per case detected. Strategies involving CXR were most costly and did not increase the overall yield compared with sputum Xpert testing alone. We also conducted a sensitivity analysis using a US$5.00 value for Xpert MTB/RIF cartridges. Under this pricing assumption for Xpert cartridges, Strategy 1 remained the most effective and efficient strategy.\n\nWe believe this further supports our primary analysis (Supplementary Appendix). Together, these results suggest that testing all inmates able to produce sputum using Xpert may be an effective and affordable strategy in high-burden prisons.\nThere has been debate over the reliability of symptom screening to triage the use of diagnostics in high-risk populations, due to limitations in both sensitivity and specificity [22, 23]. We found that approximately 39.5% of all inmates reported at least 1 WHO-defined TB symptom [13], of which cough was the most common. Several studies show that mass screening using a cough-based strategy has moderate sensitivity [22, 24]. In this study, 81.3% of individuals with TB had at least 1 TB-related symptom, such that an algorithm beginning with symptom-screening would detect the majority of cases, while reducing the number of individuals who require testing. An additional 19 patients (8.8% of all cases) were detected by screening all individuals, irrespective of symptoms (Strategy 1); this required screening an additional 289 participants by Xpert, as most of the 3260 participants without symptoms also didn\u2019t produce sputum.\nThe high number of symptomatic participants may be due to the high frequency of smoking and illicit drug use in our population. Studies of the general population [25\u201327] show a lower prevalence of symptoms than studies with prison inmates [7, 22, 24]. Our study used cough of any duration as a symptom, rather than cough >\u20092 weeks as is commonly done in other studies, increasing sensitivity at the expense of specificity. Symptombased screening in the context of mass screening may perform better in populations with a lower prevalence of smoking, in whom the specificity of cough is higher.\nAt the beginning of the study, we defined a CAD4TB threshold of \u2265\u200960 based on preliminary data indicating a sensitivity and specificity of approximately 80%. In this study, we found overall sensitivity to be 77.1% and specificity to be 82.8%. This sensitivity was slightly lower than that of symptom screening (81.3%), but specificity was much higher (60.5% for symptom screening). The cost of radiographic screening or\n\nTable 2.\u2003 Predictive Value of World Health Organization Tuberculosis Symptom Screen, Cough, and Computer-Aided Detection for Tuberculosis Score in 5387 Screened Inmates\n\nSymptoms\n\nCough\n\nCAD4TB Score\n\nNo. of Individuals (% of Total Cohort)\n\nAbsent\n\nNo\n\n<\u200960\n\n2793 (51.8)\n\n\u2265\u200960\n\n467 (8.7)\n\n\u2003Total\n\n3260 (60.5)\n\nPresent\n\nNo\n\n<\u200960\n\n498 (9.2)\n\n\u2265\u200960\n\n102 (2.0)\n\nYes\n\n<\u200960\n\n1189 (22.0)\n\n\u2265\u200960\n\n338 (6.3)\n\n\u2003Total\n\n2127 (39.5)\n\nAbbreviations: CAD4TB, Computer-Aided Detection for Tuberculosis; TB, tuberculosis.\n\nNo. of Cases (TB Prevalence)\n7 (0.2) 33 (7.0) 40 (1.2)\n6 (1.2) 17 (16.6) 36 (3.0) 115 (34.0) 174 (8.2)\n\n% of all TB Cases Detected\n3.3 15.4 18.7\n2.8 7.9 16.8 53.8 81.3\n\nMass Screening for TB in Prisons\u2002 \u2022\u2002 cid\u20022021:72\u2002(1 March)\u2002\u2022\u2002775\n\nDownloaded from https://academic.oup.com/cid/article/72/5/771/5736588 by guest on 20 August 2024\n\nTable 3.\u2003 Total and Unit Cost for Each Screening or Diagnostic Procedure\n\nCategory\n\nTotal Cost of the Item, 2018 US$\n\nUnit Cost, 2018 US$\n\nInterview (n = 5387)\n\n\u2003 Fixed costs\n\nEquipmenta\n\n\u2003 Variable costs\n\nHuman resources\n\nInputs\n\n\u2003Total\n\nClinical evaluation (n = 764)\n\n\u2003 Variable costs\n\nHuman resources\n\n\u2003Total\n\nRadiograph (n = 5387) \u2003 Fixed costs\n\nEquipmenta\n\nCAD4TB software\n\nTransport mobile diagnostic unit\n\n\u2003 Variable costs\n\nHuman resources\n\nCAD4TB score\n\n\u2003Total\n\nXpert MTB/RIF assay (n = 1743)\n\n\u2003 Fixed costs\n\nEquipmenta\n\nMaintenance\n\n\u2003 Variable costs\n\nHuman resources\n\nInputs\n\n\u2003Total\n\n127.63 10 027.80\n105.21 10 260.64\n1986.40 1986.40\n9840.73 667.00\n3875.96\n10 027.80 9427.25\n33 838.74\n4954.98 5167.95 5013.90 18 339.23 33 476.06\n\n0.02 1.86 0.02 1.9\n2.60 2.60\n1.83 0.12 0.72\n1.86 1.75 6.28\n2.84 2.97 2.87 10.52 19.20\n\nAbbreviations: CAD4TB, Computer-Aided Detection for Tuberculosis; MTB, Mycobacterium tuberculosis; RIF, rifampicin; US, United States. aProjected cost for a useful life of 10 years, based on the examinations made for 1 year.\n\nsymptom screening followed Xpert assay and CXR were considerably higher than that of symptom screening and Xpert for all individuals. The cost per case diagnosed for screening with Xpert assay for all individuals was lower. Alternative thresholds could be used to increase sensitivity of CXR with CAD4TB, at the expense of specificity, and further work is needed to identify optimal thresholds to maximize cost-effectiveness.\nThe strengths of our study include a representative sample of prisoners in Mato Grosso do Sul. The 3 prisons we screened\n\nhouse 32% of the state\u2019s prison population [19, 21]. Our participation rate of 88.9% of the study\u2019s target population is similar to previous recruitments performed by our group [7] and other mass screening initiatives [19]. We undertook a rigorous microcosting analysis to derive \u201creal-world\u201d costs of implementing various components of triage and diagnosis in prisons, which are critical to decisions of scaling up systematic screening in these settings.\nThere are several limitations to this study. A major challenge was that only 27.2% of participants were able to produce a sputum sample in initial visit, and sputum induction was not possible in this setting. As a result, we likely underestimated the true prevalence of TB. However, our estimates for the yield and cost per case diagnosed when screening all participants reflect this limitation in prisons, which is not just a study challenge but a real-world obstacle to screening. While sputum induction would likely improve yield, it is possible that the efficiency (prevalence among tested individuals) would be lower, and the cost per case diagnosed would likely be higher. Our findings do, however, underscore the need for non-sputum-based diagnostics to reach patients earlier in the TB disease spectrum [28\u201331].\nWe do not use testing for TB infection, either through a interferon gamma release assays (IGRA) or tuberculin skin test, in our diagnostic algorithms. Previous tuberculin skin test conversion studies in Brazilian prisons have demonstrated hyperendemic rates of transmission with an annual conversion above 25% [7, 12]. In a setting with such a high force of infection, it is unclear how tuberculin skin (or IGRA) testing would accurately discriminate TB disease.\nWe estimated costs assuming that diagnostic infrastructure (Xpert machines, radiography equipment) was not present; for prisons in which such investments have been made for routine diagnostic purposes, incremental costs per case diagnosed via mass screening may be lower. Finally, we evaluated a limited combination of commonly used diagnostics (symptom screening, Xpert, radiography); while many more combinations\n\nTable 4.\u2003 Yield and Cost per Case Diagnosed for 4 Tuberculosis Screening Strategies\n\nStrategies\n\nCases Diagnosed Missed Cases % Yield (95% CI)\n\nParticipants Screened With Xpert, No.\n\nMean Cost per Case Detected,\nUS$\n\nAll cases\n\n214\n\n\u2026\n\n\u2026\n\n\u2026\n\n485\n\nComparator groups\n\n\u2003 Strategy 1: Sputum Xpert for all participants\n\n160\n\n54\n\n74 (68\u201380)\n\n1452\n\n249\n\n\u2003 Strategy 2: Symptom screening\n\n141\n\n73\n\n65 (59\u201371)\n\n1163\n\n255\n\n\u2003\u2003 If positive: Xpert\n\n\u2003 Strategy 3: Chest radiography (CAD4TB)\n\n138\n\n76\n\n64 (57\u201370)\n\n383\n\n370\n\n\u2003\u2003 If score \u2265\u200960: Xpert\n\n\u2003 Strategy 4: Symptom screening\n\n163\n\n51\n\n76 (70\u201381)\n\n1248\n\n395\n\n\u2003\u2003 If positive, Xpert\n\n\u2003\u2003 If negative, CXR (CAD4TB) followed by\n\nXpert if score \u226560\n\nAbbreviations: CAD4TB, Computer-Aided Detection for Tuberculosis; CI, confidence interval; CXR, chest radiograph; MTB, Mycobacterium tuberculosis; RIF, rifampicin; US, United States; Xpert, MTB/RIF assay.\n\n776\u2002\u2022\u2002cid\u20022021:72\u2002(1 March)\u2002\u2022\u2002Santos et al\n\nDownloaded from https://academic.oup.com/cid/article/72/5/771/5736588 by guest on 20 August 2024\n\nor algorithms are possible using, for example, different criteria for interpretation of these screening tools, we selected these to be simple and scalable for use in resource-constrained settings.\nIn summary, our results suggest that mass TB screening in high-burden prisons, conducted by sputum Xpert testing of all inmates or those with symptoms, is an effective approach to case detection at a modest cost per case detected. Chest radiography, while it has higher overall accuracy than symptom screening, was more costly and did not substantially improve yield compared with sputum-based screening of all participants. Active case finding by sputum testing with Xpert MTB/RIF assay should be scaled up in Brazilian prisons and other high-burden countries to address TB in incarcerated populations.\nSupplementary Data\nSupplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.\nNotes\nAuthor contributions. A. S. S., E. F. L., C. R., A. K., J. R. A., and J. C. were involved in the study conception and design. A. S. S., E. F. L., and F. L. were involved in the data collection. A. S. S., R. D. O., E. F. L., F. L., O. C., and L. M. were involved in the data analysis and manuscript drafting. C. G., T. C., A. K., J. R. A., and J. C. were involved in the study design and manuscript review. All authors read and approved the final manuscript.\nAcknowledgments. The authors thank the State Health Department of Mato Grosso do Sul and State Agency of Administration Prisons for their full support during the study period; the study participants for their kind cooperation during the data collection process; and the Central Laboratory of the state of Mato Grosso do Sul for the support in the accomplishment of the laboratory tests. All data generated or analyzed during this study are included in this published article and its supplementary information files.\nDisclaimer. All participants provided written informed consent prior to study participation. The study was approved by Federal University of Grande Dourados, the National Committee on Research Ethics (#2.195.047), and the Institutional Review Board and Stanford University (#40285).\nFinancial support. This study was supported by the US National Institutes of Health (grant number R01 AI130058).\nPotential conflicts of interest. The authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest.\nReferences\n1. Kyu HH, Maddison ER, Henry NJ, et al. The global burden of tuberculosis: results from the Global Burden of Disease Study 2015. Lancet Infect Dis 2018; 18:261\u201384.\n2. Zumla A, George A, Sharma V, Herbert RHN, Oxley A, Oliver M. The WHO 2014 global tuberculosis report\u2014further to go. Lancet Glob Health 2015; 3:e10\u20132.\n3. Mabud TS, de Lourdes Delgado Alves M, Ko AI, et al. Evaluating strategies for control of tuberculosis in prisons and prevention of spillover into communities: An observational and modeling study from Brazil. PLoS Med 2019; 16:e1002737.\n4. Cudahy PGT, Andrews JR, Bilinski A, et al. Spatially targeted screening to reduce tuberculosis transmission in high-incidence settings. Lancet Infect Dis 2019; 19:e89\u201395.\n5. Baussano I, Williams BG, Nunn P, Beggiato M, Fedeli U, Scano F. Tuberculosis incidence in prisons: a systematic review. PLoS Med 2010; 7:e1000381.\n6. Mamani M, Mahmudian H, Majzoobi MM, Poorolajal J. Prevalence and incidence rates of latent tuberculous infection in a large prison in Iran. Int J Tuberc Lung Dis 2016; 20:1072\u20137.\n\n7. Pai\u00e3o DS, Lemos EF, Carbone AD, et al. Impact of mass-screening on tuberculosis incidence in a prospective cohort of Brazilian prisoners. BMC Infect Dis 2016; 16:533.\n8. Arroyave L, Keynan Y, L\u00f3pez L, Marin D, Arbel\u00e1ez MP, Rueda ZV. Negative latent tuberculosis at time of incarceration: identifying a very high-risk group for infection. Epidemiol Infect 2017; 145:2491\u20139.\n9. Bourdillon PM, Gon\u00e7alves CCM, Pelissari DM, et al. Increase in tuberculosis cases among prisoners, Brazil, 2009\u20132014. Emerg Infect Dis 2017; 23:496\u20139.\n10. Legrand J, Sanchez A, Le Pont F, Camacho L, Larouze B. Modeling the impact of tuberculosis control strategies in highly endemic overcrowded prisons. PLoS One 2008; 3:e2100.\n11. Brazil Ministry of Justice and Public Security. National survey of prison information: INFOPEN update June/2016. Bras\u00edlia: Ministry of Justice and Public Security, 2017. Available at: http://depen.gov.br/DEPEN/depen/sisdepen/ infopen/relatorio_2016_22-11.pdf. Accessed 22 September 2018.\n12. Carbone A da SS, Pai\u00e3o DSG, Sgarbi RVE, et al. Active and latent tuberculosis in Brazilian correctional facilities: a cross-sectional study. BMC Infect Dis 2015; 15:24.\n13. Bone A, Aerts A, Grzemska M, et al. Tuberculosis control in prisons: a manual for programme managers. WHO/CDS/TB/2000.281.191. 2000. Available at: https:// apps.who.int/iris/handle/10665/66823. Accessed 27 June 2019.\n14. World Health Organization. Implementing the End TB strategy: the essentials. Geneva, Switzerland: WHO, 2015.\n15. Delft Imaging Systems. CAD4TB: Computer-Aided Detection for Tuberculosis. Available at: https://www.delft.care/cad4tb. Accessed 26 June 2019.\n16. Brazil Ministry of Health. Guidelines for tuberculosis control in Brazil. Bras\u00edlia: Ministry of Health: Secretariat of Health Surveillance: Department of Epidemiological Surveillance, 2018.\n17. Brazil Ministry of Health. Epidemiological bulletin 11. Bras\u00edlia: Ministry of Health, 2018. Available at: http://portalarquivos2.saude.gov.br/images/pdf/2018/ marco/26/2018-009.pdf. Accessed 6 February 2019.\n18. Valen\u00e7a MS, Possuelo LG, Cezar-Vaz MR, et al. Tuberculosis in Brazilian prisons: an integrative literature review. Ci\u00eanc Amp Sa\u00fade Coletiva 2016; 21:2147\u201360.\n19. Pelissari DM, Kuhleis DC, Bartholomay P, et al. Prevalence and screening of active tuberculosis in a prison in the south of Brazil. Int J Tuberc Lung Dis 2018; 22:1166\u201371.\n20. Vinkeles Melchers NV, van Elsland SL, Lange JM, Borgdorff MW, van den Hombergh J. State of affairs of tuberculosis in prison facilities: a systematic review of screening practices and recommendations for best TB control. PLoS One 2013; 8:e53644.\n21. Lemos AC, Matos ED, Bittencourt CN. Prevalence of active and latent TB among inmates in a prison hospital in Bahia, Brazil. J Bras Pneumol 2009; 35:63\u20138.\n22. Sanchez A, Gerhardt G, Natal S, et al. Prevalence of pulmonary tuberculosis and comparative evaluation of screening strategies in a Brazilian prison. Int J Tuberc Lung Dis 2005; 9:633\u20139.\n23. Fournet N, Sanchez A, Massari V, et al. Development and evaluation of tuberculosis screening scores in Brazilian prisons. Public Health 2006; 120:976\u201383.\n24. Sanchez A, Larouz\u00e9 B, Espinola AB, et al. Screening for tuberculosis on admission to highly endemic prisons? The case of Rio de Janeiro State prisons. Int J Tuberc Lung Dis 2009; 13:1247\u201352.\n25. Hoa NB, Sy DN, Nhung NV, Tiemersma EW, Borgdorff MW, Cobelens FG. National survey of tuberculosis prevalence in Viet Nam. Bull World Health Organ 2010; 88:273\u201380.\n26. Federal Republic of Nigeria. Report first national TB prevalence survey 2012, Nigeria. Abuja, Nigeria: Ministry of Health, 2012. Available at: https://www.who. int/tb/publications/NigeriaReport_WEB_NEW.pdf. Accessed 28 June 2019.\n27. Soemantri S, Senewe FP, Tjandrarini DH, et al. Three-fold reduction in the prevalence of tuberculosis over 25 years in Indonesia. Int J Tuberc Lung Dis 2007; 11:398\u2013404.\n28. Walzl G, McNerney R, du Plessis N, et al. Tuberculosis: advances and challenges in development of new diagnostics and biomarkers. Lancet Infect Dis 2018; 18:e199\u2013210.\n29. Denkinger CM, Kik SV, Cirillo DM, et al. Defining the needs for next generation assays for tuberculosis. J Infect Dis 2015; 211(Suppl 2):S29\u201338.\n30. Keeler E, Perkins MD, Small P, et al. Reducing the global burden of tuberculosis: the contribution of improved diagnostics. Nature 2006; 444(Suppl 1):49\u201357.\n31. Calligaro GL, Zijenah LS, Peter JG, et al. Effect of new tuberculosis diagnostic technologies on community-based intensified case finding: a multicentre randomised controlled trial. Lancet Infect Dis 2017; 17:441\u201350.\n\nMass Screening for TB in Prisons\u2002 \u2022\u2002 cid\u20022021:72\u2002(1 March)\u2002\u2022\u2002777\n\n\n",
    "authors": [
      "Andrea Da Silva Santos",
      "Roberto Dias De Oliveira",
      "Everton Ferreira Lemos",
      "Fabiano Lima",
      "Ted Cohen",
      "Olivia Cords",
      "Leonardo Martinez",
      "Crhistinne Gon\u00e7alves",
      "Albert Ko",
      "Jason R Andrews",
      "Julio Croda"
    ],
    "doi": "10.1093/cid/ciaa135",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://academic.oup.com/cid/article/72/5/771/5736588"
  },
  {
    "key": "SKRA72NC",
    "title": "Cost analysis of nucleic acid amplification for diagnosing pulmonary tuberculosis, within the context of the Brazilian Unified Health Care System",
    "abstract": "We estimated the costs of a molecular test for Mycobacterium tuberculosis and resistance to rifampin (Xpert MTB/RIF) and of smear microscopy, within the Brazilian Sistema \u00danico de Sa\u00fade (SUS, Unified Health Care System). In SUS laboratories in the cities of Rio de Janeiro and Manaus, we performed activity-based costing and micro-costing. The mean unit costs for Xpert MTB/RIF and smear microscopy were R$35.57 and R$14.16, respectively. The major cost drivers for Xpert MTB/RIF and smear microscopy were consumables/reagents and staff, respectively. These results might facilitate future costeffectiveness studies and inform the decision-making process regarding the expansion of Xpert MTB/RIF use in Brazil.",
    "full_text": "J Bras Pneumol. 2015;41(6):536-538 http://dx.doi.org/10.1590/S1806-37562015000004524\n\nBRIEF COMMUNICATION\n\nCost analysis of nucleic acid amplification for diagnosing pulmonary tuberculosis, within the context of the Brazilian Unified Health Care System\nM\u00e1rcia Pinto1, Aline Piovezan Entringer1, Ricardo Steffen2, Anete Trajman2,3\n\n1. Instituto de Sa\u00fade da Mulher, da Crian\u00e7a e do Adolescente Fernandes Figueira, Funda\u00e7\u00e3o Oswaldo Cruz, Rio de Janeiro (RJ) Brasil.\n2. Programa de P\u00f3s-Gradua\u00e7\u00e3o em Cl\u00ednica M\u00e9dica, Faculdade de Medicina, Universidade Federal do Rio de Janeiro, Rio de Janeiro (RJ) Brasil.\n3. Montreal Chest Institute, McGill University, Montreal (QC) Canada.\nSubmitted: 4 December 2014. Accepted: 14 July 2015.\nStudy carried out at the Instituto Nacional de Sa\u00fade da Mulher, da Crian\u00e7a e do Adolescente Fernandes Figueira, Funda\u00e7\u00e3o Oswaldo Cruz, Rio de Janeiro (RJ) Brasil.\n\nABSTRACT\nWe estimated the costs of a molecular test for Mycobacterium tuberculosis and resistance to rifampin (Xpert MTB/RIF) and of smear microscopy, within the Brazilian Sistema \u00danico de Sa\u00fade (SUS, Unified Health Care System). In SUS laboratories in the cities of Rio de Janeiro and Manaus, we performed activity-based costing and micro-costing. The mean unit costs for Xpert MTB/RIF and smear microscopy were R$35.57 and R$14.16, respectively. The major cost drivers for Xpert MTB/RIF and smear microscopy were consumables/reagents and staff, respectively. These results might facilitate future costeffectiveness studies and inform the decision-making process regarding the expansion of Xpert MTB/RIF use in Brazil.\nKeywords: Costs and cost analysis; Tuberculosis; Nucleic acid amplification techniques.\n\nBrazil is among the 22 countries with the highest tuberculosis burden.(1) One of the major obstacles to controlling the disease is the delay in diagnosis. Smear microscopy is a routine test within the Brazilian Sistema \u00danico de Sa\u00fade (SUS, Unified Health Care System); however, it has low sensitivity.(2) Xpert\u00ae MTB/RIF (Cepheid; Sunnyvale, CA, USA), which is performed in the GeneXpert\u00ae system (Cepheid), is a rapid molecular test for detecting Mycobacterium tuberculosis and its rifampin-resistant mutants.(3) In 2010, Xpert\u00ae MTB/RIF was recommended by the World Health Organization for initial diagnosis in patients with tuberculosis and suspected multidrug resistance or HIV infection.(4) The accuracy of the test is high,(5) and studies have demonstrated that it is cost-effective(6-8) in other scenarios. In Brazil, Xpert\u00ae MTB/RIF was approved by the National Committee for Health Technology Incorporation in September of 2013 for use within the SUS.(9)\nThe objective of the present study was to estimate the unit cost of Xpert\u00ae MTB/RIF, since it does not yet have a reference value on the SUS Sigtap unified pricing list of the System for the Management of the Pricing List of Procedures, Drugs, and OPM (orthoses, prostheses, and materials). In addition, we intended to contribute information to support other economic evaluations in this field.\nThis was a descriptive study, which conducted a partial economic evaluation to estimate the cost of performing\n\nXpert\u00ae MTB/RIF and smear microscopy (Ziehl-Neelsen method) within the SUS, performed in parallel with a pilot study of implementation of GeneXpert\u00ae use for the diagnosis of pulmonary tuberculosis in two cities in Brazil.(10) We selected two laboratories in the city of Rio de Janeiro, both of which are affiliated with the Secretaria Municipal de Sa\u00fade e Defesa Civil do Rio de Janeiro (SMSDC/RJ, Rio de Janeiro Municipal Department of Health and Civil Defense), and one laboratory in the city of Manaus, which is affiliated with the Secretaria de Estado de Sa\u00fade do Amazonas (SES/AM, Amazonas State Department of Health). This selection was based on the organization of the health care facility network, the level of decentralization, and the production output. In Rio de Janeiro, \u201claboratory 1\u201d is a polyclinic with a medium production output, and \u201claboratory 2\u201d is a family clinic with a small production output. In Manaus, the selected laboratory (\u201claboratory 3\u201d) is responsible for 71% of all smear microscopy examinations conducted in the city, it being considered to have a large production output.\nWe performed activity-based costing and micro-costing, on the basis of standard operating procedures for smear microscopy(2) and for Xpert\u00ae MTB/RIF.(11) The selected cost items were as follows: administrative costs (electricity, water, cleaning, and safety); staff costs, which included training (only for Xpert\u00ae MTB/RIF); laboratory consumable costs; and equipment costs. The Xpert\u00ae MTB/RIF cartridge cost R$20.46 (US$9.98),(12) and the GeneXpert\u00ae system\n\nCorrespondence to: M\u00e1rcia Pinto. Avenida Rui Barbosa, 716. Flamengo, CEP 22250-020, Rio de Janeiro, RJ, Brasil. Tel.: 55 21 2554-1915. E-mail: mftpinto@gmail.com Financial support: This study received financial support from the project known as Inova\u00e7\u00e3o para o Controle da Tuberculose (INCo-TB, Innovation for Tuberculosis Control), which is a partnership between the Brazilian National Tuberculosis Control Program and the Ataulpho de Paiva Foundation, with funding from the Bill and Melinda Gates Foundation. The funding sources did not affect the content of the present study or the decision to publish it. The authors take full responsibility for the study design and execution, as well as for the opinions expressed here.\n\n536\n\n\u00a9 2015 Sociedade Brasileira de Pneumologia e Tisiologia\n\nISSN 1806-3713\n\nPinto M, Entringer AP, Steffen R, Trajman A\n\ncost R$34,850.00 (US$17,000)\u2014purchase prices for the study of implementation.(10) The Brazilian National Ministry of Health is exempt from the costs of taxes and nationalization regarding the cartridges and GeneXpert\u00ae.\nStaff costs for Xpert\u00ae MTB/RIF use were calculated by a time and motion study undertaken at two separate time points in order to assess the learning curve of professionals: the first was fifteen days after GeneXpert\u00ae was installed in the laboratory, which is the time required for professionals to adapt to it; and the second was three months after the equipment was installed, at which point tests were routinely performed. This time was taken into account in the calculation of staff costs. Data collection for estimating the cost of smear microcopy occurred in a single step, because smear microscopy is a well-established procedure within the SUS. The administrative costs refer to the mean quarterly expenditures of the health care facilities where the laboratories are located and were apportioned according to floor space and production. Depreciation was applied when necessary, according to the useful life of the equipment.(13) Costs of calibration and preventive/corrective maintenance on GeneXpert\u00ae were taken into account.(3)\nThe prices of the consumables were obtained from official sources, namely Comprasnet, Banco de Pre\u00e7os em Sa\u00fade, SMSDC/RJ, and SES/AM, and from the manufacturer (Cepheid). The mean cost of the tests is expressed in 2012 Brazilian reals. The costs of the cartridges and of GeneXpert\u00ae were converted from Brazilian reals to US dollars at a rate of R$2.05 = US$1.00.\nThe study was approved by the Brazilian National Research Ethics Committee (Protocol no. 493/2011), by the SMSDC/RJ Research Ethics Committee (Protocol no. 445A/11), and by the Research Ethics Committee of the Funda\u00e7\u00e3o de Medicina Tropical de Manaus Dr. Heitor Vieira Dourado (Dr. Heitor Vieira Dourado Tropical Medicine Foundation of Manaus) in November of 2011.\nWe observed the production process of 230 smear microscopy examinations and 463 tests with Xpert\u00ae MTB/RIF. There was a 30% reduction in mean completion time for Xpert\u00ae MTB/RIF between the first and second observations (9.87 min vs. 7.57 min). The largest reductions were observed in laboratories 2 (54%) and 3 (35%). In the second observation, the mean completion time was 6.20 min (range, 4.87-7.53 min)\n\nin laboratories 1 and 2 and 4.30 min (range, 3.53-5.07 min) in laboratory 3.\nThe mean cost of Xpert\u00ae MTB/RIF use was R$35.57 (range, R$33.70-R$39.40), and the mean cost of smear microscopy was R$14.16 (range, R$11.30-R$ 21.00). The major cost drivers for Xpert MTB/RIF were consumables and reagents (62%), especially the cartridges, whereas the major cost driver for smear microscopy was staff (58%). There was great variability in staff costs between the two cities (Table 1). Therefore, the cost of two smear microscopy examinations, which are recommended by the Brazilian National Tuberculosis Control Program and required to achieve a sensitivity of 70%,(2) represents 80% of the cost of an Xpert\u00ae MTB/RIF test, which has a sensitivity of 88%.(5)\nDuring the data collection process, the production of Xpert\u00ae MTB/RIF tests increased relative to that of smears of the first sample, especially in RJ. The work hours of professionals remained unchanged, which suggests that the introduction of Xpert\u00ae MTB/RIF represented a technical efficiency gain in the routine of the laboratories (Table 2).\nXpert\u00ae MTB/RIF is considered a promising technology for tuberculosis control because it provides fast, accurate, and cost-effective results.(5-8) The present study conducted a partial economic evaluation, which describes exclusively the costs of performing the two technologies for the diagnosis of tuberculosis. Although we did not conduct a complete economic evaluation, the results detailed herein, together with comparative effectiveness data for the tests performed under routine conditions in the same cities where the pilot study was conducted,(10) served as the basis for the estimates of cost-effectiveness ratios.(14) We concluded that the cost of two smear microscopy examinations, which are usually required when tuberculosis is suspected, is close to (i.e., 80% of) the cost of an Xpert\u00ae MTB/RIF test.\nOne of the advantages of the present study was that it was carried out in parallel with the Xpert\u00ae MTB/RIF implementation study,(10) which allowed us to observe the incorporation of the new technology into the use of resources and into the learning process of health professionals within the SUS. One study also estimated the cost of the test during a study of implementation, with results ranging from R$46.40 to R$56.48 (US$22.63 to US$27.55), much higher than ours.(8) However, the price of the cartridge was higher than that used in the present study (R$39.77-US$19.40). The value added by\n\nTable 1. Unit costs for Xpert\u00ae MTB/RIF and smear microscopy in the laboratories studied, Rio de Janeiro and Manaus (in Brazilian Reals, 2012).a\n\nCost item\n\nSmear microscopy\n\nXpert\u00ae MTB/RIF\n\nLab 1\n\nLab 2\n\nLab 3\n\nLab 1\n\nLab 2\n\nLab 3\n\nStaff costs\n\n5.18\n\n3.76\n\n15.87\n\n3.71\n\n3.01\n\n13.27\n\nConsumable and reagent costs\n\n2.35\n\n2.35\n\n2.35\n\n22.01\n\n22.01\n\n22.01\n\nEquipment costs\n\n1.34\n\n0.85\n\n0.97\n\n4.07\n\n3.96\n\n2.42\n\nAdministrative costs\n\n2.51\n\n2.81\n\n2.14\n\n4.18\n\n4.03\n\n2.04\n\nUnit cost\n\n11.38\n\n9.77\n\n21.33\n\n33.97\n\n33.01\n\n39.74\n\nLab: laboratory. aConversion rate used for the cartridge and GeneXpert\u00ae: US$1.00 = R$2.05 (2012).\n\nJ Bras Pneumol. 2015;41(6):536-538\n\n537\n\nCost analysis of nucleic acid amplification for diagnosing pulmonary tuberculosis, within the context of the Brazilian Unified Health Care System\n\nTable 2. Mean daily number of Xpert\u00ae MTB/RIF tests and smears of the first sample, produced in the laboratories studied, Rio de Janeiro and Manaus.\n\nUnit\n\nSmear\n\nXpert\u00ae % increase\n\nmicroscopy MTB/RIF\n\nLaboratory 1\n\n10\n\n13\n\n30\n\nLaboratory 2\n\n7\n\n9\n\n29\n\nLaboratory 3\n\n31\n\n34\n\n10\n\nthe other cost items was similar. Other studies reported costs ranging from R$30.61 (US$14.93) to R$54.41 (US$26.54).(15,16) It is of note that all those studies were conducted in countries with different structures from that of the SUS, which limits the comparison.\nThe advantage of activity-based costing is the possibility of observing a significant number of tests, which makes it possible to identify a standard completion time and to perform a detailed inventory of the cost items. However, the method limits the possibilities of generalization, because of the organizational characteristics and the functioning of the laboratories studied.\nThe reduction in test completion time between the two observations was lower in laboratory 1, since, during the second data collection time point, the trained technician was replaced with a less experienced one. In order to minimize the effects of this event, we observed a larger number of tests. It is believed that, with the incorporation of Xpert\u00ae MTB/RIF into the routine of the laboratories, test completion time will decrease and production will increase. Therefore,\n\nit will be possible to increase technical efficiency and reduce the unit cost.\nAmong the limitations of the present study is the mean salary value, which does not reflect the Brazilian reality given the diversity of contractual arrangements in operation in the country. In order to minimize this diversity, we adopted the salaries of professionals affiliated with the state of AM and with the city of Rio de Janeiro, on the basis of different salary ranges. A second limitation relates to laboratory floor space, used for estimating cost per square meter. The physical structure varies among the facilities in terms of size and location; therefore, we included three laboratories with extremely different configurations, located in two Brazilian states.\nThe results of the present study might facilitate future cost-effectiveness studies and contribute to the establishment of a reference value on the Sigtap pricing list. However, since the adoption and use of technologies are dynamic and the results of the present study refer to the initial stage of the incorporation of Xpert\u00ae MTB/RIF, it is important to observe whether there will be changes in its use.\nIn conclusion, the present study aimed to provide subsidies so that health care managers can identify the major cost drivers for Xpert\u00ae MTB/RIF as well as possible gains in efficiency and effectiveness from its adoption. In this regard, our results might facilitate both programming and planning measures targeted at tuberculosis control in Brazilian cities.\n\nREFERENCES\n1. Brasil. Minist\u00e9rio da Sa\u00fade. Secret\u00e1ria de Vigil\u00e2ncia \u00e0 Sa\u00fade. O controle da tuberculose no Brasil: avan\u00e7os, inova\u00e7\u00f5es e desafios. Bras\u00edlia: o Minist\u00e9rio. Boletim Epidemiol\u00f3gico. 2014;45(2)1-12.\n2. Minist\u00e9rio da Sa\u00fade. Secretaria de Vigil\u00e2ncia em Sa\u00fade. Manual nacional de vigil\u00e2ncia laboratorial da tuberculose e outras micobact\u00e9rias. Bras\u00edlia: Minist\u00e9rio da Sa\u00fade; 2008.\n3. Cepheid. [homepage on the Internet] Sunnyvale (CA): Cepheid; [cited 2014 Dec 3] The New GeneXpert\u00ae System. New Systems. Same game-changing performance. Available from: http://www. cepheidinternational.com/systems-and-software/genexpert-system\n4. World Health Organization. [homepage on the Internet] Geneva: WHO; [cited 2014 Dec 3]. WHO endorses new rapid tuberculosis test. Available from: http://www.who.int/mediacentre/news/ releases/2010/tb_test_20101208/en/index.html\n5. Steingart KR, Sohn H, Schiller I, Kloda LA, Boehme CC, Pai M, et al. Xpert\u00ae MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev. 2013;1:CD009593. http://dx.doi.org/10.1002/14651858.cd009593.pub2\n6. Choi HM, Miele K, Dowdy D, Shah M. Cost-effectiveness of Xpert\u00ae MTB/RIF for diagnosing pulmonary tuberculosis in the United States. Int J Tuberc Lung Dis. 2013;17(10):1328-35. http://dx.doi. org/10.5588/ijtld.13.0095\n7. Menzies NA, Cohen T, Lin HH, Murray M, Salomon JA. Population health impact and cost-effectiveness of tuberculosis diagnosis with Xpert MTB/RIF: a dynamic simulation and economic evaluation PLoS Med. 2012;9(11):e1001347 http://dx.doi.org/10.1371/journal. pmed.1001347\n8. Vassal A, van Kampen S, Sohn H, Michael JS, John KR, den Boon S, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: a cost-effectiveness analysis. PLoS Med. 2011;8(11):e1001120. http://dx.doi.org/10.1371/journal. pmed.1001120\n9. Brasil. Minist\u00e9rio da Sa\u00fade. Biblioteca Virtual em Sa\u00fade [homepage\n\non the Internet]. Bras\u00edlia: o Minist\u00e9rio; [cited 2014 Dec 3]. Portaria MS no. 48 de 10 de setembro de 2013. Available from: http://www.bvsms. saude.gov.br/bvs/saudelegis/sctie/2013/prt0048_10_09_2013.html\n10. Durovni B, Saraceni V, van den Hof S, Trajman A, Cordeiro-Santos M, Cavalcante S, et al. Impact of replacing smear microscopy with Xpert MTB/RIF for diagnosing tuberculosis in Brazil: a stepped-wedge cluster-randomized trial. PLoS Med. 2014;11(12):e1001766. http:// dx.doi.org/10.1371/journal.pmed.1001766\n11. Cepheid. Manual Cepheid Rev C [CD-ROM]. Sunnyvale (CA): Cepheid; 2009.\n12. World Health Organization. [homepage on the Internet] Geneva: WHO; [cited 2013 Aug 4]. Tuberculosis Diagnostics - Xpert MTB/RIF Test. [Adobe Acrobat document, 2p.]. Available from: http://www. who.int/tb/publications/Xpert_factsheet.pdf\n13. Minist\u00e9rio da Sa\u00fade. Secretaria de Ci\u00eancia e Tecnologia. Departamento de Tecnologia e Insumos Estrat\u00e9gicos. Diretrizes Metodol\u00f3gicas: estudos de avalia\u00e7\u00e3o econ\u00f4mica de tecnologias em sa\u00fade. Bras\u00edlia: Minist\u00e9rio da Sa\u00fade; 2009.\n14. Funda\u00e7\u00e3o Ataulpho de Paiva. Projeto Bill & Melinda Gates. Rio de Janeiro: a Funda\u00e7\u00e3o; [cited 2013 Ago 04]. Estudos econ\u00f4micos da incorpora\u00e7\u00e3o do teste molecular GeneXpert\u2122 MTB/Rif para o diagn\u00f3stico de tuberculose pulmonar no Sistema \u00danico de Sa\u00fade. [Adobe Acrobat document, 36p.]. Available from: http:// www.fundacaoataulphodepaiva.com.br/2013/07/Relatoriotecnico_25042013.pdf\n15. Shah M, Chihota V, Coetzee G, Churchyard G, Dorman SE. Comparison of laboratory costs of rapid molecular tests and conventional diagnostics for detection of tuberculosis and drugresistant tuberculosis in South Africa. BMC Infect Dis. 2013;13:352. http://dx.doi.org/10.1186/1471-2334-13-352\n16. Meyer-Rath G, Schnippel K, Long L, MacLeod W, Sanne I, Stevens W, et al. The impact and cost of scaling up GeneXpert MTB/RIF in South Africa. PLoS One. 2012;7(5):e36966. http://dx.doi.org/10.1371/ journal.pone.0036966\n\n538\n\nJ Bras Pneumol. 2015;41(6):536-538\n\n\n",
    "authors": [
      "M\u00e1rcia Pinto",
      "Aline Piovezan Entringer",
      "Ricardo Steffen",
      "Anete Trajman"
    ],
    "doi": "10.1590/s1806-37562015000004524",
    "year": null,
    "item_type": "journalArticle",
    "url": "http://www.scielo.br/scielo.php?script=sci_arttext&pid=S1806-37132015000600536&lng=en&tlng=en"
  },
  {
    "key": "3KKI6PSA",
    "title": "Point of care Xpert MTB/RIF versus smear microscopy for tuberculosis diagnosis in southern African primary care clinics: a multicentre economic evaluation",
    "abstract": "Background Rapid on-site diagnosis facilitates tuberculosis control. Performing Xpert MTB/RIF (Xpert) at point of care is feasible, even when performed by minimally trained health-care workers, and when compared with point-of-care smear microscopy, reduces time to diagnosis and pretreatment loss to follow-up. However, whether Xpert is cost-effective at point of care remains unclear.",
    "full_text": "Articles\n\nPoint of care Xpert MTB/RIF versus smear microscopy for tuberculosis diagnosis in southern African primary care clinics: a multicentre economic evaluation\n\nAnil Pooran, Grant Theron, Lynn Zijenah, Duncan Chanda, Petra Clowes, Lawrence Mwenge, Farirai Mutenherwa, Paul Lecesse, John Metcalfe, Hojoon Sohn, Michael Hoelscher, Alex Pym, Jonny Peter, David Dowdy*, Keertan Dheda*\n\nSummary\nBackground Rapid on-site diagnosis facilitates tuberculosis control. Performing Xpert MTB/RIF (Xpert) at point of care is feasible, even when performed by minimally trained health-care workers, and when compared with point-of-care smear microscopy, reduces time to diagnosis and pretreatment loss to follow-up. However, whether Xpert is cost-effective at point of care remains unclear.\n\nMethods We empirically collected cost (US$, 2014) and clinical outcome data from participants presenting to primary health-care facilities in four African countries (South Africa, Zambia, Zimbabwe, and Tanzania) during the TB-NEAT trial. Costs were determined using an bottom-up ingredients approach. Effectiveness measures from the trial included number of cases diagnosed, initiated on treatment, and completing treatment. The primary outcome was the incremental cost-effectiveness of point-of-care Xpert relative to smear microscopy. The study was performed from the perspective of the health-care provider.\n\nFindings Using data from 1502 patients, we calculated that the mean Xpert unit cost was lower when performed at a centralised laboratory (Lab Xpert) rather than at point of care ($23\u221900 [95% CI 22\u221912\u201323\u221988] vs $28\u221903 [26\u221919\u201329\u221987]). Per 1000 patients screened, and relative to smear microscopy, point-of-care Xpert cost an additional $35\u2009529 (27\u2009054\u201340\u2009025) and was associated with an additional 24\u22193 treatment initiations ([\u201320\u22190 to 68\u22195]; $1464 per treatment), 63\u22194 same-day treatment initiations ([27\u22193\u201399\u22194]; $511 per same-day treatment), and 29\u22194 treatment completions ([\u20136\u22199 to 65\u22196]; $1211 per completion). Xpert costs were most sensitive to test volume, whereas incremental outcomes were most sensitive to the number of patients initiating and completing treatment. The probability of point-of-care Xpert being cost-effective was 90% at a willingness to pay of $3820 per treatment completion.\n\nInterpretation In southern Africa, although point-of-care Xpert unit cost is higher than Lab Xpert, it is likely to offer good value for money relative to smear microscopy. With the current availability of point-of-care nucleic acid amplification platforms (eg, Xpert Edge), these data inform much needed investment and resource allocation strategies in tuberculosis endemic settings.\n\nFunding European Union European and Developing Countries Clinical Trials Partnership and the South African Medical Research Council.\n\nCopyright \u00a9 2019 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license.\n\nIntroduction\nEarly screening and diagnosis is a key component of tubercu\u00adlosis control and underpins the post-2015 END TB Strategy aimed at substantially reducing the burden of disease.1,2 The Xpert MTB/RIF (Xpert) assay is a rapid molecular-based test that has consistently shown its superior sensitivity over smear microscopy in diagnosing pulmonary,3 extrapulmonary,4 and paediatric tubercu\u00ad losis.5,6 As such, it has been endorsed by WHO7 and is undergoing a large-scale global rollout.8\nHowever, where Xpert should optimally be placed within national tuberculosis programmes (NTPs) rem\u00adains unclear. Should Xpert, to exploit its portability and userfriendly format, be situated in centralised laborat\u00adories or within more peripheral clinics at point of care? WHO endorses implementation at centralised health facilities\n\n(district and subdistrict levels), but this implementation limits the potential benefits of Xpert as a rapid diagnostic tool. Indeed, later diagnosis and reporting of results, as a consequence of centralised placement, can delay clinical decisions and hence treat\u00adment initiation.9,10 Moreover, up to 40% of patients in tuberculosis endemic areas contribute to pretreatment loss to follow-up (ie, they do not return to the clinic to start treatment after being informed of a positive result).11\u201313 A large randomised controlled trial14 showed that placing Xpert at point of care within primary care clinics was not only feasible, when performed by a minimally trained health-care worker, but significantly reduced pretreatment loss to follow-up. Given these con\u00ad sider\u00adations, a strategic and ideological drive has occurred to move to point-of-care diagnosis\u2014as occurred with HIV, sexually transmitted diseases, and diabetes. Indeed, Xpert\n\nLancet Glob Health 2019; 7: e798\u2013807\nThis online publication has been corrected. The corrected version first appeared at thelancet.com/lancetgh on May 31, 2019\nSee Comment page e692\n*Contributed equally\nCentre for Lung Infection and Immunity, Division of Pulmonology, Department of Medicine and University of Cape Town (UCT) Lung Institute and South African MRC/UCT Centre for the Study of Antimicrobial Resistance, UCT, Cape Town, South Africa (A Pooran PhD, G Theron PhD, K Dheda PhD); Department of Medicine, UCT, Cape Town, South Africa (J Peter PhD); Faculty of Infectious and Tropical Diseases, Department of Immunology and Infection, London School of Hygiene & Tropical Medicine, London, UK (K Dheda); Department of Science and Technology\u2013 National Research Foundation Centre of Excellence for Biomedical Tuberculosis Research, and South Africa Medical Research Council Centre for Tuberculosis Research, Faculty of Medicine and Health Sciences, Stellenbosch University, Tygerberg, South Africa (G Theron); Department of Immunology, University of Zimbabwe College of Health Sciences, Harare, Zimbabwe (L Zijenah PhD); University Teaching Hospital, Lusaka, Zambia (D Chanda MD, L Mwenge MSc); National Institute of Medical Research, Mbeya Medical Research Centre, Mbeya, Tanzania (P Clowes MD); Division of Infectious Diseases and Tropical Medicine, Medical Centre of the University of Munich, Munich, Germany (P Clowes, M Hoelscher MD); Biomedical Research and Training Institute,\n\nwww.thelancet.com/lancetgh Vol 7 June 2019\n\ne798\n\nArticles\n\nHarare, Zimbabwe (F Mutenherwa MSc); Denver Health Residency in Emergency\nMedicine, Denver Health Medical Center, Denver, CO, USA (P Lecesse MD); Division of Pulmonary and Critical Care\nMedicine, University of California San Francisco School of Medicine, San Francisco, CA,\nUSA (J Metcalfe MD); Department of Epidemiology,\nJohns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA\n(H Sohn PhD, D Dowdy PhD); German Centre for Infection Research, Munich, Germany\n(M Hoelscher); and South African Medical Research Council, Africa Health Research\nInstitute, and Durban, South Africa (A Pym PhD)\nCorrespondence to: Prof Keertan Dheda Division of\nPulmonology and UCT Lung Institute, Department of\nMedicine, Cape Town 7925, South Africa\nkeertan.dheda@uct.ac.za\n\nResearch in context\nEvidence before this study We searched PubMed for all studies published between Jan 1, 2010, and Sept 1, 2018, using the search terms \u201ctuberculosis\u201d OR \u201c(TB)\u201d AND (\u201ccost\u201d OR \u201ccost effectiveness\u201d) AND (\u201cXpert MTB/RIF\u201d OR \u201cGeneXpert\u201d). Many studies investigated the cost-effectiveness of Xpert MTB/RIF (Xpert) for tuberculosis diagnosis in several contexts and settings using modelling approaches. Only one South African study estimated the cost-effectiveness of Xpert in a real-world context with the use of primary economic and clinical data. This study found that Xpert was cost neutral and did not improve the cost-effectiveness of routine tuberculosis diagnosis. However, no published studies assessed the cost-effectiveness of Xpert when performed at the point of care with prospectively collected clinical and cost-related data.\nAdded value of this study This study used empirical cost data nested within a randomised controlled trial of point-of-care Xpert versus\nis already being used at the point of care in high burden clinics, mines, and prisons in countries such as Zimbabwe and South Africa.\nAlthough deployment of Xpert at the point of care can deliver same-day diagnosis14\u201316 and other benefits, as out\u00ad lined, the associated diagnostic test and clinical infra\u00ad structure upgrade costs are not insignificant.9,17,18 Thus, crucial questions for policy makers, and of prime import\u00ad ance to resource allocation planning, are (1) how does the cost of Xpert performed by a minimally trained nurse at point of care compare with when performed by a trained technician at a centralised laboratory; and (2) is point-ofcare placement of Xpert cost-effective? Although multiple studies17,19\u201326 have examined the economic implications of using Xpert in endemic settings, few have focused on the costs or cost-effectiveness when deployed at point of care,22,25,26 and no studies have calculated cost-effectiveness using clinical outcome data obtained from a pragmatic real-world prospective study.19\u201322,24,25 To address these questions, we analysed prospectively collected cost and clinical outcome data from a large randomised control parent trial that recruited patients from primary care clinics in four southern African countries (South Africa, Zambia, Zimbabwe, Tanzania).14\n\nMethods\nClinical trial design We used data obtained from the TB-NEAT trial,14 which has been described in detail elsewhere. The trial was a random\u00adised, two-group, parallel-group study of 1502 participants with presumptive tuberculosis re\u00ad cruited from periurban primary health-care clinics located in four southern African countries, including South Africa (Cape Town and Durban), Zimbabwe (Harare), Zambia (Lusaka), and Tanzania (Mbeya). Briefly, patients\n\nsputum smear microscopy in four African countries, suggesting that point-of-care Xpert is a cost-effective option for tuberculosis diagnosis in settings willing to pay at least US$3820 per additional patient with tuberculosis completing treatment. The volume of testing in each clinic was the most important determinant of cost and cost-effectiveness of the point-of-care Xpert strategy.\nImplications of all the available evidence Data on the cost-effectiveness of Xpert situated at a centralised laboratory remain discordant. However, the available evidence suggests that, in clinics where volume of testing is sufficiently high to offset implementation costs, point-of-care Xpert testing is likely to provide good value for money in high-burden settings in sub-Saharan Africa.\npresenting at the clinics with symptoms suggestive of tuberculosis between April 12, 2011, and March 30, 2012, were recruited into the study and randomly assigned to either same-day smear microscopy (n=758; the smear microscopy group) or Xpert MTB/RIF performed at the point of care (n=744; the Xpert group). Specific inclusion and exclusion criteria for patient recruitment can be found in the TB-NEAT paper.14 Two spot expectorated sputum samples were collected for the index test (smear microscopy or Xpert) and mycobacteria growth indicator tube liquid culture (BD Diagnostics, Sparks, MD USA). Smear microscopy was performed by a qualified technician in a laboratory linked to the clinic. In Cape Town, smear microscopy was done at a centralised laboratory close to the clinic in accordance with South African national diagnostic practices. Auramine fluorescence smear microscopy was instituted at all study sites except Tanzania, which instead used direct light microscopy with Ziehl-Neelsen staining. Xpert was done by a trained nurse (except in Zimbabwe where national policy required Xpert to be performed by a certified technician) using a four-module GeneXpert machine that was situated at each clinic specifically for the trial. An additional Xpert was performed on a stored sputum sample at a centralised laboratory (Lab Xpert) by a qualified technician, and liquid culture was performed at a reference laboratory at each study site. Patients were asked to wait until smear microscopy or Xpert results became available. If results were positive, patients were referred directly to the tuberculosis treatment office in the clinic. Patients with negative results were referred for routine clinical assessment and chest x-ray. Initiation of empirical treatment was decided by the attending clinician. Patients were subsequently followed up for 6 months after diagnosis.\n\ne799\n\nwww.thelancet.com/lancetgh Vol 7 June 2019\n\nArticles\n\nTest costs\nSmear microscopy\n\nMbeya, Tanzania\n\nConsumables\n\n$1\u221924\n\nStaff\n\n$0\u221997\n\nEquipment\n\n$0\u221952\n\nQuality control\n\n$0\u221903\n\nOverhead\n\n$0\u221920\n\nTransport\n\n$0\u221900\n\nTotal\n\n$2\u221996\n\nLusaka, Zambia\n\nConsumables\n\n$1\u221905\n\nStaff\n\n$0\u221933\n\nEquipment\n\n$0\u221935\n\nQuality control\n\n$0\u221905\n\nOverhead\n\n$0\u221912\n\nTransport\n\n$0\u221900\n\nTotal\n\n$1\u221990\n\nHarare, Zimbabwe\n\nConsumables\n\n$1\u221929\n\nStaff\n\n$0\u221931\n\nEquipment\n\n$0\u221939\n\nQuality control\n\n$0\u221935\n\nOverhead\n\n$0\u221921\n\nTransport\n\n$0\u221900\n\nTotal\n\n$2\u221955\n\nCape Town, South Africa\n\nConsumables\n\n\u2219\u2219\n\nStaff\n\n\u2219\u2219\n\nEquipment\n\n\u2219\u2219\n\nQuality control\n\n\u2219\u2219\n\nOverhead\n\n\u2219\u2219\n\nTransport\n\n\u2219\u2219\n\nTotal\n\n$2\u221952\u2020\n\nAll\n\nTotal (95% CI)\u2021\n\n$2\u221939 (2\u221928\u2013 2\u221950)\n\nXpert MTB/RIF at clinic (point-of-care Xpert)\n\nXpert MTB/RIF at centralised laboratory (Lab Xpert)\n\n$11\u221910 $0\u221975 $22\u221959 $0\u221922 $1\u221904 $0\u221900 $35\u221970\n\n$11\u221910 $0\u221958 $9\u221984 <$0\u221901 $0\u221931 $1\u221956 $23\u221940\n\n$11\u221903 $1\u221902 $10\u221961 $0\u221910 $1\u221998 $0\u221900 $24\u221974\n\n$11\u221903 $0\u221957 $9\u221978 $0\u221902 $0\u221916 $1\u221963 $23\u221918\n\n$10\u221972 $0\u221948 $18\u221997 $0\u221912 $0\u221965 $0\u221900 $30\u221993\n\n$10\u221931 $0\u221951 $10\u221912 $0\u221916 $3\u221952 $5\u221995 $30\u221959\n\n$10\u221944 $1\u221968 $13\u221951 $0\u221933 $1\u221928 $0\u221900 $27\u221925\n\n\u2219\u2219 \u2219\u2219 \u2219\u2219 \u2219\u2219 \u2219\u2219 \u2219\u2219 $17\u221991\u2020\n\n$28\u221903 (26\u221919\u201329\u221987)\n\n$23\u221900 (22\u221912\u201323\u221988)\n\nChest x-ray\n$1\u221980 $0\u221966 $2\u221934 $0\u221900 $1\u221974 $0\u221900 $6\u221954\n$0\u221900 $0\u221947 $5\u221963 $0\u221900 $0\u221921 $0\u221900 $6\u221931\n\u2219\u2219 \u2219\u2219 \u2219\u2219 \u2219\u2219 \u2219\u2219 \u2219\u2219 $5\u221948*\n$1\u221903 $1\u221927 $8\u221934 $0\u221900 $3\u221940 $0\u221900 $14\u221904\n$9\u221929 (7\u221994\u201310\u221964)\n\nClinic costs\nTuberculosis screening\n\nHIV testing and Treatment\n\ncounselling\n\ninitiation\n\n$0\u221900 $2\u221968 $0\u221900 $0\u221900 $0\u221908 $0\u221900 $2\u221976\n$0\u221900 $4\u221991 $0\u221900 $0\u221900 $0\u221936 $0\u221900 $5\u221927\n$0\u221900 $1\u221988 $0\u221900 $0\u221900 $0\u221948 $0\u221900 $2\u221936\n$0\u221900 $1\u221942 $0\u221900 $0\u221900 $1\u221937 $0\u221900 $2\u221979\n$3\u221934 (3\u221929\u20133\u221938)\n\n$1\u221927 $3\u221956 $0\u221900 $0\u221900 $0\u221963 $0\u221900 $5\u221946\n$0\u221974 $2\u221906 $0\u221900 $0\u221900 $0\u221929 $0\u221900 $3\u221908\n$1\u221919 $1\u221932 $0\u221900 $0\u221900 $0\u221917 $0\u221900 $2\u221968\n$1\u221939 $2\u221933 $0\u221900 $0\u221900 $1\u221971 $0\u221900 $5\u221943\n$4\u221908 (3\u221998\u20134\u221918)\n\n$0\u221900 $2\u221928 $0\u221900 $0\u221900 $3\u221910 $0\u221900 $5\u221937\n$0\u221900 $1\u221981 $0\u221900 $0\u221900 $0\u221924 $0\u221900 $2\u221905\n$0\u221900 $1\u221970 $0\u221900 $0\u221900 $2\u221959 $0\u221900 $4\u221929\n$0\u221900 $4\u221995 $0\u221900 $0\u221900 $3\u221912 $0\u221900 $8\u221907\n$5\u221932 (4\u221990\u20135\u221973)\n\nCosts given in US$, 2014. *X-ray costs in Zimbabwe were taken from the literature32 because empirical data collection was not possible. \u2020The costs of smear microscopy and Xpert MTB/RIF performed at a centralised laboratory in South Africa were taken as the per-test charge from the National Health Laboratory Services because empirical data collection was not possible. \u2021Weighted average of costs across all sites (95%CI), weighted by the number of patients screened at each site.\nTable 1: Component and total costs of diagnostic tests and clinic visits in four primary care clinics in southern Africa\n\nEconomic evaluation overview We used clinical and cost data empirically collected from each study site to compare the unit cost of point-of-care Xpert and Lab Xpert at different test volume capacities and to assess the cost-effectiveness of point-of-care Xpert com\u00adpared with smear microscopy. We performed the economic analysis according to well established cost-effectiveness analysis guidelines.27,28 A completed check\u00adlist of the essen\u00adtial components27 required for doing an economic analysis is provided in the appendix.\n\nMeasures of cost We calculated tuberculosis diagnosis and treatment costs from the health-care provider perspective in each trial country. We calculated the cost per test for smear mi\u00ad croscopy, point-of-care Xpert, Lab Xpert, and chest x-ray. At the time of the study, South Africa was using Xpert for routine tuberculosis diagnosis within the NTP; we meas\u00ad ured Lab Xpert costs at the remaining study sites (Zambia, Zimbabwe, and Tanzania) in the respective clinics, but after trial completion (within the context of other ongoing\n\nSee Online for appendix\n\nwww.thelancet.com/lancetgh Vol 7 June 2019\n\ne800\n\nArticles\n\n100\n\nAll sites\n\n90\n\nSouth Africa\n\nZimbabwe\n\n80\n\nZambia\n\n70\n\nTanzania\n\n60\n\n50\n\n40\n\n30\n\n20\n\nCost per Xpert test (US$)\n\n10 50 250 500\n\n1000\n\n1500\n\n2000\n\n2500\n\nNumber of Xpert tests performed annually\n\n3000\n\n3500\n\nFigure 1: Estimated unit cost per Xpert test Costs (US$, 2014) estimated according to annual test volume either at the clinic (point-of-care Xpert; dashed lines) or central laboratory (Xpert Lab; solid lines) for each individual study site. The overall weighted average of all sites is shown by the grey lines. The cost per Xpert test on the y-axis is expressed on a logarithmic scale. The observed annual number of Xpert tests done at the laboratory (circles) and at the clinic (diamond) in each of the four study sites are indicated. The Xpert Lab is not presented for South Africa as no empirical cost data were collected for this site. For individual and all sites, the observed annual testing frequency was greater at the laboratory compared with the clinic, resulting in a lower Xpert Lab unit cost. However, if the annual number of Xperts performed annually were the same for a given site, the unit cost of point-of-care Xpert would be less than Xpert Lab. Xpert=Xpert MTB/RIF.\n\nFor Oanda historical exchange rates see https://www.oanda. com/fx-for-business/historicalrates\nFor World Bank consumer prices see http://data.worldbank.\norg/indicator/FP.CPI.TOTL.ZG\n\nstudies). We also assessed the cost per clinic visit for tuberculosis screening, HIV testing, and counselling and tuberculosis treatment initiation. We estimated the weighted mean (weighted by volume of testing) for the cost per test and cost per clinic visit across all trial sites. We converted local costs to 2014 US$ at exchange rates of $1=R9\u221965 (South African rand), K5\u221935 (Zambia kwacha), or TSh1584\u221905 (Tanzanian shilling) according to Oanda historical exchange rates. At the time of the study, Zimbabwe had already adopted US$ as their national currency. We adjusted costs to the year of ana\u00adlysis as necessary using country-specific consumer price indices provided by The World Bank. We annualised capital costs (building, vehicles, and equipment) at a dis\u00adcount rate of 3%. We estimated expected lifeyears of buildings at 50 years, whereas the expected lifeyears of vehicles and equipment ranged from 3\u201310 years depending on their frequency of replacement as indicated by staff. Further details on costing methods can be found in the appendix.\n\nMeasures of effectiveness We derived effectiveness measures from clinical out\u00ad comes reported in the TB-NEAT trial.14 We reported outcomes for each individual trial site and subsequently combined for all sites. We only included participants with a valid culture result in the analysis. The measures\n\nof effectiveness calculated in each study group (smear microscopy and Xpert) included the number of culturepositive tuberculosis cases: (1) diagnosed by the index test, (2) initiating antituberculosis treatment, (3) initi\u00ad ating antituberculosis treatment on the same day as diag\u00ad nosis, (4) completing antituberculosis treatment, and (5) having improved morbidity (measured by a numerical tubercu\u00adlosis score). Completing antituberculosis treat\u00ad ment refers to patients who completed a full 6-month course of antituberculosis treatment and excluded those who were not treatment-adherent, who had died, or who were lost to follow-up (participants started on anti\u00ad tuberculosis treatment who were not retained in the study). Improved morbidity refers to patients who started antituberculosis treatment and showed a 25% or more decrease in the well validated tuberculosis score at the end of treatment compared with baseline (see TB-NEAT14 and Wejse and colleagues29 for more details on tubercu\u00ad losis score determination). The TB-NEAT trial was not powered to examine differences in mortality between the two groups. As such, this measure was not included in our analysis. We also reported effectiveness measures 1\u20135 as a proportion of all individuals clinically suspected of having tuberculosis based on symptom screening. Out\u00adcomes were normalised to 1000 people with sus\u00ad pected tuberculosis screened in each study group. Incremental effectiveness (per 1000 people with sus\u00ad pected tuberculosis screened) was also reported. We used incre\u00admental costs and outcomes to calculate the incre\u00ad mental cost-effectiveness ratio (ICER) for selected out\u00ad comes among culture-positive cases. Further details regarding assumptions used in the analysis can be found in the appendix.\nSensitivity analysis We did univariate sensitivity analyses to calculate the effect of varying specific parameter inputs on the cost per test of point-of-care Xpert and the incremental cost per culture-positive patient starting treatment. We also did a probabilistic sensitivity analysis to calculate the un\u00ad certainty around ICERs given the challenges in esti\u00ad mating their confidence intervals.30 This analysis involves simultaneously varying cost and effectiveness parameter inputs with the use of 10\u2009000 randomly sampled estimates drawn from their defined probability distributions. We confirmed that 10\u2009000 simulations would be sufficient for model convergence around the uncertainty using a previously published approach.31 Briefly, we generated ICERs for each outcome using two separate sets of 10\u2009000 randomly sampled estimates to ensure that the mean values of each set of simulations fell within the 95% CI range of the corresponding set (appendix). We also doubled the number of simulations from 10\u2009000 to 20\u2009000 to confirm the width of the 95% CIs were effectively unchanged (appendix). We calcul\u00adated ICERs for each estimate and used them to construct a cost-effectiveness acceptability curve to establish the\n\ne801\n\nwww.thelancet.com/lancetgh Vol 7 June 2019\n\nArticles\n\nprobability that point-of-care Xpert would be considered cost-effective compared with smear microscopy over a range of willingness-to-pay thresholds.\nStatistical analysis was done using GraphPad Prism version 6.0 and Microsoft Excel 2016.\nRole of the funding source The funders of the study had no role in study design, data collection, data analysis, data interpretation, or the writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.\nResults\nThe cost per Xpert assay performed at the clinic (ie, point-of-care Xpert) and at a centralised laboratory (ie, Lab Xpert) was calculated at each study site to allow for cost com\u00adparisons at different levels of programmatic implemen\u00adtation. Component costs included con\u00adsum\u00ad ables, staff, equip\u00adment, overheads, quality control, and transport. Under observed conditions, the cost of pointof-care Xpert ranged from $24\u221974 in Zambia to $35\u221970 in Tanzania, whereas Lab Xpert costs ranged from $17\u221991 in South Africa to $30\u221959 in Zimbabwe (table 1). Across all sites, the weighted mean Lab Xpert unit cost was lower than for the point-of-care Xpert ($23\u221900 [95% CI 22\u221912\u201323\u221988] vs $28\u221903 [26\u221919\u201329\u221987]). Under observed conditions, Lab Xpert test volumes were 2\u20135 times higher than the point-of-care Xpert at any given testing facility. Point-of-care Xpert became less costly than Lab Xpert under the assumption that annual test volume estimates were equiva\u00adlent (figure 1; appendix). Unit test and clinic costs, including the cost breakdown, are reported in table 1.\nFigure 2A and 2B (appendix) show differences in clinical outcomes comparing point-of-care Xpert to smear mi\u00adcroscopy, both among culture-positive patients and symp\u00adtom\u00adatic patients. For example, when comparing point-of-care Xpert to smear microscopy, the difference in culture-confirmed tuberculosis cases starting anti\u00ad tuberculosis treatment was 24\u22193 cases per 1000 people screened (95% CI \u201320\u00b70 to 68\u00b75) and in those com\u00ad pleting antituberculosis treatment was 29\u22194 cases per 1000 people screened (\u20136\u00b79 to 65\u00b76). In terms of incre\u00ad mental costs, point-of-care Xpert was more expensive than smear microscopy, with costs ranging from $28\u2009503 per 1000 people screened in Zimbabwe to $53\u2009280 in Tanzania. Across all study sites, the incre\u00ad mental cost amounted to $35\u2009528 per 1000 people screened (27\u2009053\u201340\u2009024) and was strongly associated with the unit cost of point-of-care Xpert (figure 2C).\nThe cost-effectiveness of point-of-care Xpert for the indi\u00ad vidual and combined study sites is reported as the incre\u00admental cost per selected outcome among culture posi\u00ad tive tuberculosis cases (table 2). Cost-effectiveness esti\u00ad mates varied widely across study sites, with cost per treatment initiation ranging from $984 in South Africa to\n\nIncremental number of culture-positive patients (per 1000 symptomatic patients screened)\n\nA\nTanzania All sites\n300 250 200 150 100 50\n0 \u201350 \u2013100 \u2013150 \u2013200\n\nZambia\n\nZimbabwe\n\nSouth Africa\n\nIncremental number of symptomatic patients (per 1000 screened)\n\nB\n300 250 200 150 100 50\n0 \u201350 \u2013100 \u2013150 \u2013200 \u2013250\nC\n70 000 60 000 50 000 40 000 30 000 20 000 10 000\n0\n\nTreatment initiated\n\nSame day treatment initiated\n\nTreatment completed\n\nImproved morbidity\n\nXpert unit cost Doubled\n\nHalved\n\nIncremental costs per 1000 symptomatic patients screened (US$)\n\nFigure 2: Differences in clinical outcomes comparing point-of-care Xpert to smear microscopy in culture-positive patients and symptomatic patients Error bars indicate 95% CIs. All outcomes and costs are normalised to 1000 patients in each study group. (A) The estimated incremental outcomes for patients ultimately found to have culture-confirmed tuberculosis. (B) Corresponding outcomes in all patients. (C) The estimated incremental costs (US$, 2014); estimates are shown under the assumption that the unit cost of Xpert at all sites can be doubled or halved. Xpert=Xpert MTB/RIF.\n\n$2699 in Zambia (weighted mean of $1464 per treat\u00ad ment initiation) and cost per treatment completion ranging from $465 per treatment completed in Zambia to $8485 in South Africa (weighted mean $1211 per treatment com\u00ad pleted). We also compared our cost-effectiveness estimates to other estimates of tuberculosis inter\u00adventional strategies, in terms of disability-adjusted life-years (DALYs) averted,\n\nwww.thelancet.com/lancetgh Vol 7 June 2019\n\ne802\n\nArticles\n\nTanzania\n\nDiagnosed by index test $4254\n\nStarting treatment\n\n$1554\n\nStarting treatment on $1107 same day as diagnosis\n\nCompletingtreatment $521\n\nImproved morbidity\n\n$508\n\nZambia\nPoint-of-care Xpert\u2020 $2699\n$785\n\nZimbabwe SouthAfrica\n\n$1675 $1685 $399\n\n$1373 $984\n$46\n\nAll sites*\n$4186 $1464\n$561\n\n$465 $2024\n\n$4309 $1710\n\n$8485 $3101\n\n$1211 $1918\n\nCosts are in $US, 2014. *Weighted average of costs and outcomes across all sites, weighted by the number of patients screened and number of clinical outcomes observed at each site. \u2020Indicates that point-of-care Xpert was both more expensive and less effective than smear microscopy for that particular clinical outcome.\nTable 2: Incremental cost-effectiveness ratios, defined as the incremental cost per selected clinical outcome among culture positive cases\n\nfrom the published literature (appendix).24,33,34 In this com\u00ad parison, our baseline cost-effectiveness estimates, in most cases, decreased to less than 3 times the gross domestic product (GDP) per capita per DALY averted in each country.\nIn one-way sensitivity analysis, the expected useful life, purchase price, and annual test volume of the Gene Xpert machine had the greatest influence on the point-of-care Xpert unit cost (figure 3A). A similar pattern was obs\u00ad erved on the incremental cost per treatment initiation and treat\u00ad ment completion among culture-positive patients. How\u00ad ever, the largest influence on cost per treat\u00adment initiation was the proportion of culture-positive patients starting treatment. In the point-of-care Xpert group of the trial,\n\nA\nELY of GeneXpert machine (1\u20135) Number of Xpert MTB/RIF tests run annually (568\u20131136)\nCost of GeneXpert machine (8750\u201335 000) Cost of Xpert cartridge (4\u00b799\u201319\u00b796)\nNumber of GeneXpert modules replaced annually (0\u20134) Discount rate (1\u201310)\nTime (min) to perform Xpert (4\u00b756\u201318\u00b724) ELY of building infrastructure (25\u201375) 0\n\n$24\u00b773 $20\u00b763\n$23\u00b764 $24\u00b764 $27\u00b721 $28\u00b794\n\n$41\u00b759 $39\u00b761 $36\u00b789 $32\u00b730\n\n$29\u00b722 $30\u00b745 $29\u00b756 $29\u00b788\n\n20\n\n25\n\n30\n\n35\n\n40\n\n45\n\nCost per point-of-care Xpert assay (US$)\n\n$47\u00b761 50\n\n$54\u00b714\n\nHigh input value Low input value\n\n55\n\n60\n\nB\n\nProportion of culture-positive patients starting treatment (85\u2013100) $954\n\nELY of GeneXpert machine (1\u20135)\n\n$1257\n\n$2469\n\nNumber of Xpert tests run annually (568\u20131136) $1089\n\n$2200\n\nCost of GeneXpert machine (8750\u201335000)\n\n$1213\n\n$1952\n\nCost of Xpert cartridge (4\u00b799\u201319\u00b796)\n\n$1254\n\n$1870\n\nNumber of GeneXpert modules replaced annually (0\u20134)\n\n$1360\n\n$1758\n\nDiscount rate (1\u201310)\n\n$1437 $1542\n\nTime (min) to perform Xpert (4\u00b756\u201318\u00b724)\n\n$1442 $1493\n\nELY of building infrastructure (25\u201375)\n\n$1458 $1465\n\n$3561\n\n750 1000 1250 1500 1750 2000 2250 2500 2750 3000 3250 3500 3750 Incremental cost per treatment initiation among culture-positive patients (US$)\n\nC\n\nProportion of culture-positive patients completing treatment (50\u201380) $684 Proportion of culture-positive patients who default or LTF (15\u201340) $710 ELY of GeneXpert machine (1\u20135)\n\n$1040\n\n$2207 $2042\n\n$3445\n\nNumber of Xpert tests run annually (568\u20131136) Proportion of culture-positive patients starting treatment (80\u2013100)\nCost of GeneXpert machine (8750\u201335000) Cost of Xpert cartridge (4\u00b799\u201319\u00b796)\nNumber of GeneXpert modules replaced annually (0\u20134)\n\n$900 $1069\n$1003 $1037\n$1124\n\n$1819 $1954\n$1614 $1547 $1454\n\nDiscount rate (1\u201310) Time (min) to perform Xpert (4\u00b756\u201318\u00b724)\n\n$1186 $1285 $1193 $1235\n\nELY of building infrastructure (25\u201375)\n\n$1205 $1213\n\n500 750 1000 1250 1500 1750 2000 2250 2500 2750 3000 3250 3500 Incremental cost per treatment completion among culture-positive patients (US$)\n\nFigure 3: Univariate sensitivity analysis Tornado diagrams showing the effect of changing individual cost parameters on the cost (US$, 2014) per Xpert assay performed at the point of care (A), the incremental cost per treatment initiation (B), and the incremental cost per treatment completion among culture-positive patients across all sites (C). Low and high estimates of each input parameter are shown in parentheses on the left side of each figure. ELY=expected life years. LTF=lost to follow up. Xpert=Xpert MTB/RIF.\n\ne803\n\nwww.thelancet.com/lancetgh Vol 7 June 2019\n\nArticles\n\n91% of culture-positive patients started treatment. If this proportion fell to 85%, then the estimated incremental cost per treatment initiation increased from $1464 to $3561 (figure 3B). Similarly, reducing the proportion of culture-positive patients completing treatment from 60% to 50% increased the estimated cost per treatment completion from $1211 to $3445. A similar pattern was also observed if the number of patients who were lost to follow-up was varied (figure 3C).\nFigure 4 shows the willingness-to-pay thresholds at which point-of-care Xpert would be preferred over smear mi\u00adcroscopy across 10\u2009000 probabilistic simulations. Assuming that a 90% probability of cost-effectiveness might make an effective decision threshold, point-of-care Xpert would be preferred at this threshold in settings willing to pay $9450 per culture positive patient diag\u00ad nosed, $4450 per patient starting treatment, $1600 per patient starting treatment on the same day as diagnosis, $3820 per patient completing treatment, or $5840 per patient with improved morbidity. Willingness-to-pay thresholds at other cost-effectiveness probability esti\u00ad mates are shown in the appendix.\nDiscussion\nThe primary findings of this economic evaluation, nested within a clinical trial across four Southern African coun\u00adtries, is that point-of-care Xpert, while more costly than smear microscopy, is likely to be cost-effective (in >90% of simulations) in settings willing to pay at least $4500 per treatment initiation or $3800 per treatment completion among culture-positive patients. High test volume capacity at primary care clinics is probably necessary for Xpert point-of-care placement to be eco\u00ad nomically feasible in many settings.\nFew studies have estimated the economic effect of implementing Xpert at the point of care22,25,26 and all have used modelling analyses rather than prospectively obtained trial data to assess cost-effectiveness.19,20,24,25,35 In most of these modelling studies, Xpert was cost-effective either as a full replacement19,24,25,35 or in conjunction with other diag\u00ad nostic tests20,21 compared with the standard of care (smearmicroscopy). However, some of these studies did not incorporate empirical treatment and thus might have over\u00ad estimated the effectiveness of Xpert.24,25 Our estimates of cost-effectiveness account for levels of empirical treatment observed in the TB-NEAT trial and require fewer modelling assumptions than these previous analyses but might, there\u00adfore, underestimate the effectiveness of point-of-care Xpert by not explicitly accounting for effects on secondary transmission. Conversely, clinical trials can provide direct and real-world data on patient important outcomes related to Xpert implementation. For example, the XTEND trial15 assessed the effect of Xpert relative to smear microscopy on patient morbidity and mortality when placed at central laboratories within the context of the South Africa national Xpert rollout. Similar to TB-NEAT, no significant differ\u00ad ences were observed in terms of patient morbidity or\n\nProbability of cost-e\ufb00ectiveness\n\nProbability of cost-e\ufb00ectiveness\n\nA\n\nB\n\n1\u00b70 0\u00b79 0\u00b78 0\u00b77 0\u00b76 0\u00b75 0\u00b74 0\u00b73 0\u00b72 0\u00b71\n0 0\n\n$9450 2500 5000 7500 10000 12500 15000\n\n$4450 0 2500 5000 7500 10000 12500 15000\n\nCost per culture-positive patient diagnosed by index test (US$)\n\nCost per culture-positive patient starting treatment (US$)\n\nC\n\nD\n\n1\u00b70 0\u00b79 0\u00b78 0\u00b77 0\u00b76 0\u00b75 0\u00b74 0\u00b73 0\u00b72 0\u00b71\n0 0\n\n$1600 2500 5000 7500 10000 12500 15000\n\n$3820 0 2500 5000 7500 10000 12500 15000\n\nCost per culture-positive patient starting treatment on the same day as diagnosis (US$)\n\nCost per culture-positive patient completing treatment (US$)\n\nE\n1\u00b70 0\u00b79 0\u00b78 0\u00b77 0\u00b76 0\u00b75 0\u00b74 0\u00b73 0\u00b72 0\u00b71\n0 0\n\n$5840\n2500 5000 7500 10000 12500 15000 Cost per culture-positive patient with improved morbidity (US$)\n\nProbability of cost-e\ufb00ectiveness\n\nFigure 4: Cost-effectiveness acceptability curves for selected incremental cost-effectiveness ratios The probability of point-of-care Xpert, relative to smear microscopy, being cost-effective was plotted as a function of willingness to pay per culture-positive patient diagnosed by the index test (A), starting treatment (B), starting treatment on same day as diagnosis (C), completing treatment (D), and experiencing improved morbidity for all sites combined (E). The arrow indicates the willingness-to-pay threshold, at which the probability of cost-effectiveness is 90%. Xpert=Xpert MTB/RIF.\n\nmortality. A subsequent follow-up cost-effectiveness ana\u00ad lysis showed that the South African Xpert rollout was cost neutral and failed to improve the cost-effectiveness of tubercu\u00adlosis diagnosis, probably due to the trial finding no benefits to mortality.23 Conversely, in our study we used clin\u00adical endpoints from the TB-NEAT trial rather than mortality or health utility (eg, DALYs as more direct meas\u00ad ures of effectiveness). We estimated that, relative to smear microscopy, point-of-care Xpert is likely to cost $1464 per treatment initiation and $1211 per treatment completion among patients with culture-confirmed tubercu\u00adlosis.\nIn deciding whether point-of-care Xpert would be cost-effective, there is no consensus on appropriate\n\nwww.thelancet.com/lancetgh Vol 7 June 2019\n\ne804\n\nArticles\n\nwillingness-to-pay thresholds, and the use of generic thresholds have been largely criticised.36\u201338 Cost-effective\u00ad ness thresholds provide no information on affordability and do not have the disease-specific context in which scale-up of point-of-care Xpert could have major impli\u00ad cations on resource allocation within the NTP. Nonetheless, a commonly used metric for highly costeffective interventions is the per-capita GDP (or gross national income) per DALY averted (or 3 times this value for cost-effective interventions). To keep our analysis faithful to trial-measured outcomes (with a minimum of modelling assumptions), we did not measure outcomes in terms of DALYs averted. However, we did compare our results to those of other model-based economic evaluations from which a ratio could be calcu\u00adlated of clinical outcomes (as calculated in this study) to DALYs averted (appendix).24,33,34 Our estimates in these comp\u00ad arisons were substantially less favourable toward point-of-care Xpert (eg, incremental cost-effectiveness ratio 10 times higher than reported by Vassall and col\u00adleagues),24 probably because our estimates incorporate the levels of empirical treatment observed in the TB-NEAT trial. Nevertheless, our estimates of point-of-care Xpert cost-effectiveness in most cases come below a threshold of 3 times per-capita GDP per DALY averted, suggesting cost-effectiveness of this intervention according to this classical willingness-topay threshold. These results do not speak to the affordability of point-of-care Xpert under existing budget constraints, but they do suggest that the cost-effectiveness of point-of-care Xpert is likely to be at least equivalent to that of many other interventions that have been characterised as cost-effective in the scientific literature up to now.\nAnother useful metric to assess cost-effectiveness and willingness to pay is the potential health gains if provided with a fixed monetary sum. For example, if an additional $10\u2009000 was provided to the NTP for the Xpert point-of-care strategy, the expected gains over smear microscopy would be an additional four tuberculosis cases diagnosed, 12 cases starting treatment, 30 cases starting same-day treatment, and 14 cases completing treatment (appendix). Thus, policy makers can directly compare these values to other tuberculosis diagnostic strategies to assess the relative expected value of investment in point-of-care Xpert.\nThe TB-NEAT trial showed an increase in the number of Xpert-positive culture-negative individuals that were placed on treatment. Conversely, a Brazilian study35 showed a reduction in these Xpert false positives (with culture as a gold standard) compared with smear mi\u00ad croscopy. Such discrepancies between Xpert and culture could represent false-negative culture results but might also reflect false-positive Xpert results due to residual Mycobacterium tuberculosis DNA, particularly in people previously (and successfully) treated for tuberculosis.39 If some of these individuals who start treatment do indeed represent Xpert false-positives, it will be important to establish the extent of such Xpert-based overtreatment\n\nin various settings because the cost of such overtreatment ($113 per patient in our analysis) is not inconsequential.\nThe TB-NEAT study also showed much higher smear microscopy-based empirical treatment decisions compared with Xpert and probably explains why no incremental morbidity benefit was observed in the trial. To the extent that such empirical diagnoses represent people without underlying tuberculosis, the true cost-effectiveness of point-of-care Xpert might be even more favourable than reported here.\nThis study also speaks to the cost implications of placing Xpert at the point of care versus a centralised facility. In our analysis, test capacity was a major influence driving the unit cost of Xpert. However, despite annual test volumes being 2\u20135 times higher when Xpert was positioned in the laboratory, point-of-care Xpert was only slightly more expensive on a per-test basis in some settings due to reductions in sample transport and overhead costs (table 1). Additionally, the variation in Lab Xpert test costs across study settings reflects the different laboratory setups at each site. For example, more GeneXpert machines were in use at the centralised laboratory in Zambia accounting for the higher Lab Xpert costs at that site.\nIn most countries, Xpert has been positioned at subdistrict level laboratories within the NTP rather than at the peripheral level, probably due to the financial and logistical limitations of point-of-care placement. A 2011 South African study17 projected that national implemen\u00ad tation of Xpert at the point of care would cost 51% more than lab placement, equivalent to an estimated $36 million per year. This cost represents a major hurdle to point-ofcare placement, especially in other high burden countries where NTP budgets are under severe financial con\u00ad straints.40 However, these costs should be interpreted with\u00adin the overall context of the economic effect of tubercu\u00adlosis; one report41 estimated that, over the next 15 years, economic losses due to tuberculosis would amount to about $300 billion in the African region (equating to about 2\u20133% of the GDP in the case of some African countries, including South Africa) and close to $1 trillion globally. The cost implications on patients can also be substantial, especially in low-income countries where tuberculosis disease can consume close to 60% of an individual\u2019s income.42 Additional concerns of point-ofcare placement include the need for a stable electricity supply, temperature control, and adequate storage facil\u00ad ities.9,18 However, placement of Xpert at centralised facilities diminishes its potential to improve patient out\u00ad comes.9 One potential solution might involve targeting point-of-care Xpert placement at specific primary care facilities where cost and health benefits can be maximised. For example, point-of-care implementation of Xpert might first be prioritised to clinics (1) in tuberculosis hotspots\u2014ie, periurban slums where the disease burden is high, (2) where transportation of samples to central labora\u00adtory facilities is difficult and delays in result\n\ne805\n\nwww.thelancet.com/lancetgh Vol 7 June 2019\n\nArticles\n\nreporting are common, (3) where empirical treatment initi\u00adation is uncommon, and (4) where the incidence of drug resistance and rates of loss to follow up are high. Any implementation strategy will need to be assessed in the context of newer Xpert technologies, such as the more point-of-care-friendly Xpert Edge instrument (recently released and uses the more sensitive Xpert Ultra cart\u00ad ridge),43 and point-of-care molecular platforms in develop\u00ad ment (eg, QuantuMDx. etc).44\nOur study had several limitations. First, we did not account for rifampicin (RIF) resistance detection, a major advantage of Xpert, in our analysis. The parent trial was not powered for detection of drug resistance, and the additional effectiveness gained by RIF resistance de\u00ad tection is not directly comparable to smear microscopy without incorporating another method of drug resistance testing, such as line probe assay, which was not per\u00ad formed in the study. Incorporation of RIF resistance de\u00ad tection might make Xpert more cost-effective by reducing the time to treatment in positive cases but might also favour smear microscopy because of the possibility for false-positive diagnosis of RIF resistance by Xpert. Second, our results are difficult to directly compare with those that used single utility metrics, such as DALYs. However, we chose to compare costs using hard data and real-world clinical outcomes (obtained from several settings) rather than to estimate a measure (eg, DALYs), which requires extensive assumptions about the down\u00ad stream consequences of a diverse array of clinical out\u00ad comes based on sparse data. We also did not attempt to estimate effects on secondary transmission for similar reasons; thus, our findings, like for resistance detection, might be biased against point-of-care Xpert, which significantly shortened time to diagnosis in the trial. Third, economic evaluation within the context of a clin\u00ad ical trial has inherent limitations. Although able to provide direct data in specific settings compared with modelling, resource use and patient recruitment is often restricted to the selection criteria of the trial protocol.45 However, TB-NEAT was designed with pragmatic implemen\u00adtation in mind, which might mitigate this concern to some degree, and empirical cost data was collected in a standardised way from multiple highburden settings, which might not have been possible outside the context of a clinical trial. The cost of Lab Xpert used in our analysis (figure 1) was taken from the National Health Laboratory Service (NHLS) in South Africa but might be lower than if estimated with the use of empirically collected cost data.46 The NHLS Lab Xpert cost estimate was chosen because it represents the cost charged to the South African government and thus represents the actual cost incurred from the health-care provider perspective. Several other studies47\u201349 have used this estimate for similar reasons. Finally, some limi\u00ad tations were also related to measurement uncertainty of costs and outcomes. The large differences in incre\u00ad mental cost-effectiveness observed between the different\n\nstudy sites was primarily driven by differences in effective\u00adness measures. However, these differences should be interpreted with caution because of the low recruit\u00adment number at any given site (eg, wide 95% CIs were reported for patient outcomes in Tanzania) and that the TB-NEAT clinical trial was not powered to detect differences across the various study site.\nIn summary, we have estimated the cost-effectiveness of implementing Xpert at the point of care in four different African settings. Overall, our results indicate that a point-of-care-based Xpert can offer good value for money relative to other tuberculosis diagnostic strategies, though the cost-effectiveness of this strategy is likely to be even higher given that transmission reduction and drug resistance detection were not factored into the analysis. These findings will facilitate decision making about public health strategy and resource allocation by NTPs so that cost savings and health benefits can be maximised.\nContributors APo, GT, LZ, DC, PC, HS, MH, APy, JP, KD, and DD conceived and designed the study. APo, GT, LZ, DC, PC, HS, MH, APy, JP, KD, and DD implemented the study. APo, LZ, DC, PC, LM, FM, PL, and JM collected economic data. APo, GT, HS, and DD analysed the data. All authors interpretated the data and gave important intellectual input. APo, GT, KD, and DD wrote the first draft of the manuscript and all authors provided input on the initial and subsequent drafts of the manuscript.\nDeclaration of interests We declare no competing interests.\nAcknowledgments The study is part of the TB-NEAT study (Evaluation of multiple novel and emerging technologies for tuberculosis diagnosis, in smear-negative and HIV-infected persons, in high burden countries), which was funded by the European and Developing Countries Clinical Trials Partnership (EDCTP grant code: IP2009.32040.009). KD and the work presented here was also supported by the South African National Research Foundation, the South African MRC (RFA-EMU-02-2017), and the EDCTP (TMA-2015SF-1043 & TMA- 1051-TESAII). We thank the TB-NEAT study team and the clinical and administrative staff at the primary care facilities where data was collected at each participating study country.\nReferences 1 WHO. Global Tuberculosis Report 2016. Geneva: World Health\nOrganization, 2016. http://www.who.int/iris/handle/10665/250441 (accessed March 29, 2019). 2 Dheda K, Barry CE 3rd, Maartens G. Tuberculosis. Lancet 2016; 387: 1211\u201326. 3 Steingart K, Sohn H, Schiller I, et al. Xpert MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev 2013; 1: CD009593. 4 Tortoli E, Russo C, Piersimoni C, et al. Clinical validation of Xpert MTB/RIF for the diagnosis of extrapulmonary tuberculosis. Eur Respir J 2012; 40: 442\u201347. 5 Nicol MP, Workman L, Isaacs W, et al. Accuracy of the Xpert MTB/RIF test for the diagnosis of pulmonary tuberculosis in children admitted to hospital in Cape Town, South Africa: a descriptive study. Lancet Infect Dis 2011; 11: 819\u201324. 6 Zar HJ, Workman L, Isaacs W, Dheda K, Zemanay W, Nicol MP. Rapid diagnosis of pulmonary tuberculosis in African children in a primary care setting by use of Xpert MTB/RIF on respiratory specimens: a prospective study. Lancet Glob Health 2013; 1: e97\u2013104. 7 WHO. Automated real-time nucleic acid amplification technology for rapid and simultaneous detection of tuberculosis and rifampicin resistance: Xpert MTB/RIF System. Policy statement. Geneva: World Health Organization, 2011. https://www.ncbi.nlm. nih.gov/pubmed/26158191 (accessed March 29, 2019).\n\nwww.thelancet.com/lancetgh Vol 7 June 2019\n\ne806\n\nArticles\n\ne807\n\n8 WHO, The Stop TB Department. Update\u2014Implementation and roll-out of Xpert MTB/RIF May 2013. Geneva: World Health Organization, 2013. http://www.stoptb.org/wg/gli/assets/ documents/Xpert%20MTB-RIF%20UPDATE%20May%202013.pdf (accessed March 11, 2019).\n9 Lawn SD, Kerkhoff AD, Wood R. Location of Xpert MTB/RIF in centralised laboratories in South Africa undermines potential impact. Int J Tuberc Lung Dis 2012; 16: 701\u201302.\n10 Cohen GM, Drain PK, Noubary F, Cloete C, Bassett IV. Diagnostic delays and clinical decision making with centralized Xpert MTB/RIF testing in Durban, South Africa. J Acquir Immune Defic Syndr 2014; 67: e88\u201393.\n11 Boehme CC, Nicol MP, Nabeta P, et al. Feasibility, diagnostic accuracy, and effectiveness of decentralised use of the Xpert MTB/RIF test for diagnosis of tuberculosis and multidrug resistance: a multicentre implementation study. Lancet 2011; 377: 1495\u2013505.\n12 Botha E, Den Boon S, Verver S, et al. Initial default from tuberculosis treatment: how often does it happen and what are the reasons? Int J Tuberc Lung Dis 2008; 12: 820\u201323.\n13 Squire SB, Belaye AK, Kashoti A, et al. Lost smear-positive pulmonary tuberculosis cases: where are they and why did we lose them? Int J Tuberc Lung Dis 2005; 9: 25\u201331.\n14 Theron G, Zijenah L, Chanda D, et al. Feasibility, accuracy, and clinical effect of point-of-care Xpert MTB/RIF testing for tuberculosis in primary-care settings in Africa: a multicentre, randomised, controlled trial. Lancet 2014; 383: 424\u201335.\n15 Churchyard GJ, Stevens WS, Mametja LD, et al. Xpert MTB/RIF versus sputum microscopy as the initial diagnostic test for tuberculosis: a cluster-randomised trial embedded in South African roll-out of Xpert MTB/RIF. Lancet Glob Health 2015; 3: e450\u201357.\n16 Hanrahan CF, Selibas K, Deery CB, et al. Time to treatment and patient outcomes among TB suspects screened by a single point-of-care Xpert MTB/RIF at a primary care clinic in Johannesburg, South Africa. PLoS One 2013; 8: e65421.\n17 Schnippel K, Meyer-Rath G, Long L, et al. Scaling up Xpert MTB/RIF technology: the costs of laboratory- vs. clinic-based roll-out in South Africa. Trop Med Int Health 2012; 17: 1142\u201351.\n18 Trebucq A, Enarson DA, Chiang CY, et al. Xpert(R) MTB/RIF for national tuberculosis programmes in low-income countries: when, where and how? Int J Tuberc Lung Dis 2011; 15: 1567\u201372.\n19 Andrews JR, Lawn SD, Rusu C, et al. The cost-effectiveness of routine tuberculosis screening with Xpert MTB/RIF prior to initiation of antiretroviral therapy: a model-based analysis. AIDS 2012; 26: 987\u201395.\n20 Langley I, Lin HH, Egwaga S, et al. Assessment of the patient, health system, and population effects of Xpert MTB/RIF and alternative diagnostics for tuberculosis in Tanzania: an integrated modelling approach. Lancet Glob Health 2014; 2: e581\u201391.\n21 Shah M, Dowdy D, Joloba M, et al. Cost-effectiveness of novel algorithms for rapid diagnosis of tuberculosis in HIV-infected individuals in Uganda. AIDS 2013; 27: 2883\u201392.\n22 Van Rie A, Page-Shipp L, Hanrahan C, et al. Point-of-care Xpert MTB/RIF for smear-negative tuberculosis suspects at a primary care clinic in South Africa. Int J Tuberc Lung Dis 2013; 17: 368\u201372.\n23 Vassall A, Siapka M, Foster N, et al. Cost-effectiveness of Xpert MTB/RIF for tuberculosis diagnosis in South Africa: a real-world cost analysis and economic evaluation. Lancet Glob Health 2017; 5: e710\u201319.\n24 Vassall A, van Kampen S, Sohn H, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: a cost-effectiveness analysis. PLoS Med 2011; 8: e1001120.\n25 Zwerling AA, Sahu M, Ngwira LG, et al. Screening for tuberculosis among adults newly diagnosed with hiv in sub-Saharan Africa: a cost-effectiveness analysis. J Acquir Immune Defic Syndr 2015; 70: 83\u201390.\n26 Hsiang E, Little KM, Haguma P, et al. Higher cost of implementing Xpert((R)) MTB/RIF in Ugandan peripheral settings: implications for cost-effectiveness. Int J Tuberc Lung Dis 2016; 20: 1212\u201318.\n27 Sanders GD, Neumann PJ, Basu A, et al. Recommendations for conduct, methodological practices, and reporting of cost-effectiveness analyses: second panel on cost-effectiveness in health and medicine. JAMA 2016; 316: 1093\u2013103.\n\n28 iDSl. Reference case for economic evaluation. 2018. http://www. idsihealth.org/resource-items/idsi-reference-case-for-economicevaluation/ (acccessed March 11, 2019).\n29 Wejse C, Gustafson P, Nielsen J, et al. TBscore: Signs and symptoms from tuberculosis patients in a low-resource setting have predictive value and may be used to assess clinical course. Scand J Infect Dis 2008; 40: 111\u201320.\n30 Fenwick E, O\u2019Brien BJ, Briggs A. Cost-effectiveness acceptability curves\u2014facts, fallacies and frequently asked questions. Health Econ 2004; 13: 405\u201315.\n31 Karnon J, Vanni T. Calibrating models in economic evaluation: a comparison of alternative measures of goodness of fit, parameter search strategies and convergence criteria. Pharmacoeconomics 2011; 29: 51\u201362.\n32 Guillaume J, Bygrave H. Cost-effectiveness study of pre and post Xpert TB diagnosis. MSF Briefing Document. Medecins Sans Frontieres. https://www.msf.org.za/about-us/publications/briefingdocuments/cost-effective-study-pre-and-post-xpert-tb-diagnosis (accessed March 29, 2019).\n33 Yadav RP, Nishikiori N, Satha P, Eang MT, Lubell Y. Cost-effectiveness of a tuberculosis active case finding program targeting household and neighborhood contacts in Cambodia. Am J Trop Med Hyg 2014; 90: 866\u201372.\n34 Baltussen R, Floyd K, Dye C. Cost effectiveness analysis of strategies for tuberculosis control in developing countries. BMJ 2005; 331: 1364.\n35 Pinto M, Steffen RE, Cobelens F, van den Hof S, Entringer A, Trajman A. Cost-effectiveness of the Xpert(R) MTB/RIF assay for tuberculosis diagnosis in Brazil. Int J Tuberc Lung Dis 2016; 20: 611\u201318.\n36 Bertram MY, Lauer JA, De Joncheere K, et al. Cost-effectiveness thresholds: pros and cons. Bull World Health Organ 2016; 94: 925\u201330.\n37 Dowdy DW, Cattamanchi A, Steingart KR, Pai M. Is scale-up worth it? Challenges in economic analysis of diagnostic tests for tuberculosis. PLoS Med 2011; 8: e1001063.\n38 Marseille E, Larson B, Kazi DS, Kahn JG, Rosen S. Thresholds for the cost-effectiveness of interventions: alternative approaches. Bull World Health Organ 2015; 93: 118\u201324.\n39 Theron G, Venter R, Calligaro G, et al. Xpert MTB/RIF results in patients with previous tuberculosis: can we distinguish true from false positive results? Clin Infect Dis 2016; 62: 995\u20131001.\n40 Albert H, Nathavitharana RR, Isaacs C, Pai M, Denkinger CM, Boehme CC. Development, roll-out and impact of Xpert MTB/RIF for tuberculosis: what lessons have we learnt and how can we do better? Eur Respir J 2016; 48: 516\u201325.\n41 KPMG. The global economic impact of tuberculosis. October, 2017. https://big.assets.huffingtonpost.com/GlobalEconomicImpactTB.pdf (accessed March 11, 2019).\n42 Tanimura T, Jaramillo E, Weil D, Raviglione M, Lonnroth K. Financial burden for tuberculosis patients in low- and middle-income countries: a systematic review. Eur Respir J 2014; 43: 1763\u201375.\n43 Dorman SE, Schumacher SG, Alland D, et al. Xpert MTB/RIF Ultra for detection of Mycobacterium tuberculosis and rifampicin resistance: a prospective multicentre diagnostic accuracy study. Lancet Infect Dis 2018; 18: 76\u201384.\n44 UNITAID. Tuberculosis\u2014Diagnostics technology landscape, 5th Edition. May, 2017. https://unitaid.eu/assets/2017-Unitaid-TBDiagnostics-Technology-Landscape.pdf (accessed March 11, 2019).\n45 Ramsey SD, Willke RJ, Glick H, et al. Cost-effectiveness analysis alongside clinical trials II\u2014An ISPOR Good Research Practices Task Force report. Value Health 2015; 18: 161\u201372.\n46 Cunnama L, Sinanovic E, Ramma L, et al. Using top-down and bottom-up costing approaches in LMICs: the case for using both to assess the incremental costs of new technologies at scale. Health Econ 2016; 25 (suppl 1): 53\u201366.\n47 Meyer-Rath G, Schnippel K, Long L, et al. The impact and cost of scaling up GeneXpert MTB/RIF in South Africa. PLoS One 2012; 7: e36966.\n48 Cleary SM, McIntyre D, Boulle AM. The cost-effectiveness of antiretroviral treatment in Khayelitsha, South Africa\u2014a primary data analysis. Cost Eff Resour Alloc 2006; 4: 20.\n49 Pooran A, Pieterson E, Davids M, Theron G, Dheda K. What is the cost of diagnosis and management of drug resistant tuberculosis in South Africa? PLoS One 2013; 8: e54587.\n\nwww.thelancet.com/lancetgh Vol 7 June 2019\n\n\n",
    "authors": [
      "Anil Pooran",
      "Grant Theron",
      "Lynn Zijenah",
      "Duncan Chanda",
      "Petra Clowes",
      "Lawrence Mwenge",
      "Farirai Mutenherwa",
      "Paul Lecesse",
      "John Metcalfe",
      "Hojoon Sohn",
      "Michael Hoelscher",
      "Alex Pym",
      "Jonny Peter",
      "David Dowdy",
      "Keertan Dheda"
    ],
    "doi": "10.1016/S2214-109X(19)30164-0",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://linkinghub.elsevier.com/retrieve/pii/S2214109X19301640"
  },
  {
    "key": "WDAKQ9DZ",
    "title": "Pooled testing of sputum with Xpert MTB/RIF and Xpert Ultra during tuberculosis active case finding campaigns in Lao People\u2019s Democratic Republic",
    "abstract": "Introduction\u2002 Active case finding (ACF) of individuals with tuberculosis (TB) is a key intervention to find the 30% of people missed every year. However, ACF requires screening large numbers of individuals who have a low probability of positive results, typically <5%, which makes using the recommended molecular tests expensive.\nMethods\u2002 We conducted two ACF surveys (in 2020 and 2021) in high TB burden areas of Lao PDR. Participants were screened for TB symptoms and received a chest X-\u00adray. Sputum samples of four consecutive individuals were pooled and tested with Xpert Mycobacterium tuberculosis (MTB)/ rifampicin (RIF) (Xpert-\u00adMTB/RIF) (2020) or Xpert-\u00adUltra (2021). The agreement of the individual and pooled samples was compared and the reasons for discrepant results and potential cartridge savings were assessed.\nResults\u2002 Each survey included 436 participants, which were tested in 109 pools. In the Xpert-\u00adMTB/RIF survey, 25 (sensitivity 89%, 95%\u2009CI 72.8% to 96.3%) of 28 pools containing MTB-\u00adpositive samples tested positive and 81 pools containing only MTB-\u00adnegative samples tested negative (specificity 100%, 95%\u2009CI 95.5% to 100%). In the Xpert-\u00adUltra survey, all 32 (sensitivity 100%, 95%\u2009CI 89.3% to 100%) pools containing MTB-\u00adpositive samples tested positive and all 77 (specificity 100%, 95%\u2009CI 95.3% to 100%) containing only MTB-\u00adnegative samples tested negative. Pooling with Xpert-\u00ad MTB/RIF and Xpert-\u00adUltra saved 52% and 46% (227/436 and 199/436, respectively) of cartridge costs alone.\nConclusion\u2002 Testing single and pooled specimens had a high level of agreement, with complete concordance when using Xpert-\u00adUltra. Pooling samples could generate significant cartridge savings during ACF campaigns.",
    "full_text": "BMJ Glob Health: first published as 10.1136/bmjgh-2021-007592 on 14 February 2022. Downloaded from http://gh.bmj.com/ on August 20, 2024 by guest. Protected by copyright.\n\nOriginal research\nPooled testing of sputum with Xpert MTB/ RIF and Xpert Ultra during tuberculosis active case finding campaigns in Lao People\u2019s Democratic Republic\n\nVibol Iem\u2002 \u200d \u200d ,1,2 Phonenaly Chittamany,2 Sakhone Suthepmany,2\n\nSouvimone Siphanthong,2 Phitsada Siphanthong,2 Silaphet Somphavong,3\n\nKonstantina Kontogianni,1 James Dodd,1 Jahangir AM Khan,4 Jose Dominguez,5\n\nTom Wingfield,6,7 Jacob Creswell,8 Luis E Cuevas\u2002 \u200d\n\n9\n\u200d\n\nTo cite: Iem V, Chittamany P, Suthepmany S, et al. Pooled testing of sputum with Xpert MTB/RIF and Xpert Ultra during tuberculosis active case finding campaigns in Lao People\u2019s Democratic Republic. BMJ Global Health 2022;7:e007592. doi:10.1136/ bmjgh-2021-007592\nHandling editor Alberto L Garcia-B\u00ad asteiro JC and LEC are joint senior authors. Received 2 October 2021 Accepted 1 December 2021\n\u00a9 Author(s) (or their employer(s)) 2022. Re-\u00aduse permitted under CC BY-\u00adNC. No commercial re-\u00aduse. See rights and permissions. Published by BMJ. For numbered affiliations see end of article. Correspondence to Professor Luis E Cuevas; \u200bluis.\u200bcuevas@l\u200b stmed.\u200bac.u\u200b k\n\nABSTRACT Introduction\u2002 Active case finding (ACF) of individuals with tuberculosis (TB) is a key intervention to find the 30% of people missed every year. However, ACF requires screening large numbers of individuals who have a low probability of positive results, typically <5%, which makes using the recommended molecular tests expensive. Methods\u2002 We conducted two ACF surveys (in 2020 and 2021) in high TB burden areas of Lao PDR. Participants were screened for TB symptoms and received a chest X-\u00adray. Sputum samples of four consecutive individuals were pooled and tested with Xpert Mycobacterium tuberculosis (MTB)/ rifampicin (RIF) (Xpert-\u00adMTB/RIF) (2020) or Xpert-\u00adUltra (2021). The agreement of the individual and pooled samples was compared and the reasons for discrepant results and potential cartridge savings were assessed. Results\u2002 Each survey included 436 participants, which were tested in 109 pools. In the Xpert-\u00adMTB/RIF survey, 25 (sensitivity 89%, 95%\u2009CI 72.8% to 96.3%) of 28 pools containing MTB-\u00adpositive samples tested positive and 81 pools containing only MTB-\u00adnegative samples tested negative (specificity 100%, 95%\u2009CI 95.5% to 100%). In the Xpert-\u00adUltra survey, all 32 (sensitivity 100%, 95%\u2009CI 89.3% to 100%) pools containing MTB-\u00adpositive samples tested positive and all 77 (specificity 100%, 95%\u2009CI 95.3% to 100%) containing only MTB-\u00adnegative samples tested negative. Pooling with Xpert-\u00ad MTB/RIF and Xpert-\u00adUltra saved 52% and 46% (227/436 and 199/436, respectively) of cartridge costs alone. Conclusion\u2002 Testing single and pooled specimens had a high level of agreement, with complete concordance when using Xpert-\u00adUltra. Pooling samples could generate significant cartridge savings during ACF campaigns.\nINTRODUCTION Despite being treatable and curable, tuberculosis (TB) remains one of the main infectious killers in the world, as ten million people fall ill and 1.4\u2009million die from the disease each year.1 Its diagnosis is usually reliant on passive case finding (PCF), in which health services wait for individuals with symptoms of TB to attend a\n\nKey questions\nWhat is already known? \u25ba Our study adds to the emerging body of evidence\nthat the pooling methods for testing with molecular assays can improve the efficiency of testing for tuberculosis (TB), potentially enabling the screening and testing of larger numbers of people more cost-\u00adeffectively.\nWhat are the new findings? \u25ba These findings contribute to recognised gaps in\nfunding sources for the procurement of sufficient cartridges for testing all individuals with presumptive TB, which jeopardises access to high sensitivity WHO-r\u00adecommended rapid molecular diagnostic tests, such as the GeneXpert Xpert MTB/RIF and Xpert-\u00adUltra.\nWhat do the new findings imply? \u25ba The method has not been tested during a real health\ncrisis situation, such as the COVID-\u00ad19 pandemic. The study took place at a time laboratory resources were being diverted and healthcare workers were repurposed for SARS-\u00adCoV2 testing. The higher efficiency of the pooling method can contribute to cope with these challenging times.\nhealth facility to initiate the diagnostic process. Although PCF identifies most people with TB in locations with adequate access to health services, it misses those unwilling or unable to attend the clinics and is a major reason only seven of the ten million people with TB are diagnosed and notified.2 Individuals missed by passive approaches often include vulnerable populations such as internally displaced, migrant or rural populations, women, the unemployed and ethnic minorities,3 4 who may face multiple societal and economic barriers to attend the service, including catastrophic costs.5 6 It is,\n\nIem V, et al. BMJ Global Health 2022;7:e007592. doi:10.1136/bmjgh-2021-007592\n\n \n\n1\n\nBMJ Glob Health: first published as 10.1136/bmjgh-2021-007592 on 14 February 2022. Downloaded from http://gh.bmj.com/ on August 20, 2024 by guest. Protected by copyright.\n\nBMJ Global Health\nthus, recognised that, to be inclusive and reduce the socioeconomic impact of TB,7 health services need to include active case finding (ACF) approaches that involve proactive interventions to extend the reach of TB services for diagnosis8 and treatment.9 Although ACF interventions can be very effective,10 11 they are less standardised than PCF, as they address the specific barriers of multiple target populations and are more resource and time intensive than PCF.12\nThe WHO recommends testing all individuals with presumptive TB with molecular assays, such as the Xpert Mycobacterium tuberculosis/rifampicin (Xpert MTB/ RIF) and Xpert-U\u00ad ltra (Cepheid Sunnyvale, California),13 with the latter being preferred given its higher sensitivity.14 Although the use of these assays is expanding, the assay cartridge unit costs of US$ 9.98 per test15 remains one of the main hurdles for its wider implementation in low-\u00adincome and middle-i\u00adncome countries. Diagnostic test costs can limit the expansion of ACF activities, as they require testing large numbers of individuals with relatively lower yields than PCF.16\nSince 2015, the Lao National Tuberculosis Control Center (NTC) has conducted ACF by implementing intensified case finding activities to increase the detection of individuals with TB in high burden districts of the country. These activities include the sensitisation of the population, the local provision of chest X-\u00adrays for screening (independently of symptoms) and the identification of individuals with symptoms of TB who have not attended health facilities. Participants with abnormal chest X-\u00adrays or symptoms of TB are tested using Xpert MTB/RIF or Xpert-U\u00ad ltra.17 The activities have increased case detection, although the cost of the Xpert cartridges is considered high and is the main limiting factor to implement the intervention on a larger scale.\nOne approach that could increase the affordability of Xpert testing is to test several samples together using the pooling method.18 This procedure combines (or pools) the sputum of several individuals into one pot and tests them together with a single test. If the test is positive, the pool\u2019s samples are retested individually to identify the positive sample(s) while if the test is negative, all samples in the pool are considered negative, resulting in 30%\u201340% savings in Xpert cartridge costs alone depending on the prevalence of TB in the population tested.19 Therefore, pooling may hold great promise for ACF, but there are few reports of its performance under operational conditions.20\nHere, we report a prospective study to assess the sensitivity and specificity of the pooling method using Xpert MTB/RIF and Xpert-\u00adUltra during intensified case finding interventions, and its potential to increase the affordability of Xpert testing in Lao PDR.\nMETHODS We conducted two independent prospective surveys embedded within the ACF activities of Lao\u2019s NTC, from March to April 2020 and from January to March 2021. Both surveys were cross-\u00adsectional and used the same\n\nrecruitment and testing procedures. The 2020 survey aimed to assess the performance of the pooling method when testing samples with Xpert MTB/RIF, while the 2021 survey assessed the method when using Xpert-\u00adUltra, after its release for routine use by Lao\u2019s NTC.\nACF was conducted in Lao\u2019s high TB burden areas, which are programmatically defined as TB incidence \u2265100 cases per 100\u2009000 population. The 2020 survey was conducted in Vientiane Capital, Luang Prabang and Savannakhet provinces with estimated populations of 890,129, 468\u2009375 and 1\u2009051\u2009675 inhabitants, respectively, and TB notification rate of 134, 88 and 102 cases per 100\u2009000 population in 2020, respectively. The 2021 survey was conducted in Saravane and Oudomxay, with 430\u2009428 and 333\u2009934 population and TB notification rates of 127 and 110 cases per 100\u2009000 population in 2020, respectively.\nBoth surveys were conducted in the same fashion. Before an ACF activity, the NTC team met the province and district health authorities and conducted preparation visits with the provincial TB coordinator, district TB manager and village authorities, distributed health education materials, obtained the addresses of individuals with TB and line listed household contacts. At an agreed date, the NTC team set up a digital chest X-r\u00aday machine and a four-m\u00ad odule GeneXpert platform in the village and invited all residents to complete a questionnaire on signs and symptoms, history and treatment of TB and offered chest X-\u00adrays for screening, independently of the presence of symptoms. Individuals with abnormal chest X-\u00adrays and those who indicated having cough >2 weeks duration were asked to provide sputum samples for Xpert testing and were managed according to the decision tree shown in figure 1. Sputum samples were tested with Xpert following the manufacturer\u2019s instructions.21\nSputum samples tested individually with Xpert MTB/ RIF or Xpert-U\u00ad ltra were processed in the village GeneXpert platform. Consecutive samples with remnant volumes \u22650.5\u2009mL were included in the pooling studies and were transported to the National TB Reference Laboratory in Vientiane using a cold chain. Samples were transported after the sample reagent had been added. Turned around time to testing was <48\u2009hours after the sample reagents had been added and samples were maintained in a cold chain at all times. Sputum samples from four participants were pooled together, with a volume of 0.5\u2009mL of sputum each added to a pot, to obtain an aggregated volume of 2\u2009mL.21 Samples for a pool were selected consecutively and staff were blind to the individual Xpert test results and the pooled specimen was tested using one new Xpert cartridge.\nStatistical analysis Categorical data were summarised using descriptive statistics and \u03c72 tests were used to test for statistically significant differences. Individuals unable to produce sputum were excluded from the analysis. The pooled samples were compared with the four Xpert MTB/RIF and Xpert-U\u00ad ltra individual results and their agreement was tested using\n\n2\n\nIem V, et al. BMJ Global Health 2022;7:e007592. doi:10.1136/bmjgh-2021-007592\n\nBMJ Global Health\n\nBMJ Glob Health: first published as 10.1136/bmjgh-2021-007592 on 14 February 2022. Downloaded from http://gh.bmj.com/ on August 20, 2024 by guest. Protected by copyright.\n\nFigure 1\u2003 Flow diagram of the sputum processing.\n\nkappa statistics. The CT values and grades (trace, very low, low, medium and high) of individual and pooled tests were compared with describe the effect of combining the samples. Cost differences were calculated on the bases of the number of cartridges required to test all specimens using pooled and individual testing.\nSample size for the surveys was not formally estimated as we were limited by the expected number of participants attending the campaigns before the COVID-\u00ad19 lockdown, the capacity of staff to conduct additional testing to their routine activities and the number of spare cartridges available for research purposes.\nPatient and public involvement It was not appropriate or possible to involve patients or the public in the design, or conduct, or reporting or dissemination plans of our research.\nRESULTS The 2020 survey included 436 participants, 334 (76.6%) men and 102 (23.4%) women, and 29 (6.7%, 95%\u2009CI 4.7% to 9.4%) were Xpert MTB/RIF MTB-p\u00ad ositive. The 2021 survey also included 436 participants, 222 (50.9%) men and 214 (49.1%) women, and 37 (8.5%, 95%\u2009CI 6.5% to 11.5%) were Xpert-U\u00ad ltra MTB-\u00adpositive (p value >0.1, table 1). Men were more likely to be MTB-\u00adpositive than women in 2020 (26/334 (7.8%) men vs 3/102\n\n(2.9%) women, respectively, p=0.014); but women were more likely to be MTB-\u00adpositive than men in 2021 (12/222 (5.4%) vs 12/214 (11.7%), respectively, p<0.008). Each survey included 109 pools of four patients.\nXpert MTB/RIF survey In 2020, 28 (25.7%) pools contained one or more Xpert MTB/RIF MTB-p\u00ad ositive sample (27 pools with one and one pool with two MTB-p\u00ad ositive samples) and 81 (74.3%) pools contained solely MTB-n\u00ad egative samples (table 2). The pool with two MTB-p\u00ad ositive and 24 of 27 pools with one MTB-\u00adpositive sample tested MTB-p\u00ad ositive and three tested MTB-n\u00adegative, resulting in a sensitivity of 89% (25/28, 95%\u2009CI 72.8% to 96.3%). All 81 pools containing solely MTB-\u00adnegative samples tested MTB-n\u00ad egative in the pooled assay (specificity 100%, 95%\u2009CI 95.5% to 100%). Therefore, the accuracy performance of the 109 pools in correlation to the 436 individual results resulted in 97.3% agreement (kappa: 0.925). Among the 27 pools containing single MTB-p\u00ad ositive samples, five contained very low, 15 low, 6 medium and one high MTB grades. The pooled MTB grade was similar to the individual test in four (14.8%), one grade lower in 21 (77.8%), two grades lower in one (3.7 %) and one grade higher in one (3.7%) of the pools. Of the five pools containing very low individual MTB-\u00adgrades, three tested MTB-n\u00ad ot detected and two very low MTB-\u00adgrade in the pooled assay (table 3).\n\nIem V, et al. BMJ Global Health 2022;7:e007592. doi:10.1136/bmjgh-2021-007592\n\n3\n\nBMJ Glob Health: first published as 10.1136/bmjgh-2021-007592 on 14 February 2022. Downloaded from http://gh.bmj.com/ on August 20, 2024 by guest. Protected by copyright.\n\nBMJ Global Health\n\nTable 1\u2003 Baseline characteristics of participants and Xpert MTB/RIF and Xpert Ultra results\n\nXpert MTB/RIF\n\nXpert Ultra\n\nIndividual n (%)\n\nPool n (%)\n\nIndividual n (%)\n\nSex Male Female Age Mean (SD) (range) <35 35\u201354 \u226555 Xpert MTB result Detected/\u22651\u2009MTB included Not detected/\u22651\u2009MTB included Not detected/only MTB-\u00adnegative Xpert MTB result by sex Male Female MTB grade Trace Very low Low Medium High RIF resistance Detected Not detected Indeterminate\n\n436 334 (76.6) 102 (23.4) 436\n45 (16.1) (12-8\u00ad 9) 131 (30.0) 184 (42.2) 121 (27.8) 436\n29 (6.7) \u2013 407 (93.3)\n26/334 (7.8) 3/102 (2.9)\n29 NA\n6 (20.7) 16 (55.2)\n6 (20.7) 1 (3.4) 29 0 27 (93.1) 2 (6.9)\n\n\u2013 \u2013 \u2013 \u2013 \u2013 \u2013 \u2013 \u2013 109\n25/28 (22.9) 3 (2.7)\n81 (74.3)\n\u2013 \u2013\n25 NA\n15 (60.0) 9 (37.5) 1 (4.2) 0 (0.0)\n25 0\n24 (96.0) 1 (4.0)\n\n436 222 (50.9) 214 (49.1) 436\n54 (13.7) (10-\u00ad90) 41 (9.4) 159 (36.5) 236 (54.1) 436 37 (8.5) \u2013 399 (91.5)\n12/222 (5.4) 25/214 (11.7) 37\n6 (16.2) 8 (21.6) 15 (40.5) 2 (5.4) 6 (16.2) 37 0 30 (81.1) 7 (18.9)\n\nBold figures are frequencies and do not indicte statistical significance. Xpert MTB/RIF, Xpert Mycobacterium tuberculosis/rifampicin.\n\nPool n (%) \u2013 \u2013 \u2013 \u2013 \u2013 \u2013 \u2013 \u2013 109\n32/32 (29.4) \u2013\n77 (70.6)\n\u2013 \u2013\n32 21 (65.6) 11 (34.4)\n0 (0.0) 0 (0.0) 0 (0.0) 32 0 10 (31.2) 22 (68.8)\n\nThe CT values for the Xpert MTB/RIF probes for both individual and pooled testing are shown in table 4. The median CT values for probes A-E\u00ad ranged from 23.4 to 24.8 for the individual tests and from 30.6 to 33.6 for the\n\npooled tests, with an increase in CT values ranging from 5.4 to 7.1.\nTwo of the MTB-\u00ad positive samples were RIF-\u00ad indeterminate and 27 RIF-\u00adnegative. Of the 28 pools with\n\nTable 2\u2003 Number of pools with 0, 1, 2, 3, 4 positive results\n\nIndividual Xpert results included in a pool\n\nAll negative One positive Two positive\n\nn (%)\n\nn (%)\n\nn (%)\n\nThree positive n (%) Four positive n (%) All\n\nPooled Xpert MTB/RIF 81\n\n27\n\n1\n\n0\n\nDetected\n\n0\n\n24 (89%)\n\n1 (100%)\n\n0\n\nNot detected\n\n81 (100%) 3 (11%)\n\n0\n\n0\n\nPooled Xpert Ultra\n\n77\n\n27\n\n5\n\n0\n\nDetected\n\n0\n\n27 (100%) 5 (100%)\n\n0\n\nNot detected\n\n77 (100%)\n\n0\n\n0\n\n0\n\n0\n\n109\n\n0\n\n25 (23%)\n\n0\n\n84 (77%)\n\n0\n\n109\n\n0\n\n32 (29%)\n\n0\n\n77 (71%)\n\nBold figures are frequencies and do not indicte statistical significance. Xpert MTB/RIF, Xpert Mycobacterium tuberculosis/rifampicin.\n\n4\n\nIem V, et al. BMJ Global Health 2022;7:e007592. doi:10.1136/bmjgh-2021-007592\n\nBMJ Glob Health: first published as 10.1136/bmjgh-2021-007592 on 14 February 2022. Downloaded from http://gh.bmj.com/ on August 20, 2024 by guest. Protected by copyright.\n\nBMJ Global Health\n\nTable 3\u2003 Correlation of Individual and pooled Xpert MTB grades (positive pools only include those with only one positive Xpert)\n\nIndividual Xpert grade included in pool\n\nNot detected n (%)\n\nTrace n (%)\n\nVery low n (%)\n\nLow n (%)\n\nMedium n (%)\n\nHigh n (%)\n\nPooled Xpert MTB/RIF\n\n81\n\nNA\n\n5\n\n15\n\n6\n\n1\n\nNot detected\n\n81 (100%)\n\nNA\n\n3 (60%)\n\n0\n\n0\n\n0\n\nVery low\n\n0\n\nNA\n\n2 (40%)\n\n12 (80%)\n\n0\n\n0\n\nLow\n\n0\n\nNA\n\n0\n\n2 (13.3%) 6 (100%)\n\n1 (100%)\n\nMedium\n\n0\n\nNA\n\n0\n\n1 (6.7%)\n\n0\n\n0\n\nHigh\n\n0\n\nNA\n\n0\n\n0\n\n0\n\n0\n\nPooled Xpert Ultra\n\n77\n\n2\n\n5\n\n13\n\n1\n\n6\n\nNot detected\n\n77 (100 %)\n\n0\n\n0\n\n0\n\n0\n\n0\n\nTrace\n\n0\n\n2 (100%)\n\n4 (80%)\n\n7 (54%)\n\n1 (100%)\n\n2 (33%)\n\nVery low Low\n\n0\n\n0\n\n1 (20%)\n\n6 (46%)\n\n0\n\n4 (67%)\n\n0\n\n0\n\n0\n\n0\n\n0\n\n0\n\nMedium High\n\n0\n\n0\n\n0\n\n0\n\n0\n\n0\n\n0\n\n0\n\n0\n\n0\n\n0\n\n0\n\nBold values are frequencies and do not indicte statistical significance. Xpert MTB/RIF, Xpert Mycobacterium tuberculosis/rifampicin.\n\nMTB-\u00adpositive samples, 25 pools contained one MTB-\u00ad positive RIF-\u00adnegative sample, one had two MTB-p\u00ad ositive RIF-\u00adnegative samples and two had one MTB-\u00adpositive RIF-\u00adindeterminate samples. Of the 25 MTB-p\u00adositive RIF-\u00adnegative pools, three tested MTB-n\u00ad egative and did not report RIF results and 22 tested RIF-n\u00ad egative. The pool containing two MTB-p\u00ad ositive RIF-\u00adnegative samples tested RIF-n\u00ad egative and the two pools containing RIF-\u00ad indeterminate samples tested RIF-n\u00ad egative in one and RIF-\u00adindeterminate in the other.\n\nXpert-Ultra survey In 2021, 32 (29.4%) pools contained MTB-\u00adpositive samples and 77 (70.6%) solely MTB-\u00adnegative samples. Twenty-\u00ad seven of the 32 MTB-\u00adpositive pools contained one and five contained two MTB-\u00adpositive samples and all tested positive in the pooled assay (sensitivity 100%, 95%\u2009CI 89.3% to 100%). All 77 pools containing only MTB-\u00adnegative samples tested MTB-\u00adnegative (specificity 100%, 95%\u2009CI 95.3% to 100%), resulting in 100% agreement (Kappa: 1). Among the 27\n\nTable 4\u2003 Median CT values of individual and pooled Xpert MTB/RIF and Xpert Ultra probe results\n\nXpert MTB RIF\n\nIndividual results n=29\n\nPooled results n=25\n\nProbe\n\nCT median IQ range\n\nMin\u2013max\n\nCT median IQ range\n\nProbe D Probe C Probe E Probe B Probe A Xpert Ultra\nProbe Probe IS1081 Probe rpoB1 Probe rpoB2 Probe rpoB3 Probe rpoB4\n\n24.6 (22.7\u201327.3) 23.7 (22.0\u201326.7) 24.8 (23.0\u201328.1) 24.6 (22.9\u201327.3) 23.3 (22.0\u201326.0)\n\n19.1\u201335.2 18.9\u201334.7 20.5\u201336.2 20.1\u201333.8 20.5\u201334.1\n\nIndividual results n=37 CT median IQ range 19.6 (17.1\u201322.6) 21.6 (17.9\u201324.9) 21.3 (17.9\u201325.8) 23.3 (19.2\u201327.0) 25.7 (21.2\u201329.7)\n\nMin\u2013max 16.0\u201332.0\n0\u201332.0 0\u201332.1 0\u201333.7 0\u201335.7\n\n32.3 (28.5\u201334.2) 30.3 (27.4\u201331.8) 33.9 (29.3\u201334.9) 30.1 (27.3\u201332.5) 31.1 (27.1\u201332.9)\nPooled results n=32 CT median IQ range 24.9 (22.2\u201326.5)\n0 (0\u201330.2) 0 (0\u201329.8) 0 (0\u201332.9) 0 (0\u201333.5b)\n\nXpert MTB/RIF, Xpert Mycobacterium tuberculosis/rifampicin.\n\nMin\u2013max 21.0\u201338.2 20.0\u201335.7 21.2\u201339.3 20.9\u201334.7 19.7\u201334.6\nMin\u2013max 19.9\u201329.3\n0\u201334.9 0\u201335.3 0\u201339.8 0\u201337.7\n\nIem V, et al. BMJ Global Health 2022;7:e007592. doi:10.1136/bmjgh-2021-007592\n\n\u0394CT 7.1 6.5 7.1 5.4 6.6\n\u0394CT 4.6 NA NA NA NA\n5\n\nBMJ Glob Health: first published as 10.1136/bmjgh-2021-007592 on 14 February 2022. Downloaded from http://gh.bmj.com/ on August 20, 2024 by guest. Protected by copyright.\n\nBMJ Global Health\n\nTable 5\u2003 Costs and savings to screen consecutive patients using the pooling method and number of patients that could be tested with Xpert MTB/RIF and Xpert Ultra cartridges\n\nIndividual Xpert\n\nPooled Xpert\n\nMTB/RIF\n\nUltra\n\nMTB/RIF\n\nUltra\n\nNumber of individuals tested \u2003 Sensitivity\n\n436 Reference\n\n436 Reference\n\n436 89%*\n\n436 100%*\n\n\u2003 Specificity\n\nReference\n\nReference\n\n100%*\n\n100%*\n\n\u2003 Proportion positive\n\n6.7%\n\n8.5%\n\n22.9%\n\n29.4%\n\n\u2003 Bacteriologically confirmed\n\n29\n\n37\n\n26\n\n37\n\n\u2003 Cartridges required\n\n436\n\n436\n\n209\n\n237\n\n\u2003 Cartridge costs (USD)\n\n4351.28\n\n4351.28\n\n2085.82\n\n2365.26\n\n\u2003 Cartridge savings (USD)\n\nNA\n\nNA\n\n2265.46(52%)\n\n1986.02(46%)\n\nNumbers tested with 436 cartridges\n\n\u2003 Number tested\n\n436\n\n436\n\n909\n\n802\n\n\u2003 Cartridge cost per patient (USD)\n\n9.98\n\n9.98\n\n4.78\n\n5.42\n\n*Assumes pools of 1:4; proportion positive taken from the surveys\u2019 findings.\n\npools with single MTB-\u00adpositive samples, two contained trace,\n5 very low, 13 low, 1 medium and 6 high MTB grades. The pooled MTB grades were the same as the individual grades in three (11%), one grade lower in 10 (42%), two grades lower in seven (29%), three grades lower in five (21%) and four grades lower in two (8%) of the pooled assays (table 3). The Xpert-\u00adUltra probes CT values are shown in . Probe IS1081/IS6110 had median CT of 19.6 for individual and 24.9 for pooled results, with a median increase of 4.6. Probes rpoB1-\u00adB4 median CT values ranging from 19.6 to 25.7 for the individual tests, but CT values were not available for the pools. Among the 37 MTB-\u00adpositive samples, 30 (81%) were RIF-\u00adnegative and 7 (18.9%) were RIF-\u00adindeterminate and were distributed in 32 pools. Twenty-\u00adfive of the 32 pools contained only RIF-\u00adnegative and 7 contained RIF-\u00ad indeterminate samples. Fifteen of the 25 pools containing only RIF-\u00adnegative samples tested RIF-\u00adindeterminate and 10 RIF-\u00adnegative, while all seven pools containing RIF-\u00ad indeterminate samples tested pooled RIF-\u00adindeterminate.\nXpert MTB/RIF and Xpert Ultra costs The cartridges cost for testing individually the 436 participants with Xpert at US$9.98 per test was US$4351.28 for each survey, as shown in table 5. The pooling method in 2020 required 109 Xpert MTB/RIF cartridges to test 109 pools and 100 cartridges to test individual samples of 25 MTB-\u00adpositive pools. The total of 209 (109+100) cartridges for pool testing would cost US$2085.82, resulting in US$2265.46 (52%) saving in cartridge costs. Similarly, testing 109 pools with Xpert-U\u00ad ltra in 2021 required 109 cartridges to test the pools and 128 cartridges to test individually the 32 positive pools. The total of 237 cartridges would cost US$2365.26, resulting in US$1986.02 (46%) savings in cartridge costs. If the number of cartridges is kept fixed, the pooling method could test more patients than testing samples individually, as 436 cartridges would allow testing 909 and 802 individuals with Xpert MTB/\n\nRIF and Xpert-U\u00ad ltra, respectively\u2014an effective test per patient cost of \u200bUS$\u200b4.7\u200b 8.\u200band 5.42, respectively (table 5).\nDISCUSSION Our surveys compared pooling with single testing during ACF for TB in a low-\u00adincome country. Our results confirm that testing individual and pooled samples with the GeneXpert platform can achieve a high level of concordance. Concordance was higher with Xpert-\u00adUltra than with Xpert MTB/RIF, which is in agreement to regional studies evaluating pooling with Xpert-\u00adUltra in Cambodia20 and Vientiane, Lao PDR (Iem et al, in press). Discrepancies between individual and pooled Xpert MTB/RIF tests only occurred among pauci-b\u00ad acillary samples with high Xpert CT values, suggesting that some samples with low DNA concentrations fall below the assay\u2019s limit of detection and that the better agreement of Xpert-U\u00ad ltra is due to its higher sensitivity. Consequently, some patients with paucibacillary disease could be missed by pooling, especially if testing is based on Xpert MTB/RIF.\nThe pooling strategy can lead to significant cost savings and facilitate testing of more individuals for a given number of cartridges. In our setting, pooling samples would double the number of people tested with the same number of cartridges. This is higher than in PCF studies, where pooling is reported to save up to 40% of cartridges.19 Cartridge savings are a function of the proportion of people with MTB-\u00adpositive results and their distribution within the pools. If the proportion positive is low, a low number of pools would need to be retested, resulting in higher cartridge savings. For example, in a survey in Lao\u2019s district clinics, 12% of individuals tested Xpert-\u00adpositive, and pooling resulted in 38.3% and 41.7% cartridge saving costs with Xpert MTB/RIF and Xpert-\u00ad Ultra, respectively (Iem et al, in press), while in our survey setting, the proportion of positives was 8.5%, which led\n\n6\n\nIem V, et al. BMJ Global Health 2022;7:e007592. doi:10.1136/bmjgh-2021-007592\n\nBMJ Glob Health: first published as 10.1136/bmjgh-2021-007592 on 14 February 2022. Downloaded from http://gh.bmj.com/ on August 20, 2024 by guest. Protected by copyright.\n\nto higher savings. The proportion of participants with positive tests in ACF is often lower than reported from studies using PCF, typically below 5% depending on the target population,16 22 and lower to 10%\u201320% of individuals attending TB clinics in PCF.23 24 We have, thus, shown that pooling could be highly efficient when testing populations using ACF, and further studies among such populations are warranted. Since the pooling method is a laboratory change, it would not affect the screening algorithm and can be easily instituted without any major modifications.\nPrevious systematic reviews have highlighted that individual and pooled RIF results are often discordant, with pools containing RIF-n\u00ad egative samples often returning RIF-\u00adindeterminate pooled results,19 and our findings are in agreement with these observations. Although samples with pooled RIF results would be routinely confirmed at the time of retesting, the samples of a positive pool to identify the individual MTB-p\u00adositive samples, it is important to highlight that pooled RIF results are unreliable and should not be used for clinical management.\nFurther studies could explore ways to further improve the efficiency of pooling when combined with other screening tools, such as C-r\u00adeactive protein (CRP)25 and digital chest X-r\u00adays with Computer-\u00adaided diagnosis (CAD).26 27 Both tools can identify individuals with and without the traditional symptoms of TB, although their relatively lower specificity requires confirming the diagnosis with more specific molecular assays. Although using tests combinations could increase assay costs, individuals with a positive CRP or abnormal chest X-\u00adrays CAD could be confirmed using the pooling method, and its efficiency gains could increase the affordability of tests combinations.\nIn conclusion, we have shown pooling samples for TB diagnoses during ACF campaigns, which can replicate testing samples with individual tests. The approach can facilitate testing higher numbers of patients with lower cartridge costs, increasing the affordability of testing with molecular assays. The high level of agreement between individual and pooled samples obtained with Xpert-U\u00ad ltra demonstrates that pooling can be reliable and contribute to achieve the WHO End TB strategy targets in resource-\u00ad limited settings.\n\nBMJ Global Health\n9Clinical Sciences and Recsearch Centre for Drugs and Diagnostics, Liverpool School of Tropical Medicine, Liverpool, UK\nTwitter Jacob Creswell @Jacob_Creswell\nAcknowledgements\u2002 The authors would like to thank the National Tuberculosis Control Center (NTC) of Lao PDR for authorising the study and facilitating samples transportation from the field until the National TB Reference Laboratory (NTRL) in Vientiane. We are also grateful to the local health authorities that allowed access to the study sites and for their coordination locally. We are thankful to the active case finding team for their engagement, screening and enrolling patients, collecting, and shipping samples from the collection sites and to laboratory technicians of the NTRL for their dedication to carefully perform these additional laboratory procedures at the time of the COVID-\u00ad19 pandemic. We also thank Jim Read and the Global Health Trial Unit, LSTM for generating databases and data curation.\nContributors\u2002 The study was designed by LEC, JC, VI, KK and JAMK. Patients screening and enrolling, individual sample testing on Xpert MTB/RIF and Xpert Ultra were conducted by PC, SaS, SoS, PS, and SiS in the field during the ACF campaign. VI performed the pooled testing. Data collection and analysis were conducted by VI, KK, LEC, JD and JC. VI, JD, TW, JC and LEC wrote the first draft of the manuscript. All authors made substantial contributions to the writing and editing of the manuscript. LEC is responsible for the overall content and is the guarantor of the study The final version has been read and approved by all named authors.\nFunding\u2002 This research was funded in part by a TB REACH, Stop TB Partnership grant supported by Global Affairs Canada (STBP/TBREACH/GSA/2020-\u00ad04, awarded to LEC), the Global Fund to Fight AIDS, Tuberculosis and Malaria (LAO-\u00adT-\u00adGFMOH grant) and the National Tuberculosis Control Program budget from the Ministry of Health of Lao PDR. TW is supported by grants from: the Wellcome Trust, UK (209075/Z/17/Z); the Medical Research Council, Department for International Development, and Wellcome Trust, UK (Joint Global Health Trials, MR/V004832/1), the Medical Research Council, UK (MR/V028618/1); the Academy of Medical Sciences, UK; and the Swedish Health Research Council, Sweden.\nCompeting interests\u2002 The authors have no conflicts of interest to declare.\nPatient and public involvement\u2002 Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.\nPatient consent for publication\u2002 Not applicable.\nEthics approval\u2002 The study was approved by the Lao NTC and the Liverpool School of Tropical Medicine Research Ethics Committee, UK (Ethical waiver 20\u2013037), and informed consent waiver was obtained.\nProvenance and peer review\u2002 Not commissioned; externally peer reviewed.\nData availability statement\u2002 Data are available upon reasonable request. Not applicable.\nOpen access\u2002 This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-\u00adNC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-c\u00ad ommercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-\u00adcommercial. See: http://creativecommons.org/licenses/by-nc/4.0/.\nORCID iDs Vibol Iem http://orcid.org/0000-0001-6155-1402 Luis E Cuevas http://orcid.org/0000-0002-6581-0587\n\nAuthor affiliations 1Clinical Sciences, Liverpool School of Tropical Medicine, Liverpool, UK 2National Tuberculosis Control Center, XJ7F+P5F, Vientiane, Lao People's\nDemocratic Republic 3Center of Infectology Lao Christophe Merieux, Vientiane, Lao People's Democratic\nRepublic 4Health Economics and Policy Unit, School of Public Health and Community\nMedicine, University of Gothenburg, Gothenburg, Sweden 5Institut d'Investigaci\u00f3 Germans Trias i Pujol, CIBER Enfermedades Respiratorias,\nand Universitat Aut\u00f2noma de Barcelona, Barcelona, Spain 6Department of International Public Health and Clinical Sciences, Liverpool School\nof Tropical Medicine, Liverpool, UK 7Department of Global Public Health, WHO Collaborating Centre for Tuberculosis\nand Social Medicine, Karolinska Institutet, Stockholm, Sweden 8Stop TB Partnership, Geneva, Switzerland\n\nREFERENCES\n1 World Health Organization. Global tuberculosis report; 2020. 2020. 2 Shewade HD, Gupta V, Satyanarayana S, et al. Patient\ncharacteristics, health seeking and delays among new sputum smear positive TB patients identified through active case finding when compared to passive case finding in India. PLoS One 2019;14:e0213345. 3 World Health Organization. Who consolidated guidelines on tuberculosis. Module 2: screening \u2013 systematic screening for tuberculosis disease 2021. 4 de Vries SG, Cremers AL, Heuvelings CC, et al. Barriers and facilitators to the uptake of tuberculosis diagnostic and treatment services by hard-\u00adto-\u00adreach populations in countries of low and medium tuberculosis incidence: a systematic review of qualitative literature. Lancet Infect Dis 2017;17:e128\u201343. Electronic).\n\nIem V, et al. BMJ Global Health 2022;7:e007592. doi:10.1136/bmjgh-2021-007592\n\n7\n\nBMJ Glob Health: first published as 10.1136/bmjgh-2021-007592 on 14 February 2022. Downloaded from http://gh.bmj.com/ on August 20, 2024 by guest. Protected by copyright.\n\nBMJ Global Health\n5 Wingfield T, Boccia D, Tovar M, et al. Defining catastrophic costs and comparing their importance for adverse tuberculosis outcome with multi-\u00addrug resistance: a prospective cohort study, Peru. PLoS Med 2014;11:e1001675.\n6 Chittamany P, Yamanaka T, Suthepmany S, et al. First national tuberculosis patient cost survey in Lao people's Democratic Republic: assessment of the financial burden faced by TB-\u00adaffected households and the comparisons by drug-\u00adresistance and HIV status. PLoS One 2020;15:e0241862.\n7 Gurung SC, Dixit K, Rai B, et al. The role of active case finding in reducing patient incurred catastrophic costs for tuberculosis in Nepal. Infect Dis Poverty 2019;8:99.\n8 Ho J, Fox GJ, Marais BJ. Passive case finding for tuberculosis is not enough. Int J Mycobacteriol 2016;5:374\u20138.\n9 Chan G, Triasih R, Nababan B, et al. Adapting active case-\u00adfinding for TB during the COVID-\u00ad19 pandemic in Yogyakarta, Indonesia. Public Health Action 2021;11:41\u20139.\n10 Mhimbira FA, Cuevas LE, Dacombe R. Interventions to increase tuberculosis case detection at primary healthcare or community\u2010 level services. Cochrane Database of Systematic Reviews 2017;11.\n11 Arshad A, Salam RA, Lassi ZS, et al. Community based interventions for the prevention and control of tuberculosis. Infect Dis Poverty 2014;3:27.\n12 Datiko DG, Yassin MA, Theobald SJ, et al. Health extension workers improve tuberculosis case finding and treatment outcome in Ethiopia: a large-\u00adscale implementation study. BMJ Glob Health 2017;2:e000390\u20137908.\n13 Chakravorty S, Simmons AM, Rowneki M, et al. The New Xpert MTB/RIF Ultra: Improving Detection of Mycobacterium tuberculosis and Resistance to Rifampin in an Assay Suitable for Point-\u00adof-C\u00ad are Testing. mBio 2017;8. doi:10.1128/mBio.00812-17. [Epub ahead of print: 29 08 2017].\n14 World Health Organization. Who meeting report of a technical expert consultation: non-\u00adinferiority analysis of Xpert MTF/RIF ultra compared to Xpert MTB/RIF 2017.\n15 Foundation for Innovative New Diagnostics. Genexpert negotiated prices 2021.\n16 Gurung SC, Dixit K, Rai B, et al. Comparative yield of tuberculosis during active case finding using GeneXpert or smear microscopy for diagnostic testing in Nepal: a cross-\u00adsectional study. Trop Med Infect\n\nDis 2021;6. doi:10.3390/tropicalmed6020050. [Epub ahead of print: 14 Apr 2021]. 17 Dorman SE, Schumacher SG, Alland D, et al. Xpert MTB/RIF ultra for detection of Mycobacterium tuberculosis and rifampicin resistance: a prospective multicentre diagnostic accuracy study. Lancet Infect Dis 2018;18:76\u201384. 18 Abdurrahman ST, Mbanaso O, Lawson L, et al. Testing pooled sputum with Xpert MTB/RIF for diagnosis of pulmonary tuberculosis to increase affordability in low-\u00adincome countries. J Clin Microbiol 2015;53:2502\u20138. 19 Cuevas LE, Santos VS, Lima SVMA, et al. Systematic review of pooling sputum as an efficient method for Xpert MTB/RIF tuberculosis testing during the COVID-1\u00ad 9 pandemic. Emerg Infect Dis 2021;27:719\u201327. 20 Chry M, Smelyanskaya M, Ky M, et al. Can the high sensitivity of Xpert MTB/RIF ultra be harnessed to save cartridge costs? results from a pooled sputum evaluation in Cambodia. Trop Med Infect Dis 2020;5. doi:10.3390/tropicalmed5010027. [Epub ahead of print: 15 Feb 2020]. 21 Lawn SD, Nicol MP, Xpert NMP. Xpert\u00ae MTB/RIF assay: development, evaluation and implementation of a new rapid molecular diagnostic for tuberculosis and rifampicin resistance. Future Microbiol 2011;6:1067\u201382. 22 Yassin MA, Datiko DG, Tulloch O, et al. Innovative community-\u00ad based approaches doubled tuberculosis case notification and improve treatment outcome in southern Ethiopia. PLoS One 2013;8:e63174\u20136203. Electronic). 23 National Tuberculosis Control Center of Lao PDR. Lao PDR district health information software 2 (DHIS2) 2020. 24 National tuberculosis and leprosy control programme of Nigeria. 2019 Annual TB Report;2019. 25 Yoon C, Chaisson LH, Patel SM, et al. Diagnostic accuracy of C-\u00ad reactive protein for active pulmonary tuberculosis: a meta-\u00adanalysis. Int J Tuberc Lung Dis 2017;21:1013\u20139. 26 Qin ZZ, Sander MS, Rai B, et al. Using artificial intelligence to read chest radiographs for tuberculosis detection: a multi-\u00adsite evaluation of the diagnostic accuracy of three deep learning systems. Sci Rep 2019;9:15000. 27 MacPherson P, Webb EL, Kamchedzera W, et al. Computer-\u00adAided X-r\u00aday screening for tuberculosis and HIV testing among adults with cough in Malawi (the prospect study): a randomised trial and cost-\u00ad effectiveness analysis. PLoS Med 2021;18:e1003752.\n\n8\n\nIem V, et al. BMJ Global Health 2022;7:e007592. doi:10.1136/bmjgh-2021-007592\n\n\n",
    "authors": [
      "Vibol Iem",
      "Phonenaly Chittamany",
      "Sakhone Suthepmany",
      "Souvimone Siphanthong",
      "Phitsada Siphanthong",
      "Silaphet Somphavong",
      "Konstantina Kontogianni",
      "James Dodd",
      "Jahangir Am Khan",
      "Jose Dominguez",
      "Tom Wingfield",
      "Jacob Creswell",
      "Luis E Cuevas"
    ],
    "doi": "10.1136/bmjgh-2021-007592",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://gh.bmj.com/lookup/doi/10.1136/bmjgh-2021-007592"
  },
  {
    "key": "TC3M828E",
    "title": "Cost\u2013benefit analysis of Xpert MTB/RIF for tuberculosis suspects in German hospitals",
    "abstract": "Our objective was to assess the cost\u2013benefit of enhancing or replacing the conventional sputum smear with the real-time PCR Xpert MTB/RIF method in the inpatient diagnostic schema for tuberculosis (TB).",
    "full_text": "ORIGINAL ARTICLE TUBERCULOSIS\nCost\u2013benefit analysis of Xpert MTB/RIF for tuberculosis suspects in German hospitals\nRoland Diel1, Albert Nienhaus2, Doris Hillemann3 and Elvira Richter3,4\nAffiliations: 1Institute for Epidemiology, University Medical Hospital Schleswig-Holstein, Kiel, Airway Research Center North (ARCN), Member of the German Center for Lung Research, Germany. 2Institute for Health Services Research in Dermatology and Nursing, University Medical Center Hamburg-Eppendorf, Hamburg, Germany. 3National Reference Center for Mycobacteria, Borstel, Germany. 4Labor Limbach, Heidelberg, Germany.\nCorrespondence: Roland Diel, Institute for Epidemiology, University Medical Hospital Schleswig-Holstein, Niemannsweg 11, 24015 Kiel, Germany. E-mail: roland.diel@epi.uni-kiel.de\nABSTRACT Our objective was to assess the cost\u2013benefit of enhancing or replacing the conventional sputum smear with the real-time PCR Xpert MTB/RIF method in the inpatient diagnostic schema for tuberculosis (TB).\nRecent data from published per-case cost studies for TB/multidrug-resistant (MDR)-TB and from comparative analyses of sputum microscopy, mycobacterial culture, Xpert MTB/RIF and drug susceptibility testing, performed at the German National Reference Center for Mycobacteria, were used. Potential cost savings of Xpert MTB/RIF, based on test accuracy and multiple cost drivers, were calculated for diagnosing TB/MDR-TB suspects from the hospital perspective.\nImplementing Xpert MTB/RIF as an add-on in smear-positive and smear-negative TB suspects saves on average \u20ac48.72 and \u20ac503, respectively, per admitted patient as compared with the conventional approach. In smear-positive and smear-negative MDR-TB suspects, cost savings amount to \u20ac189.56 and \u20ac515.25 per person, respectively. Full replacement of microscopy by Xpert MTB/RIF saves \u20ac449.98. In probabilistic Monte-Carlo simulation, adding Xpert MTB/RIF is less costly in 46.4% and 76.2% of smear-positive TB and MDR-TB suspects, respectively, but 100% less expensive in all smear-negative suspects. Full replacement by Xpert MTB/RIF is also consistently cost-saving.\nUsing Xpert MTB/RIF as an add-on to and even as a replacement for sputum smear examination may significantly reduce expenditures in TB suspects.\n@ERSpublications Enhancing/replacing conventional sputum smear testing in TB with Xpert MTB/RIF may result in cost savings http://ow.ly/Stapk\n\nThis article has supplementary material available from erj.ersjournals.com Received: Aug 11 2015 | Accepted after revision: Sept 12 2015 | First published online: Dec 03 2015 Conflict of interest: Disclosures can be found alongside the online version of this article at erj.ersjournals.com Copyright \u00a9ERS 2016\n\nEur Respir J 2016; 47: 575\u2013587 | DOI: 10.1183/13993003.01333-2015\n\n575\n\nTUBERCULOSIS | R. DIEL ET AL.\n\nIntroduction\nAlthough tuberculosis (TB) incidence rates are steadily and globally in decline, multidrug-resistant (MDR)-TB, defined as simultaneous resistance to at least isoniazid and rifampicin (RIF), remains a major public health challenge in the World Health Organization (WHO) European region, where, in 2012, the prevalence of MDR-TB among 90 127 new TB cases was 15% [1]. This trend of increasing MDR-TB prevalence has important health economics implications: 1) it strengthens the long-established requirement of national guidelines that all patients admitted to healthcare facilities with suspected TB be maintained in costly respiratory isolation until it can be assumed that they are no longer contagious [2], and 2) it increases the need for emphasis to be placed on defining and speedily implementing case-appropriate treatment [3, 4].\nThus, the rapid diagnosis of TB disease and determination of drug resistance profiles is essential not only for early treatment and the associated prevention of TB transmission, but also highly relevant to the management of scarce economic resources, Since January 1, 2004, hospital costs in Germany have been based on the German Diagnosis Related Group (G-DRG) system, which assigns each TB case to one of two categories (E76B or E76C), depending on the severity of comorbidities. This imposes a fixed \u201cbase rate\u201d of payment for 13 days treatment; if hospital treatment of >13 days is required (category E76A), the statutory health insurances pay locally negotiated daily rates. In most cases, these fall below the average daily reimbursement for the first 14 days. Accordingly, for TB patients under statutory health insurance, hospitals do well to keep the total number of patients treated high, but to keep the duration of their hospital stays as short as possible [5, 6].\nThe Xpert MTB/RIF test (Cepheid, Sunnyvale, CA, USA) is a real-time PCR assay for simultaneous detection of Mycobacterium tuberculosis complex and of mutations in the rpoB gene that are associated with resistance to RIF as a proxy for MDR-TB from clinical samples. Results are obtained in 2 h [7]. After having recommended the use of Xpert MTB/RIF in December 2010, the WHO updated its guidance in October 2013, suggesting that the use of Xpert MTB/RIF to diagnose pulmonary TB, paediatric TB, extrapulmonary TB and RIF resistance [8] should be considered. Most recently, the US Food and Drug Administration suggested removing patients with suspected pulmonary TB from airborne infection isolation units after one or two negative Xpert MTB/RIF results. This guidance was based on an in-house clinical validation study that demonstrated negative predictive values (NPVs) of 99.7% for a single negative acid-fast bacilli (AFB) smear and 100% for two consecutive negative Xpert MTB/RIF results [9].\nAs data on the economic impact of implementing Xpert MTB/RIF in low-incidence countries are sparse, we undertook to assess the consequences of routine use of Xpert MTB/RIF with respect to confirmation or exclusion of TB disease and the timing of TB treatment in German TB wards. It is to this setting that the overwhelming majority (78.6%) of subsequently diagnosed TB cases are primarily admitted [10]. Our aim was to clarify the possible advantages of Xpert MTB/RIF, either performed at the hospital itself or in easy-to-reach local laboratories, as an add-on to conventional smear examinations or alternatively by replacing serial sputum smear microscopy with single-sample Xpert MTB/RIF examination. For our calculations, we used previously unpublished data of the German National Reference Center (NRC) for Mycobacteria in Borstel. Our model was parameterised to receive data on sensitivity, specificity, positive predictive value (PPV) and NPV of Xpert MTB/RIF testing, and to use sputum culture as the reference method. The data used were from untreated TB suspects whose sputa were sent to the NRC between January 1, 2012 and December 31, 2013.\nAt the NRC, Ziehl\u2013Neelsen microscopy and Xpert MTB/RIF had been performed and one liquid culture (BACTEC MGIT 960; BD, Franklin Lakes, NJ, USA) as well as two solid cultures (L\u00f6wenstein\u2013Jensen and Stonebrink TB Medium from NRC production) had been started on the same day. Stratified by smear status, the time (in days) until the first of the cultures became positive and until drug susceptibility testing (DST) results were available was assessed and compared with Xpert MTB/RIF results of the respective patients; negative cultures were uniformly read after 56 days. Culture results were considered as the \u201cgold standard\u201d to which Xpert MTB/RIF results were compared.\nMaterial and methods\nEthical considerations Ethical approval was not necessary as only fully anonymised secondary data were used.\nModel approach The economic analysis included the incremental costs of operating expenditures in diagnosing and treating pulmonary TB suspects for three different strategies. The perspective taken was that of the hospitals themselves, i.e. from admission through to patient discharge. Two Xpert MTB/RIF add-on algorithms and one Xpert MTB/RIF-only algorithm were studied. In the first Xpert MTB/RIF add-on algorithm, all sputum smear-positive individuals were tested with Xpert MTB/RIF on a single sputum specimen and in\n\n576\n\nDOI: 10.1183/13993003.01333-2015\n\nTUBERCULOSIS | R. DIEL ET AL.\n\nthe Xpert MTB/RIF add-on algorithm only smear-negative TB suspects were tested with Xpert MTB/RIF. In the Xpert MTB/RIF-only algorithm, a single sputum specimen was collected for performing Xpert MTB/RIF, followed by a culture; microscopy was not performed.\nAs hospital costs of the add-on algorithms were calculated separately for each of the two patient groups, i.e. those thought to have fully susceptible TB and those suspected of having MDR-TB, five scenarios were considered (figures 1\u20135; refer also to online supplementary material).\nFull descriptions of the various diagnostic and treatment assumptions of the model are provided in the online supplementary material.\nModel structure A deterministic, patient-based decision-analytic model was developed simulating the costs of the three approaches as described above for adult German TB and MDR-TB suspects based on German country-specific modalities. We used TreeAge software (TreeAge, Williamstown MA, USA) for model building and analysis. Univariate sensitivity analysis was performed using all variables (with some noted exceptions) to examine the extent to which our calculations were affected by varying selected assumptions. It also revealed the relative importance of the individual variables in each of the five different scenarios. Variation was done at random using either 1) the lower and upper bounds of a parameter\u2019s standard deviation or 2) those of its confidence interval. Where these were not applicable, our model simply caused parameter values to vary by \u00b120% according to international practice, unless otherwise stated. Furthermore, in order to capture the interactions between multiple inputs we provided a probabilistic sensitivity analysis (PSA) by assigning an appropriate statistical ( probability) distribution for all parameters which were randomly drawn in a second-order Monte-Carlo simulation (n=1000). All costs are reported in 2013 Euros (\u20ac).\nAs a result of a lack of valid data, we have not included in our model the effects of transmission by TB or MDR-TB patients to co-patients or healthcare workers.\nInput parameters are shown together with their probabilistic distributions in table 3.\nModel input Laboratory parameters Laboratory results for Xpert MTB/RIF compared with sputum smears (NRC) A total of 707 sputa from untreated TB suspects were investigated in the NRC in 2012/2013 with the Xpert MTB/RIF test. Prevalence of TB in that collective proved to be 19.66% (95% CI 16.91\u201322.94).\n\nSputum+\n\nXpert MTB/RIF\n\nTB Culture+\n\nTB+\n\nPPV_Xpert_SP\n\nXpert MTB/RIF+\n\n0.9474\n\nTB Culture\u2013\n\nTB\u2013\n\n#\n\nTB Culture+\n\nTB+\n\n#\nXpert MTB/RIF\u2013\n\n#\n\nTB Culture\u2013\n\n(NTM Culture+)\n\nTB\u2013\n\nNPV_Xpert_SP\n\nUnnecessary isolation and unnecessary \ufb01rst-line treatment pending through discharge\nUnnecessary diagnostic work-up pending positive culture result\nUnnecessary isolation (1 day)\n\nNo Xpert MTB/RIF\n\nTB Culture+\n\nTB+\n\nPPV_Smear\n\nTB Culture\u2013 (NTM Culture+) TB\u2013\n#\n\nUnnecessary isolation and unnecessary \ufb01rst-line treatment pending positive culture result\n\nFIGURE 1 Xpert MTB/RIF replacing the sputum-based approach in tuberculosis (TB) suspects. NTM: nontuberculous mycobacteria.\n\nDOI: 10.1183/13993003.01333-2015\n\n577\n\nTUBERCULOSIS | R. DIEL ET AL.\n\n578\n\nTB Culture+\n\nTB+\n\nPPV_Xpert_SP\n\nXpert MTB/RIF+ 0.9474\n\nXpert MTB/RIF\n\nTB Culture\u2013\n\nTB\u2013\n\n#\n\nSputum+\n\nTB Culture+\n\nTB+\n\n#\n\nXpert MTB/RIF\u2013\n\n#\n\nTB Culture\u2013\n\n(NTM Culture+)\n\nTB\u2013\n\nNPV_Xpert_SP\n\nMDR-TB+ pMDR_TB\nMDR-TB\u2013\n#\n\nXpert RIF resistance correctly positive\nPPV_Xpert_Res\nXpert RIF resistance falsely negative\n#\nXpert RIF resistance correctly negative\nNPV_Xpert_Res\nXpert RIF resistance falsely positive\n#\n\nWHO standard regimen appropriate\npMDR_Stand\nMore complex resistance patterns\n#\n\nIneffective standard second-line treatment pending DST result. Contact investigation\n\nIneffective \ufb01rst-line medication pending DST result. Contact investigation\n\nUnnecessary second-line treatment pending DST result\n\nRIF resistance assigned\n#\nRIF resistance not assigned 1\nMDR-TB+ pMDR_TB\n\nUnnecessary isolation (14 days) and unnecessary second-line treatment through disharge\nUnnecessary isolation (14 days) and unnecessary \ufb01rst-line medication through discharge\nUnnecessary diagnostic work-up and ineffective \ufb01rst-line treatment between culture and DST result. Contact investigation\n\nMDR-TB\u2013\n#\n\nUnnecessary diagnostic work-up\n\nUnnecessary isolation (1 day)\n\nDOI: 10.1183/13993003.01333-2015\n\nXpert MTB/RIF\n\nTB Culture+ PPV_Smear\n\nMDR-TB+ pMDR_TB\nMDR-TB\u2013\n#\n\nTB Culture\u2013 (NTM Culture+) TB\u2013\n#\n\nIneffective \ufb01rst-line treatment pending DST result. Contact investigation\nUnnecessary isolation and unnecessary \ufb01rst-line treatment pending positive NTM culture\n\nFIGURE 2 Xpert MTB/RIF versus the conventional approach in smear-positive multidrug-resistant tuberculosis (MDR-TB). NTM: nontuberculous mycobacteria; DST: drug susceptibility testing; RIF: rifampicin; WHO: World Health Organization.\n\nTUBERCULOSIS | R. DIEL ET AL.\n\nXpert MTB/RIF\n\nTB Culture+\n\nTB+\n\nXpert MTB/RIF+\n\nPPV_Xpert_SN\n\n0.070\n\nTB Culture\u2013\n\nTB\u2013\n\nSputum\u2013\n\nXpert MTB/RIF\u2013\n#\n\nTB Culture+\n#\nTB Culture\u2013 NPV_Xpert_SN\n\nNo Xpert MTB/RIF\n\nTB+\n\nTB+\n\n#\n\nTB\u2013\n\nTB\u2013\n\nNPV_Smear\n\nTB+\nTB\u2013\nIsolation (3 days), three sputum smears and unnecessary diagnostic work-up\nIsolation (3 days) and three sputum smears\n\nFIGURE 3 Xpert MTB/RIF versus the conventional approach in sputum-negative tuberculosis (TB) suspects.\n\nIsolation (1 day)\nIsolation (14 days) and \ufb01rst-line treatment through discharge\nIsolation (1 day) and unnecessary diagnostic work-up\nIsolation (1 day)\n\n95 specimens were smear-positive and 612 smear-negative compared with 133 specimens tested Xpert MTB/RIF-positive and 574 Xpert MTB/RIF-negative (table 1). Accuracy of microscopy and Xpert MTB/ RIF compared with culture results were calculated as well as PPV and NPV, and directly transferred as probabilities into the model. Of note, in our study the overall sensitivity (90.65%; 95% CI 84.54\u201394.93) and specificity (98.77; 95% CI 97.48\u201399.50) of the Xpert MTB/RIF assay for the diagnosis of pulmonary TB was slightly higher than the pooled sensitivity of 88% (95% CI 83\u201392) and pooled specificity of 98% (95% CI 97\u201399) reported in a recent Cochrane meta-analysis [13].\nBased on the NRC data, the time (mean\u00b1SD) to report positive cultures of M. tuberculosis among Xpert MTB/RIF-positives was 10.32\u00b14.7 days (median 9 days, interquartile range (IQR) 7\u201311.75 days) for smear-positives and 13.17\u00b15.15 days (median 13 days, IQR 9\u201317 days) for smear-negatives (table 2). The time required from sample acquisition to DST results was registered; the waiting period was 23.9\u00b18.9 days (median 21 days, IQR 18\u201327 days) for smear-positives and 29.6\u00b17.26 days (median 29 days, IQR 24.75\u2013 35 days) for smear-negatives.\nIn concurrence with the findings of STEINGART et al. [13], Xpert MTB/RIF was able to correctly distinguish between TB and nontuberculous mycobacteria (NTM) in smear-positive samples.\nCalculation of the PPV and NPV of the Xpert MTB/RIF RIF resistance test follows the following definitions: PPV=sen\u00d7pre/(sen\u00d7pre+(1\u2212spe)\u00d7(1\u2212pre)) and NPV=spe\u00d7(1\u2212pre)/((1\u2212sen)\u00d7pre+spe\u00d7(1\u2212pre)), where sen=sensitivity, spe=specificity and pre=prevalence. According to a review by WEYER et al. [14], sensitivity and specificity of Xpert MTB/RIF for RIF resistance is 95% and 98%, respectively, using culture as the reference method.\nAlthough we assume the same prevalence for MDR-TB in our collective of MDR-TB suspects as for fully susceptible TB in the base case, that estimate is varied between 0% and 30% in our sensitivity analysis.\nSusceptibility of MDR-TB strains The determination of RIF resistance alone is not sufficient information for the establishment of case-appropriate therapy. The WHO standard concept for MDR-TB is effective in many RIF-resistant cases and it may be credibly started based on RIF resistance determination. However, DST must be performed in parallel and the therapy reconsidered once the full resistance pattern is known. If a change in therapy proves necessary, expensive but inappropriate treatment produces extraordinary costs as do the days spent with ineffective second-line treatment. In a recently published German cost analysis including the resistance patterns of MDR-TB strains [11], most strains (51/55, 92.72%) were in vitro susceptible to at least four drugs of the WHO standard scheme. We use that estimate as the base case value ( pMDR_Stand).\nEconomic parameters The analysis includes the drug and laboratory costs as well as the opportunity costs arising from revenue losses for the hospital (table 3): according to US and German guidelines [2, 15], sputum smear-positive\n\nDOI: 10.1183/13993003.01333-2015\n\n579\n\nTUBERCULOSIS | R. DIEL ET AL.\n\n580\n\nTB Culture+\n\nTB+\n\nPPV_Xpert_SN\n\nXpert MTB RIF+ 0.070\n\nXpert MTB/RIF\n\nTB Culture\u2013\n\nTB\u2013\n\n#\n\nSputum\u2013\n\nTB Culture+\n\nTB+\n\n# Xpert MTB RIF\u2013\n\n#\n\nTB Culture\u2013\n\nTB\u2013\n\nNPV_Xpert_SN\n\nXpert RIF resistance correctly positive\nPPV_Xpert_Res\n\nMDR-TB+ pMDR_TB\nMDR-TB\u2013 #\nUnnecessary isolation (14 days) and unnecessary \ufb01rst-line treatment through discharge\n# MDR-TB+\npMDR_TB MDR-TB\u2013\n# Isolation (1 day)\n\nXpert RIF resistance falsely negative\n# Xpert RIF resistance correctly negative\nNPV_Xpert_Res\nXpert RIF resistance falsely positive\n#\nIsolation (1 day) and unnecessary work-up\nIsolation (1 day) and unnecessary work-up\n\nWHO standard regimen appropriate\npMDR_Stand\nMore complex resistance patterns\n# Isolation (1 day) and ineffective \ufb01rst-line treatment pending DST result. Contact investigation\n\nIsolation (1 day)\nIsolation (1 day) and ineffective standard WHO second-line treatment pending DST result. Contact investigation\n\nIsolation (1 day)\nIsolation (1 day) and unnecessary second-line treatment pending DST result\n\nNo Xpert MTB/RIF\n\nTB Culture+\n\nTB+\n\n#\n\nTB Culture\u2013\n\nTB\u2013\n\nNPV_Smear\n\nMDR-TB+ pMDR_TB\nMDR-TB\u2013 #\nIsolation (3 days) and three sputum smears\n\nIsolation (3 days), three sputum smears, unnecessary diagnostic work-up. Ineffective \ufb01rst-line treatment between positive culture result and discharge. Contact investigation\nIsolation (3 days), three sputum smears, unnecessary work-up\n\nDOI: 10.1183/13993003.01333-2015\n\nFIGURE 4 Xpert MTB/RIF versus the conventional approach in sputum-negative multidrug-resistant tuberculosis (MDR-TB) suspects. RIF: rifampicin; WHO: World Health Organization; DST: drug susceptibility testing.\n\nTUBERCULOSIS | R. DIEL ET AL.\n\nXpert MTB/RIF\n\nXpert MTB/RIF+ 0.1881\n\nXpert versus sputa\nConventional approach\n\nXpert MTB/RIF\u2013\n#\nSputum smear+ 0.134\n\nSputum smear\u2013\n#\n\nTB Culture+\n\nTB+\n\nPPV_Xpertoverall\n\nTB Culture\u2013\n\nTB\u2013\n\n#\n\nTB Culture+\n\nTB+\n\n#\n\nTB Culture\u2013\n\n(NTM Culture+)\n\nTB\u2013\n\nNPV_Xpertoverall\n\nTB Culture+\n\nTB+\n\nPPV_Smear\n\nTB Culture\u2013\n\n(NTM Culture+)\n\nTB\u2013\n\n#\n\nTB Culture+\n\nTB+\n\n#\n\nTB Culture\u2013\n\nTB\u2013\n\nNPV_Smear\n\nUnnecessary isolation (14 days) and unnecessary \ufb01rst-line treatment through discharge\nUnnecessary diagnostic work-up pending positive TB culture result\nUnnecessary isolation and unnecessary \ufb01rst-line treatment pending positive NTM culture result Isolation (3 days), three sputum smears, unnecessary diagnostic work-up\n\nFIGURE 5 Xpert MTB/RIF versus the conventional approach in sputum-negative multidrug-resistant tuberculosis (MDR-TB) suspects. NTM: nontuberculous mycobacteria.\n\nTB suspects must be isolated from the moment of admission through the first 14 days of treatment to counteract the nosocomial spread of TB, as must be smear-negative TB suspects until they have produced negative sputum samples on 3 subsequent days. Under the premise that most TB patients are accommodated in a twin-bedded room and that TB wards in Germany are working at full capacity, the loss of the use of one bed per day (cOpp) is incurred by the hospital during the isolation period.\nThe handling of smear-negative TB suspects or of those with a negative Xpert MTB/RIF result usually requires additional diagnostic work-up, the procedures and costs (cDcs) of which are presented in table 4.\nIn cases of unidentified MDR-TB, ineffective first-line treatment is administered either 1) following the conventional approach in the waiting period between positive TB culture and DST result or 2) in TB culture- and Xpert MTB/RIF-positives when no RIF resistance is assigned erroneously pending the DST result. Furthermore, ineffective second-line treatment is administered when the later DST result reveals more complex resistance patterns. Consequently, as in all of these cases when isolation is ended prematurely, intrahospital contact investigation must be performed (cContact). For full details and derivation of the cost parameters, see the online supplementary material.\nResults\nIn base case analysis, performing Xpert MTB/RIF is cost-saving for every one of the five scenarios, when compared with the conventional procedure (sputum smear followed by culture), although to widely varying degrees (table 5).\nImplementing Xpert MTB/RIF as an add-on in smear-positive TB suspects saves on average \u20ac48.72 per admitted patient. That this saving is relatively small is due to the fact that, according to our data, a positive TB culture result can be expected in as little as 10 days time; in non-TB smear-positives a NTM-positive culture result is also usually received. Thus, unnecessary isolation and unnecessary first-line treatment are promptly terminated. Potential cost savings strongly depend on the PPV of sputum smears and that of Xpert MTB/RIF as well as on the actual delay until the sputum culture result can be received. Univariate sensitivity analysis, in which all variables in the decision trees receive assigned values within their respective ranges, reveals that due to the low proportion of NTM (5.3%) in sputum smear-positives, even a minimal increase in the PPV of sputum smears (by 1.5% from 94.7% to 96.2%) or a minimal decrease in the PPV of the Xpert MTB/RIF (by 1.1% to 98.9%) results in a reversion of cost savings by utilising Xpert MTB/RIF as an add-on (online supplementary table S6a). The same is true if the waiting period for a positive culture result falls below a threshold of 7.5 days.\n\nDOI: 10.1183/13993003.01333-2015\n\n581\n\nTUBERCULOSIS | R. DIEL ET AL.\n\nTABLE 1 Comparison of results by sputum smear culture on tuberculosis (TB) and real-time PCR Xpert MTB/RIF in a total of 707 patients in 2012/2013\n\nTest and result\n\nTB\n\nTB\n\nSensitivity\n\nSpecificity\n\nPPV\n\nNPV\n\nculture-positive culture-negative\n\nSputum smear\n\nPositive\n\n90\n\nNegative\n\n49\n\nXpert MTB/RIF\n\nOverall\n\nPositive\n\n126\n\nNegative\n\n13\n\nSmear-positive\n\nPositive\n\n90\n\nNegative\n\n0\n\nSmear-negative\n\nPositive\n\n36\n\nNegative\n\n13\n\n64.75 (56.20\u201372.66) 99.12 (97.96\u201399.71) 94.74 (88.14\u201398.27) 91.99 (89.55\u201394.02) 5# 563\n90.65 (84.54\u201394.93) 98.77 (97.48\u201399.50) 94.74 (89.46\u201397.86) 97.74 (96.16\u201398.79) 7 561\n100.0 (95.98\u2013100.00) 100.00 (87.66\u2013100.00) 100.00 (95.98\u2013100.0) 100.00 (47.82\u2013100.00) 0 5\n73.47 (58.92\u201385.05) 98.76 (97.46\u201399.50) 83.72 (69.30\u201393.19) 97.72 (96.12\u201398.78) 7 556\n\nData are presented as n or % (95% CI). NPV: negative predictive value; PPV: positive predictive value. #: samples from patients suffering from nontuberculous mycobacterial disease.\n\nAn increase in the cost of Xpert MTB/RIF by 20% reduces the average cost saving to a small total amount of only \u20ac26.12 and a 20% lower per-day cost of blocking a twin bed marginalises it to \u20ac17.15. Thus, in PSA with Monte-Carlo simulation of 1000 TB suspects, adding Xpert MTB/RIF in smear-positives suspected of having fully susceptible TB is less costly in only 46.4% of cases (online supplementary table S7).\nIn smear-positive MDR-TB cases, as compared with fully susceptible TB cases, additional cost savings may occur. These are attributable to the principal advantage of Xpert MTB/RIF of recognising RIF resistance and to the associated avoidance of ineffective first-line treatment during a wait for the result of a culture-based DST report. Furthermore, with Xpert MTB/RIF, a clearly lower number of contact individuals has to be screened for latent MDR-TB infection compared with the conventional approach, where MDR-TB may be spread to healthcare workers in the lag between termination of isolation after 14 days of ineffective treatment and the later DST report. Here, in the base case analysis, cost savings amount to \u20ac189.56 per person.\nDepending on the PPV of assignment of RIF resistance by Xpert MTB/RIF, a variable percentage of cases receives ineffective, but costly second-line treatment, whilst the phenotypical DST result is initially lacking. MDR-TB prevalence strongly influences that PPV and consequently the amount of saving will be reduced to only \u20ac48.30 if MDR-TB prevalence goes to zero (online supplementary table S6b). Cost savings will also decrease in line with a decreasing number of contacts investigated and does not exceed \u20ac105.44 if only a total of five healthcare workers has to be screened, as compared with the 10 assumed in our base case (see online supplementary section 2.e).\nIn smear-positive MDR-TB cases, the probability that the WHO standard second-line treatment is applicable following the assignment of RIF resistance by Xpert MTB/RIF also plays a role. If the DST reveals more complex resistance strain patterns, cost savings in favour of Xpert MTB/RIF decrease remarkably to \u20ac79.08, assuming the standard regimen is effective in only 74% as a lower bound in univariate sensitivity analysis (online supplementary table S6b). PSA demonstrates that adding Xpert MTB/RIF in smear-positives suspected of having MDR-TB is less costly in 76.2% of cases (online supplementary table S7).\n\nTABLE 2 Delays in reporting positive cultures and drug susceptibility testing results for Mycobacterium tuberculosis isolates separated by sputum smear status\n\nSputum smear-positive\n\nSputum smear-negative\n\nTime for reporting positive culture for M. tuberculosis days\nTime to drug susceptibility testing report for isolates days\n\n10.32\u00b14.7 (7\u201311.75) 23.9\u00b18.9 (8\u201327)\n\n13.17\u00b15.15 (9\u201317) 29.6\u00b17.26 (24.75\u201335)\n\nData are presented as mean\u00b1SD (interquartile range).\n\n582\n\nDOI: 10.1183/13993003.01333-2015\n\n583\n\nDOI: 10.1183/13993003.01333-2015\n\nTABLE 3 Input for cost\u2013benefit analysis\nVariables category\n\nVariable name Distribution#\n\nCosts of first-line drugs per day \u20ac Costs of WHO standard MDR-TB drugs per day \u20ac Costs of Ziehl\u2013Neelsen microscopy \u20ac Costs of mycobacterial culture \u20ac Costs of Xpert MTB/RIF \u20ac Opportunity costs of blocking twin bed \u20ac Costs of diagnostic work-up \u20ac Latency pending sputum culture result in\nsmear-positive day Latency pending DST result in smear-positives day Latency pending sputum culture result in\nsmear-negatives day Latency pending DST result in smear-negatives day Probability of MDR-TB in TB patients PPV of positive sputum smear NPV of negative sputum smear PPV of Xpert MTB/RIF in smear-positives PPV of Xpert MTB/RIF in smear-negatives NPV of Xpert MTB/RIF in smear-positives NPV of Xpert MTB/RIF in smear-negatives\n\ncTBD cMDR_TB\ncZN cCulture cXpert\ncOpp cDcs dCulture_SP dResistance_SP dCulture_SN dResistance_SN pMDR_TB PPV_Smear NPV_Smear PPV_Xpert_SP PPV_Xpert_SN NPV_Xpert_SP NPV_Xpert_SN\n\nTriangular Triangular Triangular Triangular Triangular Triangular Triangular\nNormal Normal Normal Normal Linear Linear Linear Linear Linear Linear Linear\n\nValue (base case)\n6.3 101.04\n6.41 23.31 110.75 314.71 306.81 10.32 23.9 13.17 29.2 0.1966 0.9474 0.9199\n1.0 0.8372\n1.0 0.9772\n\nRelative change (range)\n\u00b120% (5.04\u20137.56)\n\u00b120% (80.83\u2013121.25)\n\u00b120% (5.12\u20137.69)\n\u00b120% (18.65\u201327.97)\n\u00b120% (88.6\u2013130.9)\n\u00b120% (253.97\u2013380.95)\n\u00b120% (245.45\u2013368.17)\n\u00b1SD 8.9 (5.62\u201315.02)\n\u00b1SD 8.9 (15.0\u201332.8) \u00b1SD 5.15 (8.02\u201318.32)\n\u00b1SD 7.2 (18.94\u201336.46)\n0%/30% (0.0\u20130.3) \u00b195% CI (0.8814\u20130.9827) \u00b195% CI (0.8955\u20130.9402) \u00b195% CI (0.9598\u20131.0) \u00b195% CI (0.6930\u20130.9319) \u00b195% CI (0.4782\u20131.0) \u00b195% CI (0.9623\u20130.9878)\n\nReference\n[6] [6] GO\u00c4 no. 4513 (online supplementary section 2.b) GO\u00c4 no. 4540 (online supplementary section 2.b) GO\u00c4 nos. 4780 and 4784 (online supplementary section 2.b) Calculated from InEK (http://www.g-drg.de/cms/) data (online supplementary section 2.a) Table 4 (online supplementary section 2.d) Assessed (table 2) Assessed (table 2) Assessed (table 2) Assessed (table 2) Calculated (table 1) Calculated (table 1) Calculated (table 1) Calculated (table 1) Calculated (table 1) Calculated (table 1) Calculated (table 1)\nContinued\n\nTUBERCULOSIS | R. DIEL ET AL.\n\nTUBERCULOSIS | R. DIEL ET AL.\n\n584\n\nTABLE 3 Continued\nVariables category\n\nVariable name Distribution#\n\nValue (base case)\n\nRelative change (range)\n\nReference\n\nPPV of Xpert MTB/RIF irrespective of smear status\nNPV of Xpert MTB/RIF irrespective of smear status\nProbability that WHO standard regimen is effective\nTime left to discharge from hospital days\nCosts of intrahospital contact investigation per TB index case \u20ac\nNumber of contacts to be investigated\n\nPPV_Xpertoverall\n\nLinear\n\nNPV_Xpertoverall\n\nLinear\n\npMDR_Stand\n\nLinear\n\ndDischarge\n\nTriangular\n\ncContact\n\nTriangular\n\npContact\n\nTriangular\n\n0.9474 0.9747 0.9272 26.78 105.81\n10\n\n\u00b195% CI (0.8946\u20130.9786)\n\u00b195% CI (0.9612\u20130.9879)\n\u00b120% (0.741\u20131)\n\u00b120% (21.24\u201332.14)\n\u00b120% (84.65\u2013126.97)\n5\u201320\n\nCalculated (table 1) Calculated (table 1)\n[11] [6] Adapted from [12] Assumption (online supplementary section 2.e)\n\nWHO: World Health Organization; MDR: multidrug-resistant; TB: tuberculosis; DST: drug susceptibility testing; PPV: positive predictive value; NPV: negative predictive value; GO\u00c4: Geb\u00fchrenordnung f\u00fcr \u00c4rzte (German medical fee schedule). #: in probabilistic sensitivity analysis.\n\nDOI: 10.1183/13993003.01333-2015\n\nTUBERCULOSIS | R. DIEL ET AL.\n\nTABLE 4 Costs of diagnostic work-up in sputum smear-negative or Xpert MTB/RIF-negative tuberculosis suspects\n\nProcedure\n\nn\n\nGO\u00c4 (1.0 times rate) \u20ac\n\nComputed tomography thorax Contrast agent, injection intravenously, high pressure Bronchoscopy including lavage Cytological investigation Interferon-\u03b3 release assay test peripheral blood Ziehl\u2013Neelsen smear (lavage) PCR (lavage) Culture mycobacteria (lavage) Total\n\n5371 346 678 4852 3694 4513 4783 4540\n\n134.06 17.49 52.46 10.14 33.80 9.33 29.41 33.31 306.81\n\nGO\u00c4: Geb\u00fchrenordnung f\u00fcr \u00c4rzte (German medical fee schedule).\n\nIn contrast, and predominantly due to a shorter isolation period (2 days fewer blockage of a twin bed by using Xpert MTB/RIF), the cost saving in fully susceptible smear-negative TB suspects is \u20ac503 and remains constantly high, even when the main cost driver, the per-day hospital opportunity cost of blocking a twin bed, is assumed to be 20% lower (in this case reducing the saving to \u20ac385.42). Increasing the cost of Xpert MTB/RIF within its given range or decreasing the higher NPV that its 9% sensitivity advantage over sputum smear brings (73.47% versus 64.75%) diminishes the total amount of savings at most by 8.44% and 1.24%, respectively, and has only marginal impact (online supplementary table S6c).\nFor smear-negative MDR-TB cases, the figure for the cost saving is slightly higher at \u20ac515.25 because, as is the case in smear-positive MDR-TB patients, it is predominantly the costs of contact investigation of healthcare workers that have to considered. Due to the high NPV (92%) of sputum smears, however, only very few MDR-TB cases will be detected in smear-negatives which induce hospital contact investigations by possible transmission in the latency between positive culture and the later DST result. Therefore, neither variations of MDR-TB prevalence nor of the number of contact individuals in sensitivity analysis have a substantial impact on cost savings (online supplementary table S6d) and in PSA cost savings remain at 100% in all scenarios involving smear-negative TB or MDR-TB suspects (online supplementary table S7).\nFull replacement of the conventional approach by Xpert MTB/RIF saves \u20ac449.98. This is primarily due to the elimination of unnecessary isolation in the 87% smear-negatives of our cohort of TB suspects (612/707; table 1). Accordingly, reducing the figure of per-day opportunity cost to its lower bound (\u221220%) will reduce the amount of cost saving by 59% to only \u20ac184.67 (online supplementary table S6e).\n\nTABLE 5 Results of base case analysis (five scenarios)\n\nBase case analysis and comparator\n\nMean cost per patient\n\nIncremental cost #\n\nSputum smear-positive TB suspects Xpert MTB/RIF as an add-on Conventional approach\nSputum smear-positive MDR-TB suspects Xpert MTB/RIF as an add-on Conventional approach\nSputum smear-negative TB suspects Xpert MTB/RIF as an add-on Conventional approach\nSputum smear-negative MDR-TB suspects Xpert MTB/RIF as an add-on Conventional approach\nXpert MTB/RIF replacing smears Xpert MTB/RIF Conventional approach\n\n157.2 205.47\n240.93 430.49\n512.17 1015.17\n518.03 1033.28\n440.97 890.95\n\n0 48.27\n0 189.56\n0 503.0\n0 515.25\n0 449.98\n\nCosts are presented as \u20ac. TB: tuberculosis; MDR: multidrug-resistant. #: increase in total costs resulting from using the conventional approach alone versus including Xpert MTB/RIF as an add-on or as a replacement.\n\nDOI: 10.1183/13993003.01333-2015\n\n585\n\nTUBERCULOSIS | R. DIEL ET AL.\n\nDiscussion\nThe present study is a differentiated cost\u2013benefit analysis of the implementation of the real-time PCR Xpert MTB/RIF method in hospitalised patients with suspected TB, either as an adjunct to or a replacement for sputum smear microscopy. PCR, with its ability to very rapidly confirm or exclude infectious pulmonary TB, has the potential to minimise the duration of isolation and/or to avoid unnecessary isolation.\nTo date, only very few cost studies on the routine use of nucleic acid amplification tests have been published and their findings are unfortunately not applicable to German conditions. ADELMANN et al. [16] found significant cost savings in a US urban public hospital (US$2003 per suspected smear-positive TB case) when using the amplified MDT (Mycobacterium tuberculosis Direct) test (Gen-Probe, San Diego, CA, USA) among predominantly African-American AFB smear-positive TB suspects. These subjects had a high prevalence of HIV-1 infection and for their cases the AFB smear had a very low PPV (27%) for culture-confirmed TB. Germany\u2019s HIV prevalence is low (2009: 0.1% of individuals aged 15\u201349 [17]) and the data we used from the German NRC give the AFB smear a PPV for culture-confirmed TB of 95%. In ADELMANN et al. [16], one main cost driver was the cost of unnecessary contact investigations, which were begun not at the time of culture confirmation, but immediately upon recognition of the positive AFB smear. A second notable driver was the per-day cost difference between isolation and nonisolation rooms. In Germany, special airborne infection isolation rooms providing negative pressure are rare. Also in MILLMANN et al.\u2019s cost\u2013benefit analysis [18], the incremental cost of respiratory isolation per day in a special room and the reduction regarding the length of stay of on average from 2.7 to 1.4 days per patient saved US$2278 per admission of suspected pulmonary TB. Our hospital costs are significantly lower.\nOur analysis shows that in base case analysis, performing Xpert MTB/RIF as an add-on for TB suspects admitted to a German TB ward as well as complete replacement of sputum smears by Xpert MTB/RIF is consistently cost-saving, even when the economic perspective is restricted to hospitalisation. Driving this is a reduction in the number of isolation days per case, each of which results in a blocked twin bed and corresponding revenue loss for the hospital. Smear-positive and smear-negative patients on average wait 10 and 13 days, respectively, for their positive culture results in Germany; for phenotypical DST results the average waiting periods are 24 and 30 days. There is only a marginal difference of specificity and overlapping 95% confidence interval in the PPV of fully susceptible smear-positive TB patients. This is due to a low observed proportion of 5.3% of NTM, resulting, as sensitivity analysis reveals, in a fragile economic advantage for Xpert MTB/RIF in this group of patients.\nParadoxically, and counter-intuitively, one strikingly favourable feature of Xpert MTB/RIF has its drawbacks. The immediate detection of RIF resistance in MDR-TB patients, in comparison with the much later reportable, culture-based DST, brings additional costs that occur in those patients with complex resistance patterns and for which the preliminary treatment with the recommended WHO standard second-line treatment is not effective. The treatment, ongoing for a number of weeks, must be adjusted to correspond with the pattern identified by phenotypical DST. In these cases, treatment duration is ultimately as long as it would have been under the conventional approach. However, DST cannot fully be replaced by line probe assays as the sensitivity of these for resistance to ethambutol, ofloxacin and injectable drugs is limited even in smear-positives [19]. Even in the case of a positive resistance determination, susceptibility to the other drugs must be clarified by DST in order to implement a definitely appropriate therapeutic regimen.\nAnother drawback of RIF detection by Xpert MTB/RIF is that unnecessary second-line treatment will be prescribed in a number of cases that corresponds to 1\u2212PPV of the test. This rate of false resistance determinations logically increases with decreasing MDR-TB prevalence.\nPSA that considers all realistic assumptions of uncertainty confirms the different degrees of potential savings between the five scenarios and underlines that Xpert MTB/RIF is most likely to be cost-saving in smear-negative TB suspects. This is of particular relevance as in 2013 in Germany 79.6% (2624/3298) of all reported pulmonary TB cases were \u201copen\u201d (culture-confirmed), of which 55% (1443/2624) were sputum smear-negative, and also in pulmonary MDR-TB cases only 63.5% (54/85) were sputum smear-positive (B. Brodhun, Robert Koch Institute, personal communication, 2015).\nOur study also has some limitations that must be considered when interpreting our results. First, the general limitations of retrospective, single-centre studies have to be considered. Our use of extensive sensitivity analyses is an effort that addresses those limitations. To validate our estimates, more cost studies, preferably with a multicenter and prospective study design, are required.\nSecond, from an economic point of view, replacing smear examination by Xpert MTB/RIF rather than using it as an add-on option basically combines the advantages of implementing Xpert MTB/RIF for the\n\n586\n\nDOI: 10.1183/13993003.01333-2015\n\nTUBERCULOSIS | R. DIEL ET AL.\nsingle smear-positive or smear-negative categories. However, it must be emphasised that graduated contact tracing according to the degree of infectiousness of the index case [2, 15, 20], as is usually practiced in low-incidence countries, could no longer be done were sputum examination to lapse. In practice, then, as all index cases have to be considered potentially sputum-positives, the initial circle of contact individuals to be investigated by the public health departments would have to be drastically expanded to be on the safe side, resulting in an inestimable increase in total costs.\nConclusion The utilisation of Xpert MTB/RIF in Germany, a high-income country, is likely to reduce overall costs in cases of suspected TB, especially in MDR-TB and smear-negative patients. As such, routine use of Xpert MTB/RIF may have a direct and positive impact on the control of TB disease. Prospective clinical studies should be undertaken to further evaluate its economic advantages in the immediate future.\nReferences\n1 European Centre for Disease Prevention and Control/WHO Regional Office for Europe. Tuberculosis surveillance and monitoring in Europe 2014. Stockholm, European Centre for Disease Prevention and Control, 2014.\n2 Centers for Disease Control and Prevention. Guidelines for preventing the transmission of Mycobacterium tuberculosis in health-care settings, 2005. MMWR Recomm Rep 2005; 54: RR-17, 1\u2013141.\n3 Migliori GB, Sotgiu G. Assessing tuberculosis management: what really happens to patients? Lancet Infect Dis 2015; 15: 1249\u20131251.\n4 L\u00f6nnroth K, Migliori GB, Abubakar I, et al. Towards tuberculosis elimination: an action framework for low-incidence countries. Eur Respir J 2015; 45: 928\u2013952.\n5 Vogl M. Assessing the DRG cost accounting with respect to resource allocation and tariff calculation: the case of Germany. Health Econ Rev 2012; 2: 15.\n6 Diel R, Rutz S, Castell S, et al. Tuberculosis: cost of illness in Germany. Eur Respir J 2012; 40: 143\u2013151. 7 Raizada N, Sachdeva KS, Sreenivas A, et al. Feasibility of decentralised deployment of XPERT MTB/RIF test at\nlower level of health system in India. PLoS One 2014; 9: e89301. 8 WHO. Automated real-time nucleic acid amplification technology for rapid and simultaneous detection of\ntuberculosis and rifampicin resistance: Xpert MTB/RIF system for the diagnosis of pulmonary and extrapulmonary tuberculosis in adults and children. http://apps.who.int/iris/handle/10665/112472 Date last accessed: August 5, 2015. 9 US Food and Drug Administration. Decision summary. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfPMN/ denovo.cfm?ID=DEN130032 Date last accessed: August 5, 2015. 10 Robert Koch-Institut. Bericht zur Epidemiologie der Tuberkulose in 2012 [Report on the epidemiology of tuberculosis in Germany in 2013]. Berlin, Robert Koch-Institut, 2014. 11 Diel R, Nienhaus A, Lampenius N, et al. Cost of multidrug resistance tuberculosis in Germany. Respir Med 2014; 108: 1677\u20131687. 12 Diel R, Nienhaus A, Lange C, et al. Cost-optimisation of screening for latent tuberculosis in close contacts. Eur Respir J 2006; 28: 35\u201344. 13 Steingart KR, Sohn H, Schiller I, et al. XPERT\u00ae MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev 2013; 1: CD009593. 14 Weyer K, Mirzayev F, Migliori GB, et al. Rapid molecular TB diagnosis: evidence, policy making and global implementation of Xpert MTB/RIF. Eur Respir J 2013; 42: 252\u2013271. 15 Ziegler R, Just HM, Castell S, et al. Tuberculosis infection control \u2013 recommendations of the DZK. Gesundheitswesen 2012; 74: 337\u2013350. 16 Adelman MW, Kurbatova E, Wang YF, et al. Cost analysis of a nucleic acid amplification test in the diagnosis of pulmonary tuberculosis at an urban hospital with a high prevalence of TB/HIV. PLoS One 2014; 9: e100649. 17 Trading Economics. Prevalence of HIV \u2013 total (% of population ages 15\u201349) in Germany. www.tradingeconomics. com/germany/prevalence-of-hiv-total-percent-of population-ages-15-49-wb-data.html Date last accessed: August 5, 2015. 18 Millman AJ, Dowdy DW, Miller CR, et al. Rapid molecular testing for TB to guide respiratory isolation in the U.S.: a cost-benefit analysis. PLoS One 2013; 8: e79669. 19 Theron G, Peter J, Richardson M, et al. The diagnostic accuracy of the GenoType\u00ae MTBDRsl assay for the detection of resistance to second-line anti-tuberculosis drugs. Cochrane Database Syst Rev 2014; 10: CD010705. 20 Diel R, Loytved G, Nienhaus A, et al. New recommendations for contact tracing in tuberculosis. German Central Committee against Tuberculosis. Pneumologie 2011; 65: 359\u2013378.\n\nDOI: 10.1183/13993003.01333-2015\n\n587\n\n\n",
    "authors": [
      "Roland Diel",
      "Albert Nienhaus",
      "Doris Hillemann",
      "Elvira Richter"
    ],
    "doi": "10.1183/13993003.01333-2015",
    "year": null,
    "item_type": "journalArticle",
    "url": "http://erj.ersjournals.com/lookup/doi/10.1183/13993003.01333-2015"
  },
  {
    "key": "UYCR8J4P",
    "title": "Strengthening health systems to improve the value of tuberculosis diagnostics in South Africa: A cost and cost-effectiveness analysis",
    "abstract": "Background In South Africa, replacing smear microscopy with Xpert-MTB/RIF (Xpert) for tuberculosis diagnosis did not reduce mortality and was cost-neutral. The unchanged mortality has been attributed to suboptimal Xpert implementation. We developed a mathematical model to explore how complementary investments may improve cost-effectiveness of the tuberculosis diagnostic algorithm.",
    "full_text": "PLOS ONE\n\na1111111111 a1111111111 a1111111111 a1111111111 a1111111111\n\nRESEARCH ARTICLE\nStrengthening health systems to improve the\nvalue of tuberculosis diagnostics in South\nAfrica: A cost and cost-effectiveness analysis\nNicola FosterID1,2,3*, Lucy CunnamaID1, Kerrigan McCarthyID4,5, Lebogang Ramma6, Mariana Siapka3, Edina Sinanovic1, Gavin Churchyard5,7, Katherine Fielding3,5, Alison D. Grant3,5,8, Susan Cleary1\n1 Health Economics Unit, School of Public Health and Family Medicine, University of Cape Town, Cape Town, South Africa, 2 Division of Health Research, Lancaster University, Lancaster, United Kingdom, 3 TB Centre, London School of Hygiene & Tropical Medicine, London, United Kingdom, 4 Division of Public Health, Surveillance and Response, National Institute for Communicable Disease of the National Health Laboratory Service, Johannesburg, South Africa, 5 School of Public Health, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa, 6 Department of Health and Rehabilitation Sciences, University of Cape Town, Cape Town, South Africa, 7 Aurum Institute, Johannesburg, South Africa, 8 Africa Health Research Institute, School of Nursing and Public Health, University of KwaZulu-Natal, Durban, South Africa\n* nicola.foster@lshtm.ac.uk\n\nOPEN ACCESS\nCitation: Foster N, Cunnama L, McCarthy K, Ramma L, Siapka M, Sinanovic E, et al. (2021) Strengthening health systems to improve the value of tuberculosis diagnostics in South Africa: A cost and cost-effectiveness analysis. PLoS ONE 16(5): e0251547. https://doi.org/10.1371/journal. pone.0251547\nEditor: Frederick Quinn, The University of Georgia, UNITED STATES\nReceived: June 12, 2020\nAccepted: April 28, 2021\nPublished: May 14, 2021\nPeer Review History: PLOS recognizes the benefits of transparency in the peer review process; therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. The editorial history of this article is available here: https://doi.org/10.1371/journal.pone.0251547\nCopyright: \u00a9 2021 Foster et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.\nData Availability Statement: All relevant data are within the paper and it\u2019s Supporting information files.\n\nAbstract\nBackground\nIn South Africa, replacing smear microscopy with Xpert-MTB/RIF (Xpert) for tuberculosis diagnosis did not reduce mortality and was cost-neutral. The unchanged mortality has been attributed to suboptimal Xpert implementation. We developed a mathematical model to explore how complementary investments may improve cost-effectiveness of the tuberculosis diagnostic algorithm.\nMethods\nComplementary investments in the tuberculosis diagnostic pathway were compared to the status quo. Investment scenarios following an initial Xpert test included actions to reduce pre-treatment loss-to-follow-up; supporting same-day clinical diagnosis of tuberculosis after a negative result; and improving access to further tuberculosis diagnostic tests following a negative result. We estimated costs, deaths and disability-adjusted-life-years (DALYs) averted from provider and societal perspectives. Sensitivity analyses explored the mediating influence of behavioural, disease- and organisational characteristics on investment effectiveness.\nFindings\nAmong a cohort of symptomatic patients tested for tuberculosis, with an estimated active tuberculosis prevalence of 13%, reducing pre-treatment loss-to-follow-up from ~20% to ~0% led to a 4% (uncertainty interval [UI] 3; 4%) reduction in mortality compared to the Xpert scenario. Improving access to further tuberculosis diagnostic tests from ~4% to 90%\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n1 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\n\nFunding: The XTEND project was carried out with the support of the Bill and Melinda Gates Foundation (Grant OPP1034523). This analysis contributes to NF\u2019s PhD, funded by the Medical Research Council of South Africa in terms of the National Health Scholars Programme from funds provided by the Public Health Enhancement Fund. This paper was prepared with support from the Collaboration for Health Systems Analysis and Innovation (www.chesai.org) that receives funding from the International Development Research Centre Ottawa Canada. The funders had no involvement in the study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.\nCompeting interests: The authors have declared that no competing interests exist.\n\namong those with an initial negative Xpert result reduced overall mortality by 28% (UI 27; 28) at $39.70/ DALY averted. Effectiveness of investment scenarios to improve access to further diagnostic tests was dependent on a high return rate for follow-up visits.\nInterpretation\nInvesting in direct and indirect costs to support the TB diagnostic pathway is potentially highly cost-effective.\nIntroduction\nGlobally, there is renewed interest in understanding how disease-specific investments function in the context of broader health system challenges [1]. Alongside this interest is re-invigorated enquiry into how best to support policy makers to assess joint technology and health systems strengthening investments when introducing new technologies. A recent example of an investment with global importance is the roll-out of Xpert MTB/RIF (Xpert).\nIn 2011, South Africa started the national roll-out of Xpert as first-line tuberculosis diagnostic test, following the World Health Organization (WHO) recommendation [2]. The rollout was anticipated to result in more people starting tuberculosis treatment because of Xpert\u2019s higher sensitivity, thus reducing mortality [3]. In addition Xpert was expected to reduce the time to MDR tuberculosis treatment start [4, 5]. However, in practice no significant impact on tuberculosis-related morbidity, mortality, pre-treatment loss-to-follow-up (iLTFU) or time-totreatment for patients starting drug-sensitive tuberculosis (DS-TB) has been observed [6, 7]. Studies examining the impact on patients with multi-drug resistant (MDR) tuberculosis found that Xpert reduced time-to-appropriate-treatment, although not to same day or same week, as had been expected [8, 9]. Furthermore, an economic evaluation based on a pragmatic trial following the roll-out in South Africa (the XTEND trial) found that Xpert implementation was both effect- and cost-neutral and was unlikely to improve the cost-effectiveness of the tuberculosis diagnostic algorithm [10]. The study concluded that implementation constraints may have mediated the impact of Xpert under programmatic conditions [7]. Other countries reported similar experiences with Xpert implementation. Placement of the test in the health system, it\u2019s integration into the laboratory infrastructure and diagnostic algorithm, as well as patient linkages to treatment were found to be important mediators of costs and effects [11\u2013 18].\nFor South Africa and beyond, policy makers need support to determine which complementary investments are required to strengthen the tuberculosis diagnostic pathway. To inform this need and illustrate a potential approach to assessing combined diagnostic technology and health systems investments, we fitted a purpose-built mathematical model to empirical data from the XTEND trial [7]. We then explored which investments complementary to the Xpertbased tuberculosis diagnostic algorithm would be most cost-effective in South Africa, and used the model to identify drivers of the cost-effectiveness of these investments.\n\nMethods\nWe conducted cost-effectiveness analyses of investments in health systems to support tuberculosis diagnosis. This analysis builds on previous modelling work that explored investments in\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n2 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\npatient pathways [19\u201322], by using patient-level cohort data from a pragmatic cluster-randomised controlled trial (described in S1 Text).\nOverview\nHealth systems investments are typically conceptualised as investments in health care infrastructure, clinical guidelines, technology or human resources, with less emphasis on how the relational aspect of health systems [23] may affect the costs and outcomes of an investment. Clinical discretionary decision-points in patient care can be conceptualised as transactions between patients and providers, occurring within a given organisational system. One may consider these transactions as interactions between the hardware (technology, infrastructure and finances) and software (formal or informal rules of practice, and beliefs that explain behaviour) components of health systems [24]. While investment costs have been estimated by analysing how the production of healthcare responds to an increase in need [25, 26], here we identified patterns of provider behaviour and then modelled this behaviour as a function of resource availability, process and relational interactions. This is implemented in the model by the mediation of decisions along the patient pathway. A simplified visual representation of the model and the decision points is shown in Fig 1 and is referred to in Table 2 [27]. The costs of decision-making processes includes the cost of regulating the decision as well as the opportunity cost of the benefits forgone in the time taken to make the decision or in making the wrong decision, the transaction cost [28: 86]. We modelled the value of additional investments to strengthen these decision-making processes.\nMathematical model. A state-transition model with time-dependent Markov processes was developed, simulating disease progression and interactions with the health system in a symptomatic population being investigated for tuberculosis. Secondary benefits to the\n\nFig 1. Simplified schematic of the model. The figure is a simplified representation of the model, with the circular boxes representing health states and the square boxes representing intermittent states used to model shorter time step process. The model structure is presented in more detail in S1 Text. \u201cA\u201d refers to the decision-making process from a negative test result to starting treatment without bacteriological confirmation; \u201cB\u201d represents the decision to continue testing for tuberculosis (negative pathway) in those with a negative test result; \u201cC\u201d is the behaviour around starting treatment after a positive test results; \u201cD\u201d refers to the decision (based on an interpretation of the further diagnostic tests) to start treatment; and \u201cE\u201d refers to the decision to start tuberculosis (TB) treatment after being \u2018out of care\u2019. \ufffd The model structures following the tuberculosis test are replicated for each of the six patient types, those HIV negative (with and without tuberculosis), HIV positive not on antiretroviral therapy (with and without tuberculosis), and those HIV positive on antiretroviral therapy (with and without tuberculosis). \ufffd\ufffd The treatments states are replicated for drug-sensitive and multi-drug resistant tuberculosis treatment.\nhttps://doi.org/10.1371/journal.pone.0251547.g001\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n3 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\npopulation due to tuberculosis transmission reduction are not included [29]. The analytical timeframe is three years, representing the time until the population is either cured of this tuberculosis episode, or dead. Patients move through the model in one-monthly steps to represent movement through treatment and from out of care, with additional structure added to model shorter diagnostic processes. The model was implemented in TreeAge Pro 2018 and datasets analysed using STATA 13.\nIn the model, six patient types defined by HIV, anti-retroviral therapy and true tuberculosis status move through health states until reaching an absorbing state (cure or death). Patients are symptomatic when entering the model, transition through a series of diagnostic processes, and then move to one of four possible health states (1) \u2018out of care\u2019 if not started on treatment; (2) drug-sensitive or multi-drug resistant tuberculosis treatment; (3) death; or (4) cured. The \u2018cured\u2019 state can be entered either after treatment or based on a self-cure rate.\nParameter estimation and model fitting\nTransition probabilities and resource use were estimated from trial data (see Table 1). Where treatment-related events occurred after the six-month trial period, data from published cohorts and meta-analyses were used to construct the patient pathway until the end of the treatment episode. The pragmatic nature of the trial did not allow for definitive confirmation of TB diagnosis among trial participants. Unobservable parameters include the true TB prevalence in the population, and the predictive value of decisions to start treatment or request further investigations. These parameters were estimated by calibrating the model\u2019s mortality and treatment outputs against those observed in the trial [45]. We estimated a plausible range of values for the unobserved parameters and then iteratively fitted the mortality and time-totreatment curves from model outputs to trial outcomes until the shape of the respective curves fitted using a range of goodness-of-fit measures [46: 260] (see S1 Text).\nCost analyses\nThe costs of providing and accessing care were estimated alongside the trial, using a combination of top-down and ingredients costing approaches [10, 47, 48]. HIV-care costs were extracted from published sources (Table 1). Costs were estimated by multiplying unit costs by the number of events incurred from data collected during the trial. Patient costs included travel- and time-costs incurred by patients and caregivers when accessing care. Additionally, income loss, the cost of caregiver\u2019s time, interest on loans as well as the cost of nutritional supplements were included. The opportunity cost of time was valued by multiplying time loss by the pre-illness mean income of the cohort [44]. All costs were estimated in local currency using 2013 prices and converted to US dollars using the average 2013 exchange rate of US $1 = R9.62 (www.Oanda.com).\nInvestments\nThe pragmatic nature of the trial allowed us to identify gaps between ideal movement along different decision nodes of the pathway and mediating variables of effectiveness in routine care settings. Table 2 summarises the investment scenarios modelled and how they were implemented in the model, with a visual representation of the model and decision points provided in Fig 1. We modelled five investment strategies to support the tuberculosis diagnostic pathway. These included 1) reducing initial pre-treatment loss-to-follow-up (iLTFU), 2) supporting same-day clinical diagnosis of tuberculosis after a negative test result (TfN), and 3) improving access to further tuberculosis diagnostic tests following an initial negative result (NP). In addition, two combination scenarios were modelled (iLTFU and TFN; iLTFU and\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n4 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\n\nTable 1. Summary of parameters and distributions.\n\nDefinition\n\nMean and stratification\n\nDistribution Comments. References are listed as name of first author, year (Reference).\n\nPopulation\n\nGender\n\n59.9% female\n\nRepresents trial population. Churchyard. 2015 [7]\n\nAge (IQR)\n\n37 (29\u201348) years\n\nFixed\n\nRepresents trial population. Churchyard. 2015 [7]\n\nInitial population disease characteristics\n\nHIVneg 0.314 (0.030); HIVpos 0.531 (0.015);\nART 0.155 (0.005)\n\nDirichlet\n\nFrom trial population. Churchyard 2015 [7] Those with unknown self-reported HIV status are assumed to be HIV positive, not on ART.\n\nCD4 count in those with HIV (IQR)\n\n315 (192\u2013480) cells/\u03bcL\n\nRepresents the microscopy arm of the trial population. Churchyard. 2015 [7]\n\nTrue TB prevalence (includes bacteriologically confirmed -, clinical\u2014and undiagnosed TB) in the microscopy arm of the study.\n\n13.0%\n\nFixed\n\nEstimated from XTEND trial and model calibration. Churchyard. 2015 [7]\n\nProportion of patients diagnosed with drugresistant TB, any diagnosis\n\n4.0% (8/195)\n\nRepresents trial population. Churchyard. 2015 [7].\n\nProportion of patients starting MDR-TB treatment\n\n2.0% (3/195)\n\nRepresents what was observed in the XTEND trial. Churchyard. 2015 [7]. Time to starting MDR TB treatment was 11 and 33 days respectively.\n\nDiagnosis, transition probabilities\n\nProbability of a positive Xpert test result if symptomatic and able to provide a sputum sample, mean (standard deviation)\n\nHIVneg 0.077 (0.03); HIVpos 0.132 (0.05);\nART 0.135 (0.03)\n\nDirichlet Estimated from XTEND trial. Churchyard. 2015 [7]\n\nProbability of TB if patient had a positive test result\n\nHIVneg 0.877; HIVpos 0.936;\nART 0.938\n\nFixed\n\nEstimated based on GX sensitivity 0.86 in HIVneg; 0.79 in HIVpos, 0.94 for Rif resistance, and GX specificity of 0.99 in HIVneg, HIVpos, 0.98 for Rif resistance. Steingart 2014 [30], Steingart 2006 [31], and Boehme 2011 [32].\n\nProbability of TB if patient had a negative test result\n\nHIVneg 0.012; HIVpos 0.113;\nART 0.114\n\nFixed\n\nUnobserved parameter, estimated from model calibration. Based on GX sensitivity 0.86 in HIVneg; 0.79 in HIVpos, 0.94 for Rif resistance, and GX specificity of 0.99 in HIVneg, HIVpos, 0.98 for Rif resistance. Steingart 2014 [30], Steingart 2006 [31], and Boehme 2011 [32]. This includes a probability of a false negative test result; HIVneg 0.012; HIVpos pre-ART 0.038; HIVpos ART 0.039 as well as a probability of \u2018undiagnosed TB\u2019. Undiagnosed TB includes those who provide pauci-bacillary sputum or have extrapulmonary TB. Probability of undiagnosed, \u201chard-todiagnose\u201d TB estimated to be 0.075 in those HIVpos pre-ART and 0.075 those HIVpos.\n\nProbability of starting treatment within 30 days of a positive test result, mean (standard deviation)\n\nHIVneg 0.882 (0.325); HIVpos 0.802 (0.400);\nART 0.944 (0.236)\n\nDirichlet\n\nEstimated from XTEND trial. Churchyard et al. 2015 [7]\n\nProbability of starting treatment within one month of a negative test result without further diagnostic tests\n\nHIVneg_TB 0.535; HIVneg 0.002; HIVpos_TB 0.072; HIVpos 0.009;\nART_TB 0.017; ART 0.003\n\nFixed\n\nProbability of starting treatment was estimated from XTEND trial, whether this clinical decision was correct (treatment started in those with TB vs those without) was estimated through model calibration. Churchyard et al. 2015 [7]. We therefore assume that clinicians are unlikely to start treatment empirically in those HIV negative.\n\nProbability of receiving further investigations after a negative test result\n\nHIVpos_TB 0.041; HIVpos 0.041; ART_TB 0.073; ART 0.073\n\nFixed\n\nEstimated from XTEND trial. Churchyard. 2015 [7]. McCarthy. 2016 [33].\n\nProbability of starting TB treatment after further diagnostic tests\n\nHIVpos_TB 0.212; HIVpos 0.027; ART_TB 0.217; ART 0.037\n\nFixed\n\nEstimated from XTEND trial and the model calibration. Churchyard. 2015 [7]. McCarthy. 2016 [33].\n\nProbability of starting TB treatment from \u2018out of care\u2019, by month: from all who do not start TB treatment within one month of the diagnostic test\n\n(Continued )\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n5 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\n\nTable 1. (Continued)\n\nDefinition\n\nMean and stratification\n\nMonth 2\nMonth 3\nMonth 4\nMonth 5\nMonth 6\nProbability of starting MDR-TB treatment if diagnosed with MDR-TB Treatment, transition probabilities Probability of drug sensitive TB regimen started if TB treatment started Probability of MDR-TB regimen started if TB treatment started, mean (standard deviation) Disease progression, transition probabilities Average life expectancy at birth, South Africa\n\nHIVneg_TB 0.928; HIVneg 0.005; HIVpos_TB 0.164; HIVpos 0.000;\nART_TB 0.100; ART 0.000 HIVneg_TB 0.756; HIVneg 0.000; HIVpos_TB 0.066; HIVpos 0.000;\nART_TB 0.207; ART 0.000 HIVneg_TB 0.000; HIVneg 0.005; HIVpos_TB 0.146; HIVpos 0.000;\nART_TB 0.148; ART 0.000 HIVneg_TB 0.000; HIVneg 0.015; HIVpos_TB 0.064; HIVpos 0.000;\nART_TB 0.000; ART 0.000 HIVneg_TB 0.000; HIVneg 0.010; HIVpos_TB 0.060; HIVpos 0.000;\nART_TB 0.000; ART 0.000 HIVneg 0.025; HIVpos 0.019; ART 0.000\nHIVneg 0.952; HIVpos 0.969; ART_TB 0.834\nHIVneg 0.039 (0.208); HIVpos 0.023 (0.002); ART 0.000 (0.000);\n63 years\n\nAll-cause mortality in those without TB, monthly, mean (standard deviation)\n\nHIVneg 0.001 (0.0005); HIVpos 0.002 (0.000); ART 0.001 (0.001)\n\nStandardised mortality ratio for all-cause\n\n3.76\n\nmortality in patients post-TB treatment\n\nMonthly mortality if living with TB, not currently receiving treatment, mean (standard deviation)\nMonthly mortality on treatment for those with TB, mean (standard deviation)\n\nHIVneg 0.018 (0.020); HIVpos 0.132 (0.005); ART 0.039 (0.005)\nHIVneg 0.002 (0.001); HIVpos 0.046 (0.002); ART 0.006 (0.003)\n\nDisability weights, mean (standard deviation)\nThe disability weight is a factor reflecting the severity of disease.\n\nHIVneg_TB 0.331 (0.057); HIVpos_TB 0.399 (0.070); HIVpos 0.221 (0.041); ART\n0.053 (0.011); ART_TB 0.331 (0.057)\n\nCost and resource use Microscopy, mean (standard deviation)\n\n$6.30 ($1.34)\n\nDistribution Fixed Fixed Fixed Fixed Fixed\nDirichlet\n\nComments. References are listed as name of first author, year (Reference).\nCurve estimated from XTEND trial. Assume that the behaviour from out of care remains the same. Churchyard. 2015 [7].\nCurve estimated from XTEND. Assume that the behaviour from out of care remains the same. Churchyard. 2015 [7].\nCurve estimated from XTEND. Assume that the behaviour from out of care remains the same. Churchyard. 2015 [7].\nCurve estimated from XTEND trial. Assume that the behaviour from out of care remains the same. Churchyard. 2015 [7].\nCurve estimated from XTEND trial. Assume that the behaviour from out of care remains the same. Churchyard. 2015 [7].\nEstimated from XTEND trial. Churchyard. 2015 [7].\n\nDirichlet Dirichlet\n\nEstimated from XTEND trial. Churchyard. 2015 [7]. Estimated from XTEND trial. Churchyard. 2015 [7].\n\nFixed\nDirichlet Fixed\nChanges over time\nChanges over time\nBeta\n\nFrom the rapid mortality surveillance report 2014. Assumes that HIVpos patients who are on ART when they enter the model would have the same life expectancy as the general population (varied in the sensitivity analysis). HIV specific mortality considered in model through probabilities. Dorrington. 2015 [34]. Years of life remaining at death is estimated from the difference between current age in model (mean age of cohort + time in model) and the average life expectancy at birth.\nFrom Statistics South Africa report (P0309.3), mortality and causes of death in South Africa: findings from death notification [35].\nIncreased all-cause mortality in those with a previous episode of TB [36]. Estimated as part of a systematic review and meta-analysis.\nBased on Tiemersma. 2011 [37]. Used half-cycle correction to adjust for earlier movement into treatment in month 1 of the model.\nAndrews 2012 [38]. Mohr 2015 [39]. Monthly mortality reduction due to TB treatments added as distribution over time, where mortality reduces to 10% of the mortality of those with TB not on treatment at month 5 on treatment. Based on comparison with mortality on treatment observed in the XTEND trial. Churchyard 2015 [7].\nSalomon. 2015 [40]. Kastien-Hilka. 2017 [41]. Assuming that disability weights are not cumulative, thus those on ART with TB have the same disability weight as someone with TB disease only.\n\nGamma Cunnama 2016 [42].\n\n(Continued )\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n6 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\n\nTable 1. (Continued)\n\nDefinition\nXpert, mean (standard deviation) Sputum liquid culture, mean (standard deviation) Digital radiograph, mean (standard deviation) First-line drug sensitivity test, mean (standard deviation) Second-line drug sensitivity test, mean (standard deviation) Provider cost of clinic visit for initial diagnosis and monitoring Provider cost of clinic visit for treatment Patient cost of clinic visit Guardian cost per clinic visit Cost of caregiver per day Resource use along the diagnostic pathway\nProvider cost of drug sensitive TB treatment, episode\nProvider cost of multi-drug resistant TB treatment, episode\nPatient cost of drug sensitive TB treatment, episode\nPatient cost of multi-drug resistant TB treatment, episode\n\nMean and stratification\n$16.90 ($6.10) $12.90 ($2.26)\n$15.17 ($7.74) $20.30 ($7.28)\n$25.10 ($20.22)\n$8.63\n$3.89 $2.90 $10.04 $0.69 Detailed input available from S1 Text.\n$192.99\n$10 802.66\nCost of accessing care associated $459.16; Cost of illness $135.94\nCost of accessing care associated $3 592.27;\nCost of illness $2 442.03\n\nDistribution\nGamma Gamma\n\nComments. References are listed as name of first author, year (Reference). Cunnama 2016 [42]. Cunnama 2016 [42].\n\nGamma Gamma\n\nFoster. Unpublished. Cunnama 2016 [42].\n\nGamma Cunnama 2016 [42].\n\nFixed\n\nVassall 2017 [43].\n\nFixed\n\nVassall 2017 [43].\n\nFixed\n\nFoster 2015 [44].\n\nFixed\n\nFoster 2015 [44].\n\nFixed\n\nFoster 2015 [44].\n\nGamma\n\nEstimated by disease progression. Reported in Vassall 2017 [43]. Foster 2015 [44].\n\nFixed\n\nEstimated based on patient movements through care observed in the trial. Reported in Vassall 2017 [43]. Foster 2015 [44].\n\nFixed\n\nEstimated based on patient movements through care observed in the trial. Reported in Vassall 2017 [43]. Foster 2015 [44].\n\nTime-dependent Foster 2015 [44]. functions\n\nTime-dependent Foster 2015 [44]. functions\n\nIn the Table, a fixed distribution refers to a distribution one where no uncertainty interval is estimated in keeping with calibration practice in complex models. Furthermore, IQR = interquartile range; TB = tuberculosis; MDR-TB = multi-drug resistant tuberculosis; Xpert = Xpert MTB/RIF; HIVpos = individuals HIV positive not yet started on anti-retroviral therapy; HIVpos_TB = individuals HIV positive with tuberculosis; ART = individuals HIV positive started on anti-retroviral therapy; ART_TB = individuals HIV positive on anti-retroviral therapy with tuberculosis.\n\nhttps://doi.org/10.1371/journal.pone.0251547.t001\n\nNP) to observe the additive effects of the scenarios. Investments were modelled by altering parameters at key stages in the patient pathway and how these will increase the count of utilisation that increases costs and affects outcomes. The cost of facilitating change through changing behaviour, which we refer to as the transaction cost is shown in Fig 3.\n\nEconomic analyses\nThe cost-effectiveness of investment scenarios was estimated from the societal perspective, which includes provider and patient-incurred costs. Disability adjusted life years (DALYs) averted were estimated using model estimates of years of life lost (YLL) due to premature mortality and years lived with disability (YLD). YLL were estimated based on progression through the model, assuming an average life expectancy of 63 years and the mean age (38 years) of patients in the trial [7, 34]. Disability weights from the 2010 Global Burden of Disease study were attached to model states [52]. For people on ART with tuberculosis, we assumed the\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n7 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\n\nTable 2. Summary of the investment scenarios modelled.\n\nReduction in initial LTFU (in Fig 1; decision-point C and E)\nEmpirical treatment from negative test result (in Fig 1; decision-point A)\n\nInvestment\nAll patients with positive TB test results start treatment within one month of testing\u2014simulating a point-of-care or a track-and-trace scenario with active follow-up of people with a positive TB test result. Synergies with investment in a community health worker programme.\nThe ability of healthcare workers to correctly act based on continued clinical symptoms, on the same day as the results visit (by giving TB treatment to those with test negative TB expressed as the sensitivity and specificity of that decision). This was based on the behaviour estimated from the microscopy arm of the model calibration and was applied to behaviour after a negative Xpert test result.\n\nModel implementation\n\nParameter, events or Assumptions resource changes\n\nptxfpos = 1\u2014pMort_m1 (stratified by HIV and TB status)\nThe probability of starting treatment from a positive test result was the\nremainder of all patients in that state after those who would die in that month had been subtracted. The\nmortality rate was stratified by HIV and TB status.\n\nProbability of starting treatment after positive\n(in month 1), from: HIVneg: 0.882 to 1; HIVneg_TB: 0.882 to 1; HIVpos: 0.802 to 1; HIVpos_TB: 0.802 to 1;\nART: 0.944 to 1; ART_TB: 0.944 to 1\n\nMonthly conditional probabilities of starting treatment from \u2018out of care\u2019 were estimated from the trial in the base scenario (reported in Table 1). In this investment scenario, patients shift from moving to the \u2018out of care\u2019 state if not started on treatment within one month, to the treatment state immediately, thus probabilities of starting treatment from \u2018out of care\u2019 approximate zero. The relative proportions of those starting various treatment types is kept the same as observed in the trial.\n\npnegpathfeg = 0\nptreatfneg = value estimated from reported behaviour in the control arm of the XTEND study [49], under the assumption that behaviour observed\nafter the implementation will revert back to pre-implementation levels.\n\nProbability of the negative pathway after a negative test result, from:\nHIVpos: 0.027 to 0.000\n\nGiven the differences in health care worker behaviour after a microscopy test compared to a Xpert test result observed in the XTEND trial, we use the transition probabilities estimated from the microscopy arm of the trial [50, 51].\n\nAssumed that all have at least one visit to a public health clinic (and associated\ncosts) after a negative test result for treatment initiation.\n\nHIVpos_TB: 0.212 to 0.000\nART: 0.037 to 0.000 ART_TB: 0.217 to 0.000\n\nProbability of starting treatment after a\nnegative test result, from:\n\nHIVneg: 0.002 to 0.040\n\nHIVneg_TB:0.054 to 0.270\n\nHIVpos: 0.009 to 0.180\n\nHIVpos_TB: 0.072 to 0.360\n\nART: 0.003 to 0.060\n\nART_TB: 0.017 to 0.090\n\n(Continued )\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n8 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\n\nTable 2. (Continued)\n\nImprovements in the test-negative pathway (in Fig 1 decisionpoints B and D)\n\nInvestment\nHIV-positive people with negative test results get further investigations (radiograph/ culture) for TB, and a proportion are started on TB treatment, simulating additional investment in improving access to further diagnostic tests.\n\nModel implementation\nptreatfneg = 0\npnegpathfneg = 1 (stratified by HIV and TB status)\ntreatfnegpath = 0.10 (no TB); 0.80 (with TB)\nThe probability of starting treatment is shifted from following a negative test result to the decision to order further diagnostic tests. The probability of starting treatment after the negative pathway was 10% in those without TB,\nand 80% in those with TB. Assumed that every person will accumulate two visits to the public clinic during the negative pathway, and that each person getting further tests will get at least one radiograph.\n\nParameter, events or resource changes\nProbability of starting treatment after negative test result changes from: HIVneg: 0.002 to 0.000\nHIVneg_TB:0.054 to 0.000\nHIVpos: 0.009 to 0.000\nHIVpos_TB: 0.072 to 0.000\nART: 0.003 to 0.000 ART_TB: 0.017 to 0.000\nProbability of the negative pathway after a\nnegative test result change from:\nHIVpos: 0.041 to 0.900 HIVpos_TB: 0.041 to\n0.900 ART: 0.073 to 0.900 ART_TB: 0.073 to 0.900 Probability of treatment from negative pathway\nchanges from: HIVpos: 0.027 to 0.100 HIVpos_TB: 0.212 to\n0.800 ART: 0.037 to 0.100 ART_TB: 0.217 to 0.800\n\nAssumptions\nSimilar to the previous scenario, we model a healthcare worker behaviour change scenario based on the difference in observed behaviour between the microscopy and Xpert arms of the study. This scenario simulates a situation where there is an increase in the proportion of patients who receive further investigations after a negative test result. Therefore, we reduced all empirical treatment to 0 and all eligible patients received a radiograph as part of the negative pathway.\n\nIn the Table, the individual characteristics of the patients are labelled as HIVneg for people who are HIV negative and don\u2019t have tuberculosis; HIVneg_TB for people who are HIV negative and have been diagnosed with tuberculosis; HIVpos for people who are HIV positive and don\u2019t have tuberculosis; HIVpos_TB for people who are HIV positive and have been diagnosed with tuberculosis; ART represents the individuals who are HIV positive, on anti-retroviral therapy and don\u2019t have tuberculosis; and ART_TB represents the individuals who are HIV positive, on anti-retroviral therapy and have been diagnosed with tuberculosis.\n\nhttps://doi.org/10.1371/journal.pone.0251547.t002\n\nsame disability weight as for those with tuberculosis who are HIV negative. Costs and outcomes were discounted at 3% per annum, and varied in the sensitivity analyses [53: 108\u2013112].\nTransaction costs are conceptualised as the value of resources that would support better decision-making between agents. These costs are incurred during each decision-making interaction along the patient pathway (represented by blue dots in Fig 1). Changes in the optimal investment strategy at a range of transaction costs are evaluated by plotting cost-effectiveness acceptability frontiers (Fig 3) [54]. The optimal investment option is defined as the strategy with the highest net monetary benefit at a given cost-effectiveness threshold and transaction cost level.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n9 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\nSensitivity and scenario analyses\nThe impact of model parameters changes on results was assessed through univariate sensitivity analysis. Probabilistic uncertainty analyses, simulating 100 000 samples, were used to assess the simultaneous effect of path and parameter uncertainty on the results [55].\nScenario analyses were used to explore how implementation may vary between contexts. Given the set of interactions governing decision-making in the care pathway, some of which would be harder to mediate through additional investment [56], an increase in the value of supporting investments would not lead to proportional, linear improvements in outcomes [57].\nEthics statement. The study was approved by the research ethics committees of the University of Cape Town (363/2011), University of the Witwatersrand (M110827), London School of Hygiene & Tropical Medicine (6041), and the World Health Organization (RPC462). Health department officials and facility managers provided permission to conduct the study in the selected facilities and written informed consent was obtained from respondents.\nResults\nAfter parameterising the model with data from the trial, and validating to the observed rate of TB treatment started and other secondary outcomes, we found that in order to achieve a good fit of the model to the data, we needed to also consider the limitations of sputum-based tuberculosis diagnostic modalities [58]. Undiagnosed tuberculosis may be related to the site of infection (extra-pulmonary tuberculosis), and low bacillary load in the sample, as is common in advanced HIV disease. During the model calibration, we therefore also added a parameter to capture the prevalence of extra-pulmonary tuberculosis (EPTB), varied along with the positive predictive value (PPV) and negative predictive value (NPV) to identify the best model fit. The prevalence of TB in the cohort was estimated to be 13% (see S1 Text).\nCosts, effectiveness, and cost-effectiveness analyses of the investment scenarios\nTable 3 presents the costs, effectiveness (deaths averted and DALYs averted), and cost-effectiveness of investment scenarios, compared with the base case of Xpert as observed during the trial. The uncertainty interval (UI) is shown in brackets. From the provider\u2019s perspective, the incremental cost-effectiveness ratios (ICERs) ranged between $17.42 and $39.70 per DALY averted. We estimated a provider cost of tuberculosis services of $89.66 (UI: $87 - $92) per symptomatic person tested using an Xpert-based diagnostic algorithm. The societal cost per person was estimated to be $169.94 (UI: $167 - $173).\nReducing iLTFU by starting all individuals who test positive on treatment increased the cost of treatment and patient cost of accessing care per patient by $2.76 and $8.25 respectively. This scenario reduced time-to-treatment but has a comparatively small effect on the total number of people starting treatment and on health outcomes. Assuming that 100% start treatment in month one shifts the time-to-treatment started curve to the left, starting people on treatment who would have never started as well as those who would have started within the next couple of months. Since TB treatment does not instantly reduce mortality for patients who have TB, the proportion of patients who start treatment in month one in the reduction in iLTFU investment option only increases by 12% in those HIV negative, 10% in those HIV negative with TB, 20% in the HIV positive group, 7% in those HIV positive with TB, 6% in those on ART, and 2% in those on ART with TB.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n10 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\n\nTable 3. Costs (US$), outcomes and ICERs over three years (36 one-month cycles) in a cohort with an estimated TB prevalence of 13%.\n\nStatus quo and five investment\nscenarios\n\nIn cohort of 10 000, true TB\ntreated (range)\n\nTB service costs per symptomatic individual (US$)\nProvider costs Societal costs\n\nOutcomes per symptomatic individual\n\nICERs: compared against the status quo\n\nDALYs and DALYs Deaths and Deaths\n\naverted\n\naverted\n\nProvider cost/ DALY\naverted (95% UI)\n\nSocietal cost/ DALY\naverted (95% UI)\n\nProvider cost/ death\naverted (95% UI)\n\nSocietal cost/ death\naverted (95% UI)\n\nTotal (95% UI)\n\nIncr change from base (%)\n\nTotal (95% UI)\n\nIncr % change (range)\n\nTotal Incr DALYs DALYs (95% UI) averted %\nchange (range)\n\nTotal deaths (95%\nUI)\n\nIncr deaths averted % change (range)\n\n(95% UI)\n\n(95% UI)\n\n(95% UI)\n\n(95% UI)\n\nXpert (status quo) 940 89.66 - - - 169.94 - - -\n\n4.72\n\n---\n\n0.133\n\n---\n\n---\n\n---\n\n---\n\n---\n\n(920; (87;\n\n960)\n\n92)\n\n(167; 173)\n\n(4.6; 4.8)\n\n(0.129; 0.136)\n\nXpert plus\n\n1010 92.42 2.76 178.19 8.25\n\n4.56\n\nreduction in initial\n\nLTFU (iLTFU)\n\n0.16 0.128 0.005\n\n17.42\n\n51.86\n\n601.40 1790.50\n\n(C and E)\n\n(990; (90; 1030) 95)\n\n3% (175; 5% (4.4; 4.7) 3% 181)\n\n(0.125; 0.132)\n\n4% (2.2; 117.6) (18.5; 271.0)\n\n(75.1; 3806.7)\n\n(644; 8774)\n\nXpert plus treatment from negative (TfN)\n\n1140 110.78 21.12 256.36 86.42 4.04\n\n0.68 0.115 0.018\n\n31.40\n\n128.45 1180.00 4826.60\n\n(A and E)\n\n(1120; (109; 24% (253; 51% (3.9; 4.1) 14%\n\n1160) 113)\n\n260)\n\n(0.112; 0.118)\n\n14% (24.6; 40.5) (107.2; 157.0)\n\n(905.9; 1567.3)\n\n(3939.0; 6079.2)\n\nXpert plus\n\n1210 113.55 23.89 264.60 94.66 3.88\n\nreduction in initial\n\nLTFU, and\n\ntreatment from\n\nnegative\n\n(iLTFU_TfN)\n\n0.84 0.110 0.023\n\n28.73\n\n113.82 1061.70 4205.70\n\n(A, C and E)\n\n(1190; (111; 27% (261; 56% (3.8; 4.0) 18%\n\n1230) 116)\n\n268)\n\n(0.107; 0.113)\n\n17% (23.5; 35.3) (98.3; 133.3)\n\n(853.3; 1333.3)\n\n(3569.1; 5035.1)\n\nXpert plus\n\n1420 141.01 51.35 278.87 108.93 3.42\n\nimprovements in\n\nthe negative\n\npathway (NP)\n\n1.30 0.096 0.037\n\n39.70\n\n84.19 1387.70 2943.10\n\n(B, D and E)\n\n(1390; (139; 57% (274; 64% (3.3; 3.5) 28%\n\n1450) 143)\n\n284)\n\n(0.093; 0.099)\n\n28% (35.2; 44.9) (75.0; 94.8)\n\n(1225.6; (2608.4; 1576.9) 3334.0)\n\nXpert plus\n\n1480 142.99 53.33 285.97 116.03 3.28\n\nreduction in initial\n\nLTFU, and\n\nimprovements in\n\nthe negative\n\npathway\n\n(iLTFU_NP)\n\n1.44 0.092 0.041\n\n37.02\n\n80.55 1292.97 2813.00\n\n(B, C, D and E)\n\n(1460; (141; 59% (281; 68% (3.2; 3.4) 31%\n\n1510) 145)\n\n291)\n\n(0.089; 0.094)\n\n31% (33.3; 41.3) (72.8; 89.4)\n\n(1155.8; (2527.7; 1449.9) 3139.4)\n\nIn the Table, Incr is the incremental change in costs or effectiveness from the base case. The base case in this analysis which represents the current status quo, Xpert as observed in the intervention arm of the XTEND study; dominant: less costly and more effective; dominated: more costly and less effective; The 95% uncertainty interval (UI) is shown in parentheses; ICER: Incremental cost-effectiveness ratio; DALYs: Disability Adjusted Life Years. In the scenario column, the capital letters refer to the decision points upon which the investment scenario acts, as shown in Fig 1.\n\nhttps://doi.org/10.1371/journal.pone.0251547.t003\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n11 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\n\nFig 2. Societal service-level costs (US$) per symptomatic person per episode. In the Figure, the cost of accessing care (Access) includes out of pocket and time costs incurred by patients and caregivers when accessing care; the cost of illness (Illness) includes the cost of caregiver\u2019s time, the cost of patient\u2019s time when unable to work as well as loan interest, assets sold and the cost of nutritional supplements. Xpert referes to the Xpert baseline; iLTFU (Xpert plus iLTFU) = additional investment to reduce pre-treatment loss-to-follow-up; TfN (Xpert plus TfN) = supporting clinical diagnosis of tuberculosis after a negative test results; Np (Xpert plus NP) = improving access to further tuberculosis diagnostic tests following a negative test result.\nhttps://doi.org/10.1371/journal.pone.0251547.g002\nSupporting same-day clinical diagnosis of TB after a negative tuberculosis test result increases the cost of the TB service per symptomatic person per episode by $21.12 due to the increase in patients started on TB treatment, with likewise an increase in societal costs associated with accessing treatment of $86.42 per patient (Fig 2).\nIn contrast, improving access to further diagnostic tests following a negative test result (negative pathway) increases diagnostic costs by $35 per patient due to the follow-on tests ordered, with an increase in the cost of treatment (Fig 2). This scenario increases the patient costs associated with accessing care (from $61 to $105 per patient) as patients make multiple visits for follow-on diagnostic tests and results. In addition, delays in starting treatment increase the cost of illness due to a loss of time and income.\nFor people with a negative Xpert test result, our analysis suggest that further testing (negative pathway), as conceptualised here, may be more effective at reducing mortality than empirical treatment; however the provider costs per symptomatic individual are considerably higher at $141.01 ($139 - $143) versus $110.78 ($109\u2013$113) in the negative pathway compared against treatment started following a clinical diagnosis. Similarly, societal costs are higher due to increased diagnostic visits and delays in starting treatment increase the cost of illness which is based on caregiver\u2019s time as well as patient\u2019s time unable to work.\nTransaction cost analysis\nUsing a cost-effectiveness threshold that reflects recent decisions adopted by the South African government (revealed willingness-to-pay) [59], we find that investments of up to $601 per symptomatic individual would be cost-effective. It is therefore likely that considerable\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n12 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\n\nFig 3. Provider cost-effectiveness acceptability frontiers (CEAF) at various levels of transaction costs. Where iLTFU refers to Xpert plus a reduction in initial loss to follow up scenario; TfN refers to the scenario modelling Xpert plus treatment from negative; Np refers to Xpert plus improvements in the negative pathway. The cost-effectiveness acceptability frontier (CEAF) expressing the uncertainty around the cost-effectiveness of investments, by showing which strategy is economically preferred at a range of cost-effectiveness thresholds (on the x-axis). The base case scenario for each of these comparisons is Xpert MTB/RIF, as observed in the XTEND trial. The graph is a plot of the proportion of individual runs that would be cost-effective for each intervention (y-axis) while restricting the options to only those that would be the most cost-effective (optimal) investment for at least one individual, against a range of cost-effectiveness thresholds (x-axis). As the threshold increase, the preferred option changes, the switch point being where the incremental cost-effectiveness ratio (ICER) value increases beyond the threshold [62]. The analysis is repeated at a range of transaction costs per transaction, thereby varying the costs needed to be invested to facilitate systems level change in line with the investment strategy.\nhttps://doi.org/10.1371/journal.pone.0251547.g003\ninvestments in strengthening supportive systems around TB diagnosis in South Africa would be value for money.\nFig 3 presents the cost-effectiveness acceptability frontiers, which show the optimal provider investments at a range of transaction costs and cost-effectiveness thresholds. As explained, transaction costs are modelled per transaction, and are conceptualized as the resources needed to improve decision-making within each investment scenario. Assuming no transaction costs, investing in reducing initial loss-to-follow up was the optimal investment if the cost-effectiveness threshold was below $30/ DALY averted, but at higher thresholds, the negative pathway was the optimal investment. As the investment cost per person per transaction increased, empirical treatment became the optimal investment compared to the negative pathway at lower cost-effectiveness thresholds. This is driven by a reduction in healthcare visits when patients are started on treatment empirically.\nSensitivity analyses\nDetailed results of the univariate sensitivity analyses are included in S1 Text, summarised in Fig 4.\nIntervention provider costs is dependent on population characteristics. For example, if much of the population is HIV-positive not taking antiretroviral therapy, a higher proportion with tuberculosis would test negative, leading to higher costs, though this would be mediated\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n13 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\n\nFig 4. (A-C) Results from the univariate sensitivity analyses, showing the ten parameters with the greatest influence on the (A) provider cost, (B) the societal costs, and the (C) effectiveness (DALYs) of the base case (Xpert). The full results for these analyses are presented S1 Text. In each one-way analysis, one parameter was varied by a factor of 10 from the mean to produce the low and high estimates, with all other parameters kept constant. Where DALYs are disability adjusted life years; Prov refers provider; and Soc is societal. DS treatment is drug-sensitive treatment. MDR refers to multi-drug resistant tuberculosis.\nhttps://doi.org/10.1371/journal.pone.0251547.g004\nby the expansion of universal access to antiretroviral therapy. Similarly, the effectiveness of these investments is sensitive to the health-seeking behaviours of patients and health system characteristics, specifically whether patients return for their results, the availability of chest radiographs and whether treatment is started after further diagnostic tests. The prevalence of multi-drug resistant tuberculosis and the cost of multi-drug resistant tuberculosis treatment was an important driver of costs and effectiveness of the overall results.\nDiscussion\nOur analyses build on a global body of work evaluating the use of Xpert-based diagnostic pathways [4, 10, 15, 60\u201363] by presenting the cost-effectiveness of complementary investments to strengthen the diagnostic pathway [64]. We explored how investments in health system to support the patient pathway may affect the resource use and outcomes associated with tuberculosis diagnostics. Our findings suggest that it is unlikely that a single investment or technology would dramatically improve the outcomes of symptomatic patients receiving a tuberculosis diagnostic test; instead our results suggest that investments in various parts of the care pathway could generate additional benefits, and, based on the transaction cost analysis, we show that relatively high levels of investment in health systems strengthening may be cost-effective.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n14 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\nWhen comparing across the care pathway (Table 2), our analysis finds that in a symptomatic cohort with 13% prevalence of tuberculosis, only minor reductions in mortality can be achieved by improving initial pre-treatment loss to follow up, while much larger benefits can be achieved by improving access to further tests after a negative tuberculosis test. This may be explained, in part, by the higher mortality rates observed in people who are HIV-positive with an initial negative tuberculosis test result.\nPotential drivers of investment value\nWhile the Xpert assay automates diagnostic processes and provides test results within two hours, in South Africa, Xpert machines were placed at laboratories with results delivered to health facilities in two days. Therefore, despite Xpert implementation reducing the turnaround time of results, follow-up clinic visits by patients were still required [42]. The need to improve the linkage of patients with their results has been highlighted as an important component of better tuberculosis diagnosis, however our analysis suggests that comparatively low gains in terms of mortality reduction would be achieved in such an investment scenarios. This may be explained by lower mortality rates in those with positive sputum test results, and that people-living-with-HIV who have high rates of tuberculosis-associated mortality are less likely to have a positive sputum test result. While the mortality reduction is likely to be modest, those with positive tuberculosis test results are potentially transmitting tuberculosis in communities, increasing the future burden of need at a population level [65]. These results are somewhat supported by findings from studies that highlighted the challenges of point-of-care Xpert testing at facilities in urban settings [66] and benefits in rural communities [67].\nClinical decision-making after a negative test result is important in understanding the costeffectiveness of new tuberculosis diagnostics, suggesting that greater awareness of tuberculosis symptoms among health care workers may improve outcomes and be a cost-effective intervention [10, 68, 69]. In Uganda, Hermans et al. (2017) found that tuberculosis treatment was initiated based on clinical symptoms in 17% of patients for whom an Xpert test was requested [50]. In South Africa, an evaluation of tuberculosis programmatic data found that there was a decline in the use of empirical tuberculosis treatment from 42% to 27% following the introduction of Xpert [51]. It is possible that the introduction of Xpert did not significantly reduce tuberculosis-associated mortality due, in part, to a reduction in action, including follow-on tests, after a negative test result [49]. Access to further tests such as chest radiography and mycobacterial culture of sputum after a negative result is dependent on the availability of chest radiography in close proximity to the health facility, how healthcare workers use these tests, as well as access barriers to patients [70\u201372]. Our analysis suggests that assumptions of how quickly tuberculosis treatment reduces mortality rates is a key determinant of the effectiveness of this strategy.\nInvesting in health systems strengthening\nWhile it is not possible to say whether an investment scenario is cost-effective without consensus on a cost-effectiveness threshold in South Africa, we find that investing in strengthening health systems to support the tuberculosis diagnostic algorithm is likely to be a high value investment. The outcomes of these investments are also likely to influence other disease programs and sectors [73]. We do not include these spill over benefits or costs in our analysis, and thus our estimates are conservative. Empirical work has highlighted the importance of going beyond investing in assets and technology to invest in developing agency and governance (the software capacities of health systems) [74]. Those investments are highly contextual and\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n15 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\ndifficult to cost, so while our approach highlights to decision makers the resource envelopes required, more work is needed to develop and iteratively assess context-specific investment strategies. In-depth qualitative work to understand the barriers and facilitators of health care workers\u2019 implementation of diagnostic guidelines would fill some of this gap.\nThe following limitations should be considered when interpreting our findings. Firstly, we did not model the effect of the various scenarios on the tuberculosis epidemic at a population level. While the implementation of Xpert primarily resulted in an increased identification of smear-negative tuberculosis, currently thought not to be a major driver of transmission, not including transmission in the analysis is likely to underestimate the relative benefit of reducing pre-treatment LTFU at a population level [65, 75, 76]. Secondly, while we are modelling scenarios and benefits in a nuanced way, the relationship between health system structures, health care worker -, and patient behaviour is complex and while one can observe patterns, predictions will be limited by our understanding of the mechanisms driving these patterns. Thirdly, any investment in the health system will be likely to have an impact on other associated services (externalities), the benefits of which we did not include in our analysis [42]. Lastly, while our model includes a pathway for patients to initiate multi-drug resistant tuberculosis treatment if diagnosed, and incur the associated costs, we do not attempt to estimate the true prevalence of multi-drug resistant tuberculosis or what effect the investments may have on the epidemic. In the analysis of the negative pathway, therefore, the model may be underestimating the effect of incorrectly starting an individual on drug-sensitive tuberculosis. Studies following the roll-out of Xpert have found that barriers to initiating multi-drug resistant tuberculosis persisted and that the time-to-appropriate-treatment was only slightly reduced [8, 77].\nIn conclusion, our findings suggest that within the context of a high tuberculosis prevalence setting, with a well-developed laboratory infrastructure, the implementation of new tuberculosis diagnostics should be accompanied by additional investments in the health system. Current international policy is to substantially expand and intensify tuberculosis detection, yet if this is not accompanied by investments to support decision-making after a negative test result, it is unlikely that these efforts alone will modify the tuberculosis epidemic.\nSupporting information\nS1 Text. Technical appendix. (DOCX)\nS1 Data. Parameter list accompanying manuscript Foster et al. strengthening health systems to improve the value of tuberculosis diagnostics in high-burden settings: A cost and cost-effectiveness analysis. (XLSX)\nAcknowledgments\nThis study is part of the \u201cXpert for TB: Evaluating a New Diagnostic\u201d (XTEND) project. This work would not have been possible without the many generous contributions of the study team, the respondents and health facility staff. In particular, the authors would like to thank Professor Anna Vassall for her contributions to this work. We are also grateful to the staff of the UCT Health Economics Unit; the Health Policy and Systems Division; as well as the UCT School of Public Health and Family Medicine for their insights and contributions to discussions of the work presented here.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n16 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\nAuthor Contributions\nConceptualization: Nicola Foster, Lucy Cunnama, Alison D. Grant, Susan Cleary.\nData curation: Nicola Foster, Lucy Cunnama, Kerrigan McCarthy, Lebogang Ramma, Mariana Siapka, Edina Sinanovic, Katherine Fielding.\nFormal analysis: Nicola Foster, Mariana Siapka, Susan Cleary.\nFunding acquisition: Nicola Foster, Edina Sinanovic, Gavin Churchyard.\nInvestigation: Nicola Foster, Lucy Cunnama, Kerrigan McCarthy, Lebogang Ramma, Gavin Churchyard, Katherine Fielding, Alison D. Grant, Susan Cleary.\nMethodology: Nicola Foster, Susan Cleary.\nProject administration: Nicola Foster.\nResources: Kerrigan McCarthy, Lebogang Ramma, Mariana Siapka, Gavin Churchyard.\nSupervision: Edina Sinanovic, Alison D. Grant, Susan Cleary.\nWriting \u2013 original draft: Nicola Foster.\nWriting \u2013 review & editing: Lucy Cunnama, Kerrigan McCarthy, Lebogang Ramma, Mariana Siapka, Edina Sinanovic, Gavin Churchyard, Katherine Fielding, Alison D. Grant, Susan Cleary.\nReferences\n1. Stenberg K, Axelson H, Sheehan P, Anderson I, Gu\u00a8 lmezoglu a M, Temmerman M, et al. Advancing social and economic development by investing in women\u2019s and children\u2019s health: a new Global Investment Framework. Lancet. 2014 Apr 12; 383(9925):1333\u201354. Available from: http://www.ncbi.nlm.nih. gov/pubmed/24263249 PMID: 24263249\n2. World Health Organisation. Rapid Implementation of the Xpert MTB / RIF diagnostic test. World Health. 2011.\n3. Nicol MP, Workman L, Isaacs W, Munro J, Black F, Eley B, et al. Accuracy of the Xpert MTB/RIF test for the diagnosis of pulmonary tuberculosis in children admitted to hospital in Cape Town, South Africa: a descriptive study. Lancet Infect Dis. 2011 Nov 15; 11(11):819\u201324. Available from: http://www.ncbi.nlm. nih.gov/pubmed/21764384 PMID: 21764384\n4. Vassall A, van Kampen S, Sohn H, Michael JS, John KR, den Boon S, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: a cost-effectiveness analysis. Wilson D, editor. PLoS Med. 2011 Nov 8; 8(11):e1001120. Available from: http://www.pubmedcentral.nih.gov/ articlerender.fcgi?artid=3210757&tool=pmcentrez&rendertype=abstractPMID: 22087078\n5. Menzies N a., Cohen T, Lin H-HH, Murray M, Salomon JA. Population Health Impact and Cost-Effectiveness of Tuberculosis Diagnosis with Xpert MTB/RIF: A Dynamic Simulation and Economic Evaluation. Rosen S, editor. PLoS Med. 2012 Nov 20; 9(11):e1001347. Available from: http://dx.plos.org/10. 1371/journal.pmed.1001347 PMID: 23185139\n6. Theron G, Peter J, van Zyl-Smit R, Mishra H, Streicher E, Murray S, et al. Evaluation of the Xpert MTB/ RIF assay for the diagnosis of pulmonary tuberculosis in a high HIV prevalence setting. Am J Respir Crit Care Med. 2011 Jul 1; 184(1):132\u201340. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ 21493734 PMID: 21493734\n7. Churchyard GJ, Stevens WS, Mametja LD, McCarthy KM, Chihota V, Nicol MP, et al. Xpert MTB / RIF versus sputum microscopy as the initial diagnostic test for tuberculosis: a cluster-randomised trial embedded in South African roll-out of Xpert MTB / RIF. Lancet Glob Heal. 2015; 3(8):e450\u20137. Available from: http://dx.doi.org/10.1016/S2214-109X(15)00100-X PMID: 26187490\n8. Cox HS, Mbhele S, Mohess N, Whitelaw A, Muller O, Zemanay W, et al. Impact of Xpert MTB/RIF for TB Diagnosis in a Primary Care Clinic with High TB and HIV Prevalence in South Africa: A Pragmatic Randomised Trial. PLoS Med. 2014 Nov; 11(11):e1001760. Available from: http://www.ncbi.nlm.nih. gov/pubmed/25423041 PMID: 25423041\n9. Naidoo P, du Toit E, Dunbar R, Lombard C, Caldwell J, Detjen A, et al. A Comparison of MultidrugResistant Tuberculosis Treatment Commencement Times in MDRTBPlus Line Probe Assay and\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n17 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\nXpert\u00ae MTB/RIF-Based Algorithms in a Routine Operational Setting in Cape Town. PLoS One. 2014 Jan; 9(7):e103328. Available from: http://www.ncbi.nlm.nih.gov/pubmed/25079599 PMID: 25079599\n10. Vassall A, Siapka M, Foster N, Cunnama L, Ramma L, Fielding K, et al. Cost-effectiveness of Xpert MTB/RIF for tuberculosis diagnosis in South Africa: a real-world cost analysis and economic evaluation. Lancet Glob Heal. 2017; 5(7):e710\u20139. Available from: http://linkinghub.elsevier.com/retrieve/pii/ S2214109X1730205X PMID: 28619229\n11. Hanrahan CF, Haguma P, Ochom E, Kinera I, Cobelens F, Cattamanchi A, et al. Implementation of Xpert MTB/RIF in Uganda: Missed Opportunities to Improve Diagnosis of Tuberculosis. Open Forum Infect Dis. 2016; 3(2):ofw068. Available from: http://ofid.oxfordjournals.org/lookup/doi/10.1093/ofid/ ofw068 PMID: 27186589\n12. Moyenga I, Roggi A, Sulis G, Diande S, Tamboura D, Tagliani E, et al. The impact of Xpert\u00ae MTB/RIF depends on service coordination: Experience in Burkina Faso. Int J Tuberc Lung Dis. 2015; 19(3):285\u2013 7. https://doi.org/10.5588/ijtld.14.0629 PMID: 25686135\n13. Van Den Handel T, Hampton KH, Sanne I, Stevens W, Crous R, Van Rie A, et al. The impact of Xpert MTB/RIF in sparsely populated rural settings. Int J Tuberc Lung Dis. 2015; 19(4):392\u20138. https://doi.org/ 10.5588/ijtld.14.0653 PMID: 25859993\n14. Shah SK, Kumar AMV, Dogar OF, Khan MA, Qadeer E, Tahseen S, et al. Xpert \u00ae MTB / RIF under routine conditions in diagnosing pulmonary tuberculosis: a study in two hospitals in Pakistan. 2013; I (1):20\u20132.\n15. Durovni B, Saraceni V, Van Den Hof S, Trajman A. Impact of Replacing Smear Microscopy with Xpert MTB / RIF for Diagnosing Tuberculosis in Brazil: A. 2014; 11(12).\n16. Langley I, Lin H-H, Egwaga S, Doulla B, Ku C-C, Murray M, et al. Assessment of the patient, health system, and population effects of Xpert MTB/RIF and alternative diagnostics for tuberculosis in Tanzania: an integrated modelling approach. Lancet Glob Heal. 2014 Oct; 2(10):e581\u201391. Available from: http:// linkinghub.elsevier.com/retrieve/pii/S2214109X14702918\n17. Abdurrahman ST, Emenyonu N, Obasanya OJ, Lawson L, Dacombe R, Muhammad M, et al. The hidden costs of installing xpert machines in a Tuberculosis [TB] high-burden country: Experiences from Nigeria. Pan Afr Med J. 2014; 18:277. https://doi.org/10.11604/pamj.2014.18.277.3906 PMID: 25489371\n18. Chawla KS, Kanyama C, Mbewe A, Matoga M, Hoffman I, Ngoma J, et al. Policy to practice: impact of GeneXpert MTB/RIF implementation on the TB spectrum of care in Lilongwe, Malawi. Trans R Soc Trop Med Hyg. 2016; 110(5):305\u201311. Available from: http://trstmh.oxfordjournals.org/lookup/doi/10. 1093/trstmh/trw030 PMID: 27198215\n19. Dunbar R, Naidoo P, Beyers N, Langley I. High laboratory cost predicted per tuberculosis case diagnosed with increased case finding without a triage strategy. Int J Tuberc Lung Dis. 2017; 21(9):1026\u2013 34. Available from: http://www.ingentaconnect.com/content/10.5588/ijtld.17.0156 PMID: 28826453\n20. Lin H-H, Langley I, Mwenda R, Doulla B, Egwaga S, Millington KA, et al. A modelling framework to support the selection and implementation of new tuberculosis diagnostic tools. Int J Tuberc lung Dis. 2011 Aug; 15(8):996\u20131004. Available from: http://www.ncbi.nlm.nih.gov/pubmed/21740663 PMID: 21740663\n21. Hallett TB, Eaton JW. A side door into care cascade for HIV-infected patients? J Acquir Immune Defic Syndr. 2013; 63 Suppl 2:S228\u201332. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23764640 PMID: 23764640\n22. Naidoo P, Theron G, Rangaka MX, Chihota VN, Vaughan L, Brey ZO, et al. The South African Tuberculosis Care Cascade: Estimated Losses and Methodological Challenges. J Infect Dis. 2017; 216 (suppl_7):S702\u201313. Available from: http://academic.oup.com/jid/article/216/suppl_7/S702/4595546 PMID: 29117342\n23. Lessard C, Contandriopoulos AP, Beaulieu MD. The role (or not) of economic evaluation at the micro level: Can Bourdieu\u2019s theory provide a way forward for clinical decision-making? Soc Sci Med. 2010; 70 (12):1948\u201356. https://doi.org/10.1016/j.socscimed.2010.03.013 PMID: 20382461\n24. Arago\u00b4n AO. A Case for Surfacing Theories of Change for Purposeful Organisational Capacity Development. IDS Bull. 2010 May; 41(3):36\u201346. Available from: http://doi.wiley.com/10.1111/j.1759-5436.2010. 00135.x\n25. Meyer-rath G, Over M, Klein DJ, Bershteyn A. The Cost and Cost-Effectiveness of Alternative Strategies to Expand Treatment to HIV-Positive South Africans: Scale Economies and Outreach Costs. Heal Dev. 2015;(17). Available from: http://www.bu.edu/cghd/files/2015/05/HDDP-17-TasP-cost-modelreport.pdf\n26. Bozzani FM, Mudzengi D, Sumner T, Gomez GB, Hippner P, Cardenas V, et al. Empirical estimation of resource constraints for use in model-based economic evaluation: An example of TB services in South\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n18 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\nAfrica. Cost Eff Resour Alloc. 2018; 16(1):1\u201310. Available from: https://doi.org/10.1186/s12962-0180113-z PMID: 30069166\n27. Vassall A, Mangham-jefferies L, Gomez GB, Pitt C, Foster N. Including demand and supply side constraints in economic evaluations: the case of Malaria and TB control. Health Econ. 2015;\n28. Brown C, Jackson P. Public sector economics. 3rd ed. Basil Blackwell; 1987.\n29. Lin H, Dowdy D, Dye C. The impact of new tuberculosis diagnostics on transmission: why context matters.: 1\u201322.\n30. Steingart K, Sohn H, Schiller I, Kloda L, Boehme C, Pai M, et al. Xpert \u00ae MTB / RIF assay for pulmonary tuberculosis and rifampicin resistance in adults (Review). 2013;(1).\n31. Steingart KR, Henry M, Ng V, Hopewell PC, Ramsay A, Cunningham J, et al. Fluorescence versus conventional sputum smear microscopy for tuberculosis: a systematic review. Lancet. 2006 Sep; 6(September):570\u201381. Available from: http://www.ncbi.nlm.nih.gov/pubmed/16931408 PMID: 16931408\n32. Boehme CC, Nicol MP, Nabeta P, Michael JS, Gotuzzo E, Tahirli R, et al. Feasibility, diagnostic accuracy, and effectiveness of decentralised use of the Xpert MTB/RIF test for diagnosis of tuberculosis and multidrug resistance: a multicentre implementation study. Lancet. 2011 Apr 30; 377(9776):1495\u2013505. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3085933&tool= pmcentrez&rendertype=abstract PMID: 21507477\n33. McCarthy KM, Grant AD, Chihota V, Ginindza S, Mvusi L, Churchyard GJ, et al. Implementation and Operational Research: What Happens After a Negative Test for Tuberculosis? Evaluating Adherence to TB Diagnostic Algorithms in South African Primary Health Clinics. J Acquir Immune Defic Syndr. 2016.\n34. Johnson LF, Mossong J, Dorrington RE, Schomaker M, Hoffmann CJ, Keiser O, et al. Life expectancies of South african adults starting antiretroviral treatment: collaborative analysis of cohort studies. PLoS Med. 2013 May; 10(4):e1001418. Available from: http://www.pubmedcentral.nih.gov/articlerender. fcgi?artid=3621664&tool=pmcentrez&rendertype=abstract PMID: 23585736\n35. Statistics South Africa. Mortality and causes of death in South Africa, 2013: Findings from death notification. Statistical release. 2013. http://www.statssa.gov.za/Publications/P03093/P030932010.pdf% 5Cnhttp://www.statssa.gov.za/Publications/P03093/P030932009.pdf\n36. Romanowski K, Baumann B, Basham CA, Ahmad Khan F, Fox GJ, Johnston JC. Long-term all-cause mortality in people treated for tuberculosis: a systematic review and meta-analysis. Lancet Infect Dis. 2019; 3099(19):1\u20139. Available from: http://dx.doi.org/10.1016/S1473-3099(19)30309-3 PMID: 31324519\n37. Tiemersma EW, van der Werf MJ, Borgdorff MW, Williams BG, Nagelkerke NJD. Natural history of tuberculosis: duration and fatality of untreated pulmonary tuberculosis in HIV negative patients: a systematic review. PLoS One. 2011 Jan [cited 2014 May 28]; 6(4):e17601. Available from: http://www. pubmedcentral.nih.gov/articlerender.fcgi?artid=3070694&tool=pmcentrez&rendertype=abstract PMID: 21483732\n38. Andrews JR, Noubary F, Walensky RP, Cerda R, Losina E, Horsburgh CR. Risk of progression to active tuberculosis following reinfection with Mycobacterium tuberculosis. Clin Infect Dis. 2012 Mar [cited 2014 May 24]; 54(6):784\u201391. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi? artid=3284215&tool=pmcentrez&rendertype=abstract PMID: 22267721\n39. Mohr E, Cox V, Wilkinson L, Moyo S, Hughes J, Daniels J, et al. Programmatic treatment outcomes in HIV-infected and uninfected drug-resistant tuberculosis patients in Khayelitsha, South Africa. Trans R Soc Trop Med Hyg. 2015;(May):425\u201332. Available from: http://trstmh.oxfordjournals.org/cgi/doi/10. 1093/trstmh/trv037 PMID: 25979526\n40. Salomon JA, Haagsma JA, Davis A, De Noordhout CM, Polinder S, Havelaar AH, et al. Disability weights for the Global Burden of Disease 2013 study. Lancet Glob Heal. 2015; 3(11):e712\u201323. https:// doi.org/10.1016/S2214-109X(15)00069-8 PMID: 26475018\n41. Kastien-Hilka T, Rosenkranz B, Sinanovic E, Bennett B, Schwenkglenks M. Health-related quality of life in South African patients with pulmonary tuberculosis. PLoS One. 2017; 12(4):e0174605. Available from: http://dx.doi.org/10.1371/journal.pone.0174605 PMID: 28426759\n42. Cunnama L, Sinanovic E, Ramma L, Foster N, Berrie L, Stevens W, et al. Using Top-down and Bottomup Costing Approaches in LMICs: The Case for Using Both to Assess the Incremental Costs of New Technologies at Scale. Health Econ. 2016; 25(January). https://doi.org/10.1002/hec.3295 PMID: 26763594\n43. Vassall A, Siapka M, Foster N, Cunnama L, Ramma L, Fielding K, et al. Revisitig the cost-effectiveness of Xpert MTB/RIF: lessons learned from South Africa. Lancet Glob Heal. 2017.\n44. Foster N, Vassall A, Cleary S, Cunnama L, Churchyard G, Sinanovic E. The economic burden of TB diagnosis and treatment in South Africa. Soc Sci Med. 2015; 130:42\u201350. Available from: http:// linkinghub.elsevier.com/retrieve/pii/S0277953615000726 PMID: 25681713\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n19 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\n45. Sendi PP, Craig BA, Pfluger D, Gafni A, Bucher HC. Systematic validation of disease models for pharmacoeconomic evaluations. J Eval Clin Pract. 1999; 5:283\u201395. Available from: http://dx.doi.org/10. 1046/j.1365-2753.1999.00174.x PMID: 10461580\n46. Railsback SF, Grimm V. Agent-based and individual-based modeling: a practical introduction. Princeton University Press; 2012. 1\u2013329 p.\n47. Cunnama L, Sinanovic E, Ramma L, Foster N, Berrie L, Stevens W, et al. The costs of novel TB diagnostics during scale-up: evidence form the roll-out of Xpert MTB/RIF in South Africa. J Health Econ. 2016;January.\n48. Sinanovic E, Ramma L, Vassall A, Azevedo V, Wilkinson L, Ndjeka N, et al. Impact of reduced hospitalisation on the cost of treatment for drug-resistant tuberculosis in South Africa. 2015; 19(October 2014):172\u20138.\n49. McCarthy K, Grant AD, Chihota V, Ginindza S, Mvusi L, Mametja D, et al. What happens after a negative test for Tuberculosis? Evaluating adherence to TB diagnostic algorithms in South African primary health care clinics. J Acquir Immune Defic Syndr. 2015.\n50. Hermans SM, Babirye JA, Mbabazi O, Kakooza F, Colebunders R, Castelnuovo B, et al. Treatment decisions and mortality in HIV-positive presumptive smear-negative TB in the Xpert\u00ae MTB/RIF era: a cohort study. BMC Infect Dis. 2017; 17(1):433. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ 28622763%0Ahttp://bmcinfectdis.biomedcentral.com/articles/10.1186/s12879-017-2534-2 PMID: 28622763\n51. Hermans S, Caldwell J, Kaplan R, Cobelens F, Wood R. Rapid diagnostic test roll-out, South Africa The impact of the roll-out of rapid molecular diagnostic testing for tuberculosis on empirical treatment in Cape Town, South Africa. http://www.who.int/bulletin/online_first/BLT.16.185314.pdf?ua=1\n52. Salomon J a, Vos T, Hogan DR, Gagnon M, Naghavi M, Mokdad A, et al. Common values in assessing health outcomes from disease and injury: disability weights measurement study for the Global Burden of Disease Study 2010. Lancet. 2012 Dec 15 [cited 2013 May 22]; 380(9859):2129\u201343. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23245605 PMID: 23245605\n53. Drummond M, Schulpher M, Claxton K, Stoddart G, Torrance G. Methods for the economic evaluation of health care programmes. 2015.\n54. Barton GR, Briggs AH, Fenwick EAL. Optimal cost-effectiveness decisions: The role of the cost-effectiveness acceptability curve (CEAC), the cost-effectiveness acceptability frontier (CEAF), and the expected value of perfection information (EVPI). Value Heal. 2008; 11(5):886\u201397. Available from: http:// dx.doi.org/10.1111/j.1524-4733.2008.00358.x\n55. Baltussen RMPM Hutubessy RCW, Evans DB. Uncertainty in cost-effectiveness analyses: probabilistic uncertainty analysis and stochastic league tables. Int J Technol Assess Health Care. 2002; 18(01):112\u2013 9. Available from: http://www.who.int/healthinfo/paper34.pdf\n56. Hanson K, Ranson MK, Oliveira-cruz V, Mills A. Expanding access to priority health interventions: a framework for understanding the constraints to scaling-up. J Int Dev. 2003 Jan [cited 2014 May 1]; 15 (1):1\u201314. Available from: http://doi.wiley.com/10.1002/jid.963\n57. Leslie HH, Sun Z, Kruk ME. Association between infrastructure and observed quality of care in 4 healthcare services: A cross-sectional study of 4,300 facilities in 8 countries. 2017;1\u201316. Available from: http://journals.plos.org/plosmedicine/article/file?id=10.1371/journal.pmed.1002464&type=printable\n58. Karat AS, Omar T, Von Gottberg A, Tlali M, Chihota VN, Churchyard GJ, et al. Autopsy prevalence of tuberculosis and other potentially treatable infections among adults with advanced HIV enrolled in outpatient care in South Africa. PLoS One. 2016; 11(11):1\u201316. https://doi.org/10.1371/journal.pone. 0166158 PMID: 27829072\n59. Meyer-Rath G, Van Rensburg C, Larson B, Jamieson L, Rosen S. Revealed willingness-to-pay versus standard cost-effectiveness thresholds: Evidence from the South African HIV Investment Case. PLoS One. 2017; 12(10):1\u20139. https://doi.org/10.1371/journal.pone.0186496 PMID: 29073167\n60. Steffen R, Menzies D, Oxlade O, Pinto M, de Castro AZ, Monteiro P, et al. Patients\u2019 costs and costeffectiveness of tuberculosis treatment in DOTS and non-DOTS facilities in Rio de Janeiro, Brazil. PLoS One. 2010 Jan [cited 2012 Apr 23]; 5(11):e14014. Available from: http://www.pubmedcentral.nih. gov/articlerender.fcgi?artid=2984447&tool=pmcentrez&rendertype=abstract PMID: 21103344\n61. Zwerling A, White RG, Vassall A, Cohen T, Dowdy DW, Houben RMGJ. Modeling of novel diagnostic strategies for active tuberculosis\u2014a systematic review: current practices and recommendations. PLoS One. 2014 Jan [cited 2014 Dec 28]; 9(10):e110558. Available from: http://www.pubmedcentral.nih.gov/ articlerender.fcgi?artid=4207742&tool=pmcentrez&rendertype=abstract PMID: 25340701\n62. Van\u2019t Hoog AH, Cobelens F, Vassall A, van Kampen S, Dorman SE, Alland D, et al. Optimal Triage Test Characteristics to Improve the Cost-Effectiveness of the Xpert MTB/RIF Assay for TB Diagnosis: A Decision Analysis. PLoS One. 2013 Jan [cited 2014 Feb 20]; 8(12):e82786. Available from: http://www.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n20 / 21\n\nPLOS ONE\n\nCost-effectiveness of strengthening health systems to support TB diagnosis in South Africa\npubmedcentral.nih.gov/articlerender.fcgi?artid=3867409&tool=pmcentrez&rendertype=abstract PMID: 24367555\n63. Dheda K, Theron G, Welte A. Cost-effectiveness of Xpert MTB/RIF and investing in health care in Africa. Lancet Glob Heal. 2014 Oct [cited 2014 Sep 29]; 2(10):e554\u20136. Available from: http://linkinghub. elsevier.com/retrieve/pii/S2214109X14703055 PMID: 25304623\n64. Vassall A, Siapka M, Foster N, Cunnama L, Ramma L, Fielding K, et al. Cost-effectiveness of Xpert MTB/RIF for tuberculosis diagnosis in South Africa: a real-world cost analysis and economic evaluation. Lancet Glob Heal. 2017; 5(7). https://doi.org/10.1016/S2214-109X(17)30205-X PMID: 28619229\n65. Dowdy DW, Chaisson RE, Maartens G, Corbett EL, Dorman SE. Impact of enhanced tuberculosis diagnosis in South Africa: a mathematical model of expanded culture and drug susceptibility testing. Proc Natl Acad Sci U S A. 2008 Aug 12; 105(32):11293\u20138. Available from: http://www.pubmedcentral.nih. gov/articlerender.fcgi?artid=2516234&tool=pmcentrez&rendertype=abstract PMID: 18695217\n66. Van Rie A, Page-Shipp L, Scott L, Sanne I, Stevens W. Xpert(\u00ae) MTB/RIF for point-of-care diagnosis of TB in high-HIV burden, resource-limited countries: hype or hope? Expert Rev Mol Diagn. 2010 Oct; 10 (7):937\u201346. Available from: http://www.ncbi.nlm.nih.gov/pubmed/20964612 PMID: 20964612\n67. Lessells RJ, Cooke GS, McCgrath N, Nicol MP, Newell ML, Godfrey-Faussett P. Impact of point-of-care Xpert MTB/RIF on tuberculosis treatment initiation: a cluster randomised trial. AJRCCM. 2017; xxx:1\u2013 55.\n68. Schnippel K, Long L, Stevens WS, Sanne I, Rosen S. Diagnosing Xpert MTB / RIF-negative TB: Impact and cost of alternative algorithms for South Africa. South African Med J. 2013; 103(2):101\u20136.\n69. Fairall L, Bachmann MO, Zwarenstein M, Bateman ED, Niessen LW, Lombard C, et al. Cost-effectiveness of educational outreach to primary care nurses to increase tuberculosis case detection and improve respiratory care: economic evaluation alongside a randomised trial. 2010; 15(3):277\u201386.\n70. Abimbola S, Ukwaja KN, Onyedum CC, Negin J, Jan S, Martiniuk ALC. Transaction costs of access to health care: Implications of the care-seeking pathways of tuberculosis patients for health system governance in Nigeria. Glob Public Health. 2015; 10(9):1060\u201377. Available from: http://www.tandfonline.com/ doi/full/10.1080/17441692.2015.1007470 PMID: 25652349\n71. Churchyard G, McCarthy K, Fielding K. Effect of XPert MTB/RIF on early mortality in adults with suspected TB: a pragmatic randomised trial. In: Conference on Retroviruses and Opportunistic Infections. Boston, USA; 2014. p. Oral Abstract 95.\n72. Lawn SD, Nicol MP, Corbett EL, Menzies NA, Cohen T, Murray M, et al. Effect of empirical treatment on outcomes of clinical trials of diagnostic assays for tuberculosis. Lancet Infect Dis. 2015 Jan [cited 2014 Dec 15]; 15(1):16\u20137. Available from: http://linkinghub.elsevier.com/retrieve/pii/S1473309914710266 PMID: 25541164\n73. Remme M, Martinez-Alvarez M, Vassall A. Cost-Effectiveness Thresholds in Global Health: Taking a Multisectoral Perspective. Value Heal. 2017; 20(4):699\u2013704. Available from: http://dx.doi.org/10.1016/j. jval.2016.11.009 PMID: 28408014\n74. Gilson L, Barasa E, Nxumalo N, Cleary S, Goudge J, Molyneux S, et al. Everyday resilience in district health systems: emerging insights from the front lines in Kenya and South Africa. BMJ Glob Heal. 2017; 2(2):e000224. Available from: http://gh.bmj.com/lookup/doi/10.1136/bmjgh-2016-000224 PMID: 29081995\n75. Yates TA, Khan PY, Knight GM, Taylor JG, McHugh TD, Lipman M, et al. The transmission of Mycobacterium tuberculosis in high burden settings. Lancet Infect Dis. 2016; 16(2):227\u201338. Available from: http://linkinghub.elsevier.com/retrieve/pii/S1473309915004995 PMID: 26867464\n76. Dye C. The potential impact of new diagnostic tests on tuberculosis epidemics. Indian J Med Res. 2012 May; 135(5):737\u201344. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3401708&tool= pmcentrez&rendertype=abstract PMID: 22771607\n77. Naidoo P, Dunbar R, Lombard C, du Toit E, Caldwell J, Detjen A, et al. Comparing Tuberculosis Diagnostic Yield in Smear/Culture and Xpert\u00ae MTB/RIF-Based Algorithms Using a Non-Randomised Stepped-Wedge Design. PLoS One. 2016; 11(3):e0150487. Available from: http://www.ncbi.nlm.nih. gov/pubmed/26930400 PMID: 26930400\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0251547 May 14, 2021\n\n21 / 21\n\n\n",
    "authors": [
      "Nicola Foster",
      "Lucy Cunnama",
      "Kerrigan McCarthy",
      "Lebogang Ramma",
      "Mariana Siapka",
      "Edina Sinanovic",
      "Gavin Churchyard",
      "Katherine Fielding",
      "Alison D. Grant",
      "Susan Cleary"
    ],
    "doi": "10.1371/journal.pone.0251547",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://dx.plos.org/10.1371/journal.pone.0251547"
  },
  {
    "key": "EDXGN3P9",
    "title": "Cost-effectiveness of Xpert\u00ae MTB/RIF for diagnosing pulmonary tuberculosis in the United States",
    "abstract": "OBJECTIVE\u2014We evaluated the cost-effectiveness of incorporating Xpert into TB diagnostic algorithms in the United States compared to existing diagnostics. DESIGN\u2014A decision-analysis model compared current TB diagnostic algorithms in the United States to algorithms incorporating Xpert. Primary outcomes were the costs and quality-adjusted life years (QALYs) accrued with each strategy; cost-effectiveness was represented using incremental cost-effectiveness ratios (ICER).\nRESULTS\u2014Xpert testing of a single sputum sample from TB suspects is expected to result in lower total health care costs per patient (US$2673) compared to diagnostic algorithms using only sputum microscopy and culture (US$2728) and improved health outcomes (6.32 QALYs gained per 1000 TB suspects). Compared to existing molecular assays, implementation of Xpert in the United States would be considered highly cost-effective (ICER US$39 992 per QALY gained).\nCONCLUSION\u2014TB diagnostic algorithms incorporating Xpert in the United States are highly cost-effective.",
    "full_text": "NIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nNIH Public Access\nAuthor Manuscript\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14. Published in final edited form as:\nInt J Tuberc Lung Dis. 2013 October ; 17(10): 1328\u20131335. doi:10.5588/ijtld.13.0095.\nCost-effectiveness of Xpert\u00ae MTB/RIF for diagnosing pulmonary tuberculosis in the United States\nH. W. Choi*, K. Miele*, D. Dowdy\u2020, and M. Shah*,\u2021 *Johns Hopkins University School of Medicine, Baltimore, Maryland, USA \u2020Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, USA \u2021Johns Hopkins University Department of Medicine, Division of Infectious Diseases, Baltimore, Maryland, USA\nSUMMARY\nSETTING\u2014Conventional approaches to tuberculosis (TB) diagnosis and resistance testing are slow. The Xpert\u00ae MTB/RIF assay is an emerging molecular diagnostic assay for rapid TB diagnosis, offering results within 2 hours. However, the cost-effectiveness of implementing Xpert in settings with low TB prevalence, such as the United States, is unknown.\nOBJECTIVE\u2014We evaluated the cost-effectiveness of incorporating Xpert into TB diagnostic algorithms in the United States compared to existing diagnostics.\nDESIGN\u2014A decision-analysis model compared current TB diagnostic algorithms in the United States to algorithms incorporating Xpert. Primary outcomes were the costs and quality-adjusted life years (QALYs) accrued with each strategy; cost-effectiveness was represented using incremental cost-effectiveness ratios (ICER).\nRESULTS\u2014Xpert testing of a single sputum sample from TB suspects is expected to result in lower total health care costs per patient (US$2673) compared to diagnostic algorithms using only sputum microscopy and culture (US$2728) and improved health outcomes (6.32 QALYs gained per 1000 TB suspects). Compared to existing molecular assays, implementation of Xpert in the United States would be considered highly cost-effective (ICER US$39 992 per QALY gained).\nCONCLUSION\u2014TB diagnostic algorithms incorporating Xpert in the United States are highly cost-effective.\nKeywords GeneXpert; MTD; diagnostics\nTuberculosis (TB) is the second most common cause of death due to infectious disease in the world, with over 10 000 cases of active TB disease in the United States.1 Among the challenges in controlling TB is the lack of rapid, accurate diagnostic tests. Currently, sputum smear microscopy is used as the initial test in most diagnostic algorithms, but it has poor sensitivity, leading to missed diagnoses.2 Smear microscopy is also a marker of infectiousness, and current US guidelines suggest isolation of TB suspects with smearpositive results pending results from mycobacterial culture or response to treatment.3\u20135\n\u00a9 2013 The Union Correspondence to: Maunank Shah, Johns Hopkins University, Department of Medicine, Division of Infectious Disease, Room 118, 1503 East Jefferson St, Baltimore, MD 21231, USA. Tel: (+1) 443 287 0401. Fax: (+1) 410 955 0740. mshah28@jhmi.edu. Conflict of interest: none declared.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 2\n\nHowever, smear microscopy is not specific to Mycobacterium tuberculosis and has low positive predictive value in low-prevalence settings such as the United States, leading to unnecessary treatment and prolonged hospitalizations. Sputum culture and conventional drug susceptibility testing (DST) are utilized as the reference standard in the United States, but these take weeks to provide results, leading to diagnostic and therapeutic delays.2\nThe Amplified MTD\u00ae (Mycobacterium Tuberculosis Direct) test (Gen-Probe, San Diego, CA, USA), a molecular assay that detects M. tuberculosis genetic material and can provide a rapid diagnosis of TB disease, is Food and Drug Administration (FDA) approved for use in the United States.6,7 For smear-positive samples, the Amplified MTD test has high sensitivity of between 91% and 100% and a high negative predictive value.6\u20138 It thus allows rapid confirmation of M. tuberculosis in smear-positive samples, and it can reduce the duration of respiratory isolation and prevent empiric medication expenses.6\u20138 However, its sensitivity on smear-negative samples is approximately 50% and remains suboptimal, and it is not a replacement for mycobacterial culture.3,6\u20138 The Centers for Disease Control and Prevention nonetheless recommends that at least one respiratory specimen from all TB suspects be sent for molecular testing.7 Despite these recommendations, however, broad implementation of molecular testing remains limited, as they are labor- and resourceintensive. In our local setting, for example, few hospital laboratories perform MTD, and the state mycobacteriology reference laboratory performs routine MTD testing only for smearpositive samples.3\nImproved molecular TB diagnostic systems are now commercially available that are faster and require less labor than MTD, with improved performance characteristics. The Xpert\u00ae MTB/RIF test (Cepheid Inc, Sunnyvale, CA, USA) is an automated nucleic-acid amplification test for the diagnosis of TB, offering results in 2 h. Importantly, Xpert requires minimal laboratory equipment, space and technician time and also provides rapid identification of rifampin resistance, allowing earlier treatment of drug-resistant TB. Studies have found sensitivity and specificity for TB and drug resistance to be >97% on smearpositive samples, while sensitivity on smear-negative samples may be as high as 70\u2013 80%.9,10 It was therefore endorsed by the World Health Organization (WHO) for the detection of pulmonary TB.11 Although Xpert is not yet FDA-approved in the United States, it may be implemented in US laboratories after appropriate internal laboratory validations, with results reported with a disclaimer.\nSeveral mycobacterial laboratories in the United States are currently considering the adoption of Xpert, but its cost-effectiveness in low TB prevalence settings is unknown. In low-income settings globally, Xpert is available at a negotiated discounted price; implementation in such settings with high TB incidence was found to be cost-effective.12\u201314 However, the United States does not qualify for reduced Xpert pricing and its optimal role in existing diagnostic algorithms is unclear. We thus sought to evaluate the cost-effectiveness of incorporating Xpert into TB diagnostic algorithms compared to current approaches in the United States.\n\nMETHODS\nThis economic evaluation was conducted from a health system perspective with a target population of individuals with suspected pulmonary TB disease in the United States. Target audiences include health departments, hospitals and TB control programs. A timeframe of 1 year was used and the analytic horizon extended to the life expectancy of the patients. Model development and analysis utilized TreeAge Software (TreeAge Software Inc, Williamstown, MA, USA).\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 3\n\nStudy model\nA decision-analysis model was constructed to determine if TB diagnostic algorithms that incorporate Xpert are cost-effective compared to current TB diagnostic strategies using MTD or without any molecular testing (Figure 1). In all model arms, patients submit three sputum samples for mycobacterial testing and undergo a chest radiograph and clinical evaluation; mycobacterial testing (i.e., `conventional diagnostics') includes smear microscopy and liquid culture, which are performed on all sputum samples, and DST performed on positive cultures. Treatment was assumed to be initiated and/or adjusted based on diagnostic test results. We compared five strategies for pulmonary TB diagnosis with and without the incorporation of molecular testing (the model details are described in Appendix A).*\n\u2022 Algorithm 1: a `no molecular testing' algorithm, in which sputum samples are sent for conventional diagnostics; no molecular tests are employed.\n\u2022 Algorithm 2: a `selective MTD' algorithm, in which sputum samples are sent for conventional diagnostics; MTD testing is selectively performed on one sample only if smear microscopy is positive.\n\u2022 Algorithm 3: an `intensive MTD' algorithm, in which sputum samples are sent for conventional diagnostics; MTD testing is performed on one sample, regardless of smear microscopy results.7\n\u2022 Algorithm 4: a `selective Xpert' algorithm, in which sputum samples are sent for conventional diagnostics; Xpert testing is selectively performed on one sample only if smear microscopy is positive.\n\u2022 Algorithm 5: an `intensive Xpert' algorithm, in which sputum samples are sent for conventional diagnostics; Xpert testing is performed on one sample, regardless of smear microscopy results.\n\nEpidemiologic and diagnostic parameters\nData regarding disease prevalence and diagnostic test performance are summarized in Table 1.\n\nEstimation of costs\nCosts for diagnostics are shown in Table 1, with additional details in Appendix B.29 The amount of staff time, consumable supplies and equipment utilized for each test system were determined through direct observation of testing procedures at the Maryland Department of Health and Mental Hygiene Mycobacteriology Laboratory and the Johns Hopkins Hospital Mycobacteriology Laboratory. Costs of key consumable items and equipment for each diagnostic test were obtained from manufacturer/distributor quotations and published literature; salaries and wages were based on published estimates.8 For the base case, we utilized published estimates from the manufacturer for commercial pricing of Xpert cartridges (US$71.63 per cartridge) and 4-cartridge Xpert instrument (US$78 200).13\nLittle has been published on TB treatment costs in the United States.15 We utilized local health department invoices and budget records to determine out-patient costs of TB treatment and monitoring, including staff labor, directly observed therapy and medications.30 For the base case, we assumed that 20% of TB suspects were initially\n\nNIH-PA Author Manuscript\n\n*The Appendices are available in the online version of this article at http://www.ingentaconnect.com/content/iuatld/ijtld/ 2013/00000017/00000010/art00014\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 4\n\nevaluated as in-patients, and this included initial hospitalization costs. We conducted sensitivity analysis on all key cost parameters. Costs are presented in 2012 US dollars.\nOutcome parameters\nThe primary outcomes were the expected costs per patient with suspected pulmonary TB, quality-adjusted life years (QALYs) accrued per patient, and the cost-effectiveness of the proposed diagnostic algorithms. Cost-effectiveness was represented using incremental costeffectiveness ratios (ICERs) comparing new TB diagnostic algorithms incorporating Xpert with standard approaches without molecular diagnostics or algorithms incorporating MTD. QALY calculations were based on duration with and without active TB and/or antituberculosis treatment; future QALYs were discounted at 3% (Table 1).\nWe further explored parameter uncertainty through probabilistic sensitivity analysis using Monte-Carlo simulation methods. We utilized a willingness-to-pay (WTP) threshold of US $50 000 per QALY gained to determine cost-effectiveness, and explored other thresholds using cost-effectiveness acceptability curves.11,31\nEthics statement\nThis economic evaluation was approved by the ethics committee at the Johns Hopkins University School of Medicine (Baltimore, MD, USA) as exempt and did not constitute human subjects research.\n\nRESULTS\nLaboratory and health system costs for tuberculosis diagnostic algorithms\nIncremental laboratory costs of each diagnostic algorithm are shown in Table 2. In the base case, laboratory costs for an algorithm without molecular testing were US$158 per patient with suspected pulmonary TB (Algorithm 1). Implementing `intensive' Xpert testing (Algorithm 5) for at least one sputum sample increases laboratory costs to US$256 per patient (incremental US$98 [62%] increase). `Selective' deployment of Xpert only for smear-positive samples is cheaper than `intensive' molecular testing and costs US$162 per patient (Table 2).\n\nWhen all health system costs are considered, a strategy without molecular testing was found to be the most costly approach ($2728 per patient, Algorithm 1). Molecular testing with MTD or Xpert either `selectively' or `intensively' (Algorithms 2\u20135) results in less empiric treatment and shorter hospitalizations, and were subsequently less costly ($2480 to $2673 per patient) than a strategy without molecular testing (Table 2). Overall, `intensive' Xpert testing ($2673 per patient) was expected to cost US$191 more (8% increase in total costs) than a `selective' strategy of Xpert testing of only smear-positive samples ($2482 per patient).\n\nIn one-way sensitivity analysis, the incremental cost of `intensive' Xpert implementation compared to an algorithm without molecular testing was most influenced by the cost of Xpert (incremental \u2212$133 [cost-saving] to $685 for costs per test ranging from $20 to $838), the percentage of patients hospitalized for initial evaluation (incremental \u2212$702 to $106 for percentage hospitalization ranging from 100% to 0%) and specificity of smear microscopy (incremental \u2212$561 to $198 for specificity 91% to 100%, Appendix Figure C.1).\n\nEffects\n\nOverall, a strategy without molecular testing (Algorithm 1) resulted in the fewest QALYs experienced (22.0862) per patient, while `intensive' Xpert testing resulted in the most\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 5\n\nQALYs experienced per patient (22.0925 QALYs; incremental 6.32 QALYs gained per 1000 patients, Table 2). Xpert testing was expected to reduce the time to diagnosis for TB cases with subsequent QALY gains (Table 2, estimated time to diagnosis reduced from 16.3 days without molecular testing to 2.7 days with `intensive' Xpert testing among those with TB). `Intensive' Xpert testing was associated with an incremental increase of 4.74 QALYs gained per 1000 TB suspects compared with `selective' Xpert testing (Table 2).\nIn one-way sensitivity analyses, there were no circumstances in which a diagnostic algorithm without molecular testing (Algorithm 1) led to better health outcomes than strategies incorporating molecular testing (Algorithms 2\u20135, Appendix Figure C.2).\nCost-effectiveness\nA strategy of no molecular testing (Algorithm 1) was dominated (i.e., was both more costly and less effective) by all strategies incorporating either MTD or Xpert for TB diagnosis (Table 2). Replacing MTD with Xpert was also found to be cost-effective. Compared to `selective' MTD testing of only smear-positive sputum samples (Algorithm 2), utilizing Xpert `selectively' was associated with an ICER of only $23 111 per QALY-gained (Table 2), and was considered highly cost-effective compared to the WTP for the United States.32\nThe `intensive' Xpert algorithm was associated with an ICER of $16 289 per QALY gained compared to the `intensive' MTD algorithm. For programs considering `selective' vs. `intensive' Xpert implementation, `intensive' Xpert testing was highly cost-effective compared to the `selective' Xpert algorithm (ICER of $40 312 per QALY-gained, Table 2).\nIn one-way sensitivity analysis, we found that the `intensive' Xpert algorithm dominated (i.e., negative ICER) the algorithm without molecular testing (Algorithm 1) in most circumstances; `intensive' Xpert was cost-effective at the WTP in all scenarios except when cost per Xpert test rose above $475 or Xpert specificity was lower than 96%. For out-patient evaluations, `intensive' Xpert was cost-effective compared to Algorithm 1 (ICER US$16 900 per QALY gained); by contrast, for in-patient evaluations, `intensive' Xpert dominated Algorithm 1. Additional sensitivity analysis comparing the `intensive' Xpert algorithm with `selective' molecular algorithms is shown in Appendix Figures C.3 and C.4.\nResults of a probabilistic sensitivity analysis (shown in Figure 2 and Appendix D) found that the `intensive' Xpert algorithm was cost-effective in more than 99% of simulations compared to diagnostic algorithms without molecular testing.\n\nDISCUSSION\nIn the United States, diagnostic testing for pulmonary TB includes sputum smear microscopy and mycobacterial culture, but can lead to diagnostic delays and inappropriate empiric anti-tuberculosis treatment. Following the WHO endorsement of the Xpert assay for the rapid detection of TB and drug-resistant TB, several low- and middle-income countries with a high TB burden have taken advantage of negotiated price reductions to scale up Xpert implementation.11,33 Although the costs of Xpert testing are likely to be substantially higher in the United States, we show that the incorporation of Xpert into TB diagnostic algorithms in the United States would be highly cost-effective compared to current algorithms that utilize conventional diagnostics and existing molecular assays. From a health system standpoint, we found that the implementation of Xpert testing as an addition to microscopy and culture is expected to be less costly and more effective (i.e., cost-saving) than diagnostic algorithms that rely on sputum microscopy and culture alone. Some programs have chosen to implement molecular assays selectively only for smear-positive sputum samples;3 however, we found that `intensive' implementation of Xpert for at least one sputum sample\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 6\n\nfrom all individuals with signs/symptoms of TB was highly cost-effective compared to the more selective approach.\nDespite the availability of mycobacterial culture as the reference standard in current US diagnostic algorithms, the addition of rapid Xpert testing leads to a gain in QALYs experienced by patients as a result of more rapid diagnosis and treatment of active TB, and less unnecessary treatment in cases of false-positive smear microscopy. Moreover, we found that the use of `intensive' Xpert testing was associated with increased health utilities compared with use of the currently FDA-approved molecular assay (i.e., MTD) or `selective' use of molecular assays only for smear-positive samples.\nFrom a laboratory perspective, Xpert testing of at least one sputum sample would increase costs by over 60% per patient compared to no molecular testing. Despite the higher laboratory costs, however, we found that incorporating Xpert into diagnostic algorithms in the United States would be cost-saving from a health systems perspective. These cost savings are partially attributable to the low positive predictive value of smear microscopy in a low-prevalence setting, which can lead to increased health care expenditures while awaiting mycobacterial culture results.\nOur study has several limitations. First, we did not include the potential costs and effects associated with TB transmission that may occur during diagnostic delays. Despite this, we found that Xpert implementation was cost-effective compared to conventional approaches; incorporation of transmission would be expected to further enhance cost-effectiveness. Second, the costs of Xpert testing in the United States are not yet well defined and may vary across laboratories and hospitals. Nonetheless, we conducted extensive sensitivity analysis and found that Xpert implementation would be cost-effective even at higher Xpert test costs. Interestingly, laboratory costs for countries that qualify for prices negotiated by the Foundation for Innovative New Diagnostics (Geneva, Switzerland) are as little as $20\u201330 per Xpert test;14,34,35 whether public health laboratories in the United States will qualify for lower negotiated rates in the future remains unknown. Finally, to allow generalizability, our analysis incorporated both in-patients and out-patients. However, in sensitivity analysis, we found that health system costs of incorporating molecular testing were influenced by hospitalization status. For hospitalized patients, utilizing molecular assays is highly costsaving due to the reduced need for respiratory isolation and shorter hospitalizations for individuals with false-positive smear microscopy. On the other hand, for out-patients, the use of molecular assays was associated with incremental cost increases. Nevertheless, given the health benefits associated with rapid diagnosis, we show that Xpert implementation would be cost-effective for both in-patient and out-patient settings.\nOn the other hand our study has several strengths. We are among the first to evaluate the cost-effectiveness of Xpert testing in diagnostic algorithms in a low-prevalence setting that does not qualify for the reduced Xpert pricing available in other parts of the world. We show that despite the availability of mycobacterial culture, the addition of Xpert for TB diagnosis would be beneficial and cost-effective compared to current approaches. Finally, our study is unique in attempting to determine the optimal role for molecular testing in a low-prevalence setting by evaluating multiple algorithms with either `selective' or `intensive' Xpert implementation.\n\nCONCLUSION\nThe diagnosis of TB disease remains challenging and resource-intensive. We show that Xpert implementation is cost-effective for the diagnosis of pulmonary TB in the United States.\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 7\n\nAcknowledgments\nThe project was supported by a National Institutes Health K23 grant (AI089259) to study novel TB diagnostics. The funders had no role in the study design, data collection and analysis, decision to publish or preparation of the manuscript.\nAPPENDIX A: MODEL DETAILS\nAlgorithm details\nAlgorithm 1: No molecular testing\nAll tuberculosis (TB) suspects submit three sputum specimens for mycobacterial testing. Specimens are tested using smear microscopy and culture, and drug susceptibility testing (DST) if culture-positive. Individuals with a positive sputum smear are initiated on standard treatment per current guidelines.3,5 Individuals with a negative sputum smear result are initiated on treatment if clinical suspicion is high (i.e., clinical diagnosis). For all individuals, treatment is adjusted based on final mycobacterial culture and DST results (i.e., initiated/continued for positive cultures, and discontinued for negative cultures). For hospitalized patients with positive sputum smear results, respiratory isolation is continued for a minimum of 14 days after treatment initiation per current guidelines;3,5 hospitalized patients with negative sputum smear results are assumed to be eligible for discontinuation of respiratory isolation and discharge.\nAlgorithm 2: Selective Amplified Mycobacterium Tuberculosis Direct\nAll TB suspects submit three sputum specimens for mycobacterial testing. Specimens are tested using smear microscopy and culture, and if the smear is positive, one specimen is tested using the Amplified MTD\u00ae (Mycobacterium Tuberculosis Direct) nucleic acid amplification assay (Gen-Probe, San Diego, CA, USA). A positive sputum smear with a positive MTD result is considered to be a confirmed TB diagnosis and the patient is treated with the standard regimen per current guidelines.3,5 Individuals with a positive sputum smear result with negative MTD testing were assumed to have no treatment initiated. Individuals with a negative sputum smear result are initiated on treatment if clinical suspicion is high (i.e., clinical diagnosis). For all individuals, treatment is adjusted based on final mycobacterial culture and DST results. For hospitalized patients with positive sputum smear results with negative MTD testing, respiratory isolation is discontinued and patient is considered eligible for discharge; hospitalized patients with negative sputum smear results are considered eligible for the discontinuation of respiratory isolation and discharge.\nAlgorithm 3: Intensive MTD\nAlgorithm 3 is similar to the `selective' MTD algorithm, except that MTD is performed on one sputum sample of all TB suspects, regardless of smear microscopy results. Positive sputum smears with a positive MTD result are considered to be a confirmed TB diagnosis and the patient is treated with the standardized regimen. Individuals with a positive sputum smear result with a negative MTD result were assumed to have no treatment initiated. Individuals with negative smears with a positive MTD result are started on the standardized TB regimen per current guidelines.3,5 If the smear and MTD are negative, treatment is initiated if clinical suspicion is high (i.e., clinical diagnosis). For all individuals, treatment is adjusted based on final mycobacterial culture and DST results. For hospitalized patients with positive sputum smear results with negative MTD testing, respiratory isolation is discontinued and the patient is considered eligible for discharge; hospitalized patients with negative sputum smear results are considered eligible for discontinuation of respiratory isolation and discharge.\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 8\n\nAlgorithm 4: Selective Xpert\nThe diagnostic and clinical approach is the same as for the `selective' MTD algorithm, except that Xpert\u00ae MTB/RIF is incorporated into the algorithm in lieu of the MTD test. All TB suspects submit three sputum specimens for mycobacterial testing. Specimens are tested using smear microscopy and culture, and if the smear is positive, one specimen is tested using Xpert. Treatment and clinical care is initiated as outlined in the `selective' MTD algorithm. Xpert also provides rapid DST data when it is performed. We assumed that individuals with positive Xpert results are initiated on standardized treatment regimens for drug-susceptible or drug-resistant TB (based on Xpert results) per current guidelines.3,5 If DST results are not in line with Xpert results, we assumed treatment was adjusted based on conventional DST when these results became available (performed on positive mycobacterial culture isolates).\n\nAlgorithm 5: Intensive Xpert MTB/RIF\nThe diagnostic and clinical approach is the same as for the `intensive' MTD algorithm, except that Xpert is incorporated into the algorithm in lieu of the MTD test. All TB suspects submit three sputum specimens for mycobacterial testing. All specimens are tested by smear microscopy and culture, and one specimen is tested using the Xpert assay, regardless of smear microscopy results. Treatment and clinical care is initiated as outlined in the `intensive MTD' algorithm. Xpert also provides rapid DST data when it is performed. We assumed that individuals with positive Xpert results are initiated on standardized treatment regimens for drug-susceptible or drug-resistant TB (based on Xpert test results) per current guidelines.3,5 If DST results did not correspond to Xpert results, we assumed treatment was adjusted based on conventional DST when these results became available (performed on positive mycobacterial culture isolates).\n\nAdditional model assumptions and details\nAll patients diagnosed with TB on the basis of diagnostic testing are assumed to receive complete treatment with directly observed therapy (DOT), with negli gible rates of nonadherence. For the base case, hospitalized TB suspects with positive sputum smear microscopy were assumed to require a minimum of 14 days of hospitalization with respiratory isolation before discharge per current guidelines;3,5 individuals with sputum smear negativity and/or negative rapid molecular testing were assumed to have the duration of their hospitalization shortened to 3 days.3,5,8 For drug-susceptible TB or when DST results are unavailable, anti-tuberculosis treatment comprised an initial phase of isoniazid (INH), rifampin (RMP), pyrazinamide (PZA) and ethambutol (EMB) for 2 months, followed by a continuation phase of INH and RMP for an additional 4 months.3,5 If multidrugresistant TB (MDRTB) was detected using conventional DST or molecular testing (RMP resistance on Xpert was assumed to be a marker of MDR-TB), MDR-TB treatment was assumed to consist of an injectable (amikacin), moxifloxacin (MFX), PZA, ethionamide (ETH) and cycloserine (CS) daily for 6 months, followed by 12 months of MFX, PZA, ETH and CS.3,5 Treatment efficacy was based on published estimates and Baltimore City Health Department records.1 The degree to which diagnostic delays impact TB-associated mortality is unknown; for the base case, we assumed a modest 20% relative increase in mortality (absolute increase of 1%) in TB patients without rapid diagnosis/treatment compared to those with rapid diagnosis/treatment and varied this parameter in sensitivity analysis (baseline TB mortality 5%, Table 1).\nTo estimate the time to diagnosis, we assumed that smear microscopy and Xpert results would be available within 1 day; when molecular tests were performed selectively for smear-positive samples only, we added time for sequential testing. We assumed that the\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 9\n\nmean time until culture positivity using liquid culture was 16 days, that negative cultures were held until 60 days, that DST took on average of 14 days after isolate grew in culture and that MTD test results would be available within 2 days; diagnostic times were based on discussions with local laboratory managers and based on the literature. We assumed that clinical diagnoses to initiate treatment in the absence of positive test results were made at 14 days after presentation. Given that all algorithms incorporated multiple test results, we used the earliest time to correct diagnosis (for arms with selective molecular testing, we assumed smear-positive cases had diagnostic decisions based on molecular test results); in cases when an incorrect diagnosis was made, we utilized the time until the earliest test results upon which diagnostic decisions were based. TB transmission and secondary cases were not considered in this model.\n\nAPPENDIX B\n\nCOSTS OF DIAGNOSTIC TESTS PER SAMPLE\n\nDiagnostic test\n\nCost of\n\nCost of\n\nconsumables equipment\n\nLabor cost Overhead cost\n\nUS$ (% total) US$ (% total) US$ (% total) US$ (% total)*\n\nper saTmotpallec[orsatnge]\u2020\n\nDecontamination/ concentration\n\n4.93 (66)\n\n0.17 (2)\n\n1.70 (23) 0.68 (9)\n\n7.48 [2.58\u201312.88]\n\nSmear microscopy\n\n0.92 (23)\n\n0.09 (2)\n\n2.69 (66) 0.37 (9)\n\n4.07 [2.35\u20135.95]\n\nMGIT\n\n15.02 (42)\n\n2.87 (8)\n\n14.16 (40) 3.51 (10)\n\n35.56 [17.29\u201352.60]\n\nDST\n\n57.00 (56) 23.43 (23)\n\n11.99 (12) 9.26 (9)\n\n101.68 [19.60\u2013166.37]\n\nMTD\u00ae Xpert\u00ae MTB/RIF\n\n70.37 (77) 74.60 (76)\n\n1.50 (2) 13.94 (14)\n\n11.30 (12) 4.78 (5)\n\n8.32 (9) 4.78 (5)\n\n91.49 [26.08\u2013320.42] 98.10 [20.24\u2013838.46]\u2021\n\nMGIT = Mycobacteria Growth Indicator Tube; DST = drug susceptibility testing. * Overhead was assumed to be between 5% and 10% of total costs for each test system based on discussions with laboratory\nmanagers. \u2020 Laboratory testing capacity was estimated based on laboratory records; for reference laboratories we assumed that 3000\nTB suspects are evaluated each year providing 9000 sputum samples. Low and high estimates based on estimated range of vendor pricing for consumables and equipment, variations in laboratory wages and volume of testing (low of 150 TB suspects per year, high of 3000 TB suspects per year); when component estimates were unavailable, unit costs were adjusted by \u00b175%. Baseline laboratory technician wages were estimated to be $25.47 per hour (range $18.75\u2013$31.80).8 \u2021Includes costs associated with indeterminate test results requiring repeat testing, as well as costs associated with\ninstrument calibration and maintenance; we incorporated costs for two 4-cartridge Xpert instruments (Cepheid, Sunnyvale, CA, USA) based on base-case volume. Xpert instrument cost was estimated to be $78200, with Xpert cartridge costs (per test) of $71.36 based on current manufacturer estimates.12, 14 Low and high range of costs determined by using lowest\nand highest estimates for Xpert consumable costs, Xpert equipment costs, labor costs and volume of testing.\n\nAPPENDIX C: SENSITIVITY ANALYSIS OF COSTS AND EFFECTS\nThe incremental quality-adjusted life years (QALYs) gained comparing `intensive' Xpert testing to no molecular testing were most influenced by: 1) the relative increase in probability of death with delayed diagnosis (incremental QALYs gained range from 4.9 per 1000 TB suspects if there is no increase in risk of death with diagnostic delay, to 26.7 QALYs gained per 1000 TB suspects if there is a five-fold i ncrease in risk of death with diagnostic delay), and 2) the prevalence of TB among TB suspects (incremental QALYs gained range from 5.7 per 1000 TB suspects to 22.9 per 1000 TB suspects for TB prevalence of 1% to 30% among TB suspects; Figure C.2).\nWhen comparing `intensive' Xpert implementation with `selective' Xpert implementation, the incremental cost-effectiveness ratio (ICER) was most influenced by the specificity of\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nChoi et al.\n\nPage 10\nXpert (range $18 630 to $230 471 per QALY-gained, for specificity from 100% to 95%), cost of Xpert (range $24 614 to $189 257 per QALY-gained for Xpert costs from $20 per test to $838 per test), and sensitivity of culture for smear-negative pulmonary TB ($16 440 per QALY-gained to $180 352 per QALY-gained for sensitivity range from 70% to 100%; Figure C.3).\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nFigure C.1. One-way sensitivity analysis of incremental costs comparing `intensive' Xpert (Algorithm 5) vs. no molecular testing (Algorithm 1). Line represents incremental effects when using basecase estimates of all parameters. Not all parameters tested in sensitivity analysis are shown. TB = tuberculosis; HIV = human immunodeficiency virus; RMP = rifampin; MDR-TB = multidrugr esistant TB.\nFigure C.2. One-way sensitivity analysis of incremental QALYs gained comparing `intensive' Xpert (Algorithm 5) vs. no molecular testing (Algorithm 1). Line represents incremental effects when using base-case estimates of all parameters. Not all parameters tested in the sensitivity analysis are shown. TB = tuberculosis; HIV = human immunodeficiency virus; MDR-TB = multidrug-resistant TB; QALY = quality-adjusted life-years.\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nChoi et al.\n\nPage 11\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nFigure C.3. One-way sensitivity analysis of ICER comparing `intensive' Xpert (Algorithm 5) vs. `selective' Xpert (Algorithm 4). Line represents incremental effects when using base-case estimates of all parameters. Not all parameters tested in the sensitivity analysis are shown. All key variables were included in the sensitivity analysis. Only the top 15 variables influencing the ICER are shown; among the additional variables that were evaluated included the cost of MTD\u00ae testing, utility weights for TB disease and treatment, cost of TB and MDR-TB treatment, and probability of toxicity. TB = tuberculosis; HIV = human immunodeficiency virus; RMP = rifampin; ICER = incremental cost effectiveness ratio; QALY = quality-adjusted life year.\nFigure C.4. One-way sensitivity analysis of ICER comparing `intensive' Xpert (Algorithm 5) vs. `selective' MTD (Algorithm 2). Line represents incremental effects when using base-case estimates of all parameters. Not all parameters tested in the sensitivity analysis are shown. All key variables were included in sensitivity analysis. Only the top 15 variables influencing the ICER are shown; among the additional variables that were evaluated included cost of MTD testing, utility weights for TB disease and treatment, cost of TB and MDR-TB treatment, and probability of toxicity. TB = tuberculosis; HIV = human immunodeficiency virus; RMP = rifampin; ICER = incremental cost-effectiveness ratio; QALY = qualityadjusted life year.\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 12\n\nAPPENDIX D: PROBABILISTIC SENSITIVITY ANALYSIS RESULTS\nProbabilistic sensitivity analysis was conducted using Monte-Carlo simulation methods with 1000 iterations in which all key parameters are varied simultaneously (triangular, beta and log-normal distributions were used). Median costs, effects and ICERs are shown in Table D, along with a cost-effectiveness acceptability curve and scatterplot of costs and effects comparing `intensive' Xpert algorithm (Algorithm 5) with `no molecular testing' (Algorithm 1; Figure D). An algorithm using intensive testing of at least one sputum sample with Xpert (Algorithm 5) dominated the strategy of no molecular testing (Algorithm 1) in 53% of the simulations (i.e., was cost-saving), and was considered cost-effective at a willingness to pay (WTP) of $50 000 per QALY gained over 99% of the time. Compared to a strategy of `selective' molecular testing of only smear-positive samples with MTD, `selective' molecular testing with Xpert was considered highly cost-effective and was associated with a median ICER of $7972 per QALY gained (95%CI 6146\u201350 420). `Intensive' Xpert testing on at least one sputum sample for all TB suspects was associated with a median ICER of $16 165 per QALY-gained (95%CI 4886\u201384 896) compared to `selective' MTD testing. `Intensive' Xpert implementation was also considered cost-effective compared to `selective' Xpert implementation (median ICER $16 376 per QALY gained, 95%CI 4861\u201388 733).\n\nTable D Probabilistic sensitivity analysis results\n\nAlgorithm\n1: No molecular 2: Selective MTD 3: Intensive MTD 4: Selective Xpert 5: Intensive Xpert\n\nHealth system costs/TB suspect median (95%CI)\n3172 (1728 to 6528)\n2870 (1546 to 5688)\n3114 (1750 to 5946)\n2872 (1548 to 5697)\n3153 (1770 to 5987)\n\nIncremental costs/TB suspect median (95%CI)*\nReference \u2212288 (\u22121016 to \u221253) \u221235 (\u2212837 to 332) \u2212284 (\u22121011 to \u221247) \u221211 (\u2212824 to 375)\n\nQALYs accrued per patient median (95%CI)\n22.033 (21.95 to 22.089)\n22.035 (21.95 to 22.090)\n22.050 (21.98 to 22.096)\n22.036 (21.95 to 22.090)\n22.053 (21.99 to 22.097)\n\nIncremental QALYs gained/ 1000 TB suspects (95%CI)*\nReference\n\nMedian ICER\u2014 cost/QALY gained (95%CI)*\n\u2014\n\n1.5 (0 to 4.1) Reference\n\n15.9 (5.0 to 41.0)\n\n17 487 (5 227 to 87 382)\n\n2.0 (0.10 to 4.5)\n\n7972 (\u22126 146 to 50 420)\n\n18.5 (5.6 to 47.1)\n\n16 165 (4 886 to 84 896)\n\nTB = tuberculosis; CI = confidence interval; QALY = quality-adjusted life year; ICER = incremental cost effectiveness ratio; MTD = Amplified Mycobacterium Tuberculosis Direct. * Incremental costs and effects and ICERs are calculated per simulation; median results with 95%Cls are reported. `Selective MTD' algorithm was used as Reference because the `No molecular' algorithm was dominated in the majority (53%) of simulations.\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nChoi et al.\n\nPage 13\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nFigure D. Incremental cost-effectiveness of Algorithm 5 vs. Algorithm 1 during iterations of Monte Carlo simulation. The ellipse represents 95% confidence points. Diagonal dashed line represents ICERs at a WTP threshold of $50 000. Points to the right of this dashed line are considered cost-effective. Dotted horizontal line shows incremental cost of $0. Points below this line represents i terations in which Algorithm 5 was cost-saving compared to Algorithm 1. WTP = willingness to pay; ICER = incremental cost-effectiveness ratio.\nReferences\n1. Centers for Disease Control and Prevention. Reported tuberculosis in the United States\u20142011. CDC; Atlanta, GA, USA: 2012. http://www.cdc.gov/tb/statistics/reports/2011/pdf/report 2011.pdf [Accessed July 2013]\n2. American Thoracic Society/Centers for Disease Control and Prevention/Infectious Diseases Society of America. Diagnostic standards and classification of tuberculosis in adults and children. Am J Respir Crit Care Med. 2000; 161(4 Pt 1):1376\u20131395. [PubMed: 10764337]\n3. Maryland Department of Health & Mental Hygiene. Maryland TB guidelines for prevention and treatment of tuberculosis. DHMH; Baltimore, MD, USA: 2007. http://ideha.dhmh.maryland.gov/ OIDPCS/CTBCP/CTBCPDocuments/tbguidelines.pdf [Accessed July 2013]\n4. Jensen PA, Lambert LA, Iademarco MF, Ridzon R. US Centers for Disease Control and Prevention. Guidelines for preventing the transmission of Mycobacterium tuberculosis in health care facilities. Morb Mortal Wkly Rep. 2005; 54(RR-17):1\u2013141.\n5. American Thoracic Society/Centers for Disease Control and Prevention/Infectious Diseases Society of America. Controlling tuberculosis in the United States. Am J Respir Crit Care Med. 2005; 172:1169\u20131227. [PubMed: 16249321]\n6. American Thoracic Society. Rapid diagnostic tests for tuberculosis\u2014workshop report. ATS; New York, NY, USA: 1997. http://www.thoracic.org/statements/resources/mtpi/rapidtb1-12. pdf [Accessed July 2013]\n7. Centers for Disease Control and Prevention. Update: nucleic acid amplification tests for tuberculosis. MMWR Morb Mortal Wkly Rep. 2000; 49:593\u2013594. [PubMed: 10921499]\n8. Dowdy DW, Maters A, Parrish N, Beyrer C, Dorman SE. Cost-effectiveness analysis of the GenProbe Amplified Myco-bacterium Tuberculosis Direct test as used routinely on smear-positive respiratory specimens. J Clin Microbiol. 2003; 41:948\u2013953. [PubMed: 12624014]\n9. Boehme CC, Nabeta P, Hillemann D, et al. Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med. 2010; 363:1005\u20131015. [PubMed: 20825313]\n10. Boehme CC, Nicol MP, Nabeta P, et al. Feasibility, diagnostic accuracy, and effectiveness of decentralised use of the Xpert MTB/RIF test for diagnosis of tuberculosis and multidrug\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 14\nresistance: a multicentre implementation study. Lancet. 2011; 377:1495\u20131505. [PubMed: 21507477]\n11. World Health Organization. Automated real-time nucleic acid amplification technology for rapid and simultaneous detection of tuberculosis and drug resistant tuberculosis: policy statement. WHO; Geneva, Switzerland: 2011. WHO/HTM/TB/2011.4http://whqlibdoc.who.int/publications/ 2011/9789241501545_eng.pdf [Accessed July 2013]\n12. Foundation for Innovative New Diagnostics. Negotiated prices for Xpert\u00ae MTB/RIF and FIND country list. FIND; Geneva, Switzerland: 2013. http://www.finddiagnostics.org/about/ what_we_do/successes/find-negotiated-prices/xpert_mtb_rif.html [Accessed July 2013]\n13. Cepheid. Pricing to the FIND target market of 145 countries. Cepheid; Sunnyvale, CA, USA: 2012. http://www.cepheidcares.com/tb/cepheid-vision.html [Accessed July 2013]\n14. Vassall A, van Kampen S, Sohn H, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high-burden countries: a cost-effectiveness analysis. PLoS Med. 2012; 8:e1001120. [PubMed: 22087078]\n15. Brown RE, Miller B, Taylor WR, et al. Health care expenditures for tuberculosis in the United States. Arch Intern Med. 1995; 155:1595\u20131600. [PubMed: 7618981]\n16. de Perio MA, Tsevat J, Roselle GA, Kralovic SM, Eckman MH. Cost-effectiveness of interferon gamma release assays vs tuberculin skin tests in health care workers. Arch Intern Med. 2009; 169:179\u2013187. [PubMed: 19171815]\n17. Agency for Healthcare Research and Quality. Medical Expenditure Panel survey. AHRQ; Rockville, MD, USA: 2009. http://meps.ahrq.gov/mepsweb/data_stats/summ_tables/hc/ mean_expend/2009/table2.htm [Accessed July 2013]\n18. Kaiser Family Foundation. Hospital in-patient expenses. KFF; Menlo Park, CA, USA: 2012. http:// www.statehealthfacts.org/comparemaptable.jsp?ind=273&cat=5 [Accessed August 2012]\n19. Rajbhandary SS, Marks SM, Bock NN. Costs of patients hospitalized for multidrug-resistant tuberculosis. Int J Tuberc Lung Dis. 2004; 8:1012\u20131016. [PubMed: 15305486]\n20. Lieberman D, Schlaeffer F, Boldur I, et al. Multiple pathogens in adult patients admitted with community-acquired pneumonia: a one year prospective study of 346 consecutive patients. Thorax. 1996; 51:179\u2013184. [PubMed: 8711652]\n21. Rimland D, Navin TR, Lennox JL, et al. Prospective study of etiologic agents of communityacquired pneumonia in patients with HIV infection. AIDS. 2002; 16:85\u201395. [PubMed: 11741166]\n22. Colice GL, Morley MA, Asche C, Birnbaum HG. Treatment costs of community-acquired pneumonia in an employed population. Chest. 2004; 125:2140\u20132145. [PubMed: 15189934]\n23. Tiemersma EW, van der Werf MJ, Borgdorff MW, Williams BG, Nagelkerke NJ. Natural history of tuberculosis: duration and fatality of untreated pulmonary tuberculosis in HIV negative patients: a systematic review. PLoS ONE. 2011; 6:e17601. [PubMed: 21483732]\n24. Corbett EL, Watt CJ, Walker N, et al. The growing burden of tuberculosis: global trends and interactions with the HIV epidemic. Arch Intern Med. 2003; 163:1009\u20131021. [PubMed: 12742798]\n25. Centers for Disease Control and Prevention/American Thoracic Society. Diagnostic standards and classification of tuberculosis in adults and children. Am J Respir Crit Care Med. 2000; 161:1376\u2013 1395. [PubMed: 10764337]\n26. Steingart KR, Ng V, Henry M, et al. Sputum processing methods to improve the sensitivity of smear microscopy for tuberculosis: a systematic review. Lancet Infect Dis. 2006; 6:664\u2013674. [PubMed: 17008175]\n27. Guerra RL, Hooper NM, Baker JF, et al. Use of the Amplified Mycobacterium Tuberculosis Direct test in a public health laboratory: test performance and impact on clinical care. Chest. 2007; 132:946\u2013951. [PubMed: 17573496]\n28. Guo N, Marra CA, Marra F, Moadebi S, Elwood RK, Fitzgerald JM. Health state utilities in latent and active tuberculosis. Value Health. 2008; 11:1154\u20131161. [PubMed: 18489493]\n29. Sohn H, Minion J, Albert H, Dheda K, Pai M. TB diagnostic tests: how do we figure out their costs? Expert Rev Anti Infect Ther. 2009; 7:723\u2013733. [PubMed: 19681700]\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 15\n30. Shah M, Miele K, Choi H, et al. QuantiFERON-TB Gold In-Tube implementation for latent tuberculosis diagnosis in a public health clinic: a cost-effectiveness analysis. BMC Infect Dis. 2012; 12:360. [PubMed: 23253780]\n31. Hirth RA, Chernew ME, Miller E, Fendrick AM, Weissert WG. Willingness to pay for a qualityadjusted life year: in search of a standard. Med Decis Making. 2000; 20:332\u2013342. [PubMed: 10929856]\n32. Owens DK. Interpretation of cost-effectiveness analyses. J Gen Intern Med. 1998; 13:716\u2013717. [PubMed: 9798822]\n33. World Health Organization. Implementation and roll-out of Xpert MTB/RIF. WHO; Geneva, Switzerland: 2012. May 2012 updatehttp://www.stoptb.org/wg/gli/assets/documents/ Xpert %20MTB-RIF%20UPDATE%20May%202012.pdf [Accessed July 2013]\n34. Schnippel K, Meyer-Rath G, Long L, et al. Scaling up Xpert MTB/RIF technology: the costs of laboratory- vs. clinic-based roll-out in South Africa. Trop Med Int Health. 2012; 17:1142\u20131151. [PubMed: 22686606]\n35. Meyer-Rath G, Schnippel K, Long L, et al. The impact and cost of scaling up GeneXpert MTB/RIF in South Africa. PLoS ONE. 2012; 7:e36966. [PubMed: 22693561]\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 16\n\nFigure 1. Schematic diagram of decision-analysis model. Not all branches are shown; model details are found in Appendix A. `Selective' molecular algorithms (Algorithms 2 and 4) incorporate molecular testing with either Xpert\u00ae MTB/RIF or MTD\u00ae for sputum samples positive by smear microscopy, but not for sputum samples that are smear-negative. `Intensive' molecular algorithms incorporate molecular testing of at least one sputum, regardless of sputum microscopy results. All algorithms include testing of three sputum samples by conventional diagnostics. TB = tuberculosis; \u2212 = negative; + = positive; DST = drug susceptibility testing.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 17\n\nFigure 2. Cost-effectiveness acceptability curve comparing Algorithm 1 (no molecular testing) vs. Algorithm 5 (`intensive Xpert'). Cost-effectiveness acceptability curve showing the probability that implementation of an `intensive' Xpert MTB/RIF diagnostic algorithm (Algorithm 5) will be cost-effective compared to an algorithm without molecular testing (Algorithm 1) at varying thresholds WTP. At WTP of $50K, `intensive Xpert' was costeffective in 99% of simulations; at WTP $100K, `intensive Xpert' was cost-effective in 99.8% of simulations. WTP = willingness-to-pay.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 18\n\nModel parameters and sources\n\nTable 1\n\nVariable Laboratory costs, US$*\nSmear microscopy Mycobacterial culture DST MTD\u00ae Xpert\u00ae MTB/RIF Treatment costs, US$\nAnti-tuberculosis treatment course (drug-susceptible TB, 6 months) Hospitalization costs per day\nMDR-TB treatment Epidemiology and diagnostic and treatment parameters\nPrevalence of TB among TB suspects in the United States (HIV-positive TB suspects)\u00a7\nMDR-TB prevalence among TB cases in the United States Probability of hospitalization during initial TB evaluation Mortality of untreated smear-positive TB (smear-negative TB)\u00b6 Mortality of treated drug-susceptible TB (MDR-TB)\u00b6 Sensitivity, smear microscopy (HIV-positive TB suspects) Specificity, smear microscopy Sensitivity, mycobacterial culture for smear-positive TB (smear-negative TB) Sensitivity, MTD for smear-positive TB (smear-negative TB) Specificity, MTD Sensitivity, clinical diagnosis (specificity) Sensitivity, Xpert MTB/RIF for smear-positive TB (smear-negative TB) Specificity, Xpert MTB/RIF Sensitivity, Xpert MTB/RIF rifampin resistance detection (specificity) Sensitivity, conventional DST resistance detection (specificity) Utility weight Complete health First-line treatment# MDR-TB treatment# Treated active TB disease** Untreated active TB disease** Drug-related hepatotoxicity Death\n\nBase case Low\n\nHigh\n\nReference\n\n4.07 35.56 101.68 91.49 98.10\n\n2.35 17.29 19.60 26.08 20.24\n\n5.95 52.60 166.37 320.42 838.46\n\nCalculated Calculated Calculated Calculated Calculated\n\n9 037\u2020 2 469 57 889\u2020\n\n3 069 1 161 40 133\n\n53 401 2 975 204 862\n\nBCHD\u2020, 15, 16 17, 18 BCHD\u2020, 15, 19\n\n0.02 (0.062) 0.011 0.20\n\n0.01 (0.03) 0.009 0\n\n0.30 (0.30) 0.075 1.0\n\n1, 20, 21\u2021 1 BCHD, 15, 22\n\n0.70 (0.20) 0.53 (0.10) 0.86 (0.30) 23, 24\n\n0.05 (0.08) 0.81 (0.25) 0.97 1.0 (0.9) 0.98 (0.55) 0.99 0.3 (0.9) 0.98 (0.72) 0.99 0.976 (0.98) 1.0 (1.0)\n\n0.01 (0.05) 0.45 (0.20) 0.91 0.90 (0.70) 0.91 (0.45) 0.95 0 (0.80) 0.95 (0.50) 0.95 0.90 (0.90) \u2014\n\n0.20 (0.50) 0.90 (0.50) 1.0 1.0 (1.0) 1.0 (0.90) 1.0 0.80 (1.0) 1.0 (0.90) 1.0 1.0 (1.0) \u2014\n\n1 25, 26 25, 26 Assumption 6\u20138, 27 6\u20138, 27 1, 14 9, 14 9, 14 9, 14 Assumption\n\n1\n\n16\n\n0.9\n\n0.7\n\n0.95\n\n16\n\n0.7\n\n0.5\n\n0.95\n\n16, Assumption\n\n0.85\n\n0.7\n\n0.9\n\n16, 28\n\n0.7\n\n0.5\n\n0.9\n\n16, 28\n\n0.8\n\n0.7\n\n0.95\n\n16\n\n0\n\n16\n\nDST = drug susceptibility testing; TB = tuberculosis; BCHD = Baltimore City Health Department; MDR-TB = multidrug-resistant TB; HIV = human immunodeficiency virus; DHMH = Department of Health & Mental Hygiene.\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 19\n\n*Cost of component consumables and equipment were obtained from manufacturers or distributors. Quantity of consumables and equipment as well as labor and overheads were determined through direct observation at the Johns Hopkins Hospital Mycobacteriology Laboratory and the Maryland DHMH Mycobacteriology Laboratory. Detailed breakdown of costs is shown in Appendix B.\n\u2020 The base case cost of anti-tuberculosis treatment was calculated from Baltimore City Health Department records and invoices and reflects the costs of out-patient care. The base case costs include labor, drugs and diagnostic monitoring, and assume no additional hospitalizations during treatment. Labor associated with clinicians, nurses, case managers, and directly observed therapy accounted for US$8313 (92%) of costs. For MDR-TB, costs included those associated with extended treatment course, second-line drug regimens, and estimated in-patient hospitalizations during treatment. Range was determined by using lowest and highest estimates for labor, drugs, diagnostic monitoring costs, and complications and hospitalization during treatment.\n\u2021 Kim Dionne, personal communication, Johns Hopkins Hospital Mycobacteriology Laboratory, Baltimore, MD, USA, 2012.\n\u00a7In the base case, we assumed 7% of all TB suspects were HIV-infected, and varied this between 5% and 20% in sensitivity analysis.\n\u00b6 Parameters were varied for HIV infection. For untreated smear-positive TB, mortality was assumed to be 83%; for smear-negative TB, 74%. For\ntreated drug-susceptible TB, base case mortality was assumed to be 13%; for MDR-TB, 18%.24\n# Represents utility weight associated with being placed on medication regimen and is independent of TB disease utility.\n**We assumed that the severity of disease was the same for drug-susceptible and drug-resistant TB, but the duration of disease differed between the two. We assumed that average treatment duration for drug-susceptible TB was 6 months and treatment duration for MDR-TB was 18 months.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\nNIH-PA Author Manuscript\n\nChoi et al.\n\nPage 20\n\nTable 2 Costs and cost-effectiveness of diagnostic algorithms\n\nCosts and effects\nTotal laboratory costs per TB suspect, US$ Incremental laboratory costs, US$ Total health care costs per TB suspect, US $\u2020 Incremental total costs, US$ QALYs accrued per TB suspect\nIncremental QALYs per 1000 suspects Average time to diagnosis among TB cases, days\u00a7 TB cases diagnosed by molecular testing, % ICER, US$ per QALY gained\u00b6 ICER, US$ per QALY gained ICER, US$ per QALY gained\n\nAlgorithm 1 no molecular 157.64 Reference\n2727.68 Reference 22.08622 Reference\n16.30\n0 Dominated\u00b6\n\u2014 \u2014\n\nAlgorithm 2 selective MTD\u00ae 161.80 4.16\n2479.63 \u2212248.05 22.08771\n1.49\n13.31\n75.50 Reference\n\u2014 \u2014\n\nAlgorithm 3 intensive MTD\u00ae 249.13 91.49\n2653.08 \u221274.60 22.09133\n5.11\n3.92\n88.00 47 914 Reference\n\u2014\n\nAlgorithm 4 selective Xpert\u00ae 162.10 4.46\n2481.71 \u2212245.97 22.08780\n1.58\n6.03\n75.50 23 111\n\u2014 Reference\n\nAlgorithm 5 intensive Xpert\u00ae 255.75* 98.11\n2672.79 \u221254.89 22.09254 6.32\u2021\n2.71\n92.00\n39 992 16 289 40 312\n\nTB = tuberculosis; QALY = quality-adjusted life year; ICER = incremental cost-effectiveness ratio.\n* The base case scenario assumes a volume of testing of 3000 TB suspects per year. If the volume of testing is reduced to a low estimate of 150 TB suspects per year, laboratory costs per suspect for `intensive' Xpert algorithm increase to $384 per TB suspect. Total laboratory costs include the cost of smear microscopy and culture for three samples per TB suspect and Xpert testing of one sample, as well as costs associated with preparation, clean-up and reporting.\n\u2020Total health care costs include costs associated with initial diagnostic evaluations, including laboratory testing, hospitalization, respiratory isolation, as well as costs associated with anti-tuberculosis treatment (includes labor, drugs and monitoring). For the base case, we assumed that 20% of TB suspects receive initial diagnostic evaluation in the in-patient setting.\n\u2021 `Intensive' Xpert (Algorithm 5) compared with `selective' Xpert (Algorithm 4) testing resulted in incremental 4.74 QALYs gained per 1000 TB suspects; `intensive' Xpert MTB/RIF (Algorithm 5) compared to intensive MTD (Algorithm 3) testing resulted in 1.21 QALYs gained per 1000 TB suspects.\n\u00a7 Time to diagnosis is based on estimated time to earliest correct diagnosis. We estimated that smear microscopy and Xpert results are available within 1 day, and average time to culture positivity was 16 days; algorithms with selective implementation of molecular assays assumed sequential testing with molecular assay after smear microscopy results are available.\n\u00b6 Algorithm 1 was dominated by all other algorithms (i.e., was more costly and less effective). ICERs were calculated as incremental costs divided\nby incremental effects.\n\nNIH-PA Author Manuscript\n\nNIH-PA Author Manuscript\n\nInt J Tuberc Lung Dis. Author manuscript; available in PMC 2014 January 14.\n\n\n",
    "authors": [
      "H. W. Choi",
      "K. Miele",
      "D. Dowdy",
      "M. Shah"
    ],
    "doi": "10.5588/ijtld.13.0095",
    "year": null,
    "item_type": "journalArticle",
    "url": "http://openurl.ingenta.com/content/xref?genre=article&issn=1027-3719&volume=17&issue=10&spage=1328"
  },
  {
    "key": "EQ7T93VG",
    "title": "Systematic Review of Pooling Sputum as an Efficient Method for Xpert MTB/RIF Tuberculosis Testing during the COVID-19 Pandemic",
    "abstract": "",
    "full_text": "Systematic Review of Pooling Sputum as an Efficient\nMethod for Xpert MTB/RIF Tuberculosis Testing during\nCOVID-19 Pandemic\nLuis E. Cuevas, Victor S. Santos, Shirley Ver\u00f4nica Melo Almeida Lima, Konstantina Kontogianni, John S. Bimba, Vibol Iem, Jose Dominguez, Emily Adams, Ana Cubas Atienzar, Thomas Edwards, S. Bertel Squire, Patricia J. Hall, Jacob Creswell\n\nGeneXpert-based testing with Xpert MTB/RIF or Ultra assays is essential for tuberculosis diagnosis. However, testing may be affected by cartridge and staff shortages. More efficient testing strategies could help, especially during the coronavirus disease pandemic. We searched the literature to systematically review whether GeneXpertbased testing of pooled sputum samples achieves sensitivity and specificity similar to testing individual samples; this method could potentially save time and preserve the limited supply of cartridges. From 6 publications, we found 2-sample pools using Xpert MTB/RIF had 87.5% and 96.0% sensitivity (average sensitivity 94%; 95% CI 89.0%\u201398.0%) (2 studies). Four-sample pools averaged 91% sensitivity with Xpert MTB/RIF (2 studies) and 98% with Ultra (2 studies); combining >4 samples resulted in lower sensitivity. Two studies reported that pooling achieved 99%\u2013100% specificity and 27%\u201331% in cartridge savings. Our results show that pooling may improve efficiency of GeneXpert-based testing.\nXpert MTB/RIF (Cepheid, https://www.cepheid. com) is a cartridge-based nucleic amplification assay for use with Cepheid\u2019s GeneXpert diagnostic\nAuthor affiliations: Liverpool School of Tropical Medicine, Liverpool, UK (L.E. Cuevas, K. Kontogianni, V. Iem, E. Adams, A.T. Atienzar, T. Edwards, S.B. Squire); Federal University of Alagoas, Arapiraca, Brazil (V.S. Santos); Federal University of Sergipe, Aracaju, Brazil (S.V.M. Almeida Lima); Bingham University, Karu, Nigeria (J.S. Bimba); National TB Control Program, Vientiane, Laos (V. Iem); Institut d\u2019Investigaci\u00f3 Germans Trias i Pujol and Universitat Aut\u00f2noma de Barcelona, Badalona, Spain (J. Dominguez); Centers for Disease Control and Prevention, Atlanta, Georgia, USA (P.J. Hall); Stop TB Partnership, Innovations and Grants, Geneva, Switzerland (J. Creswell)\nDOI: https://doi.org/10.3201/eid2703.204090\n\ninstrument systems that detects both Mycobacterium tuberculosis complex (MTB) and resistance to rifampin (RIF). In 2010, the World Health Organization endorsed Xpert MTB/RIF for laboratory detection of tuberculosis (TB) (1), signaling a sea change for diagnosing TB. Xpert MTB/RIF increased sensitivity over microscopy and its ability to simultaneously detect rifampin resistance led to its rapid adoption in low- and middle-income countries. Within the first 5 years, 23 million cartridges were procured at the negotiated price of $9.98/each (P. Jacon, Cepheid, pers. comm., email, April 2020). In 2017, the Cepheid Xpert MTB/RIF Ultra assay (Ultra) was released for use on GeneXpert instruments and results determined to be comparable to those from the Xpert MTB/RIF assay, with an even lower limit for detection (1).\nCoronavirus disease (COVID-19) is severely disrupting health systems and is threatening progress made by national TB control programs. The new Xpert Xpress SARS-CoV-2 test is run on the same GeneXpert instruments as those for Xpert MTB/RIF and Ultra testing; it is being expedited for large-scale production and deployment. Consequently, TB-testing capacity, already limited by the availability of necessary staff, testing modules, and Xpert MTB/RIF and Ultra cartridges, may be further reduced by the increased demand for GeneXpert for COVID-19 testing (3). There is an urgent need to develop laboratory testing approaches to expand TB diagnostic and casefinding services in preparation for crises, such as the COVID-19 pandemic.\nGeneXpert-based testing for TB requires 1 cartridge per sputum sample. However, screening for other infectious diseases has used sample pooling methods, in which samples from several patients are\n\nEmerging Infectious Diseases \u2022 www.cdc.gov/eid \u2022 Vol. 27, No. 3, March 2021\n\n719\n\nSYNOPSIS\n\npooled together for a single test to optimize processing. If a pooled-sample test is negative, all samples in the pool are considered negative; if the pooled-sample test is positive, all samples in the pool are retested individually to identify the samples that are positive. This method is routinely used in situations where the prevalence of disease is low (e.g., blood banks screening donated blood for hepatitis and syphilis) (4\u20139). The method can substantially reduce workload and cost and, for TB, could more efficiently process samples for diagnosis. We reviewed the literature to determine the accuracy of pooling for Xpert MTB/RIF and Ultra detection of pulmonary TB, with the aim of supporting TB programs as they continue to test for TB in the context of increased resource constraints during the COVID-19 pandemic.\nMethods We conducted a systematic review following the Cochrane Collaboration\u2019s Diagnosis Test Accuracy Working Group protocol (https://methods.cochrane. org). Our primary aim was to describe whether testing using GeneXpert for pulmonary TB on pooled samples would result in similar numbers of patients being confirmed with TB as testing samples individually. Secondarily, we aimed to describe the advantages and disadvantages reported, such as savings in cartridges used and time required to process samples.\nWe searched PubMed, CINAHL, Global Health, and Web of Science for publications from January 2010\u2013March 2020 with no regional or language restrictions. We used the terms \u201cGeneXpert\u201d OR \u201cXpert\u201d OR \u201cUltra\u201d AND \u201ctuberculos*\u201d AND \u201cpool*\u201d AND \u201cdiagnos*\u201d with associated subject headings and search terms without filters (Appendix Table, https://wwwnc.cdc.gov/EID/article/27/3/204090-App1.pdf). S.V.M.A.L. and K.K. eliminated duplicates, screened titles and abstracts, and read full texts to determine eligibility. We also searched for article references manually and for abstracts published at the 2019 Union World Conference of Lung Health. Studies were included if they presented original data, if data were not duplicated in other publications, and if the articles were not reviews or opinions. We excluded studies that pooled several samples from the same patient to increase the yield and those that included samples other than sputum. Given the paucity of studies, we included both those that directly processed patient samples and those that used leftover samples to prepare a specimen repository for bench evaluation of the pooling method. We read selected studies in full for data extraction; L.E.C. and V.S.S. resolved disagreements by consensus.\n\nData extracted included study identifiers (author, year, country, and setting), methods (study design, pooling methods, number of participants, pooling ratio, number of pools, and type of test), and whether the pooled positive and negative test results coincided with those obtained through individual testing. Data are presented as sensitivity and specificity values, considering the individual GeneXpert test as the reference. Sensitivity was defined as the proportion of pooled samples correctly identified as positive when the pool contained at least 1 sample with a positive individual GeneXpert test. Specificity was defined as the proportion of pooled samples correctly identified as negative when all samples in the pool were negative in individual GeneXpert tests. Data are presented with 95% confidence intervals and ranges.\nWe assessed the quality of the studies based on a further reference standard, the use of TB culture by any method, whether pooled results were recorded blind to the individual results and whether participants had been recruited consecutively to represent the range of disease severity. The quality of studies and the risk of bias were assessed by 2 independent reviewers (authors) using the QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies) guidelines (https://www.bristol.ac.uk/media-library/sites/ quadas/migrated/documents/quadas2.pdf). We used Cochrane Collaboration Rev-Man 5.3 software (https://training.cochrane.org/online-learning/coresoftware-cochrane-reviews/revman/revman-5-download) to generate the graphs on the risk of bias (Appendix Figures 1, 2). Because the studies were highly heterogeneous and most (4/6) did not present data on specificity, we were unable to perform a meta-analysis to estimate the pooled sensitivity and specificity or to explore the reasons for heterogeneity through metaregression. Institutional review board approval was not required because all data sources and publications were in the public domain and in aggregate format.\nResults We identified 33 publications through the initial publication search. After screening titles and abstracts, we assessed 5 full-text articles for eligibility and initially included 2 in data syntheses. In addition, 4 studies were identified from other sources: 1 conference report, 1 preprint article, and 2 articles from the reference lists of other studies. We included 6 articles in the final data synthesis (Figure 1). One study was conducted in South America (10), 2 in Africa (11,12), and 3 in Asia (13\u201315); all were published during 2014\u2013 2020, before the COVID-19 pandemic.\n\n720\n\nEmerging Infectious Diseases \u2022 www.cdc.gov/eid \u2022 Vol. 27, No. 3, March 2021\n\nPooling Sputum for TB Testing during COVID-19\n\nWe assessed the quality of the studies and the risk of bias (Appendix Figures 1, 2). Three studies used samples collected directly from patients with presumptive TB, and 3 studies used previously collected stored samples with known GeneXpert results. Studies pooling direct clinical samples were conducted in high-burden settings in which the proportion of patients that tested GeneXpert-positive was high (15%, 16%, and 38.6%), whereas stored samples were used to prepare pools varying the proportion of positive specimens in each pool to explore the effect on sensitivity. Pools were prepared with clinical samples from consecutive patients in 5 studies and in benchprepared spiked sputum in a laboratory setting in 1 study. The latter study had also prepared the pool using combinations of smear-positive/culture-positive and smear-negative/culture-positive samples. Generally, the studies followed a similar approach to pooling: a sample was collected from patients with presumptive TB and split into aliquots for Xpert MTB/RIF or Ultra testing following the manufacturer\u2019s guidelines. Studies that processed and homogenized sputum used the same steps for the individual and pooled GeneXpert tests. One aliquot was used to obtain an individual result, which was considered the reference result; and the second aliquot was mixed with aliquots from other patients and then tested as a pooled sample. All studies reported that laboratory technicians were blind to whether they were testing pooled versus individual samples. One study collected smear and culture results from all participants in addition to the GeneXpert result (11). Four studies tested sputum using Xpert MTB/RIF (11\u201314) and 2 with Ultra (10,15) (Table 1).\nThese 6 studies tested 1,878 individual samples. Participants were recruited from hospitals (n = 262), ambulatory clinics (n = 914), and outreach activities (n = 702). The percentage of individual patients with Xpert MTB/RIF-positive tests included in the pools ranged from 8.9% to 37%, except for 1 in vitro study, which used spiked samples and prepared pools with up to 64% of positive samples. Only 15 (0.8%) participants across all studies had rifampin resistance (Table 1). Overall, of the 690 pools tested, 117 pooled 2 samples, 28 pooled 3 samples, 364 pooled 4 samples, 37 pooled 5 samples, 16 pooled 6 samples, 36 pooled 8 samples, 16 pooled 10 samples, 36 pooled 12 samples, and 40 pooled 16 samples. Most of the pools with high numbers of samples (\u22656) per pool were in the benchbased study. Only 2 studies reported specificity, 1 in which pools were tested with Xpert MTB/RIF (99%, 95% CI 94%\u2013100%) and 1 in which pools were tested with Ultra (100%, 95% CI 96%\u2013100%; Table 2) (12,15).\n\nThe 2 studies (13,14) combining 2 sputum samples per pool reported 87.5% and 96.0% Xpert MTB/ RIF sensitivity relative to individual testing (Figure 2, panel A). The 4 studies combining 4 samples per pool reported sensitivities of 88% (10) and 96% (12) for Xpert MTB/RIF and 95% (13) and 100% (15) for Ultra (Figure 2, panel B). In 2 studies (10,13), pools combining >4 sputum samples reported lower sensitivity ranges for Xpert MTB/RIF (63%\u201381%) and for Ultra (80%\u2013100%) (Table 2).\nGiven that all studies had <200 pools, we combined the results from all studies with similar pool sizes and test type (e.g., all studies that pooled 4 samples and test them using Xpert MTB/RIF) to evaluate the effect of the number of pooled samples on accuracy. Although this approach has limitations due to variations in study design and proportion of sample positivity, we believe the benefit of this preliminary analysis of the potential use of pooling during the COVID-19 pandemic outweighs these limitations. After combination, when using Xpert MTB/RIF, 114/117 2-sputa pools\nFigure 1. Flow diagram of study selection for a systematic review of pooling sputum as an efficient method for Xpert MTB/RIF and Ultra (Cepheid, https://www.cepheid.com) testing for tuberculosis during the coronavirus disease pandemic.\n\nEmerging Infectious Diseases \u2022 www.cdc.gov/eid \u2022 Vol. 27, No. 3, March 2021\n\n721\n\nSYNOPSIS\n\nTable 1. Characteristics of the studies, number of participants, and pool size used in a systematic review of pooling sputum as an\n\nefficient method for Xpert MTB/RIF and Ultra testing for tuberculosis during the coronavirus disease pandemic*\n\nParticipants\n\nGX\n\nRIF-\n\nrecruited\n\nNo.\n\ncartridge Pooling No. GX-pos,\u2020 GX-neg,\u2020 pos,\n\nStudy Country\n\nfrom samples Culture used\n\nratio pools no. (%)\n\nno. (%)\n\nno. Comments\n\n(11) South Reference 100\n\nYes MTB/RIF 1:5\n\n20 20 (20.6) 80 (79.4)\n\n5 Culture and\n\nAfrica laboratory\n\nSM pos\n\n85\n\n1:5\n\n17\n\n17 (20)\n\n68 (32)\n\n3\n\nCulture\n\npos/SM neg\n\n(12) Nigeria\n\nOPD\n\n729\n\nNo MTB/RIF 1:4 185\u2021 115 (15.8) 614 (84.2) 4\n\nCompared\n\nactive and\n\npassive case\n\nfinding\n\n(13) Vietnam\n\nSS\n\n118\n\nNo MTB/RIF 1:2\n\n16 75 (63.6) 43 (36.4) NR\n\nNone\n\n1:4\n\n16\n\n1:6\n\n16\n\n1:8\n\n16\n\n1:10\n\n16\n\n1:12\n\n16\n\n(14) Vietnam Hospitals 262\n\nNo MTB/RIF 1:2 101\u00a7 99 (37.7) 163 (62.3) NR\n\nPools\n\nconstructed 1\n\npos/1 neg\n\n(15) Cambodia ACF\n\n584\n\nNo\n\nULTRA\n\n1:4\n\n125 91 (15.6) 493 (84.4) 3\n\nUsed chest\n\n1:3\n\n28\n\nradiograph to\n\nscreen\n\n(10)\n\nBrazil Prisons, SS 1,120 Yes\n\nULTRA\n\n1:4\n\n20 100 (8.9) 1,020 (91.1) NR\n\nNone\n\n1:8\n\n20\n\n1:12\n\n20\n\n1:16\n\n40\n\n*Xpert MTB/RIF and Ultra, Cepheid (https://www.cepheid.com). ACF, active case finding; GX, GeneXpert; hosp, hospitalized patients; neg, negative; NR,\n\nnot reported; OPD, outpatient department; pos, positive; RIF, rifampin; SM, smear; SS, spiked samples.\n\n\u2020Single tests.\n\n\u20213 had failed results.\n\n\u00a72 had failed results.\n\nand 101/201 4-sputa pools tested contained an Xpert MTB/RIF-positive sputum; when using Ultra, 93/173 4-sputa pools tested contained an Ultra-positive sputum. If only pools containing a positive sputum sample were considered, 109/114 2-sputa pools tested by Xpert MTB/RIF had a MTB-positive result (sensitivity 93.2%, 95% CI 87.0%\u201396.4%), and 94/101 4-sputa pools tested by Xpert MTB/RIF had a MTB-positive result (sensitivity 93.0%, 95% CI 86.4%\u201396.6%). Lastly, 92/93 of the 4-sputa pools tested by Ultra had an MTB-positive result (sensitivity 98.9%, 95% CI 94.1%\u201399.9%), an increase in sensitivity over those tested by Xpert MTB/RIF.\nStudies reported slight changes in the cycle threshold (Ct) values of the pooled samples compared with the individual tests. Most of the Ct changes were relatively small, although studies were not sufficiently powered to determine statistical significance. One study reported that the pooled Xpert MTB/RIF test was negative in 5/10 samples with very low individual Xpert MTB/RIF semiquantitative results (12). The South African study that used reconstituted processed sputa to generate pools reported that 20 pools containing 1 smear-positive and 4 smear-negative, but culture-positive, samples yielded a median Xpert MTB/RIF Ct value increase of 12 (IQR 0.3\u201320.0), and\n\n22 pools containing only smear-negative/culturepositive samples had a median Ct increase of 6.2 (IQR 3.2\u201316.0) (11). Another study (13) also reported that Xpert MTB/RIF Ct values increased slightly with increasing pool ratios and, although most pools had Ct values similar to the individual sample tests, pools containing >12 sputum samples had a median increase in Ct value of 2.1 (IQR 0.0\u20134.5).\nA study from South Africa (11) reported 5 fivesample pools in which 1 was smear-positive/culturepositive and RIF-resistant and 3 five-sample pools in which 1 was smear-negative/culture-positive and RIFresistant. All 8 pools containing RIF-resistant samples tested positive for RIF-resistance (11). However, in Chry et al. (15), of the 3 MTB-positive/RIF-resistant samples subjected to Ultra testing, the pools containing the samples yielded MTB-positive but RIF-sensitive results. Abdurrahman et al. (12) included MTB-positive/ RIF-resistant samples in all 4 pools, of which 3 were detected by Xpert MTB/RIF as MTB-positive/RIF-resistant and 1 as MTB-positive/RIF-sensitive.\nOnly 2 studies (12\u201315) reported on the operational effects of using a pooling method, including cartridge costs and time savings. The 2 studies (12,15) using 4 samples per pool reported savings in cartridge costs alone of 31% ($2,295 on 230 Xpert MTB/RIF\n\n722\n\nEmerging Infectious Diseases \u2022 www.cdc.gov/eid \u2022 Vol. 27, No. 3, March 2021\n\nPooling Sputum for TB Testing during COVID-19\n\nTable 2. Tuberculosis Xpert results of pools composed of positive and negative samples, with sensitivity and specificity, in a\n\nsystematic review of pooling sputum as an efficient method for Xpert MTB/RIF and Ultra testing for tuberculosis during the coronavirus\n\ndisease pandemic\n\nTest results, no.\n\nSensitivity, Specificity,\n\nStudy\n\nPooling ratio\n\nTrue pos\u2020 False pos\u2021 False neg\u2020 True neg\u2021 % (95% CI) % (95% CI)\n\n(11)\n\n1:5 (Cult neg/SM pos)\n\n20\n\nNA\n\n0\n\nNA\n\n100 (80\u2013100)\n\nNR\n\n1:5 (Cult pos/SM neg)\n\n13\n\nNA\n\n4\n\nNA\n\n76 (50\u201392)\n\nNR\n\n(12)\n\n1:4\n\n80\n\n1\n\n5\n\n96\n\n94 (87\u201398) 99 (94\u2013100)\n\n(13)\n\n1:2\n\n14\n\nNA\n\n2\n\nNA\n\n88 (62\u201398)\n\nNR\n\n1:4\n\n14\n\nNA\n\n2\n\nNA\n\n88 (62\u201398)\n\nNR\n\n1:6\n\n11\n\nNA\n\n5\n\nNA\n\n69 (41\u201398)\n\nNR\n\n1:8\n\n10\n\nNA\n\n6\n\nNA\n\n63 (35\u201385)\n\nNR\n\n1:10\n\n13\n\nNA\n\n3\n\nNA\n\n81 (54\u201396)\n\nNR\n\n1:12\n\n13\n\nNA\n\n3\n\nNA\n\n81 (54\u201396)\n\nNR\n\n(14)\n\n1:2\n\n95\n\nNA\n\n4\n\nNA\n\n96 (90\u201399)\n\nNR\n\n(15)\n\n1:4\n\n73\n\n0\n\n0\n\n80\n\n100 (95\u2013100) 100 (96\u2013100)\n\n(10)\n\n1:4\n\n19\n\nNA\n\n1\n\nNA\n\n95 (75\u2013100)\n\nNR\n\n1:8\n\n20\n\nNA\n\n0\n\nNA\n\n100 (83\u2013100)\n\nNR\n\n1:12\n\n16\n\nNA\n\n4\n\nNA\n\n80 (56\u201394)\n\nNR\n\n1:16\n\n39\n\n0\n\n1\n\n0\n\n98 (87\u2013100)\n\nNR\n\n*Xpert MTB/RIF and Ultra, Cepheid (https://www.cepheid.com). Cult, culture, NA, not applicable; neg, negative; NR, not reported; pos, positive;\n\nSM, smear.\n\n\u2020At least one of the patients included in the pool had an Xpert-positive test.\n\n\u2021All patients included in the pool were Xpert-negative in the individual tests.\n\ncartridges) and 27% ($2,092 on 202 Ultra cartridges). These 2 studies also reported reductions of 377 (62%) and 226 (26%) hours in the staff time required to process and run samples (Table 3). All 6 studies included comments indicating the pooling procedure was feasible and beneficial. The study from South Africa (11) noted the lower sensitivity found among smearnegative/culture-positive patients. Several studies mentioned the need for specific training on the pooling procedure. The only negative effect, reported anecdotally, was the need to process samples more carefully to avoid handling and reporting errors. No studies included data on patient outcomes, such as treatment initiation.\nDiscussion This systematic review synthesizes the available literature on the performance of the pooling method\n\nusing sputum for GeneXpert testing for detecting pulmonary TB. Although the number of studies is small, the studies reported high sensitivity and specificity for 1:2 and 1:4 pooling ratios, replicating single test results, but pooling >4 samples decreased sensitivity. Studies reporting Ct values consistently reported a slight increase in Ct values and corresponding lower MTB/RIF semiquantitative results for pooled samples. This result is to be expected because testing samples together necessarily dilutes individual samples. Efficiency gained by pooling samples could increase the resilience of TB diagnostic services in a time when health system resources are being challenged by the COVID-19 pandemic.\nThe Xpert MTB/RIF Ultra cartridge was expected to help improve the sensitivity of pooled tests because the new assay has a much lower limit for detection than Xpert MTB/RIF (16). Ultra\u2019s improved\n\nFigure 2. Sensitivity and specificity for pooling sputum in the ratio of 1:2 (A) and pooling sputum in the ratio of 1:4 (B) in a systematic review of pooling sputum as an efficient method for Xpert MTB/RIF and Ultra testing (Cepheid, https://www.cepheid.com) for tuberculosis during the coronavirus disease pandemic.\n\nEmerging Infectious Diseases \u2022 www.cdc.gov/eid \u2022 Vol. 27, No. 3, March 2021\n\n723\n\nSYNOPSIS\n\nTable 3. Potential cost and time savings and positive and negative effects of pooling in a systematic review of pooling sputum as an\n\nefficient method for Xpert MTB/RIF and Ultra testing for tuberculosis during the coronavirus disease pandemic*\n\nStudy\n\nCartridge savings\n\nTime savings, h (%)\n\nNegative effects\n\nPositive effects\n\n(11) Model of 1,000 patients with\n\nNR\n\nLower sensitivity for\n\nProcesses higher volume of samples with\n\nTB prevalence rate of 3%\n\nsmear-negative\n\nfewer materials; time savings\n\nfound 67.5% cartridge\n\ntuberculosis; requires\n\nsavings\n\nlaboratory infrastructure\n\nand training\n\n(12)\n\n11% cartridge savings for\n\n377 (62%)\n\nSteps involved heighten High-level agreement with individual Xpert\n\nhospital-based patients\n\npotential for errors\n\nresults at reduced cost; substantial time\n\nsavings to process hospital samples\n\n41% cartridge savings for\n\nNR\n\nNR\n\nHigher savings on cartridge cost and\n\npatients identified through\n\nprocessing time for patients identified\n\nactive case finding\n\nthrough active case finding\n\n(13)\n\nNR\n\nNR\n\nNR\n\nImproved feasibility and cost-effectiveness\n\nof large-scale testing; reduced\n\nnumber of cartridges\n\n(14)\n\nNR\n\nNR\n\nIncrease in \u201cerror\u201d results Reduced costs and number of cartridges\n\nwhen using less buffer for\n\npooling compared with\n\nstandard buffer technique\n\n(15) 27% (lower savings estimate 226/876 (26%) for\n\nNR\n\nMethod feasible; potential to reduce\n\nusing combination of\n\nall samples; 300/876\n\ncosts, increase throughput. Pooling can be\n\napproaches)\n\n(30%) if hybrid\n\nused selectively if another screening test\n\napproach used\n\n(e.g., radiograph) used for additional\n\nsavings (hybrid approach)\n\n34.5% (if used in patients\n\nNR\n\nNR\n\nHigher savings if only samples from\n\nwith normal chest x-rays)\n\npatients without abnormal chest radiographs\n\nare included\n\n(10)\n\nNR\n\nNR\n\nNR\n\nMethod sensitive and cost-effective\n\n*NR, not reported.\n\nperformance was confirmed by the higher sensitivities reported in 2 studies included in this review, suggesting that Ultra may be preferred over Xpert MTB/ RIF for pooled sample testing (10,15). Moreover, the only 2 studies reporting specificity (of 99% and 100%) indicated that almost all pools containing all negative individual samples correctly reported negative results for the pooled samples (12\u201315). This is an important consideration because the additional steps required to split sputum samples and the need to keep track of sputum batches with a link between individual samples could be prone to cross contamination and error. Further studies are needed to replicate these findings under operational conditions.\nRegarding the reproducibility of RIF resistance results in pooled samples, in 1 study from South Africa, all 8 individual RIF-resistant results were detected as pooled RIF-resistant (11). However, in a study in Cambodia, 3 samples with RIF-resistant results from individual testing were reported as RIF-susceptible in the pooled testing (15) and in a study from Nigeria, pooling missed 1 of 4 RIF-resistant results (12). Although pooling seems to be an unreliable method to detect RIF resistance, in practice all samples from MTB-positive pools would be retested individually, which should replicate RIF resistance results from individual samples.\nAlmost all studies reported anecdotal positive feedback from laboratory staff, and 2 studies (12,15)\n\nquantified savings in cartridge costs and staff time required to process samples. Although both of those studies reported substantial savings, they were conducted in populations with a high proportion of patients testing positive. If a high proportion of presumptive TB patients is expected to be positive, presumably a greater proportion of pools would test positive and require follow-up testing of individual samples. Savings therefore would be more substantial when applied within outreach case-finding activities in the community, where typically around 5% of samples are Xpert MTB/RIF-positive (12) and lower in referral and congregate centers (e.g., prisons), where patients might have a higher probability of having TB. The expected proportion of positive samples may therefore guide the pooling ratio selected for evaluation. For example, in active case finding, it is likely a pool ratio of 1:4 would be highly efficient and generate substantial savings, whereas a ratio of 1:2 would be more suitable for busy TB diagnostic centers where the proportion of samples that are positive can be as high as 15%. Pooling is not likely to be useful at a much higher prevalence than 20%, because most of the pools would be positive and samples would have to be retested individually (B.G. Williams, unpub. data, https://arxiv.org/abs/1007.4903). Moreover, there are operational issues that need further study, as it is unclear whether the timing of sputum splitting\n\n724\n\nEmerging Infectious Diseases \u2022 www.cdc.gov/eid \u2022 Vol. 27, No. 3, March 2021\n\nPooling Sputum for TB Testing during COVID-19\n\ncould affect results. For example, splitting samples before adding the GeneXpert buffers requires dividing thick and infectious samples, which are likely to have unevenly distributed bacilli, whereas splitting after adding the buffers could increase the risk of cross contamination but provide a safer and more liquid sample with more evenly distributed bacilli.\nTo inform national programs, further research is needed to determine the effects on time savings from pooled testing, from sample collection to notification and treatment initiation. Two studies quantified large reductions in testing time from pooling (12,15), which could shorten turnaround times for patient notification, but time to notification was not reported in any of the studies. Quality management of the pooling process is critical, as reflected in discussions in the studies highlighting the importance of sample management and procedure training. As with routine testing procedures, ensuring that pooling is implemented in a biosafe and quality-assured manner would help mitigate risk to laboratorians from increased sample manipulation and prevent errors in sample handling and testing, which could reduce efficiency and benefit to both patients and programs.\nOur findings are especially relevant during the ongoing global COVID-19 pandemic, which is severely disrupting health services, the availability of diagnostic and treatment resources, supply chains, and other disease control efforts. Although the diagnosis of COVID-19 takes precedence, steps can be taken to preserve key services for diagnosing and treating patients with presumptive TB. Quarantine and restriction of movement during the pandemic have limited accessibility to services and reduced the numbers of patients attending TB diagnostic and treatment centers. Confinement of the population to households and the resulting increase in contact with other household members in crowded conditions could increase TB transmission. A surge in undetected cases, together with increases in treatment interruptions, will likely lead to increases in incident cases. Demand for testing also may cause severe resource constraints. Preparing for this scenario, such as by introducing pooling strategies, may result in more efficient use of limited resources.\nBefore the COVID-19 pandemic, the World Health Organization issued guidelines promoting a rapid diagnostic test, such as a GeneXpert-based test, for all persons with presumptive TB (17). However, <20% of the GeneXpert TB tests necessary to test the estimated 100 million people who develop presumptive TB each year have been procured (2). Individual rapid molecular diagnostic testing for all patients\n\nwith presumptive TB remains the standard of care and a goal for national TB programs worldwide, but the cost of individually testing all estimated symptomatic persons using GeneXpert would have been more than US $1 billion in cartridges alone in 2018 (2), more than the total amount of funding provided by international donors globally for TB in 2019 (18). Moreover, although passive case finding has long been the standard approach in many countries, it is becoming apparent that outreach beyond health facilities is needed to identify those with TB missed by programs (19). Increasing outreach activities usually means more testing, requiring more cartridges, will be needed. However, a typically greater negative-topositive testing ratio in persons identified through outreach activities means that pooling strategies might decrease costs.\nDespite the potential usefulness of our findings, the quality of evidence we present remains insufficient to support wide adoption of the pooling method. Because the 6 studies were heterogeneous, we were unable to conduct a meta-analysis, and we considered all the studies together with bench evaluations of the technical sensitivity and specificity of the methods; our findings should therefore be considered hypothesis-generating to promote and inform further studies. Moreover, all studies were underpowered for investigating the performance of the pooled testing method in subpopulations (e.g., HIV-positive vs. HIV-negative, men vs. women), and very few samples tested rifampin resistant. Ct values also need to be interpreted with caution.\nAlthough both Xpert MTB/RIF and Ultra tests report Ct values, the test algorithms that determine their Ct and semiquantitative results differ, which impacts the interpretation of Ct-based analyses. Moreover, because Ct ranges vary between multiple tests on the same homogenized sample, it would have been preferable to describe changes in positivity relative to the semiquantitative results. However, semiquantitative results were not reported in most studies. Similarly, although culture was used in some of the studies, this information was not used to stratify analyses. A second reference method would have been useful to further investigate whether discordant results were potentially due to improper sample management, cross-contamination in the laboratory, or random variation due to the bacilli not being homogeneously distributed in the sputum sample.\nDespite these limitations, we propose that the pooling method be considered as an interim option to strengthen capacity of TB laboratories during times of crisis, such as during the COVID-19 pandemic. Our\n\nEmerging Infectious Diseases \u2022 www.cdc.gov/eid \u2022 Vol. 27, No. 3, March 2021\n\n725\n\nSYNOPSIS\n\nteam is currently conducting accelerated evaluations of the pooling method in Laos and Nigeria. We encourage the TB community to conduct studies on the pooling strategy and other resource-saving strategies for TB diagnostic testing that generates data for open access databases to inform national programs.\nThis research was funded in part by the European and Developing Countries Clinical Trial Partnership (grant no. DRIA2014-309) and its cofunders, Medical Research Council UK and Instituto de Salud Carlos III, Spain; Coordena\u00e7\u00e3o de Aperfei\u00e7oamento de Pessoal de N\u00edvel Superior, Brazil, as a travel scholarship for SVMAL (process no. 88881. 187327/2018\u201301); TB REACH grant (STBP/TBREACH/ GSA/2020-04) supported by Global Affairs Canada; UK Department for International Development, LIGHT Health Research Programme Consortium (contract pending); UK National Institute for Health Research, Health Protection Research Unit in Emerging and Zoonotic Infections, Centre of Excellence in Infectious Diseases Research; and the Alder Hey Charity.\nAbout the Author\nLuis Cuevas is professor of international health and epidemiology at the Liverpool School of Tropical Medicine. His main research focus is the evaluation of diagnostics for high burden and emerging infections for use in locations with limited resources, with a primary interest in tuberculosis.\nReferences 1. World Health Organization. WHO monitoring of Xpert\nMTB/RIF roll-out. Geneva: The Organization; 2019 [cited 2020 Apr 1]. https://www.who.int/tb/areas-of-work/ laboratory/mtb-rif-rollout/en 2. Van Deun A, Tahseen S, Affolabi D, Hossain MA, Joloba ML, Angra PK, et al. Sputum smear microscopy in the Xpert\u00ae MTB/RIF era. Int J Tuberc Lung Dis. 2019;23:12\u20138. https://doi.org/10.5588/ijtld.18.0553 3. World Health Organization. WHO: Global TB progress at risk. Geneva: The Organization; 2020. [cited 2020 Nov 1]. https://www.who.int/news/item/14-10-2020-who-globaltb-progress-at-risk 4. Emmanuel JC, Bassett MT, Smith HJ, Jacobs JA. Pooling of sera for human immunodeficiency virus (HIV) testing: an economical method for use in developing countries. J Clin Pathol. 1988;41:582\u20135. https://doi.org/10.1136/jcp.41.5.582 5. Morandi PA, Schockmel GA, Yerly S, Burgisser P, Erb P, Matter L, et al. Detection of human immunodeficiency virus type 1 (HIV-1) RNA in pools of sera negative for antibodies to HIV-1 and HIV-2. J Clin Microbiol. 1998;36:1534\u20138. https://doi.org/10.1128/JCM.36.6.15341538.1998 6. Peeling RW, Toye B, Jessamine P, Gemmill I. Pooling of urine specimens for PCR testing: a cost saving strategy for Chlamydia trachomatis control programmes. Sex Transm Infect. 1998;74:66\u201370. https://doi.org/10.1136/sti.74.1.66\n\n7. Mine H, Emura H, Miyamoto M, Tomono T, Minegishi K, Murokawa H, et al.; Japanese Red Cross NAT Research Group. High throughput screening of 16 million serologically negative blood donors for hepatitis B virus, hepatitis C virus and human immunodeficiency virus type-1 by nucleic acid amplification testing with specific and sensitive multiplex reagent in Japan. J Virol Methods. 2003;112:145\u201351. https://doi.org/10.1016/S0166-0934(03)00215-5\n8. Lindan C, Mathur M, Kumta S, Jerajani H, Gogate A, Schachter J, et al. Utility of pooled urine specimens for detection of Chlamydia trachomatis and Neisseria gonorrhoeae in men attending public sexually transmitted infection clinics in Mumbai, India, by PCR. J Clin Microbiol. 2005;43:1674\u20137. https://doi.org/10.1128/ JCM.43.4.1674-1677.2005\n9. Westreich DJ, Hudgens MG, Fiscus SA, Pilcher CD. Optimizing screening for acute human immunodeficiency virus infection with pooled nucleic acid amplification tests. J Clin Microbiol. 2008;46:1785\u201392. https://doi.org/10.1128/ JCM.00787-07\n10. Santos P, Santos A, Verma R, Oliveira R, Camioli C, Lemos E, et al. The utility of pooling sputum samples for mass screening for tuberculosis in prisons using Xpert MTB/RIF Ultra. In: 50th World Conference on International Union Against Tuberculosis and Lung Disease; 2019 Oct 30\u2013 Nov 2; Hyderabad, India. Abstract SOA-01-1001-31. Int J Tuberc Lung Dis. 2019;23:S110\u20131 [cited 2020 Mar 30]. https://hyderabad.worldlunghealth.org/wp-content/ uploads/2019/11/20191101_UNION2019_Abstracts_ Final.pdf\n11. Zishiri V, Chihota V, McCarthy K, Charalambous S, Churchyard GJ, Hoffmann CJ. Pooling sputum from multiple individuals for Xpert\u00ae MTB/RIF testing: a strategy for screening high-risk populations. Int J Tuberc Lung Dis. 2015;19:87\u201390. https://doi.org/10.5588/ijtld.14.0372\n12. Abdurrahman ST, Mbanaso O, Lawson L, Oladimeji O, Blakiston M, Obasanya J, et al. Testing pooled sputum with Xpert MTB/RIF for diagnosis of pulmonary tuberculosis to increase affordability in low-income countries. J Clin Microbiol. 2015;53:2502\u20138. https://doi.org/10.1128/ JCM.00864-15\n13. Ho J, Jelfs P, Nguyen PTB, Sintchenko V, Fox GJ, Marks GB. Pooling sputum samples to improve the feasibility of Xpert\u00ae MTB/RIF in systematic screening for tuberculosis. Int J Tuberc Lung Dis. 2017;21:503\u20138. https://doi.org/10.5588/ ijtld.16.0846\n14. Phuong NTB, Anh NT, Van Son N, Sintchenko V, Ho J, Fox GJ, et al. Effect of two alternative methods of pooling sputum prior to testing for tuberculosis with Genexpert MTB/RIF. BMC Infect Dis. 2019;19:347. https://doi.org/10.1186/s12879-019-3778-9\n15. Chry M, Smelyanskaya M, Ky M, Codlin AJ, Cazabon D, Tan Eang M, et al. Can the high sensitivity of Xpert MTB/RIF Ultra be harnessed to save cartridge costs? Results from a pooled sputum evaluation in Cambodia. Trop Med Infect Dis. 2020;5:27. https://doi.org/10.3390/ tropicalmed5010027\n16. Chakravorty S, Simmons AM, Rowneki M, Parmar H, Cao Y, Ryan J, et al. The new Xpert MTB/RIF Ultra: improving detection of Mycobacterium tuberculosis and resistance to rifampin in an assay suitable for point-of-care testing. MBio. 2017;8:e00812\u20137. https://doi.org/10.1128/ mBio.00812-17\n17. World Health Organization. Automated real-time nucleic acid amplification technology for rapid and simultaneous detection of tuberculosis and rifampicin resistance: Xpert\n\n726\n\nEmerging Infectious Diseases \u2022 www.cdc.gov/eid \u2022 Vol. 27, No. 3, March 2021\n\nPooling Sputum for TB Testing during COVID-19\n\nMTB/RIF assay for the diagnosis of pulmonary and extrapulmonary TB in adults and children: policy update. Geneva: The Organization; 2013 [cited 2020 Apr 1]. https://apps.who.int/iris/bitstream/handle/10665/ 112472/9789241506335_eng.pdf 18. World Health Organization. Global tuberculosis report 2019. Geneva: The Organization; 2019. [cited 2020 Apr 1]. https://apps.who.int/iris/bitstream/ handle/10665/329368/9789241565714-eng.pdf\n\n19. Dowdy DW, Basu S, Andrews JR. Is passive diagnosis enough? The impact of subclinical disease on diagnostic strategies for tuberculosis. Am J Respir Crit Care Med. 2013;187:543\u201351. https://doi.org/10.1164/rccm.201207-1217OC\nAddress for correspondence: Luis E. Cuevas, Liverpool School of Tropical Medicine, Pembroke Place, Liverpool L3 5QA, UK; email: Luis.Cuevas@lstmed.ac.uk\n\nJanuary 2020\nViruses\n\n\u2022  Spatial Epidemiologic Trends and Hotspots of Leishmaniasis, Sri Lanka, 2001\u20132018\n\u2022  Candidatus Mycoplasma haemohominis in Human, Japan\n\u2022  Nutritional Care for Patients with Ebola Virus Disease\n\u2022  Paid Leave and Access to Telework as Work Attendance Determinants during Acute Respiratory Illness, United States, 2017\u20132018\n\u2022  Preclinical Detection of Prions in Blood of Nonhuman Primates Infected with Variant Creutzfeldt-Jakob Disease\n\u2022  Effect of Acute Illness on Contact Patterns, Malawi, 2017\n\u2022  Outbreak of Peste des Petits Ruminants among Critically Endangered Mongolian Saiga and Other Wild Ungulates, Mongolia, 2016\u20132017\n\u2022  Elephant Endotheliotropic Herpesvirus Hemorrhagic Disease in Asian Elephant Calves in Logging Camps, Myanmar\n\u2022  Risk Factors for and Seroprevalence of Tickborne Zoonotic Diseases among Livestock Owners, Kazakhstan\n\u2022  High Azole Resistance in Aspergillus fumigatus Isolates from Strawberry Fields, China, 2018\n\u2022  Tick-Borne Encephalitis Virus, United Kingdom\n\u2022  Phenotypic and Genotypic Correlates of Penicillin Susceptibility in Nontoxigenic Corynebacterium diphtheriae, British Columbia, Canada, 2015\u20132018\n\n\u2022  High Pathogenicity of Nipah Virus from Pteropus lylei Fruit Bats, Cambodia\n\u2022  Varicella in Adult Foreigners at a Referral Hospital, Central Tokyo, Japan, 2012\u20132016\n\u2022  Geographic Distribution and Incidence of Melioidosis, Panama\n\u2022  Shigella Bacteremia, Georgia, USA, 2002\u20132012\n\u2022  Distribution of Japanese Encephalitis Virus, Japan and Southeast Asia, 2016\u20132018\n\u2022  Novel Reassortant Highly Pathogenic Avian Influenza A(H5N2) Virus in Broiler Chickens, Egypt\n\u2022  Infectivity of Norovirus GI and GII from Bottled Mineral Water during a Waterborne Outbreak, Spain\n\n\u2022  Effect of Pediatric Influenza Vaccination on Antibiotic Resistance, England and Wales\n\u2022  Locally Acquired Human Infection with Swine-Origin Influenza A(H3N2) Variant Virus, Australia, 2018\n\u2022  Use of Ambulance Dispatch Calls for Surveillance of Severe Acute Respiratory Infections\n\u2022  Hantavirus Pulmonary Syndrome in Traveler Returning from Nepal to Spain\n\u2022  Visceral Leishmaniasis, Northern Somalia, 2013\u20132019\n\u2022  Autochthonous Human Fascioliasis, Belgium\n\u2022  Recombinant Nontypeable Genotype II Human Noroviruses in the Americas\n\u2022 L  egionella pneumophila as Cause of Severe Community-Acquired Pneumonia, China\n\u2022  Training for Foodborne Outbreak Investigations by Using Structured Learning Experience\n\u2022  Emergence of Vibrio cholerae O1 Sequence Type 75 in Taiwan\n\u2022  Diabetes Mellitus, Hypertension, and Death among 32 Patients with MERS-CoV Infection, Saudi Arabia\n\u2022  Influenza D Virus of New Phylogenetic Lineage, Japan\n\u2022  Diagnosis of Syphilitic Bilateral Papillitis Mimicking Papilloedema\n\u2022  Influenza A Virus Infections in Dromedary Camels, Nigeria and Ethiopia, 2015\u20132017\n\nTo revisit the January 2020 issue, go to:\n\n\u00ae\n\nhttps://wwwnc.cdc.gov/eid/articles/issue/26/1/table-of-contents\n\nEmerging Infectious Diseases \u2022 www.cdc.gov/eid \u2022 Vol. 27, No. 3, March 2021\n\n727\n\n\n",
    "authors": [
      "Luis E. Cuevas",
      "Victor S. Santos",
      "Shirley Ver\u00f4nica Melo Almeida Lima",
      "Konstantina Kontogianni",
      "John S. Bimba",
      "Vibol Iem",
      "Jose Dominguez",
      "Emily Adams",
      "Ana Cubas Atienzar",
      "Thomas Edwards",
      "S. Bertel Squire",
      "Patricia J. Hall",
      "Jacob Creswell"
    ],
    "doi": "10.3201/eid2703.204090",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://wwwnc.cdc.gov/eid/article/27/3/20-4090_article.htm"
  },
  {
    "key": "T7FKPC83",
    "title": "Patient costs for the diagnosis of tuberculosis in Brazil: comparison of Xpert<SUP>\u00ae</SUP> MTB/RIF and smear microscopy",
    "abstract": "S E T T I N G : Manaus and Rio de Janeiro, two Brazilian state capitals with the country\u2019s fifth and sixth highest tuberculosis (TB) incidence rates (around 90/100 000 population in 2012).",
    "full_text": "INT J TUBERC LUNG DIS 18(5):547\u2013551 Q 2014 The Union http://dx.doi.org/10.5588/ijtld.13.0637\n\nPatient costs for the diagnosis of tuberculosis in Brazil: comparison of XpertW MTB/RIF and smear microscopy\n\nR. da Silva Antunes,* M. Pinto,\u2020 A. Trajman*\u2021\n*Post-graduate Program on Health Education, Gama Filho University, Rio de Janeiro, Rio de Janeiro, \u2020Instituto Fernandes Figueira, Fiocruz, Rio de Janeiro, Rio de Janeiro, Brazil; \u2021Montreal Chest Institute, McGill University, Montreal, Quebec, Canada\n\nSUMMARY\n\nS E T T I N G : Manaus and Rio de Janeiro, two Brazilian state capitals with the country\u2019s fifth and sixth highest tuberculosis (TB) incidence rates (around 90/100 000 population in 2012). O B J E C T I V E : To compare the costs of the Xpertw MTB/ RIF assay with those of standard care (two smears) in diagnosing TB from the patient\u2019s perspective. M E T H O D : We interviewed 218 patients diagnosed with TB in the previous 4 months by Xpert or smear microscopy. Information on non-medical direct costs for transportation and food, indirect costs such as time spent for diagnostic visits and socio-demographic data were gathered. R E S U LT S : The median patient income was US$390.24. Median total costs incurred by patients were 54% higher\n\nwith the smear process than with Xpert (US$25.24 vs. US$16.44, P , 0.000) due to higher indirect and direct costs. Male patients incurred higher indirect costs (U$10.27 vs. US$7.51, P \u00bc 0.038), and patients in Manaus incurred higher total costs. C O N C L U S I O N S : Although the diagnosis and treatment of TB in Brazil are free of charge, non-medical direct and indirect costs for patients may represent important barriers to accessing appropriate care. Compared to standard care, Xpert reduced the financial burden for patients. These findings support the decision to scale-up Xpert technology in the country. K E Y W O R D S : costs; diagnosis; real-time polymerase chain reaction; sputum smear; tuberculosis\n\nIN 2010, THE WORLD HEALTH ORGANIZATION endorsed the Xpertw MTB/RIF assay (Xpert, Cepheid, Sunnyvale, CA, USA) for countries with a high prevalence of multidrug-resistant tuberculosis and human immunodeficiency virus (HIV) infection1 after demonstration and validation studies had shown the high accuracy of the test for detecting both Mycobacterium tuberculosis DNA and rifampicin resistance (RMP).2,3 More recently, a metaanalysis confirmed the high accuracy of the test, and cost-effectiveness studies have shown the benefit of implementing this technology.4,5 South Africa was the first country to widely adopt Xpert technology, and some countries have plans to scale up Xpert in the near future.6 However, to the best of our knowledge, no economic analyses has been conducted from the patient\u2019s perspective to evaluate the eventual benefit of this technology on the financial burden of tuberculosis (TB) patients.\nBrazil is one of the 22 high TB burden countries, with around 70 000 new cases reported in 2012.7 Among these, 9.7% are co-infected with HIV and around 15% are not confirmed by smear microscopy\n\n(50% in Manaus and Rio de Janeiro), the standard test for diagnosing TB in the country.8 Drug resistance is relatively low: 6% to isoniazid and 1.4% to RMP (unpublished data, II National Survey on Drug Resistance, 2010 National Tuberculosis Programme [NTP]). Among retreatment cases, more than 70% have no access to culture and drug susceptibility testing.8\nIn this scenario, the Brazilian NTP decided to conduct an implementation stepped-wedge study of the Xpert assay to replace diagnostic microscopy in the two municipalities (Manaus and Rio de Janeiro) with the fifth and sixth highest incidence rates in the country \u2013 89.3 and 94.4 per 100 000 population, respectively.8 The study evaluates important operational outcomes such as the confirmation of TB among notified cases and delay to treatment initiation. In addition, a cost-effectiveness analysis from the health system perspective was conducted (unpublished data).\nIn the present study, we sought to evaluate the nonmedical direct and indirect costs of Xpert from the\n\nCorrespondence to: Anete Trajman, Gama Filho University, Rua Macedo Sobrinho 74/203, Humaita 22271-080, Rio de Janeiro, RJ, Brazil. Tel: (\u00fe55) 21 8218 9194. e-mail: atrajman@gmail.com\nArticle submitted 16 September 2013. Final version accepted 19 December 2013.\n\n548 The International Journal of Tuberculosis and Lung Disease\n\npatient\u2019s perspective in both cities where the roll-out implementation study took place.\nMETHODS\nSetting\nThe study was conducted in six primary care clinics in Rio de Janeiro and 14 in Manaus between October 2012 and June 2013. The clinics were located in different areas of the cities; those with the highest number of TB diagnoses and those where the Xpert was implemented in the stepped-wedge trial were selected. In Manaus, patients were asked to submit their samples to the reference laboratory, while in Rio de Janeiro patients provided samples to the local clinic and the health system was responsible for the logistics of transporting samples to the reference laboratory. The recommendation to provide two sputum samples was maintained during the trial, although only the first sample was tested if it was adequate for processing (no blood or pus, and sufficient volume, with a minimum of 1 ml required for the technique). In Brazil, TB diagnosis and treatment are provided free of charge.\nStudy population and data collection\nAdult patients aged .18 years with a diagnosis of pulmonary TB of no more than 4 months were eligible, regardless of test results and of the diagnostic process used \u2014 either smear microscopy (standard care, two samples recommended by the NTP)9 or Xpert (one sample recommended during the roll-out study). This time point was chosen to allow more reliable recall of the costs incurred in the prediagnostic period.\nPatients who provided signed informed consent were interviewed. We applied a standardised questionnaire, validated in a small sample of 10 patients in both municipalities, before starting the data collection. No changes in the tool were necessary. The questionnaire focused only on the costs for the tests, as costs with other items of the diagnostic process have been studied previously in Brazil.10 Interviews were conducted in a private space, without the presence of any medical staff, to guarantee privacy. The following information was gathered: sociodemographic factors such as sex, age, employment and monthly income; out-of-pocket expenditures incurred during the diagnostic visits, such as food and transportation; and time spent during the diagnostic process, such as travel, waiting time in the clinics and duration of consultation. Costs per visit were calculated and multiplied by the number of informed visits. Costs associated with treatment and with other items of the diagnostic process were not included.\n\nData analysis\nIndirect costs were calculated according to the current job activity. Monthly income was divided by 198 (44 weekly hours, 4.5 weeks/month) to calculate the hourly cost, per national legislation.11 For those without paid activity (regardless of their availability or lack of income from retirement, allowances or government cash transfer), the lost hours spent during the diagnostic process were valued based on the national minimum wage:12 678 Brazilian reais (R$), equivalent to 331 US dollars (US$) in 2012. All costs were collected in R$ and converted to $US using the mean 2012 value (US$1 \u00bc R$2.05).\nData were double-entered in SPSS 17.0 (Statistical Product and Service Solutions, Chicago, IL, USA). Patient characteristics were compared using the odds ratio (OR). Median costs (excluding, for direct costs, patients who had no costs) were compared using the Mann-Whitney test. A 0.05 significance level was considered.\nEthics\nThe study was approved by the National Ethical Committee (CONEP 782/2011) and by the Institutional Review Boards of the Rio de Janeiro Health Secretariat (445A/2011) and the Fundac\u00b8a\u02dco de Medicina Tropical Dr. Heitor Vieira Dourado (dated 31 July 2012). All patients provided written informed consent.\nRESULTS\nA total of 218 patients were invited to participate, and all accepted to be interviewed. The median age was 36 years (interquartile range [IQR] 27\u201351); 139 (64%) were male; the median income was US$390.24 (IQR 317.00\u2013634.00); 135 (62%) had ,8 years of schooling. Of the 132 (61%) patients diagnosed in Manaus, 120 (55%) were diagnosed using Xpert. Patient characteristics by type of diagnosis are given in Table 1. Apart from age (patients who underwent smear testing were younger), there was no significant difference between the two groups.\nThe median total patient costs with diagnosis by smear microscopy were US$25.24, representing a 54% increase over Xpert (US$16.44, P , 0.000). Both direct and indirect costs were higher with smear microscopy, although the proportion of the increase was higher with indirect costs (90% vs. 40%). Transportation was the main cost driver of direct costs (55%) (Table 2). Men incurred higher costs (US$22.05 vs. US$16.64, P \u00bc 0.235, Table 3) due to indirect costs (US$10.27 vs. US$7.51, P \u00bc 0.038, Table 3). This was a consequence of their higher incomes (US$439.02 vs. US$317.07, P \u00bc 0.001), as\n\nPatient costs for diagnosing TB using Xpert 549\n\nTable 1 Characteristics of 218 patients with pulmonary tuberculosis by diagnostic process, Manaus and Rio de Janeiro, 2012\u20132013\n\nXpert Smear n n (%) n (%)\n\nOR (95%CI)\n\nCity Manaus Rio de Janeiro\n\n132 70 (53) 62 (47) 0.81 (0.47\u20131.41) 86 50 (58) 36 (42) 1.0 (reference)\n\nSex Female Male\n\n79 48 (61) 31 (39) 1.44 (0.82\u20132.53) 139 72 (52) 67 (48) 1.0 (reference)\n\nAge, years ,42 !42\n\n124 59 (48) 65 (52) 0.49 (0.28\u20130.85) 94 61 (65) 33 (35) 1.0 (reference)\n\nYears of schooling Illiterate ,1 1\u20133 4\u20137 8\u201312 .12\n\n0.98 (0.56\u20131.69)* 14 8 (54) 6 (46) 17 9 (53) 8 (47) 38 21 (55) 17 (45) 66 36 (55) 30 (45) 77 41 (53) 36 (47)\n6 5 (84) 1 (17)\n\nIncome, US$\u2020 331\u2021\n.331\u2021\n\n50 39 (65) 21 (35) 1.82 (0.95\u20133.50) 107 54 (51) 53 (49) 1.0 (reference)\n\nEmployed No Yes\n\n100 51 (51) 49 (49) 0.73 (0.79\u20132.31) 118 69 (59) 49 (41) 1.0 (reference)\n\nIn activity No (n \u00bc 175) Yes (n \u00bc 43)\n\n175 98 (56) 77 (44) 1.21 (0.42\u20131.60) 43 22 (51) 21 (48) 1.0 (reference)\n\nOccupation Retired Student Salaried Informal Not working Data missing\n\n23 13 (57) 10 (43) 12 4 (33) 8 (67) 87 44 (51) 43 (49) 56 36 (64) 20 (36) 27 16 (59) 11 (41) 13 7 (54) 6 (46)\n\n*OR calculated with cut-off of 7 years. \u2020Those without income excluded (US$1 \u00bc R$2.05). \u2021Minimum wage at the time of data collection.\nOR \u00bc odds ratio; CI \u00bc confidence interval.\n\ntime spent on visits was similar (5 vs. 4 h, P \u00bc 0.543). In addition, total costs in Manaus were higher than in Rio de Janeiro (Table 3).\nIn both municipalities, indirect costs were similar to direct costs for smears, but were 30% lower than direct costs for Xpert (Table 2). The median\n\ndifference between the costs of both tests for patients represented 4% of their median income and 5% of the minimum wage.\nDISCUSSION\nThis study has shown that Xpert can significantly reduce the financial burden of patients in diagnosing TB, even in a country where TB care is free of charge and there is a universal public health care system. Xpert reduced both out-of-pocket expenditures and opportunity costs from lost hours. This was independent of the number of samples provided, as routine procedures did not change: patients were requested to provide two samples in both arms of the study. If Brazil scales up Xpert as a substitute for microscopy, only one sample will usually be necessary, and this will therefore likely further reduce costs for patients.\nPatient out-of-pocket expenses with TB diagnosis and treatment can be a barrier to appropriate care,13\u201316 and the disease has been shown to reduce income,10,17,18 perpetuating the vicious cycle of poverty, disease and disability. The financial burden of TB care in resource-poor countries can represent up to a year\u2019s income when treatment and household costs are also taken into account. Costs in most studies are mainly driven by indirect costs, particularly when hospitalisation is required.17,19\u201322\nIn our study, the proportion of costs over income was much lower than in other studies. This was partly because we only considered costs with the bacteriological test, as we were interested in specifically comparing the two same-day tests to inform the NTP about the feasibility of adopting the new technology. A previous study in Brazil has already shown that total costs in the pre-diagnostic phase involve other items such as absenteeism and the need for antibiotics and other drugs, among others;10 these costs were out of the scope of the present study. In addition, salaries in Brazil are higher than in other high-burden countries. Finally, we did not include hospitalised\n\nTable 2 Non-medical direct and indirect costs of 218 pulmonary tuberculosis patients by diagnostic process, Manaus and Rio de Janeiro, 2012\u20132013\n\nCosts\n\nXpert (n \u00bc 120) median (range)\n\nSmear microscopy (n \u00bc 98) median (range)\n\nDifference of medians P value\n\nNon-medical direct costs, US$* Transport Food\nDirect costs (total)\nIndirect costs Number of visits Hours lost per visit Total hours lost Indirect cost per visit\nIndirect cost (total)\nTotal costs (direct \u00fe indirect)\n\n5.56 (1.34\u2013243.90) (n \u00bc 84, 70%) 4.88 (1.95\u201331.22) (n \u00bc 51, 43%) 9.27 (1.34\u2013256.10) (n \u00bc 97, 81%)\n2 (1\u201315) 1.7 (0.02\u201313.4) 3.0 (0.3\u201331.7) 3.00 (0.25\u2013146.34) 6.51 (1.00\u2013365.85) 16.44 (1.50\u2013621.95)\n\n8.63 (4.88\u201397.56) (n \u00bc 68, 69%) 7.32 (1.95\u201330.73) (n \u00bc 48, 40%) 13.02 (1.95\u2013107.32) (n \u00bc 76, 78%)\n3 (1\u201310) 2.3 (0.7\u201324.9) 6.7 (1.3\u201349.8) 3.92 (1.29\u2013244.39) 12.40 (2.00\u2013353.17) 25.24 (2.00\u2013757.32)\n\n\u00c03.07 \u00c02.44 \u00c03.75\n\u00c01 \u00c00,6 \u00c03,7 \u00c00.92 \u00c05.89 \u00c08.8\n\n0.002 0.218 0.003\n,0.000 ,0.000 ,0.000 ,0.000 ,0.000 ,0.000\n\n*Those without transport and/or food costs excluded (US$1 \u00bc R$2.05).\n\nP value 0.038 0.389 0.385\n,0.000 0.078\n\n10.27 (1.00\u2013353.17) 7.51 (1.00\u2013365.85) 9.85 (1.00\u2013353.17) 9.14 (1.25\u2013365.85)\n10.14 (1.25\u2013365.85) 8.51 (1.00\u2013353.17) 6.51 (1.00\u2013365.85)\n12.40 (2.00\u2013353.17) 7.45 (1.50\u2013117.07)\n10.52 (1.00\u2013365.85)\n\nIndirect costs (US$) median (range)\n\n0.201 0.729 0.672 0.003 0.002\n\nP value\n\n9.76 (1.34\u2013121.95) (n \u00bc 110, 79%) 10.73 (1.95\u2013256.10) (n \u00bc 63, 80%) 10.24 (1.34\u201368.29) (n \u00bc 99, 80%) 10.20 (1.95\u2013256.10) (n \u00bc 74, 79%) 10.00 (1.95\u2013256.10) (n \u00bc 104, 77%) 10.73 (1.34\u2013107.32) (n \u00bc 69, 83%)\n9.27 (1.34\u2013256.10) (n \u00bc 97, 81%) 13.02 (1.95\u2013107.32) (n \u00bc 76, 78%)\n8.66 (1.34\u201353.66) (n \u00bc 64, 74%) 12.68 (1.95\u2013256.10) (n \u00bc 109, 83%)\n\nDirect costs* (US$) median (range)\n\nTable 3 Factors associated with costs incurred by 218 pulmonary tuberculosis patients according to the diagnostic process, Manaus and Rio de Janeiro, 2012/2013\n\n550 The International Journal of Tuberculosis and Lung Disease\npatients in our sample; indirect costs were thus much lower than in other studies.10,22,23 This may also explain why, unlike in other studies, indirect costs were not much higher than direct costs. However, we did observe that indirect costs were the main factor responsible for the difference between the two groups: they were 90% higher for patients who submitted samples for microscopy.\nTransportation was the driving component of direct costs, mainly in Manaus, as the health units are spread over the city, and patients\u2019 trips to the reference laboratories after their consultations to provide the sputum specimens. The organisation in Rio de Janeiro, where there is a system for the transportation of specimens from health units to laboratories, transfers patient costs to the health system.\nTransportation tickets, decentralisation of health care units and cash transfer have all been suggested as possible incentives for TB patients.14 Brazil has adopted the Family Health Strategy,24 and TB control actions have been decentralised to district clinics. However, our results show that TB patients still incur relatively high costs. Although the Brazilian Federal Government is currently evaluating a cash transfer incentive for TB patients, this is unlikely to affect individuals who submit samples for TB diagnosis, as the majority are unlikely to have TB.\nOur study has a few limitations. Recollection bias might have occurred, as we interviewed patients up to 4 months after the diagnostic process; however, this would affect both groups equally. In addition, economic, social and cultural differences between countries hamper the ability to generalise our findings. The socio-demographic characteristics of our sample are nevertheless representative of patients usually treated for TB globally, who are mostly male, of productive age, with low levels of schooling and income, and data are likely to be valid nationwide. Finally, it was out of the scope of this study to collect comprehensive data on health costs during the diagnostic phase, such as non-specific antibiotics, medical visits, vitamins and other tests. Nonetheless, given the potential of the new Xpert technology in detecting TB during earlier stages of the disease, and thus reduce transmission and morbidity due to its higher sensitivity, it is likely that its use would only further reduce health costs.\nIn conclusion, the adoption of Xpert in Brazil can help overcome barriers to accessing TB diagnosis, as it is cost saving for patients. However, these diagnostic technologies are only available to those who reach the health care system. As the technology spreads globally, it will be important to document local costs from the perspectives of both the health care system and of the patients, and to understand barriers to accessing the health care system.\n\n0.235 0.656 0.811 ,0.000 0.004\n\nP value\n\n*Those without transport and/or food costs excluded; US$1 \u00bc R$2.05.\n\n22.05 (1.50\u2013757.32) 16.64 (1.50\u2013621.95) 21.26 (1.50\u2013421.46) 17.93 (1.50\u2013757.32) 18.38 (1.50\u2013621.95) 20.95 (1.50\u2013757.32) 16.44 (1.50\u2013621.95) 25.24 (2.00\u2013757.32) 15.96 (1.50\u2013125.12) 23.89 (1.50\u2013757.32)\n\nTotal costs, US$ median (range)\n\nSex Male (n \u00bc 139) Female (n \u00bc 79)\nAge group, years ,42 (n \u00bc 124) !42 (n \u00bc 94)\nYears of schooling 7 (n \u00bc 135)\n.7 (n \u00bc 83) Diagnostic test\nXpert (n \u00bc 120) Smear microscopy (n \u00bc 98) City Rio de Janeiro (n \u00bc 86) Manaus (n \u00bc 132)\n\nVariables\n\nPatient costs for diagnosing TB using Xpert 551\n\nAcknowledgements\nThe study was supported by a partnership between the Brazilian National Tuberculosis Programme and the Fundac\u00b8a\u02dco Ataulpho de Paiva, Rio de Janeiro, RJ, Brazil, through a grant from the Bill and Melinda Gates Foundation, Seattle, WA, USA. None of these institutions is responsible for the publication decision or for the statements in this manuscript. AT received a grant from the Conselho Nacional de Desenvolvimento Cient\u0131\u00b4fico e Tecnolo\u00b4 gico, Brasilia, Brazil.\nConflict of interest: none declared.\nReferences\n1 World Health Organization. Road map for rolling out Xpert MTB/RIF for rapid diagnosis of TB and MDR-TB. Geneva, Switzerland: WHO, 2010. http://www.who.int/tb/laboratory/ roadmap_xpert_mtb-rif.pdf Accessed January 2014.\n2 Boehme C C, Nabeta P, Hillemann D, et al. Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med 2010; 363: 1005\u20131015.\n3 Blakemore R, Nabeta P, Davidow A L, et al. A multisite assessment of the quantitative capabilities of the Xpert MTB/ RIF assay. Am J Respir Crit Care Med 2011; 184: 1076\u20131084.\n4 Steingart K R, Sohn H, Schiller I, et al. Xpertw MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. In: The Cochrane Collaboration, Steingart K R, ed. Chichester, UK: John Wiley & Sons, 2013. http://onlinelibrary.wiley.com/ doi/10.1002/14651858.CD009593.pub2/full Accessed January 2014.\n5 Vassall A, van Kampen S, Sohn H, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: a cost-effectiveness analysis. PLOS Med 2011; 8: e1001120.\n6 Hanrahan C F, Selibas K, Deery C B, et al. Time to treatment and patient outcomes among TB suspects screened by a single point-of-care Xpert MTB/RIF at a primary care clinic in Johannesburg, South Africa. PLOS ONE 2013; 8: e65421.\n7 World Health Organization. Global tuberculosis control epidemiology, strategy, financing. WHO/HTM/TB/2009.411. Geneva, Switzerland: WHO, 2009. http://www.who.int/tb/ publications/global_report/2009/en/index.html Accessed January 2014.\n8 Ministe\u00b4rio da Sau\u00b4 de. Programa Nacional de Controle da Tuberculose. Bras\u0131\u00b4lia, DF, Brazil: Brasil Ministe\u00b4rio da Sau\u00b4 de, 2013. https://docs.google.com/file/d/0B0CE2wqdEaRVG1fa0JJMi1qa0U/edit\n9 Brazil Ministe\u00b4rio da Sau\u00b4 de, Secretaria de Vigila\u02c6 ncia Sanita\u00b4 ria. Programa Nacional de Controle da Tuberculose. Manual de Recomendac\u00b8o\u02dc es para o Controle da Tuberculose no Brasil. Bras\u00b4\u0131lia, DF, Brazil: Ministe\u00b4rio da Sau\u00b4 de, 2013. http://portal. saude.gov.br/portal/arquivos/pdf/manual_de_recomendacoes_ controle_tb_novo.pdf Accessed January 2014.\n10 Steffen R, Menzies D, Oxlade O, et al. Patients\u2019 costs and cost-effectiveness of tuberculosis treatment in DOTS and non-\n\nDOTS facilities in Rio de Janeiro, Brazil. PLOS ONE 2010; 5: e14014. 11 Government of Brazil. Decreto Lei 5452 de 1o de maio de 1943. Consolidac\u00b8a\u02dc o Das Leis Trab. Bras\u00b4\u0131lia, DF, Brazil: Government of Brazil, 2013. http://www.planalto.gov.br/ccivil_03/decretolei/del5452.htm Accessed January 2014. 12 Government of Brazil. Decreto n8 7872 de 26 de dezembro de 2013. Bras\u00b4\u0131lia, DF, Brazil: Government of Brazil, 2012. http:// www.planalto.gov.br/ccivil_03/_Ato2011-2014/2012/Decreto/ D7872.htm Accessed January 2014. 13 Ananthakrishnan R, Muniyandi M, Jeyaraj A, Palani G, Sathiyasekaran B W C. Expenditure pattern for TB treatment among patients registered in an urban Government DOTS program in Chennai City, South India. Tuberc Res Treat 2012; 2012: 747924. 14 Mauch V, Bonsu F, Gyapong M, et al. Free tuberculosis diagnosis and treatment are not enough: patient cost evidence from three continents. Int J Tuberc Lung Dis 2013; 17: 381\u2013 387. 15 Long Q, Smith H, Zhang T, Tang S, Garner P. Patient medical costs for tuberculosis treatment and impact on adherence in China: a systematic review. BMC Public Health 2011; 11: 393\u2013 401. 16 Mauch V, Woods N, Kirubi B, Kipruto H, Sitienei J, Klinkenberg E. Assessing access barriers to tuberculosis care with the tool to estimate patients\u2019 costs: pilot results from two districts in Kenya. BMC Public Health 2011; 11: 43\u201351. 17 Barter D M, Agboola S O, Murray M B, Ba\u00a8 rnighausen T. Tuberculosis and poverty: the contribution of patient costs in sub-Saharan Africa \u2013 a systematic review. BMC Public Health 2012; 12: 980\u20131000. 18 Ukwaja K N, Alobu I, lgwenyi C, Hopewell P C. The high cost of free tuberculosis services: patient and household costs associated with tuberculosis care in Ebonyi State, Nigeria. PLOS ONE 2013; 8: e73134. 19 Aspler A, Menzies D, Oxlade O, et al. Cost of tuberculosis diagnosis and treatment from the patient perspective in Lusaka, Zambia. Int J Tuberc Lung Dis 2008; 12: 928\u2013935. 20 John K R, Daley P, Kincler N, Oxlade O, Menzies D. Costs incurred by patients with pulmonary tuberculosis in rural India. Int J Tuberc Lung Dis 2009; 13: 1281\u20131287. 21 Ukwaja K N, Modebe O, Igwenyi C, Alobu I. The economic burden of tuberculosis care for patients and households in Africa: a systematic review. Int J Tuberc Lung Dis 2012; 16: 733\u2013739. 22 Umar N A, Abubakar I, Fordham R, Bachmann M. Direct costs of pulmonary tuberculosis among patients receiving treatment in Bauchi State, Nigeria. Int J Tuberc Lung Dis 2012; 16: 835\u2013 840. 23 Kik S V, Olthof S P J, de Vries J T N, et al Direct and indirect costs of tuberculosis among immigrant patients in the Netherlands. BMC Public Health 2009; 9: 283\u2013291. 24 Paim J, Travassos C, Almeida C, Bahia L, Macinko J. The Brazilian health system: history, advances and challenges. Lancet 2011; 377: 1778-1797.\n\nPatient costs for diagnosing TB using Xpert i\n\nC O N T E X T E : Manaos et Rio de Janeiro, deux capitales d\u2019e\u00b4tats du Bre\u00b4sil avec le cinquie`me et sixie`me taux d\u2019incidence le plus e\u00b4leve\u00b4 (autour de 90/100 000 habitants en 2012). O B J E C T I F S : Comparer les cou\u02c6 ts du Xpertw MTB/RIF avec la proce\u00b4dure standard (deux frottis) pour le diagnostic de la tuberculose (TB) du point de vue du patient. M E\u00b4 T H O D E S : Nous avons interroge\u00b4 218 patients ayant eu un diagnostic de TB dans les 4 mois pre\u00b4ce\u00b4dents soit par Xpert soit par examen microscopique de frottis. Nous avons e\u00b4galement recueilli des donne\u00b4es relatives aux cou\u02c6 ts directs non me\u00b4dicaux \u2014 de transport et de nourriture \u2014 aux cou\u02c6 ts indirects comme le temps passe\u00b4 en consultations ainsi que des donne\u00b4es sociode\u00b4 mographiques.\n\nRESUME\nR E\u00b4 S U LTAT S : Le revenu me\u00b4dian des patients e\u00b4tait de US$390,24. Les cou\u02c6 ts me\u00b4dians subis par les patients e\u00b4taient plus e\u00b4leve\u00b4s de 54% avec les frottis compare\u00b4s au Xpert (US$25,24 contre 16,44 ; P , 0,000) en raison de cou\u02c6 ts a` la fois directs et indirects plus e\u00b4leve\u00b4s. Les hommes subissaient des cou\u02c6 ts indirects plus e\u00b4leve\u00b4s (10,27$ contre 7,51 ; P \u00bc 0,038). Les patients de Manaos supportaient e\u00b4galement les cou\u02c6 ts plus importants. C O N C L U S I O N : Bien que le traitement de la TB au Bre\u00b4sil soit gratuit, les cou\u02c6 ts directs non me\u00b4dicaux et indirects impute\u00b4s aux patients constituent des obstacles importants a` l\u2019acce`s aux soins approprie\u00b4s. Compare\u00b4 au protocole standard, le test Xpert a re\u00b4duit le fardeau financier des patients. Ces re\u00b4sultats soutiennent la de\u00b4cision d\u2019e\u00b4tendre cette strate\u00b4gie dans le pays.\n\nM A R C O D E R E F E R E N C I A: Las ciudades de Manaos y R\u00b4\u0131o de Janeiro, dos capitales de estado en el Brasil, que ocupan el quinto y el sexto puesto de las tasas de incidencia de tuberculosis (TB) ma\u00b4 s altas (cerca de 90 100 000 habitantes en el 2012). O B J E T I V O S: Comparar los costos del uso de la prueba Xpertw MTB/RIF con el protocolo corriente (dos baciloscopias) en el diagno\u00b4 stico de la TB, desde el punto de vista de los pacientes. M E\u00b4 T O D O: Se entrevistaron 218 pacientes en quienes se hab\u0131\u00b4a establecido el diagno\u00b4 stico de TB en los u\u00b4 ltimos 4 meses, ya sea mediante la prueba Xpert o con el examen microsco\u00b4 pico del esputo. Se recogio\u00b4 informacio\u00b4 n sobre los costos directos diferentes de los costos me\u00b4dicos, como el transporte, la alimentacio\u00b4 n y los costos indirectos como el tiempo ocupado en las consultas diagno\u00b4 sticas, adema\u00b4 s de los datos sociodemogra\u00b4 ficos. R E S U LTA D O S: La mediana del ingreso de los pacientes\n\nRESUMEN\nfue US$390,24. La mediana del total de los costos sufragados por los pacientes fue 54% ma\u00b4 s alto con el procedimiento de las baciloscopias que con la prueba Xpert (US$25,24 contra US$16,44; P , 0,000), debido a los mayores costos directos e indirectos. Los costos indirectos fueron ma\u00b4 s altos para los hombres (US$10,27 contra US$7,51; P \u00bc 0,038). Los gastos fueron ma\u00b4 s altos para los pacientes de Manaos. C O N C L U S I O\u00b4 N: Si bien en el Brasil el diagno\u00b4 stico y el tratamiento de la TB se prestan sin costo alguno, los costos directos e indirectos diferentes de los costos me\u00b4dicos pueden representar un obsta\u00b4 culo al acceso a una atencio\u00b4 n adecuada. En comparacio\u00b4 n con el me\u00b4todo corriente, el uso de la prueba Xpert disminuyo\u00b4 la carga econo\u00b4 mica de los pacientes. Estos resultados respaldan la decisio\u00b4 n de ampliar la escala de aplicacio\u00b4 n de esta te\u00b4cnica en el pa\u0131\u00b4s.\n\n\n",
    "authors": [
      "R. Da Silva Antunes",
      "M. Pinto",
      "A. Trajman"
    ],
    "doi": "10.5588/ijtld.13.0637",
    "year": null,
    "item_type": "journalArticle",
    "url": "http://openurl.ingenta.com/content/xref?genre=article&issn=1027-3719&volume=18&issue=5&spage=547"
  },
  {
    "key": "CK9QZYMF",
    "title": "Establishing the cost of Xpert MTB/RIF mobile testing in high-burden peri-mining communities in South Africa",
    "abstract": "Background: Globally, tuberculosis remains a major cause of mortality, with an estimated 1.3 million deaths per annum. The Xpert MTB/RIF assay is used as the initial diagnostic test in the tuberculosis diagnostic algorithm. To extend the national tuberculosis testing programme in South Africa, mobile units fitted with the GeneXpert equipment were introduced to high-burden peri-mining communities.\nObjective: This study sought to assess the cost of mobile testing compared to traditional laboratory-based testing in a peri-mining community setting.\nMethods: Actual cost data for mobile and laboratory-based Xpert MTB/RIF testing from 2018 were analysed using a bottom-up ingredients-based approach to establish the annual equivalent cost and the cost per result. Historical cost data were obtained from supplier quotations and the local enterprise resource planning system. Costs were obtained in rand and reported in United States dollars (USD).\nResults: The mobile units performed 4866 tests with an overall cost per result of $49.16. Staffing accounted for 30.7% of this cost, while reagents and laboratory equipment accounted for 20.7% and 20.8%. The cost per result of traditional laboratory-based testing was $15.44 US dollars (USD). The cost for identifying a tuberculosis-positive result using mobile testing was $439.58 USD per case, compared to $164.95 USD with laboratory-based testing.\nConclusion: Mobile testing is substantially more expensive than traditional laboratory services but offers benefits for rapid tuberculosis case detection and same-day antiretroviral therapy initiation. Mobile tuberculosis testing should however be reserved for high-burden communities with limited access to laboratory testing where immediate intervention can benefit patient outcomes.",
    "full_text": "African Journal of Laboratory Medicine\nISSN: (Online) 2225-2010, (Print) 2225-2002\nPage 1 of 7\n\nOriginal Research\n\nEstablishing the cost of Xpert MTB/RIF mobile testing in high-burden peri-mining\ncommunities in South Africa\n\nAuthors: Naseem Cassim1,2 Lindi M. Coetzee1,2 Abel L. Makuraj2 Wendy S. Stevens1,2 Deborah K. Glencross1,2\nAffiliations: 1Department of Molecular Medicine and Haematology, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa\n2National Priority Programme, National Health Laboratory Service, Johannesburg, South Africa\nCorresponding author: Naseem Cassim, naseem.cassim@wits.ac.za\nDates: Received: 16 Sept. 2020 Accepted: 16 July 2021 Published: 30 Nov. 2021\nHow to cite this article: Cassim N, Coetzee LM, Makuraj AL, Stevens WS, Glencross DK. Establishing the cost of Xpert MTB/RIF mobile testing in high-burden peri-mining communities in South Africa. Afr J Lab Med. 2021;10(1), a1229 https://doi. org/10.4102/ajlm.v10i1.1229\nCopyright: \u00a9 2021. The Authors. Licensee: AOSIS. This work is licensed under the Creative Commons Attribution License.\nRead online: Scan this QR code with your smart phone or mobile device to read online.\n\nBackground: Globally, tuberculosis remains a major cause of mortality, with an estimated 1.3 million deaths per annum. The Xpert MTB/RIF assay is used as the initial diagnostic test in the tuberculosis diagnostic algorithm. To extend the national tuberculosis testing programme in South Africa, mobile units fitted with the GeneXpert equipment were introduced to high-burden peri-mining communities.\nObjective: This study sought to assess the cost of mobile testing compared to traditional laboratory-based testing in a peri-mining community setting.\nMethods: Actual cost data for mobile and laboratory-based Xpert MTB/RIF testing from 2018 were analysed using a bottom-up ingredients-based approach to establish the annual equivalent cost and the cost per result. Historical cost data were obtained from supplier quotations and the local enterprise resource planning system. Costs were obtained in rand and reported in United States dollars (USD).\nResults: The mobile units performed 4866 tests with an overall cost per result of $49.16. Staffing accounted for 30.7% of this cost, while reagents and laboratory equipment accounted for 20.7% and 20.8%. The cost per result of traditional laboratory-based testing was $15.44 US dollars (USD). The cost for identifying a tuberculosis-positive result using mobile testing was $439.58 USD per case, compared to $164.95 USD with laboratory-based testing.\nConclusion: Mobile testing is substantially more expensive than traditional laboratory services but offers benefits for rapid tuberculosis case detection and same-day antiretroviral therapy initiation. Mobile tuberculosis testing should however be reserved for high-burden communities with limited access to laboratory testing where immediate intervention can benefit patient outcomes.\nKeywords: GeneXpert test, tuberculosis screening, mobile testing, costing.\nIntroduction\nGlobally, tuberculosis is one of the top 10 causes of mortality.1 In 2017, tuberculosis infected about 10 million individuals and accounted for an estimated 1.3 million deaths among HIV-negative people, with an additional 300 000 deaths among people living with HIV.1 The epidemiology of tuberculosis varies widely between countries. In 2017, the tuberculosis incidence in most high-income countries was under 10 tuberculosis cases per 100 000 population compared to between 150 and 400 tuberculosis cases per 100 000 population in most of the top 30 high-burden countries.1,2 Countries such as South Africa (567), Mozambique (551) and the Philippines (554) reported over 500 cases per 100 000 population.1 As reported by the World Health Organization, there were 227 224 new cases of tuberculosis in South Africa in 2017. Although 322 000 cases of active tuberculosis were diagnosed in 2017 in South Africa, only 65% of the cases were bacteriologically confirmed, with a treatment coverage of 68% (95% confidence interval [CI]: 51\u201396).3\nClinically, a patient is suspected of having tuberculosis based on the following symptoms: persistent cough of 2 weeks or more, persistent cough of any duration for HIV-positive individuals, fever for over 2 weeks, night sweats, and unexplained weight loss (\u2265 1.5 kg within 1 month).4 Tuberculosis can present with different symptoms and atypical radiologic findings, and the pathological diagnosis has historically been based on acid-fast bacilli smear microscopy and bacteriological culture.5 The latter has a higher sensitivity for diagnosing and confirming active tuberculosis than acid-fast bacilli smear microscopy.5 The development of polymerase chain reaction tuberculosis assays has improved tuberculosis diagnosis and facilitates early treatment\n\nhttp://www.ajlmonline.org\n\nOpen Access\n\nPage 2 of 7\n\nOriginal Research\n\ninitiation by significantly reducing the time to result to 2 h, compared to 6 months for bacteriological culture.6 In South Africa, the Xpert MTB/RIF polymerase chain reaction assay (Cepheid, California, United States) is used routinely for tuberculosis diagnosis using patient sputum. Test results, which determine the therapeutic intervention and management in line with the diagnostic algorithm, are returned within two days.7\nTuberculosis incidence rates globally are especially high in the mining sector. In gold mines around the world, an estimated 3000 per 100 000 population are infected.8 In South Africa, the mining sector accounted for 7.5% of the national gross domestic product, employing 495 592 workers in 2014.9 Mining activities and environments are associated with a high risk of HIV and tuberculosis transmission and the migration of miners to their place of work is known to disrupt tuberculosis detection and care.10,11 Given the higher rates of tuberculosis transmission in mines, it is anticipated that the communities where miners live, the so-called peri-mining communities, would also have higher tuberculosis incidence rates.\nDue to the higher burden of disease among miners, a framework to address tuberculosis in the mining sector was developed for the Southern African Development Community in 2014.11 In March 2015, a comprehensive tuberculosis campaign targeted at inmates in correctional services prisons, mine workers and peri-mining communities was launched in South Africa under the banner \u2018Ending SA [South Africa] tuberculosis epidemic: Accelerating the response in key populations\u2019.12 In response to this call and through the support of the Global Fund, the National Health Laboratory Service and its clinical partner, the Aurum Institute, introduced a funded mobile GeneXpert testing facility to improve tuberculosis diagnosis in peri-mining communities.13 This initiative aimed to increase resources to deal with three of the world\u2019s most devastating diseases (HIV and AIDS, tuberculosis and malaria) by focusing on the areas of greatest need.13 Mobile testing was targeted at communities with a high burden of disease (high tuberculosis or HIV prevalence) and little or no access to laboratory testing facilities. These included remote areas of the North West and Limpopo provinces in South Africa between 2016 and 2019.13 The step-by-step approach to introducing mobile testing included identification of testing needs, execution of a feasibility study, procurement of funding, conducting of the necessary steps and processes to prepare for testing (setup of vehicles and equipment), assay verification, training, competency assessment, identification of measurable outcomes for monitoring, and commencement of testing.\nVarious studies have demonstrated that mobile testing is feasible, improves access to diagnostics, and may improve linkage to care and decrease time to treatment.14,15,16,17 A local study has reported that linkage to tuberculosis treatment was not associated with either sex or service type (mobile versus stand-alone), but older patients were less likely to be linked to tuberculosis treatment.15 Mobile testing\nhttp://www.ajlmonline.org\n\nfor HIV, tuberculosis and, more recently, severe acute respiratory syndrome coronavirus 2 can bring diagnostics to where it is needed in high-burden or outbreak communities.18 As previously reported in a local study to evaluate mobile versus traditional laboratory CD4 testing, mobile diagnostics could be substantially more expensive.19 Mobile testing is not widely used in South Africa, with its use limited to pilot projects or funded studies. However, it should be possible to integrate mobile testing as part of a national tiered laboratory network to extend services20 and absorb the higher cost of mobile testing into the national laboratory expenditure allocations.\nThere is limited local data on the cost to provide mobile Xpert MTB/RIF testing in high-burden communities. Only one local study reported that the cost to detect one tuberculosis case was $1117.00 United States dollars (USD) based on 1385 patients enrolled.16 The paucity of local data for mobile tuberculosis testing highlights the need for a comprehensive costing study, which could inform the modalities of providing these services and identify scenarios that are best suited for on-site testing.\nThe objective of this study was to determine the cost per result and cost per positive result of mobile Xpert MTB/RIF testing and to compare it to the cost of traditional laboratory-based testing.\nMethods\nEthical considerations\nEthics clearance was obtained from the University of the Witwatersrand (reference number: M160978). Our study did not contain any patient identifiers. No patient consent was required.\nContext\nThe National Health Laboratory Service implemented mobile testing in three high-tuberculosis-burden districts in South Africa (Kenneth Kaunda, North West, Waterberg, Limpopo, and Sekhukhune, Limpopo). Traditional laboratory-\u200b based testing was conducted at the Potchefstroom laboratory, a clinical pathology district laboratory offering a basic repertoire of testing, including tuberculosis testing, in the Kenneth Kaunda district.\nCosting methodology\nThe costing analysis was undertaken using Microsoft Excel (Redmond, Washington, United States).21 A bottom-up costing approach was used to determine the cost per result from a provider perspective; all costs are reported for the National Health Laboratory Service as the provider of mobile tuberculosis testing. All costs (excluding value-added tax) were obtained in South African rand and reported in United States dollars, with an exchange rate of R14.4838 South African rand (ZAR) to the dollar.22 The main outcome of interest was the cost per result. The ingredients-based costing approach established annual equivalent costs (AEC) for the\nOpen Access\n\nPage 3 of 7\n\nOriginal Research\n\nfollowing categories of mobile testing: staff (medical technologist and driver), reagents, external quality assurance, vehicle purchase, vehicle operations, laboratory equipment, and coordinator costs to manage testing. For the costing of the traditional laboratory-based Xpert MTB/RIF testing, we reported the following cost categories: staff (medical technologist), reagents, external quality assurance, laboratory equipment, courier logistics, and coordinator costs to manage testing. All laboratory equipment was purchased outright. For traditional laboratory testing, a placement agreement includes the costs for regular maintenance and servicing of the analyser. All data are reported for the 2018 calendar year. The Consolidated Health Economic Evaluation Reporting Standards checklist was used in the preparation of the manuscript.23 For laboratory equipment costing, useful life, which refers to the projected lifespan of depreciable equipment, was set at seven years, with a discount rate of 4%.\nFor the calculation of staff costs, we determined the full-time equivalent hours (the number of hours worked by an employee divided by the number of hours worked by a full-time employee) based on the amount of time employees were assigned to mobile testing and multiplied this by the annual cost to company salary scales to determine the AEC. Reagent and test consumable costs were obtained from quotations received from the Oracle enterprise resource planning system used by the National Health Laboratory Service, and the AEC was determined using annual test volumes.24 For external quality assurance, the frequency of panel testing and the number of samples prepared were used to calculate the AEC per site, that is, panels were sent out quarterly, with three samples per instrument. The AECs for vehicle purchase, vehicle operations, laboratory equipment and the coordinator costs were also determined and are described in more detail below. Start-up costs were defined as all AECs associated with the purchase of the mobile vehicle and laboratory equipment. The total cost per result minus the contribution of start-up costs was also determined. We reported the cost per positive result (the cost to find one tuberculosis-positive case) for both mobile and laboratory tuberculosis testing. This was calculated as the AEC divided by the number of tuberculosis-positive results. For mobile testing, it was also possible to use the clinical outcomes data to estimate the diagnostic cost per tuberculosis-positive patient, as well as the cost per patient initiated on treatment (calculated as AEC divided by the number of people that received treatment).\nMobile Xpert MTB/RIF costing\nThe costs for the initial start-up of the mobile service were determined and included the costs for the purchase of the vehicles, modifications made to the mobile units (benches, air conditioning), and purchase and placement of equipment on the mobile units. The mobile units were equipped with GeneXpert platform instruments (Cepheid, Sunnyvale, California, United States). This is an automated real-time polymerase chain reaction test for the simultaneous detection of tuberculosis and rifampicin resistance.25 Four GeneXpert\nhttp://www.ajlmonline.org\n\ninstruments, as well as one computer per analyser, were placed in each mobile unit for a combined daily testing capacity of 64 samples. Operational vehicle costs were included in the cost per result and comprised maintenance, fuel, repairs, and annual licensing costs. Additional operational costs included costs for procurement of reagents, consumables, specimen collection and quality control materials (internal and external schemes), as well as other miscellaneous costs such as for printing of results. Each mobile testing unit required a driver and a medical technologist. The percentage of time spent offering mobile testing was used for full-time equivalent calculations, ranging from 40% to 80%. The cost to company salary for a coordinator was calculated using historical expenditure data. The AECs for travel, office setup, miscellaneous costs and coordinator costs were also determined (total AEC divided by the number of mobile testing sites).\nThe test volumes and number of positive results for each mobile unit were reported using bar charts, with the total cost per result presented as a line chart on the secondary y-axis. The cost per result without start-up costs and the cost per kilometre were also reported. The number of site visits and kilometres travelled were indicated as text on the charts. For the three mobile units, we reported the correlation between the cost per result and distance travelled.\nLaboratory-based Xpert MTB/RIF comparative costing\nAs a comparator, the cost per result was determined for traditional laboratory-based Xpert MTB/RIF testing. Initial laboratory setup included the installation of the four GeneXpert systems (Cepheid, Sunnyvale, California, United States) (capacity of 64 samples per day), an air conditioner, a level two biosafety hood and a vortex mixer. Operational costs included costs to procure reagents, consumables, specimen collection materials, quality control materials (internal and external), printer cartridges and paper. The assumptions for these operational costs were similar to those for mobile testing.\nThe staff complement required to perform mobile testing included a medical technologist and a laboratory manager, who provided minimal supervision. The technologists performed other testing in addition to Xpert MTB/RIF. The costs of the business management unit (coordinator costs) in the North West province were determined and included the following personnel: business manager, secretary, quality assurance coordinator, human resources officers, training staff, and other support staff. To determine the coordinator costs per result, the AEC was divided by the annual test volume for the province. For the courier costs, the annual expenditure for the laboratory was used.\nResults\nThe three mobile units performed 4866 tuberculosis tests, of which the majority were performed by mobile\nOpen Access\n\nPage 4 of 7\n\nOriginal Research\n\nunit 1 (68.7%). The mobile units covered a total distance of 64 605 km, with mobile units 3 and 1 contributing 73.7% of all travel. A total of 258 healthcare facilities were visited, evenly distributed between the three units. There were 544 tuberculosis-positive samples reported, with an overall tuberculosis positivity of 11.2%. The tuberculosis positivity was 9.6% for mobile unit 1, 16.6% for mobile unit 2, and 10.7% for mobile unit 3. For the period reported, 11 603 tests were done at the Potchefstroom laboratory, of which 1086 were positive (9.4%).\n\nEffect of distance travelled on the cost per result\nThe three mobile units covered distances of 21 766 km, 16 985 km and 25 854 km. The estimated overall cost per kilometre was $2.34 USD, with mobile unit 2 accounting for the highest cost per kilometre ($8.91 USD). The number of health clinics visited by the mobile units ranged from 79 to 90 clinics. The correlation between the cost per result and distance travelled was not statistically significant (p = 0.053), with a perfect negative correlation reported (\u22121.0000).\n\nMobile testing costs\nThe overall cost per result for mobile testing was $49.16 USD with an AEC of $239 130.00 USD (Table 1). Without the start-up costs, the overall cost per result decreased to $31.11 USD. A breakdown of cost contributors showed that staff accounted for 30.7%, primarily due to the cost per result ($11.69 USD; 23.8%) of the medical technologist performing the test. Reagents accounted for 20.7% ($10.16 USD), while vehicle operation costs made up 3.6% ($1.76 USD) of the overall cost per result. Specimen collection and external quality control only contributed 0.5% ($0.27 USD) to the final cost per result. The AEC for reagents, staffing and laboratory equipment made up 72.2% of the total cost. The start-up costs, which comprised the costs to purchase the mobile vehicle and laboratory equipment, accounted for 36.7% ($87 804.00 USD) of the total cost of mobile testing. These initial costs need to be considered when mobile units are rolled out without links to an established laboratory network or testing programme. The cost per result for the three mobile units ranged from $30.22 USD to $154.31 USD. Without the start-up costs, the cost per result ranged from $21.47 USD to $95.06 USD (Figure 1).\n\nTest volumes Cost per result (USD)\n\n4000 3500 3000 2500 2000 1500 1000\n500 -\n\n89 site visits and 21 766 km\n79 site visits and 16 985 km\n\nMobile 1\n\nMobile 2\n\nTests performed\nNumber of posi\u019fve TB results\nCost per result (USD) Cost per result - start-up costs (USD)\nCost per km\n\n3344 320 \u03a930.22\n\u03a921.47 \u03a94.64\n\n1028 171 \u03a960.17\n\u03a931.70 \u03a93.64\n\n90 site visits and 25 854 km\nMobile 3\n\n\u03a9180.00 \u03a9160.00 \u03a9140.00 \u03a9120.00 \u03a9100.00 \u03a980.00 \u03a960.00 \u03a940.00 \u03a920.00 \u03a90.00\n\n494\n\n53\n\n\u03a9154.31\n\n\u03a995.06 \u03a92.95\n\nUSD, United States dollars; TB, tuberculosis; km, kilometre.\nFIGURE 1: Number of tuberculosis tests performed (dark blue bars) by mobile Xpert MTB/RIF testing units in high-burden peri-mining communities in South Africa, 2018. Positive results (red bars) are reported on the primary y-axis. On the secondary y-axis, the green line indicates the total cost per result in USD, the purple line indicates the total cost per result less start-up costs, and the orange line indicates the cost per kilometre travelled. The number of site visits for testing and the total distance travelled for those visits are indicated as text for each mobile unit.\n\nTABLE 1: Comparison of cost per result between mobile Xpert MTB/RIF testing in high-burden peri-mining communities and traditional laboratory-based Xpert MTB/RIF testing offered at a laboratory in the Kenneth Kaunda district in South Africa, 2018.\n\nCost category\n\nMobile testing\n\nTraditional testing\n\nCost per result\n\nn\n\n(USD)\n\n%\n\nAEC (USD) Cost per result\n\nn\n\n(USD)\n\n%\n\nAEC (USD)\n\nReagents\n\n10.16\n\n-\n\n20.7\n\n49 461.20\n\n10.16\n\n-\n\n65.8\n\n117 940.47\n\nStaffing: Medical technologist\n\n11.69\n\n-\n\n23.8\n\n56 866.12\n\n1.62\n\n-\n\n10.5\n\n18 801.30\n\nStaffing: Driver\n\n3.41\n\n-\n\n6.9\n\n16 600.98\n\n0.00\n\n-\n\n0.0\n\n0.00\n\nSpecimen collection materials\n\n0.17\n\n-\n\n0.3\n\n825.01\n\n0.34\n\n-\n\n2.2\n\n3955.03\n\nTest consumables\n\n0.27\n\n-\n\n0.5\n\n1297.16\n\n1.61\n\n-\n\n10.4\n\n18 637.87\n\nExternal quality assurance\n\n0.10\n\n-\n\n0.2\n\n472.53\n\n0.02\n\n-\n\n0.1\n\n201.40\n\nVehicle purchase\u2020\n\n7.81\n\n-\n\n15.9\n\n37 992.92\n\n0.00\n\n-\n\n0.0\n\n0.00\n\nVehicle operation costs\n\n1.76\n\n-\n\n3.6\n\n8540.21\n\n0.00\n\n-\n\n0.0\n\n0.00\n\nLaboratory equipment\u2020\n\n10.24\n\n-\n\n20.8\n\n49 811.96\n\n1.37\n\n-\n\n8.9\n\n15 873.64\n\nCourier costs\n\n0.00\n\n-\n\n0.0\n\n0.00\n\n0.26\n\n-\n\n1.7\n\n2993.39\n\nCoordinator costs\n\n3.55\n\n-\n\n7.2\n\n17 262.69\n\n0.06\n\n-\n\n0.4\n\n728.98\n\nTotal cost per result\n\n49.16\n\n-\n\n100.0\n\n239 130.78\n\n15.44\n\n100.0\n\n179 132.08\n\nLess start-up costs\n\n31.11\n\n-\n\n-\n\n-\n\n-\n\n-\n\n-\n\n-\n\nNumber of tests performed\n\n-\n\n4866\n\n-\n\n-\n\n-\n\n11 603\n\n-\n\n-\n\nTB+ results\n\n-\n\n544\n\n-\n\n-\n\n-\n\n1086\n\n9.4\n\n-\n\nCost per result for TB+ results\n\n439.58\n\n-\n\n11.2\n\n-\n\n-\n\n164.95\n\n-\n\n-\n\nOn TB treatment\n\n-\n\n300\n\n55.1\n\n-\n\nNo data\n\n-\n\n-\n\n-\n\nCost per result for TB+ patient\n\n797.10\n\n-\n\n-\n\n-\n\n-\n\n-\n\n-\n\n-\n\non treatment\n\nUSD, United States dollars; TB+, Xpert MTB/RIF positive; TB, tuberculosis; AEC, annual equivalent cost. \u2020, Start-up costs.\n\nhttp://www.ajlmonline.org\n\nOpen Access\n\nPage 5 of 7\n\nOriginal Research\n\nCost per positive tuberculosis result\nThe AEC for offering mobile testing was $239 130.78 USD to produce 4866 results. There were 544 positive results (11.2%), with 300 patients documented as having received tuberculosis treatment (55.1%). The cost to find one positive tuberculosis case using mobile testing was $439.58 USD and the cost of initiating a positive patient on treatment was $797.10 USD (Table 1).\nComparative costing analysis\nThe overall cost per result for laboratory-based Xpert MTB/RIF testing was $15.44 USD (Table 1). Equipment for laboratory testing is procured through a national tender process, that is, there are no costs for installation and maintenance of adequate testing platforms. Reagent costs were similar to that of mobile testing and accounted for 65.8% of the total cost per result. Staff costs contributed $1.62 USD (10.5%) to the cost per result. For specimen collection materials, the cost per result was $0.34 USD (2.2%); for test consumables, the cost was $1.61 USD (10.4%); for external quality assurance, the cost was $0.02 USD (0.1%); for laboratory equipment, the cost was $1.37 USD (8.9%); for the coordinator, the cost was $0.06 USD (0.4%). The courier costs contributed $0.26 USD (1.7%) to the total cost per result.\nThe AEC for laboratory-based testing was $179 132.08 USD to produce 11 603 results. The cost to find one positive tuberculosis case was $164.95 USD. Unfortunately, the number of patients with a laboratory test result who received tuberculosis treatment was not available.\nDiscussion\nMobile diagnostics for high burden diseases such as tuberculosis can provide significant public health and epidemiological value in regions where individuals do not have easy access to laboratory facilities. Overall, the average cost per result for all three mobile units was $49.16 USD. However, the cost per result ranged from $30.22 USD to $154.31 USD, highlighting differences in how and where mobile testing was offered. The biggest contributors to cost differences were test volumes and distance travelled. For example, mobile unit 1 performed the most testing with short travel distances and reported the lowest cost per result. In contrast, mobile unit 3 served a very remote area with longer travel times and had the highest cost per result.\nStaff, reagents, laboratory equipment and vehicle purchase contributed 88.1% of the total cost per result. This indicates that the majority of costs associated with mobile testing are not flexible, and suggests that the cost of mobile testing could only be reduced by increasing test volumes, reducing input costs or widening the test repertoire. Test volumes could be increased by identifying clinical settings with higher test volumes that would maximise the use of mobile testing. Test volumes are however limited by the daily throughput of the testing platform and space on the mobile units for multiple units of the test platforms. Negotiations with suppliers\nhttp://www.ajlmonline.org\n\ncould result in lower reagent and consumable pricing. In addition, by adding mobile testing to the existing traditional laboratory national tenders, the placement agreement for reagents and analysers could be extended to mobile testing. The higher test volumes would lower the unit costs of the traditional laboratory supply chain management agreements and, by extension, benefit mobile testing. Various point-of-care platforms with a very small footprint could be used to offer additional routine haematology and chemical pathology mobile testing.26 These could be used to facilitate the fast-tracking of antiretroviral therapy for patients with tuberculosis and HIV.27\nA wide range of tuberculosis positivity rates were reported for the three mobile units in this study. This highlights the importance of identifying high-burden settings with high tuberculosis prevalence for effective deployment of mobile testing. The reported cost to find a single tuberculosis-positive case would vary substantially based on the setting where testing is offered. Offering mobile testing in high-burden areas with a large population would substantially reduce the overall diagnostic cost and simultaneously offer immediate access to treatment. The higher cost of mobile testing should be weighed against the impact of earlier diagnosis, improved coverage, same-day treatment and care, as well as reduced loss to follow-up.17,28,29,30,31 Mobile testing as an extension of laboratory testing could also see its higher costs offset by high volume laboratory testing, as bulk testing is still reserved for the laboratory service. The findings of this study confirmed that mobile testing is 3.2 times more expensive than conventional laboratory testing on the same GeneXpert testing platform. Some of the reasons for the higher cost per result for mobile testing include lower test volumes, lost time due to travel to the health facility, and the impact of the clinical workflow on sample collection. An earlier study conducted to determine the cost of providing mobile CD4 testing in Pixley ka Seme in the Northern Cape of South Africa also reported a substantially higher cost for mobile testing versus laboratory testing.19 In such remote areas, the cost of mobile testing should be weighed against improving sample collection and distribution routes to the nearest testing laboratory.\nFor mobile tuberculosis testing, scenarios should be identified that match the increased costs of mobile testing with improved patient outcomes such as rapid tuberculosis case identification and same-day antiretroviral therapy initiation. A clinical outcome study should be embedded within any future mobile testing to assess the impact on patient outcomes. Similarly, detailed cost-effectiveness studies are needed to provide evidence of how mobile tuberculosis testing can save lives and fully realise the potential of targeting high-risk groups.\nLimitations\nThis study used actual costs from the 2018 calendar year that would be more accurate than a desktop exercise. However, some staffing estimates are based on the typical number of days of mobile testing and this could have underestimated\nOpen Access\n\nPage 6 of 7\n\nOriginal Research\n\nthe costs. More so, the costs reported are based on the clinical referral of patients for testing. In a different clinical scenario with higher patient volumes, the costs could be very different. There are several assumptions made for this costing analysis that could have affected the reported cost per result. The number of Xpert platforms in each mobile unit, the level and type of staff employed (technologist versus technician), fulltime equivalent assumptions, and the exclusion of some costs, such as overheads, would affect the reported cost per result.\nConclusion\nThis study reported that mobile tuberculosis testing is more expensive than traditional laboratory testing. However, mobile testing holds the potential to offer rapid tuberculosis case detection and improve coverage and diagnostics in communities with a high burden of disease. Furthermore, mobiles could be dovetailed to be used to deliver same-day antiretroviral therapy initiation. Further cost-effectiveness studies are needed using the patient outcome data reported.\nAcknowledgements\nThe authors thank the staff that operated the mobile units. We also wish to thank the Global Fund for making this project possible and the Aurum Institute (clinical partner).\nCompeting interests\nThe authors declare that they have no financial or personal relationships that may have inappropriately influenced them in writing this article.\nAuthors\u2019 contributions\nN.C. and L.M.C. designed the study, developed the methodology and conducted the research. N.C. conducted the costing analysis. A.L.M. provided the data required for the costing analysis. D.K.G. and W.S.S. provided editorial comments and technical input. D.K.G. supervised the study by providing leadership and oversight as the project leader. All authors reviewed the results and contributed to the manuscript development.\nSources of support\nNo funding was obtained for this study. The Global Fund to Fight AIDS, Tuberculosis and Malaria covered the cost of mobile testing in the peri-mining communities (ZAF-C-NDOH [National Department of Health]).\nData availability\nThe authors do not have permission to share the data used for this study.\nDisclaimer\nThe authors declare that the views expressed in the submitted article are our own and not the official position of any institution or funder.\nhttp://www.ajlmonline.org\n\nReferences\n1. World Health Organization (WHO). Global tuberculosis report [homepage on the Internet]. Geneva: World Health Organization; 2018 [cited 2019 Jul 29]. Available from: https://www.who.int/tb/publications/global_report/en/\n2. World Health Organization (WHO). Use of high burden country lists for TB by WHO in the post-2015 era: Summary [homepage on the Internet]. Geneva: World Health Organization; 2015 [cited 2019 Jul 29]. Available from: https:// www.who.int/tb/publications/global_report/high_tb_burdencountrylists\u200b20162020.pdf\n3. World Health Organization (WHO). South Africa: Tuberculosis profile [homepage on the Internet]. Geneva: World Health Organization; 2017 [cited 2019 Jul 29]. Available from: https://extranet.who.int/sree/Reports?op=Replet&name=/ WHO_HQ_Reports/G2/PROD/EXT/TBCountryProfile&ISO2=ZA&outtype=PDF\n4. National Department of Health (NDOH). National tuberculosis management guidelines [homepage on the Internet]. Pretoria: National Department of Health; 2014 [cited 2019 Aug 01]. Available from: http://www.tbonline.info/media/ uploads/documents/ntcp_adult_tb-guidelines-27.5.2014.pdf\n5. Ryu YJ. Diagnosis of pulmonary tuberculosis: Recent advances and diagnostic algorithms. Tuberc Respir Dis (Seoul). 2015;78(2):64\u201371. https://doi.org.10.4046/ trd.2015.78.2.64\n6. Singer-Leshinsky S. Pulmonary tuberculosis: Improving diagnosis and management. JAAPA. 2016;29(2):20\u201325. https://doi.org.10.1097/01.JAA.0000\u200b 476207.96819.a7\n7. National Department of Health (NDOH). National consolidated Guidelines for the prevention of mother-to-child transmission of HIV (PMTCT) and the management of HIV in children, adolescents and adults [homepage on the Internet]. Pretoria, 2015 [cited 2019 Jun 10]; p. 128. Available from: http://www.health.gov.za/index. php/2014-03-17-09-09-38/policies-and-guidelines/category/2302015p?download=937:national-art-guidelines-2015final\n8. Chang ST, Chihota VN, Fielding KL, et al. Small contribution of gold mines to the ongoing tuberculosis epidemic in South Africa: A modeling-based study. BMC Med. 2018;16(1):52. https://doi.org.10.1186/s12916-018-1037-3\n9. U.S. Department of the Interior, U.S. Geological Survey. 2014 minerals yearbook [homepage on the Internet]. U.S. Department of the Interior U.S. Geological Survey; 2017 [cited 2019 Aug 01]. Available from: https://s3-us-west-2. amazonaws.com/prd-wret/assets/palladium/production/mineral-pubs/ country/2014/myb3-2014-sf.pdf\n10. Stuckler D, Basu S, McKee M, Lurie M. Mining and risk of tuberculosis in subSaharan Africa. Am J Public Health. 2011;101(3):524\u2013530. https://doi.org.10.\u200b 2105/AJPH.2009.175646\n11. Churchyard GJ, Mametja LD, Mvusi L, et al. Tuberculosis control in South Africa: Successes, challenges and recommendations. S Afr Med J. 2014;104(3):244\u2013248. https://doi.org.10.7196/SAMJ.7689\n12. South African Government. Ending South Africa\u2019s TB epidemic: Accelerating our response in key populations [homepage on the Internet]. Pretoria: South African Government; 2015 [cited 2019 Aug 01]. Available from: https://www.gov.za/ world-tb-day-2015\n13. National Department of Health (NDOH). Global fund grants operations manual for National Department of Health [homepage on the Internet]. Pretoria: National Department of Health; 2016 [cited 2019 Aug 01]. Available from: https:// www.theglobalfund.org/media/6586/oig_gf-oig-17-014_report_en.pdf\n14. Sloot R, Glenshaw MT, Van Niekerk M, Meehan S-A. Rapid point-of-care CD4 testing at mobile units and linkage to HIV care: An evaluation of community-based mobile HIV testing services in South Africa. BMC Public Health. 2020;20(1):528. https://doi.org.10.1186/s12889-020-08643-3\n15. Meehan S-A, Sloot R, Draper HR, Naidoo P, Burger R, Beyers N. Factors associated with linkage to HIV care and TB treatment at community-based HIV testing services in Cape Town, South Africa. PLoS One. 2018;13(4):e0195208. https://doi. org.10.1371/journal.pone.0195208\n16. Kranzer K, Lawn SD, Meyer-Rath G, et al. Feasibility, yield, and cost of active tuberculosis case finding linked to a mobile HIV service in Cape Town, South Africa: A cross-sectional study. PLoS Med. 2012;9(8):e1001281. https://doi. org.10.1371/journal.pmed.1001281\n17. Bassett IV, Forman LS, Govere S, et al. Test and Treat TB: A pilot trial of GeneXpert MTB/RIF screening on a mobile HIV testing unit in South Africa. BMC Infect Dis. 2019;19(1):110. https://doi.org.10.1186/s12879-019-3738-4\n18. Abdool Karim SS. The South African response to the pandemic. N Engl J Med. 2020;382(24):e95. https://doi.org.10.1056/NEJMc2014960\n19. Coetzee LM, Cassim N, Glencross DK. A cost analyses of mobile laboratory CD4 testing in a National Health Insurance (NHI) pilot site [homepage on the Internet]. Cape Town: African Society for Laboratory Medicine; 2012 [cited 2020 Dec 12]. Available from: https://www.researchgate.net/publication/267651724_A_Cost_ A n a l y s e s _ o f _ M o b i l e _ L a b o r a t o r y _ C D 4 _ Te s t i n g _ I n _ a _ N a t i o n a l _ H e a l t h _ Insurance_NHI_Pilot_Site\n20. Glencross DK, Coetzee LM, Cassim N. An integrated tiered service delivery model (ITSDM) based on local CD4 testing demands can improve turn-around times and save costs whilst ensuring accessible and scalable CD4 services across a national programme. PLoS One. 2014;9(12):e114727. https://doi.org.10.1371/journal. pone.0114727\n21. Microsoft Corporation. Office 365. Redmond, WA: Microsoft Corporation, 2019.\n22. Standard Bank of South Africa (SBSA). Forex closing indication rates for 24 April 2019 as at 16:08 [homepage on the Internet]. Johannesburg: Standard Bank of South Africa; 2019 [cited 2019 Apr 24]. Available from: https://ws15.standardbank. co.za/finSnapShot/GetforexServlet\nOpen Access\n\nPage 7 of 7\n\nOriginal Research\n\n23. Husereau D, Drummond M, Petrou S, et al. Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement. Int J Technol Assess Health Care. 2013;29(2):117\u2013122. https://doi.org.10.1017/S0266462313000160\n24. National Health Laboratory Service (NHLS). Annual report 2017/18 [homepage on the Internet]. Johannesburg: National Health Laboratory Service; 2018 [cited 2019 Oct 08]. Available from: https://nationalgovernment.co.za/entity_ annual/1714/2018-national-health-laboratory-service-(nhls)-annual-report.pdf\n25. Schnippel K, Meyer-Rath G, Long L, et al. Scaling up Xpert MTB/RIF technology: The costs of laboratory- vs. clinic-based roll-out in South Africa. Trop Med Int Health. 2012;17(9):1142\u20131151. https://doi.org.10.1111/j.1365-3156.2012.03028.x\n26. Martin CL. i-STAT \u2013 Combining Chemistry and Haematology in PoCT. Clin Biochem Rev. 2010;31(3):81\u201384.\n27. Cassim N, Coetzee LM, Stevens WS, Glencross DK. Addressing antiretroviral therapy-related diagnostic coverage gaps across South Africa using a programmatic approach. Afr J Lab Med. 2018;7(1):681. https://doi.org.10.4102/ajlm.v7i1.681\n\n28. Govindasamy D, Van Schaik N, Kranzer K, Wood R, Mathews C, Bekker LG. Linkage to HIV care from a mobile testing unit in South Africa by different CD4 count strata. J Acquir Immune Defic Syndr. 2011;58(3):344\u2013352. https://doi.org.10.1097/ QAI.0b013e31822e0c4c\n29. Grolla A, Jones S, Kobinger G, et al. Flexibility of mobile laboratory unit in support of patient management during the 2007 Ebola-Zaire outbreak in the Democratic Republic of Congo. Zoonoses Public Health. 2012;59 Suppl 2:151\u2013157. https://doi.org.10.1111/j.1863-2378.2012.01477.x\n30. Larson BA, Schnippel K, Ndibongo B, et al. Rapid point-of-care CD4 testing at mobile HIV testing sites to increase linkage to care: An evaluation of a pilot program in South Africa. J Acquir Immune Defic Syndr. 2012;61(2):e13\u2013e17. https://doi.org.10.1097/QAI.0b013e31825eec60\n31. Van Schaik N, Kranzer K, Wood R, Bekker LG. Earlier HIV diagnosis \u2013 Are mobile services the answer? S Afr Med J. 2010;100(10):671\u2013674. https://doi.org.10.7196/ samj.4162\n\nhttp://www.ajlmonline.org\n\nOpen Access\n\n\n",
    "authors": [
      "Naseem Cassim",
      "Lindi M. Coetzee",
      "Abel L. Makuraj",
      "Wendy S. Stevens",
      "Deborah K. Glencross"
    ],
    "doi": "10.4102/ajlm.v10i1.1229",
    "year": null,
    "item_type": "journalArticle",
    "url": "http://www.ajlmonline.org/index.php/AJLM/article/view/1229"
  },
  {
    "key": "PUHKXAQ7",
    "title": "Clinical impact and cost analysis of the use of either the Xpert MTB Rif test or sputum smear microscopy in the diagnosis of pulmonary tuberculosis in Rio de Janeiro, Brazil",
    "abstract": "Introduction: The molecular test Xpert MTB/RIF (Xpert) has been recommended for use in the diagnosis of pulmonary tuberculosis (PTB); however, data on the cost of incorporating it under routine conditions in high-burden countries are scarce. The clinical impact and costs incurred in adopting the Xpert test in routine PTB diagnosis was evaluated in a prospective study conducted from November 2012 to November of 2013, in the City of Rio de Janeiro, Brazil. Methods: The diagnostic and therapeutic cascade for TB treatment was evaluated using Xpert in the first stage (S1), and sputum smear microscopy (SSM) in the second stage (S2). The mean costs associated with each diagnostic test were calculated including equipment, human resources, supplies, and infrastructure. Results: We included 232 subjects with probable TB (S1 = 87; S2 = 145). The sensitivities of Xpert and SSM were 91.7% (22/24) and 79.1% (34/43), respectively. The median time between triage and TB treatment initiation in S1 (n = 24) was 14.5 days (IQR 8-28.0) and in S2 (n = 43) it was 8 days [interquartile range (IQR) 6-12.0]. The estimated mean costs per examination in S1 and S2 were US$24.61 and US$6.98, respectively. Conclusions: Compared with SSM, Xpert test showed a greater sensitivity, but it also had a time delay with respect to treatment initiation and a higher mean cost per examination.",
    "full_text": "Rev Soc Bras Med Trop 51(5):631-637, Sep-Oct, 2018 doi: 10.1590/0037-8682-0082-2018\nMajor Article\nClinical impact and cost analysis of the use of either the Xpert MTB Rif test or sputum smear microscopy in the diagnosis of pulmonary tuberculosis in Rio de Janeiro, Brazil\nAn\u00e1lia Zuleika de Castro[1], Adriana Rezende Moreira[1],[2], Jaqueline Oliveira[2], Paulo Albuquerque Costa[2],[3], Carolyne Lalucha Alves Lima Da Gra\u00e7a[1],\nMauricio de Andrade P\u00e9rez[4],[5], Afr\u00e2nio Kritski[1],[2] and Maria Claudia Vater[4]\n[1]. Programa P\u00f3s-Gradua\u00e7\u00e3o em Cl\u00ednica M\u00e9dica, Faculdade de Medicina, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brasil. [2]. Programa Acad\u00eamico de Tuberculose, Faculdade de Medicina e do Complexo Hospitalar, Instituto de Doen\u00e7as do T\u00f3rax, Hospital Universit\u00e1rio Clementino Fraga Filho, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brasil. [3]. Policl\u00ednica Augusto Amaral Peixoto, Departamento de Sa\u00fade Municipal, Rio de Janeiro, RJ, Brasil. [4]. Instituto de Estudos em Sa\u00fade Coletiva, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brasil. [5]. Curso de Medicina, Faculdades Souza Marques, Rio de Janeiro, RJ, Brasil.\nAbstract Introduction: The molecular test Xpert MTB/RIF (Xpert) has been recommended for use in the diagnosis of pulmonary tuberculosis (PTB); however, data on the cost of incorporating it under routine conditions in high-burden countries are scarce. The clinical impact and costs incurred in adopting the Xpert test in routine PTB diagnosis was evaluated in a prospective study conducted from November 2012 to November of 2013, in the City of Rio de Janeiro, Brazil. Methods: The diagnostic and therapeutic cascade for TB treatment was evaluated using Xpert in the first stage (S1), and sputum smear microscopy (SSM) in the second stage (S2). The mean costs associated with each diagnostic test were calculated including equipment, human resources, supplies, and infrastructure. Results: We included 232 subjects with probable TB (S1 = 87; S2 = 145). The sensitivities of Xpert and SSM were 91.7% (22/24) and 79.1% (34/43), respectively. The median time between triage and TB treatment initiation in S1 (n = 24) was 14.5 days (IQR 8-28.0) and in S2 (n = 43) it was 8 days [interquartile range (IQR) 6-12.0]. The estimated mean costs per examination in S1 and S2 were US$24.61 and US$6.98, respectively. Conclusions: Compared with SSM, Xpert test showed a greater sensitivity, but it also had a time delay with respect to treatment initiation and a higher mean cost per examination.\nKeywords: Diagnostic methods. Pulmonary tuberculosis. Cost analysis. Xpert MTB/RIFtest. Sputum smear microscopy.\n\nINTRODUCTION\nTuberculosis (TB) has become one of the top 10 causes of death worldwide, being the major cause of mortality among infectious diseases in 20151,2, with about 10.4 million people becoming infected with TB, and 1.8 million dying from this disease [including 0.4 million patients with human immunodeficiency virus (HIV)]. Since 2008, the World Health Organization (WHO) has prioritized the early diagnosis of TB through new molecular technologies, which allow for the adoption of appropriate measures, for both treatment and infection control in the community3,4.\nCorresponding author: Dr. Afr\u00e2nio Kritski. e-mail: kritskia@gmail.com Received 3 March 2018 Accepted 30 July 2018\n\nSince 2010,the WHO has recommended the Xpert MTB/RIF test (Xpert) as a substitute for the sputum smear microscopy (SSM) test for TB diagnosis in adults and children1.In clinical trials that analyzed the use of Xpert in TB diagnosis in comparison with SSM, the Xpert test increased the number of bacteriologically-confirmed TB patients, reduced the time between triage and treatment initiation, but did not reduce either mortality or the number of patients lost to follow-up5.In addition, a recent survey in South Africa, using primary data obtained under routine conditions, found that the use of Xpert was not cost-effective6. These results indicate the need for studies that can evaluate the clinical impact and costs of incorporating new diagnostic technologies for TB, under routine conditions, in high-burden countries7-9.\nIn Brazil, owing to the low access to TB diagnosis in some regions10,11, higher rates of mortality per 100,000 inhabitants due to TB occur in the States of Rio de Janeiro, Pernambuco, and Amazonas, being, 5.0, 4.5, and 3.2 respectively1.\n\n631\n\nde Castro AZ et al. - Xpert MTB/RIF for TB diagnosis in Rio de Janeiro\n\nIn a clinical trial performed in the Cities of Rio de Janeiro and Manaus, using secondary data obtained from the Integrated Laboratory System (ILS) and Disease Notification system (DNS), compared to SSM, the use of Xpert resulted in a) a higher proportion of bacteriologically-confirmed TB, a shorter time to treatment initiation12, and a lower cost13, but there was no impact on TB treatment outcomes14. Under this scenario, the Xpert test was introduced into the Brazilian Health System in 201315.\nIn the present study, using primary data obtained from a health unit in Rio de Janeiro City, we compared the care process indicators related to the diagnostic and treatment cascades for patients with probable TB, and the mean costs incurred using either Xpert or SSM.\nMETHODS\nPatients: A prospective, descriptive, analytical study was conducted from November 2012 to November 2013, at the Policl\u00ednica Augusto Amaral Peixoto (PAAP), a health unit located in Guadalupe neighborhood at Programmatic Area AP 3.3, in the city of Rio de Janeiro. Following the recommendations of the Ministry of Health, eligible patients were defined as having signs and/or symptoms suggestive of pulmonary TB, were older than 12 years of age, were either male or female, had initiated or not TB treatment at PAAP, and were residents of Rio de Janeiro City16. Subjects with a history of coughing, hemoptysis, and/ or abnormalities on a chest X-ray compatible with pulmonary TB, who signed the informed consent form, and those that had their respiratory samples analyzed by Xpert or SSM were included. Subjects whose clinical samples were not submitted for mycobacterial culture, whose diagnosis was extra-pulmonary TB, or who had already started an anti-TB treatment were excluded.\nConfirmed bacteriological TB was defined as a positive culture result, and Empirical TB treatment was defined when anti-TB treatment was started by a physician based on the patient's symptoms and clinical history, before having access to the laboratory test result16.\nZiehl-Neelsen staining and culture for mycobacteria were used, following standard laboratory recommendations and the Xpert test was performed according to the manufacturer\u2019s recommendations17,18.Cultures for mycobacteria were carried out in Lowenstein-Jensen or BACTECTMMGITTM960 systems at the Mycobacteriology Laboratory of the University Hospital Clementino Fraga Filho [Hospital Univerist\u00e1rio Clementino Fraga Filho (HUCFF)], Federal University of Rio de Janeiro [Universidade Federal do Rio de Janeiro (UFRJ)], and the Evandro Chagas Clinical Research Institute [Instituto de Pesquisa Clinica Evandro Chagas, (IPEC)], Oswaldo Cruz Foundation [Funda\u00e7\u00e3o Oswaldo Cruz (FIOCRUZ)].The clinical samples were collected in two stages, stage 1 (Nov 2012-April 2013) and stage 2 (May-November 2013). In stage 1, Xpert was performed at the Souza Marques Family Clinic located in another neighborhood (Madureira), and in stage 2, owing to a lack of Xpert supplies, SSM was reintroduced for TB diagnosis at PAAP.\nAnti-HIV testing was proposed for patients who started antiTB treatment, following the guidance of the Ministry of Health Standards Manual18. Lost to follow-up cases were called when the patient did not attend the health unit after two pre-scheduled,\n\nunfilled consultations. In cases of death, the medical records were evaluated to confirm its relationship to TB.\nCost analysis\nFollowing the recommendation of the Ministry of Health for Health Technologies Evaluation19, the costs associated with Xpert and SSM were evaluated. The items evaluated were classified into four different categories: equipment, human resources, inputs, and infrastructure (electricity, safety, water, cleaning). The appropriation method was used to estimate the mean cost which was calculated by adding up all the monthly expenses of the Laboratory associated with the exam, divided by the number of exams carried out in one month. This had the advantages that it simplified calculations and the information was easy to obtain.\nIt should be noted that the calculation of the item infrastructure cost involved the use of the so-called apportionment method20. For this, the values of the unit's total assets were known, and the proportion used by each item in the laboratory was calculated. The item human resources, was calculated as follows: (salary of the professionals X amount of human resource per category X the average dedication devoted to the accomplishment of the procedure by category) / the total examinations. With regard to estimated inputs costs we used a series of parameters to compute the expenditure published elsewhere21, which inform the model, the average expenditure for each one to perform the exam. The expenditure of the input by exam was multiplied by its market price (at the time), divided by the total of exams made. In the case of equipment, the amount of equipment was multiplied by the price of the equipment used, discounting a depreciation rate of 5%, obtained from its useful life divided by the total of examinations carried out in the month. These values were converted for US$, according to Brazil's central bank exchange rate (November, 2013).22 The time (in days) from triage to TB treatment initiation was used as a measure of the effectiveness of the diagnostic procedure. A predicted cost effectiveness analysis that incorporates the loss, in terms of costs, and the gain, in terms of effectiveness, was performed using Xpert compared with the traditional technology, sputum smear microscopy.\nStatistical analysis\nWe compared the socio-demographic data and the clinical characteristics between the patients in stage 1 and stage 2. For categorical data, percentages, chi square tests, and odds ratios were used. For variables, measures of central tendency, such as median and quartiles, were used. For analysis of the time between triage and process indicators linked to the diagnostic and therapeutic cascade, analyses were performed to compare continuous variables with normal and non-normal distributions, such as the t-test and Mann Whitney test. To verify if there was a difference between proportions, the 95% confidence interval was used. The analyses were performed using the statistical package SPSS, version 20.0.\nEthical considerations\nThe Research Ethics Committee of Faculdade de Medicina, Hospital Universit\u00e1rio Clementino Fraga Filho, Universidade Federal do Rio de Janeiro approved the study under number 019/07.17.\n\n632\n\nRev Soc Bras Med Trop 51(5):631-637, Sep-Oct, 2018\n\nRESULTS\nDuring the study period, 411 subjects with probable TB were deemed eligible (146 in stage 1 and 265 in stage 2). Among these, 232 (56.4%) were included in the study; 87 in stage 1 and 145 in stage 2. TB treatment was initiated in 67 (28.8%) patients; 24 (27.6%) in stage 1 and 43 (29.6%) in stage 2. The Xpert test used in stage 1 was positive in 22/24 (91.7%) patients, and the SSM used in stage 2 was positive in 34/43 (79.1%) patients (Figure 1).\nFrom the data shown in Table 1, we observed that in both stages, there was a predominance of male patients (70.2%), a black or brown skin color (77.1%), economic class C and D (84.5%), lower levels of schooling (87.5%), and smokers or ex-smokers (65.5%). In addition, there was a high rate of contact with pulmonary TB (34.9%) and a past history of TB (22.8%). The empirical treatment ratio was higher in stage 1 (7.7%) than in stage 2 (2.4%). No significant differences were observed when comparing stage 1 and 2 with respect to sociodemographic and clinical variables.\nIn stage 1, the time from the triage and laboratory results, receipt of the laboratory results by health professionals, and initiation of TB treatment were, 4, 13, and 14.5 days, respectively. In stage 2, the time from the triage and laboratory results, receipt of the laboratory results by health professionals, and the initiation of TB treatment were, 3, 7, and 8 days, respectively. Comparing stage 1 to stage 2, the times (in days) from triage and receipt of laboratory results by health professionals and treatment initiation were significantly longer with Xpert (p <0.001), but there was no difference between\n\ntriage and sputum collection and laboratory results release (p =0.06) (Table 2).\nTable 3 shows that there was no difference between the TB treatment outcomes among the TB patients diagnosed in stage 1 or stage 2. Among the 67 TB patients, 20 (29.8%) were transferred to primary health units. Among the 47 patients who remained at PAAP, 8 patients were lost to follow-up (11.9%).\nThe exams performed in stage 1 and 2 among the 232 patients were included in the cost analysis. In stage 1 and 2, 92 Xpert and 378 SSM exams were performed as a diagnostic routine, respectively. In stage 2, 213 SSMs were performed in the first exam and 152 in the second exam. On average, 31.3 tests per month were performed by Xpert and 70.2 tests by SSM. The estimated mean costs of Xpert and SSM were US$24.61 and US$ 6.98 per exam, respectively (Table 4). The highest cost categories associated with Xpert were supplies and equipment, whereas for SSM the highest cost category was human resources (Table 4).\nA cost-effectiveness analysis was not performed, since the patients who attended in stage 1 (Xpert) took 14.5 days on average to start TB treatment, whereas the patients in stage 2 (SSM) took 8 days. In addition, the SSM has a lower mean cost relative to Xpert. That way, in this sample, a dominated analysis was characterized.\nDISCUSSION\nIn our study, under routine conditions, patients who were analyzed using Xpert (stage 1) experienced an average of 14.5 days from the time triage to the initiation of TB treatment.\n\nStage 1 November 2012 - April 2013\nTest - Xpert MTB/RIF Presumed PTB n = 87\nCulture positive n = 24 TB treatment\nn = 24\n\nStage 2 May - November 2013\nTest - SSM Presumed PTB\nn = 145\nCulture positive n = 43\nTB Treatment n = 43\n\nXpert pos = 22 Xpert neg = 2\n\nSSM pos = 34 SSM neg = 9\n\nFIGURE 1: TFBIGdUiaRgEn1o:sTisBadniadgntroesaitsmaenndttrceaastmcaednet caamscoandge pamreosnugmperdespuumlemdopnualmryontuabryertucbuelorcsuislosciassceassdesurdiunrgingthtehesstutuddyy\nperiod using Xpert MTB RIF and sputum smear microscopy.Xpert MTB RIF: Xpert Mycobacterium Rifampicin;\nperiod using XPpTeBrt:pMulmToBnaRryIFtubaenrdcuslopsuistu; mSSsMm: esaprutummicsrmosecaorpmyi.cXropsecrotpMy; TB: RtuIbFe:rcXulposeirst. Mycobacterium Rifampicin\n;PTB:pulmonary tuberculosis; SSM: sputum smear microscopy; TB: tuberculosis.\n\n633\n\nde Castro AZ et al. - Xpert MTB/RIF for TB diagnosis in Rio de Janeiro\n\nTABLE 1: Socio-demographic and clinical characteristics of patients with probable pulmonary tuberculosis treated at the health unit from November 2012 to November 2013.\n\nCharacteristics\n\nTotal (n =232) n (%)\n\nStage 1 \u2013 Xpert (n= 87) n (%)\n\nStage 2 \u2013 SSM (n = 145) n (%)\n\nP value\n\nSex\n\nmale\n\n163 (70.2)\n\n59 (67.8)\n\n104 (71.7)\n\n0.62\n\nfemale\n\n69 (29.7)\n\n28 (32.2)\n\n41 (28.3)\n\nAge in years (SD)\n\n38,7 (\u00b1 15.2)\n\n38.2 (\u00b115.1)\n\n39.6 (\u00b115.4)\n\n0.93\n\nEthnicity\n\nwhite\n\n47,0 (20.2)\n\n18 (20.7)\n\n29 (20.0)\n\n0.86\n\nblack\n\n43 (18.5)\n\n17 (19.5)\n\n26 (17.9)\n\nmixed race (white/black)\n\n136 (58.6)\n\n48 (55.2)\n\n88 (60.7)\n\nindigenous\n\n6 (2.6)\n\n4 (4.7)\n\n2 (1.4)\n\nEducation (years)\n\n\u02c3 8\n\n29 (12.5)\n\n11 (12.6)\n\n18 (12.4)\n\n0.87\n\n\u2264 8\n\n203 (87.5)\n\n76 (87.4)\n\n127 (87.6)\n\nSocioeconomic class\n\n\u2265B\n\n36 (15.5)\n\n16 (18,4)\n\n20 (13,8)\n\n0.45\n\n\u2264 C\n\n196 (84.5)\n\n71 (81.6)\n\n125 (86.2)\n\nSmoking\n\nyes, currently\n\n100 (43.1)\n\n37 (42.5)\n\n63(43.5)\n\n0.89\n\nex- smoker\n\n52 (22.4)\n\n19 (21.8)\n\n33 (22.8)\n\nno\n\n80 (34.5)\n\n31 (35.6)\n\n49 (33.8)\n\nAlcoholism*\n\nyes\n\n31 (25.2)\n\n18 (35.3)\n\n13 (18.1)\n\n0.24\n\nno\n\n92 (74.8)\n\n33 (64.7)\n\n59 (81.9)\n\nHIV testing\n\npositive\n\n5 (4.9)\n\n0 (0,0)\n\n5 (8.5)\n\n0.32\n\nnegative\n\n96 (95.1)\n\n37 (100.0)\n\n59 (91.5)\n\nHistory of hospital admission\n\nyes\n\n19 (8.2)\n\n5 (5.8)\n\n14 (9.7)\n\n0.62\n\nno\n\n213 (91.8)\n\n82 (94.3)\n\n131 (90.3)\n\nIncarcerated\n\nyes\n\n7 (3.0)\n\n5 (5.8)\n\n2 (1.4)\n\n0.42\n\nno\n\n225 (97.0)\n\n82 (94.3)\n\n143 (98.6)\n\nContact with TB\n\nyes\n\n81 (36.2)\n\n32 (39.5)\n\n49 (34.3)\n\n0.68\n\nno\n\n143 (63.8)\n\n49 (60.5)\n\n94 (65.7)\n\nPrevious TB treatment\n\nyes\n\n53 (22.8)\n\n21 (24.1)\n\n32 (22.1)\n\n0.88\n\nno\n\n179 (77.2)\n\n66 (75.9)\n\n113 (77.9)\n\nAnti-TB treatment initiation\n\nyes\n\n67 (28.9)\n\n24 (27.6)\n\n43 (29.6)\n\n0.92\n\nno\n\n165 (71.1)\n\n63 (72.4)\n\n102 (70.4)\n\nEmpirical treatment**\n\nyes\n\n3 (4.5)\n\n2 (7.7)\n\n1 (2.4)\n\n0.84\n\nno\n\n64 (95.5)\n\n22 (92.3)\n\n42 (97.6)\n\nSSM: sputum smear microscopy.* Empirical treatment was initiated before the laboratory test results; TB: tuberculosis. SD: standar deviation. ** Alcoholism analyzed by the test CAGE (Cutting down, Annoyance by criticism, Guilty feeling, and Eye-openers).\n\n634\n\nRev Soc Bras Med Trop 51(5):631-637, Sep-Oct, 2018\n\nTABLE 2: Time (in days) between triage and other process indicators for TB diagnosis and treatment cascade among 232 patients with probable pulmonary TB, from November 2012 to November 2013.\n\nProcess indicators Sputum collection\n\ntotal(n=87) median (IQR)\n0 (0\u20131)\n\nStage 1 (Xpert) TB active(n = 24)\nmedian (IQR) 0\n(0\u20131)\n\nno TB(n =63) median (IQR)\n0 (0\u20131)\n\ntotal(n =145) median (IQR)\n0 (0\u20130)\n\nLaboratory test result released\n\n4,0 (2\u20137)\n\n6 (2\u20138)\n\n4 (2\u20137)\n\n3 (2\u20135)\n\nLaboratory results received by HCWs\n\n13 (8\u201322)\n\n10.5 (7\u201315)\n\n14 (8\u201324)\n\n7 (6\u20138)\n\nTB treatmentinitiation\n\n14,5\n\n(8\u00ad\u201328)\n\nTB: tuberculosis;SSM: sputum smear microscopy ;IQR: interquartile range. HCWs: health care workers.\n\nStage 2 (SSM) TB active(n = 41)\nmedian (IQR) 0\n(0\u20130) 3\n(2\u20136) 7\n(6\u20138) 8\n(6\u201312)\n\nno TB (n =104) median (IQR)\n0 (0\u20130)\n3 (2\u20135) 7 (6-9)\n\nTABLE 3: Treatment outcomes in patients diagnosed by Xpert in stage 1 and by SSM in stage 2, from November 2012 to November 2013.\n\nEffectiveness anti-TB treatment\n\nStage 1 (n=24)\n\nFavorable (healing/complete treatment)\n\n12 (75.0%)\n\nUnfavorable\n\n4 (25.0%)\n\nLost to follow-up\n\n4\n\nDeath due to TB\n\n0\n\nMDR-TB\n\n0\n\nOthers\n\n8\n\nDeath not related to TB\n\n1\n\nTransferred\n\n7\n\nSSM: sputum smear microscopy; TB: tuberculosis; MDR: multi drug resistance.\n\nStage 2 (n=43) 23 (79.3%) 6 (20.7%) 4 1 1 14 3 11\n\nTABLE 4: - Average cost of sputum smear microscopy and Xpert (values in US$1=2.32 Reals, 2012) according to the different categories.\n\nCost item Equipment Human resources Supplies Infrastructure Estimated average cost SSM: sputum smear microscopy\n\nSSM 0.08 4.61 2.24 0.05 6.98\n\nXpert 6.81 2.06 15.34 0.39 24.61\n\nThis was also associated with a higher mean cost (US$24.61). In contrast, patients who were analyzed using SSM (stage 2), experienced an average of 8 days from time of triage to the initiation of TB treatment, and this was associated with a lower mean cost (US$6.98). These results are different from those described in the literature5,7,12. This discrepancy is probably related to the fact that the primary data here were collected\n\nroutinely and not during a clinical trial. The health unit where the data were collected participated in a previous pragmatic trial conducted in Rio de Janeiro and Manaus. Using secondary data collected in the Integrated Laboratory System and Disease Notification system, they found that the time elapsed between triage and the beginning of TB treatment was lower for patients assessed using Xpert (8 days) compared to the SSM (11 days)12. In this study, the Xpert and SSM tests were performed at Policl\u00ednica Augusto Amaral Peixoto, but not mycobacterial culture. When we conducted the current study, only patients with culture results were included in the analysis. Delays were observed for patients assessed using Xpert results, since the equipment was moved to another health unit, whereas the SSM analysis was retained at Policl\u00ednica Augusto Amaral Peixoto . These results reinforce the argument that, in any analysis of the incorporation of new diagnostic technologies, it is fundamental to take into account all of the care process indicators related to the diagnostic and treatment cascade, in order to analyze the patients and clinical sample flow-through of locally collected primary data, as has been reported by other authors5,7-9.\n\n635\n\nde Castro AZ et al. - Xpert MTB/RIF for TB diagnosis in Rio de Janeiro\n\nIn our study, since we used only the mean cost, and not the activity based cost (ABC), we adopted the apportionment method to cope with any distortions, when we consider the proportion of the laboratory expenses with the exam considered. We observed the cost with Xpert (US$24.61) to be higher than the cost reported by Pinto et al.(US$17.35)13 in our country, and by Shah et al.23 ($14.93), but lower than the cost (US$60-61) reported in South Africa24. However, since it was not possible to perform a costeffectiveness analysis, we cannot compare our data with the unfavorable data reported by Vassal et al.l6 and by Pinto et al25.\nIn stages 1 and 2, the sensitivities of Xpert and SSM were 91.7% (22/24) and 79.1% (34/43).\n(34/43), respectively, similar to those described in a recent meta-analysis, which analyzed studies comparing the two technologies, where the sensitivity of Xpert (88%) was higher than that of SSM (70.0%)26.\nThe proportion of TB (29.7%) identified in our study among the probable pulmonary TB cases is similar to that which has been described previously in the same health unit27. The proportion of empirical treatments was higher with Xpert (7.7%), but was lower than that described in other series5,12.\nThe adoption of the Xpert molecular test, with a higher diagnostic yield, did not provide more favorable TB treatment outcomes, as has been described in other studies5,7,14. The proportion of patients lost to follow-up (11.9%) was high, but similar to that described in the City of Rio de Janeiro16.\nIn the present study, the majority of the patients were young male adults, of low socioeconomic status with low education levels, similar to the patient population described previously1,16,28,29.In our pragmatic, prospective study, primary data were collected either through interviewing subjects with probable pulmonary tuberculosis, as well as through secondary laboratory data and the TB treatment outcomes, available from the laboratory electronic system and the Disease Notification System, respectively.\nThe limitations of this study include the following: a) a small number of subjects with probable pulmonary tuberculosis attended only one health unit in the City of Rio de Janeiro; b) no data was collected on patient costs related to TB diagnosis; these costs should be evaluated in the future, because of significant and growing socioeconomic inequality, the free diagnosis and treatment of TB is not sufficient to alleviate the financial limitations to which most patients are exposed. For many patients, expenditures related to TB care (transportation, food, etc.) can be catastrophic, so this should be evaluated in future studies; and c) we did not perform a qualitative evaluation to analyze the barriers or facilitators factors to incorporating the use of the Xpert molecular test compared to smear microscopy.\nIn addition, our results call attention to the importance of rapid tests such as Xpert, when routinely incorporated. A shorter time is required to release the data for the detection of the Mycobacterium tuberculosis complex. The test also allows for the detection of rifampicin resistance (a molecular marker of MDR-TB). The SSM test only detects the presence or absence of mycobacteria, and cannot differentiate non-tuberculous\n\nMycobacteria from the Mycobacterium tuberculosis complex. However, failures in the operation of health services, such as those observed in our study, may negatively affect the potential utilization of new technologies such as Xpert. With a new promising technology that has been recommended by WHO, and before its incorporation into the health system, the results of our study highlight the need to evaluate the clinical impacts and the costs of any new technology in association with care processes adopted by the local health system, under field conditions7-10.\nIn conclusion, in high-burden countries, it is necessary to identify, through primary data, at both the regional and local level, the most effective and cost-effective diagnostic strategies that can expedite the TB diagnosis and the initiation of appropriate TB treatment, thereby lowering the transmission of TB in the community and promote the sustainability of the procedures adopted.\nConflict of interest\nThe authors declare that there is no conflict of interest.\nFinancial support\nThis work was funded by Conselho Nacional Pesquisa (CNPq) /Institutos Nacionais de Ci\u00eancia e Tecnologia (INCT) \u2013 Process: 465318/2014-2.\nREFERENCES\n1. World Health Organization (WHO). Global tuberculosis report 2016. Geneva: WHO; 2016.\n2. World Health Organization (WHO). Integrating collaborative TB and HIV services within a comprehensive package of care for people who inject drugs: consolidated guidelines. WHO/HTM/TB/2016.02. http://apps.who.int/iris/bitstream/10665/204484/1/9789241510226_ eng.pdf?ua=1\n3. World Health Organization (WHO). Use of liquid TB culture and drug susceptibility testing (DST) in low and medium income settings. Summary Report of the expert group meeting on the use of liquid culture media. Geneva: World Health Organization, 2007.\n4. World Health Organization (WHO). Moving research findings into new WHO policies reference: World Health Organization. Moving research findings into new WHO policies. http://www.who.int/tb/ dots/laboratory/policy/en/index4.html. 2008.\n5. Boyles TH. Why do clinical trials of Xpert\u00ae MTB/RIF fail to show an effect on patient relevant outcomes? Int J Tuberc Lung Dis. 2017;21(3):249-50.\n6. Vassall A, Siapka M, Foster N, Cunnama L, Ramma L, Fielding K, et al. Cost-effectiveness of Xpert MTB/RIF for tuberculosis diagnosis in South Africa: a real-world cost analysis and economic evaluation. Lancet Glob Health. 2017;5(7):e710-e719.\n7. Churchyard GJ, Stevens WS, Mametja LD, McCarthy KM, Chihota V, Nicol MP, et al. Xpert MTB/RIF versus sputum microscopy as the initial diagnostic test for tuberculosis: a cluster randomised trial embedded in South African roll-out of Xpert MTB/RIF. Lancet Glob Health. 2015;3(8):450-7.\n8. Langley, I, Squire SB, Dacombe R, Madan J, Lapa e Silva, JR, Barreira D, et al. Developments in impact assessment of new diagnostic algorithms for tuberculosis control. Clin Infect Dis. 2015;61(Suppl 3):S126-34.\n\n636\n\nRev Soc Bras Med Trop 51(5):631-637, Sep-Oct, 2018\n\n9. Albert H, Nathavitharana RR, Isaacs C, Pai M, Denkinger CM, Boehme CC. Development, roll-out and impact of Xpert MTB/RIF for tuberculosis: what lessons have we learnt and how can we do better? Eur Respir J. 2016;48(2):516-25.\n10. Scatena LM, Villa TCS, Ruffino Netto A, Kritski AL, Figueiredo RMRM, Vendramini SHF, et al. Dificuldades de acesso a servi\u00e7os de sa\u00fade para diagn\u00f3stico de tuberculose em munic\u00edpios do Brasil. Rev Sa\u00fade P\u00fablica. 2009;43(3):389-97.\n11. Minist\u00e9rio da Sa\u00fade (MS). Secretaria de Vigil\u00e2ncia em Sa\u00fade. Indicadores priorit\u00e1rios para o monitoramento do Plano Nacional pelo Fim da Tuberculose como Problema de Sa\u00fade P\u00fablica no Brasil. Boletim Epidemiol\u00f3gico.2017;48(8):1-11.http://portalarquivos. saude.gov.br/images/pdf/2017/marco/23/2017-V- 48-N-8Indicadoresprioritrios-para-o-monitoramento-do-Plano-Nacionalpelo-Fim-da-Tuberculose-como- Problema-de-Saude-Publica-noBrasil.pdf.\n12. Durovni B, Saraceni V, van denHof S, Trajman A, Cordeiro-Santos M, Cavalcante S, et al. Impact of replacing smear microscopy with Xpert MTB/RIF for diagnosing tuberculosis in Brazil: a steppedwedge cluster-randomized trial. PLoS Med. 2014; 9:11(12):e1001766.\n13. Pinto M. Entringer AP, Steffen R, Trajman A. Cost analysis of nucleic acid amplification for diagnosing pulmonary tuberculosis, within the context of the Brazilian Unified Health Care System. J Bras Pneumol. 2015;41(6):536-8.\n14. Trajman A, Durovni B, Saraceni V, Menzies A, Cobelens F, van den Hof S. Impact on patient\u00b4s treatment outcomes of Xpert MTB/ RIF implementation for the diagnosis of tuberculosis: followup of a stepped-wedge randomized clinical trial. Plos one. 2015; 27:10(4):e0123252.\n15. Minist\u00e9rio da Sa\u00fade (MS). Biblioteca Virtual em Sa\u00fade. Portaria MS n\u00ba 48 de 10 de setembro de 2013. Bras\u00edlia:MS; 2013. Citado:3 dez 2014. Dispon\u00edvel em:http://www.bvsms.saude.gov.br/bvs/ saudelegis/sctie/2013/prt0048_10_09_2013.html.\n16. Minist\u00e9rio da Sa\u00fade (MS). Secretaria de Vigil\u00e2ncia em Sa\u00fade. Departamento de Vigil\u00e2ncia Epidemiol\u00f3gica. Manual de recomenda\u00e7\u00f5es para o controle da tuberculose no Brasil. Bras\u00edlia:MS;2011.288p.\n17. Kent PT, Kubica GP. Public Health Mycobacteriology: aGuide for the Level III Laboratory. National Technical Reports Library. U.S. Department of Commerce. Washington, DC: 1985. 226p.\n18. Cepheid Brasil. A better way. The New GeneXpert\u00ae System. New Systems. Same game-changing performance. http://www.cepheid.\n\ncom/administrator/components/com_ productcatalog/libraryfiles/9a.\n19. Minist\u00e9rio da Sa\u00fade (MS). Secretaria de Vigil\u00e2ncia em Sa\u00fade. Diretrizes Metodol\u00f3gicas: estudos de avalia\u00e7\u00e3o econ\u00f4mica de tecnologias em sa\u00fade Minist\u00e9rio da Sa\u00fade, Secretaria de Ci\u00eancia, Tecnologia e Insumos Estrat\u00e9gicos, Departamento de Ci\u00eancia e Tecnologia. Bras\u00edlia: MS;2009. 145p.\n20. Alinski ML, Young HP. A new method for congressional apportionment. Proc Natl Acad Sci USA.1974;71(11):4602-6.\n21. Minist\u00e9rio da Sa\u00fade (MS). Secretaria de Vigil\u00e2ncia em Sa\u00fade. Departamento de Vigil\u00e2ncia Epidemiol\u00f3gica. Manual Nacional de Vigil\u00e2ncia Laboratorial da Tuberculose e outras Micobact\u00e9rias. Bras\u00edlia:MS; 2008. 436p. Dispon\u00edvel em: http://portal.saude.gov.br/ portal/arquivo/pdf/manual_laboratorio_tb_8_05_13.pdf.\n22. Minist\u00e9rio da Fazenda (MF). Banco Central do Brasil. Cota\u00e7\u00f5es e boletins. Bras\u00edlia:MF; 2018. Acesso: 26 fev 2018.http://www4.bcb. gov.br/pec/taxas /port /ptaxnpesq.asp.\n23. Shah M, Chihota V, Coetzee G, Churchyard G, Dorman SE. Comparison of laboratory costs of rapid molecular tests and conventional diagnostics for detection of tuberculosis and drug resistant tuberculosis in South Africa. BMC Infect Dis. 2013;13:352.\n24. Meyer-Rath G, Schnippel K, Long L, W, Sanne I, Stevens W, Pillay S, et al. The impact and cost of scaling up Gene Xpert MTB/RIF in South Africa. PLoS One. 2012;7(5):e36966.\n25. Pinto M, Steffen RE, Cobelens F, van den Hof S, EntringerA, Trajman A. Cost effectiveness of the Xpert\u00ae MTB/RIF assay for tuberculosis diagnosis in Brazil. Int J TubercLung Dis.2016;20(5):611-8.\n26. Steingart KR, Schiller I, Horne DJ, Pai M, Boehme CC, Dendukuri N. Xpert\u00ae MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev. 2014;21;(1):CD009593.\n27. Castro CB, Costa PA, Ruffino-Netto A, Maciel EL, Kritski AL. Assessment of a clinical score for screening suspected pulmonary tuberculosis cases. Rev Sa\u00fade P\u00fablica. 2011;45(6):1110-6.\n28. L\u00f6nnroth K, Castro KG, Chakaya JM, Chauhan LS, Floyd K, Glaziou P, et al. Tuberculosis control and elimination 2010-50: cure, care, and social development. Lancet. 2010;375(9728):1814-29.\n29. Yamamura M, Santos Neto M, Freitas IM, Rodrigues LBB, Popolin MP, Uchoa SAC, et al. Tuberculose e iniquidade social em sa\u00fade: uma an\u00e1lise ecol\u00f3gica utilizando t\u00e9cnicas estat\u00edsticas multivariadas, S\u00e3o Paulo, Brasil. Rev Panam Salud Publica. 2014;35(4):2707.\n\n637\n\n\n",
    "authors": [
      "An\u00e1lia Zuleika De Castro",
      "Adriana Rezende Moreira",
      "Jaqueline Oliveira",
      "Paulo Albuquerque Costa",
      "Carolyne Lalucha Alves Lima Da Gra\u00e7a",
      "Mauricio De Andrade P\u00e9rez",
      "Afr\u00e2nio Kritski",
      "Maria Claudia Vater"
    ],
    "doi": "10.1590/0037-8682-0082-2018",
    "year": null,
    "item_type": "journalArticle",
    "url": "http://www.scielo.br/scielo.php?script=sci_arttext&pid=S0037-86822018000500631&tlng=en"
  },
  {
    "key": "69TXZA77",
    "title": "Testing Pooled Sputum with Xpert MTB/RIF for Diagnosis of Pulmonary Tuberculosis To Increase Affordability in Low-Income Countries",
    "abstract": "ABSTRACT\n            \n              Tuberculosis (TB) is a global public health problem, with the highest burden occurring in low-income countries. In these countries, the use of more sensitive diagnostics, such as Xpert MTB/RIF (Xpert), is still limited by costs. A cost-saving strategy to diagnose other diseases is to pool samples from various individuals and test them with single tests. The samples in positive pool samples are then retested individually to identify the patients with the disease. We assessed a pooled testing strategy to optimize the affordability of Xpert for the diagnosis of TB. Adults with presumptive TB attending hospitals or identified by canvassing of households in Abuja, Nigeria, were asked to provide sputum for individual and pooled (4 per pool) testing. The agreement of the results of testing of individual and pooled samples and costs were assessed. A total of 738 individuals submitted samples, with 115 (16%) being\n              Mycobacterium tuberculosis\n              positive. Valid Xpert results for individual and pooled samples were available for 718 specimens. Of these, testing of pooled samples detected 109 (96%) of 114 individual\n              M. tuberculosis\n              -positive samples, with the overall agreement being 99%. Xpert semiquantitative\n              M. tuberculosis\n              levels had a positive correlation with the smear grades, and the individual sample-positive/pooled sample-negative results were likely due to the\n              M. tuberculosis\n              concentration being below the detection limit. The strategy reduced cartridge costs by 31%. Savings were higher with samples from individuals recruited in the community, where the proportion of positive specimens was low. The results of testing of pooled samples had a high level of agreement with the results of testing of individual samples, and use of the pooled testing strategy reduced costs and has the potential to increase the affordability of Xpert in countries with limited resources.",
    "full_text": "Testing Pooled Sputum with Xpert MTB/RIF for Diagnosis of Pulmonary Tuberculosis To Increase Affordability in Low-Income Countries\nSaddiq T. Abdurrahman,a Omezikam Mbanaso,b Lovett Lawson,c,d Olanrewaju Oladimeji,c Matthew Blakiston,b Joshua Obasanya,d,e Russell Dacombe,b Emily R. Adams,b Nnamdi Emenyonu,c Suvanand Sahu,f Jacob Creswell,f Luis E. Cuevasb Department of Public Health, Federal Capital Territory, Abuja, Nigeriaa; Liverpool School of Tropical Medicine, Liverpool, United Kingdomb; Zankli Medical Centre, Abuja, Nigeriac; Bingham University, Abuja, Nigeriad; Nigerian Centres for Disease Control, Abuja, Nigeriae; Stop TB Partnership Secretariat, Geneva, Switzerlandf\nTuberculosis (TB) is a global public health problem, with the highest burden occurring in low-income countries. In these countries, the use of more sensitive diagnostics, such as Xpert MTB/RIF (Xpert), is still limited by costs. A cost-saving strategy to diagnose other diseases is to pool samples from various individuals and test them with single tests. The samples in positive pool samples are then retested individually to identify the patients with the disease. We assessed a pooled testing strategy to optimize the affordability of Xpert for the diagnosis of TB. Adults with presumptive TB attending hospitals or identi\ufb01ed by canvassing of households in Abuja, Nigeria, were asked to provide sputum for individual and pooled (4 per pool) testing. The agreement of the results of testing of individual and pooled samples and costs were assessed. A total of 738 individuals submitted samples, with 115 (16%) being Mycobacterium tuberculosis positive. Valid Xpert results for individual and pooled samples were available for 718 specimens. Of these, testing of pooled samples detected 109 (96%) of 114 individual M. tuberculosis-positive samples, with the overall agreement being 99%. Xpert semiquantitative M. tuberculosis levels had a positive correlation with the smear grades, and the individual sample-positive/pooled sample-negative results were likely due to the M. tuberculosis concentration being below the detection limit. The strategy reduced cartridge costs by 31%. Savings were higher with samples from individuals recruited in the community, where the proportion of positive specimens was low. The results of testing of pooled samples had a high level of agreement with the results of testing of individual samples, and use of the pooled testing strategy reduced costs and has the potential to increase the affordability of Xpert in countries with limited resources.\n\nTuberculosis (TB) is a signi\ufb01cant global public health problem (1). Despite the availability of curative treatment, TB sits behind only human immunode\ufb01ciency virus (HIV) as the major cause of mortality associated with infectious disease worldwide (1). In 2013 there were an estimated 9 million new cases and 1.5 million deaths from TB, most of which occurred in low- and middle-income countries (LMICs) (1). The highest rates of TB per capita and the highest proportion of cases with HIV coinfection occur in sub-Saharan Africa (1).\nIn most low-income countries, direct sputum smear microscopy is the mainstay of TB diagnostics (2), as this test is inexpensive and highly speci\ufb01c, but it has a low to moderate sensitivity (2). The sensitivity of direct sputum smear microscopy is lower in patients with paucibacillary disease associated with HIV coinfection and in children, due to lower bacillary loads (3), and it cannot provide information on drug susceptibility (4). Conversely, sputum culture, in particular, automated liquid culture, is the most sensitive and speci\ufb01c diagnostic tool available for TB and facilitates drug susceptibility testing (2). However, culture requires a laboratory infrastructure, including biosafety equipment, not widely available in low-resource settings, and results typically take 2 to 6 weeks and, therefore, are rarely helpful for initial treatment decisions (2, 4).\nThe Xpert MTB/RIF (Xpert) assay (Cepheid Inc., Sunnyvale, CA, USA) is a self-contained, fully automated, real-time PCR assay that facilitates rapid semiquantitative detection of Mycobacterium tuberculosis and rifampin (RIF) resistance with minimal laboratory requirements compared to those needed for culture and other manually operated nucleic acid ampli\ufb01cation tests (NAATs)\n\n(4). Xpert is highly speci\ufb01c (99%) and substantially more sensitive than smear microscopy (4). The assay\u2019s turnaround time is less than 2 h, greatly shortening the time to TB diagnosis in locations where the machine is available, and it detects markers of RIF resistance (4). For low-income countries, the single-use cartridges cost $9.98 (FIND, 2013). However, despite this concessionary pricing, the cost involved to purchase and run the tests is still a limiting factor for widespread sustainable adoption of Xpert by TB control programs in LMICs (4, 5).\nThe high costs of diagnostics are not con\ufb01ned to TB, and the more cost-effective use of diagnostic tests for other infectious diseases has been explored. One approach that can reduce costs is to pool (put together) specimens from several patients and test them using a single test (6, 7). If a pool tests positive, then each specimen is tested individually to detect the positive sample(s), whereas if\nReceived 30 March 2015 Returned for modi\ufb01cation 26 April 2015 Accepted 18 May 2015 Accepted manuscript posted online 27 May 2015 Citation Abdurrahman ST, Mbanaso O, Lawson L, Oladimeji O, Blakiston M, Obasanya J, Dacombe R, Adams ER, Emenyonu N, Sahu S, Creswell J, Cuevas LE. 2015. Testing pooled sputum with Xpert MTB/RIF for diagnosis of pulmonary tuberculosis to increase affordability in low-income countries. J Clin Microbiol 53:2502\u20132508. doi:10.1128/JCM.00864-15. Editor: C.-A. D. Burnham Address correspondence to Luis E. Cuevas, luis.cuevas@lstmed.ac.uk. Copyright \u00a9 2015, American Society for Microbiology. All Rights Reserved. doi:10.1128/JCM.00864-15\n\n2502 jcm.asm.org\n\nJournal of Clinical Microbiology\n\nAugust 2015 Volume 53 Number 8\n\nPooled Testing of Sputum with Xpert MTB/RIF\n\nthe pooled specimens test negative, all individuals are considered infection free (6, 7). A pooling strategy appears to be cost-effective and accurate when NAATs are used to screen blood for HIV (8) and blood-borne hepatitis viruses (9), detect Chlamydia trachomatis and Neisseria gonorrhoeae in urine specimens (7), and identify in\ufb02uenza virus in nasopharyngeal swab samples (4). A potential disadvantage of pooled testing, however, is a decrease in test sensitivity through dilution of positive specimens beyond an assay limit of detection (10). The cost savings of pooled testing are determined by the prevalence of disease in the tested population, the number of samples per pool, and the degree of clustering of positive individuals in the tested population (6, 11).\nTherefore, as a means to optimize the use of Xpert cartridges, we explored whether a pooling strategy could be applied to sputum samples from individuals being screened for TB in a lowincome, high-HIV-prevalence setting. This study evaluated the agreement and cost savings of a two-stage pooled testing approach, whereby sputum samples from four consecutive patients were tested using a single Xpert cartridge with follow-on individual testing of positive pools and the results were compared to those obtained by individual Xpert testing of each sample. We also evaluated whether the rate of detection of positive samples varied with smear microscopy grade and assessed the impact of specimen dilution and the relationship between smear grade and the Xpert semiquantitative M. tuberculosis level.\nMATERIALS AND METHODS\nThe study took place in the Federal Capital Territory (FCT) of Nigeria. New adult patients with suspected pulmonary TB (PTB), presumed on the basis of a cough for more than 2 weeks, were recruited consecutively using two strategies. First, all adults with suspected PTB who presented to \ufb01ve outpatient departments of district hospitals in the FCT (Wuse, Bwari, Kwali, Kuje, and the university teaching hospital) were asked to participate. Second, patients with suspected PTB (symptomatic individuals) were recruited by community health extension workers canvassing consecutive households in slum areas and rural villages of the \ufb01ve FCT local government area councils (Abaji, Bwari, Kuje, Kwali, and Gwagwalada). These two recruitment strategies were used, as it was expected that the TB prevalence would be higher among hospital patients than those identi\ufb01ed in the community. Each participating individual provided at least two spot sputum samples for standard diagnostic practice, and the \ufb01rst one was also used for the evaluation in this study. Patients were asked to provide at least 5 ml of sputum in sputum cups with a wide mouth and a line to mark the amount. However, some patients had dif\ufb01culty producing this amount of sputum. Patients submitting specimens with less than 2 ml were asked to produce further specimens because it would not have allowed testing of specimens in duplicate (Xpert requires a minimum of 2 ml per test).\nThe two sputum samples were tested using Ziehl-Neelsen staining and smear microscopy and were graded per the World Health Organization (WHO) criteria (12). After smear preparation, the \ufb01rst sputum specimen was mixed with the Xpert MTB/RIF sample reagent (SR) in equal amounts per the manufacturer\u2019s guidelines. Two milliliters of this volume was transferred into a separate container with three other specimens. Each container of pooled sputa was manually shaken for approximately 1 min, and 2 ml of the pool was transferred to an Xpert MTB/RIF cartridge. Two milliliters of each remaining specimen processed with the SR was added to an individual Xpert cartridge. The individual and pooled samples were then tested simultaneously (Fig. 1). Xpert tests producing nonvalid results (error, an invalid result, or no result) were retested, if suf\ufb01cient sample was available.\nFor the purposes of this analysis, the individual Xpert MTB/RIF result was considered an individual\u2019s de\ufb01nitive TB status. Categorical data were\n\nsummarized using frequency counts and percentages, with the chi-square test being used to test for signi\ufb01cant differences where appropriate. Continuous data were summarized using the median and range. The agreement of the results obtained by use of the pooled and individual testing strategies was determined, with tests of agreement being performed (the kappa statistic was calculated). The relationship between the smear grade and the individual Xpert M. tuberculosis concentration and between the individual and pooled Xpert M. tuberculosis concentrations was evaluated using the Spearman rank order correlation. Cost differences were calculated on the basis of the number of cartridges that would have been required to test all specimens when using either a pooled or an individual testing strategy at a cartridge price of $9.98. Theoretical cost savings for pools of different sizes were also calculated using the formula proposed by Raboud et al. that estimates the number of positive pools for a given disease prevalence (13).\nEthical approval was obtained from the Liverpool School of Tropical Medicine Research Ethics Committee and the FCT Health Research Ethics Committee. All participants gave informed consent.\nRESULTS\nA total of 738 individuals with suspected PTB were recruited and supplied spot sputum samples for the study. Of the 738 individuals, 488 (66%) were recruited in the community and 250 (34%) were from district hospitals. The participants\u2019 baseline characteristics are described in Table 1. The 738 sputum samples were tested individually, and 183 pools of four samples plus 2 pools that contained only three samples (185 pools) were tested with the Xpert MTB/RIF assay.\nOne hundred \ufb01fteen (16%) of 738 samples were individual Xpert M. tuberculosis positive, 614 (83%) were negative, and 9 (1%) had failed results (error, invalid result, no result). Thirtytwo (4%) samples had an initial failed result, and 23 of these were successfully retested. Four (3%) of 115 M. tuberculosis-positive specimens were resistant to rifampin (Xpert RIF positive). As expected, there was a strong positive correlation between smear microscopy grade and Xpert semiquantitative M. tuberculosis level (rho \u03ed 0.694, P \u03fd 0.001) (Table 2). Only 2 (6%) of 31 samples with M. tuberculosis levels that were very low (cycle threshold [CT] value, \u03fe28) or low (CT value, 22 to 28) were smear positive, 20 (63%) of 32 samples with a medium M. tuberculosis level (CT value, 16 to 22) were smear positive, and 33 (92%) of 36 samples with a high M. tuberculosis level (CT value, \u03fd16) were smear positive.\nEighty-one (44%) of the 185 pools were Xpert positive for M. tuberculosis, 101 (55%) were negative, and 3 (2%) had a failed result reported. Six (3%) pools had failed results reported initially; however, three were successfully retested. There was no signi\ufb01cant difference (P \u03ed 0.47) in the number of failed Xpert results reported (before retesting) with testing of pooled (6 of 185, 3%) and individual (32 of 738, 4%) samples. Ninety-eight of 185 pools contained only samples collected in the community, and of these, 33 (34%) were positive; 42 pools contained only samples collected from district hospitals, and of these, 27 (64%) were positive; and 45 pools contained a mixture of community and hospital samples, and of these, 21 (47%) were positive.\nEighty (99%) of 81 M. tuberculosis-positive pools had at least one M. tuberculosis-positive sample from individual testing (true positives), with 1 M. tuberculosis-positive pool (1%) containing only negative samples (false positive). Conversely, 96 (95%) of the 101 M. tuberculosis-negative pools contained only M. tuberculosisnegative samples from individual testing (true negatives); 5 M.\n\nAugust 2015 Volume 53 Number 8\n\nJournal of Clinical Microbiology\n\njcm.asm.org 2503\n\nAbdurrahman et al.\n\nFIG 1 Flow diagram of the sputum processing scheme used.\n\ntuberculosis-negative pools (5%) had one sample that was M. tuberculosis positive upon individual testing (false negatives). Fiftysix (70%) of the 80 true-positive pools had one M. tuberculosispositive sample, 20 (25%) pools had two M. tuberculosis-positive samples, 3 (4%) pools had three M. tuberculosis-positive samples, and 1 (1%) pool had four M. tuberculosis-positive samples. Resistance to rifampin was detected in 3 (4%) of the 81 M. tuberculosispositive pools. All three (100%) pools had one or more RIF-resistant samples upon individual testing. Seventy-seven (99%) of the\n\n78 RIF-negative pools contained only rifampin-sensitive samples, with 1 RIF-sensitive pool (1%) containing one RIF-resistant sample.\nA cross tabulation of 61 pools containing only one M. tuberculosis-positive sample is shown in Table 3. As expected, there was a strong positive correlation between the Xpert M. tuberculosis concentration in the individual and pooled tests (rho \u03ed 0.799, P \u03fd 0.001). Five (50%) of 10 samples with very low individual M. tuberculosis levels were negative when tested in a pool. Likewise, samples with low, medium, and high individual M. tuberculosis\n\nTABLE 1 Baseline characteristics of participantsa\nCharacteristic No. of participants Median (range) age (yr)\n\nCommunity\n488 38 (12\u201385)\n\nDistrict hospital\n250 35 (15\u201390)\n\nTotal\n738 37 (12\u201390)\n\nNo. (%) participants Femaleb GeneXpert-con\ufb01rmed PTB cases Con\ufb01rmed PTB cases that were smear positive\n\n230 (54) 51 (11) 20 (45)\n\n99 (47) 64 (26) 37 (65)\n\n329 (51) 115 (16) 57 (56)\n\na Only 101 of 115 Xpert M. tuberculosis-positive PTB cases had a valid smear result. For both Xpert M. tuberculosis-positive and smear-positive cases, the percentages given use the\nnumber of cases with available data as the denominator. b Gender data were available for only 641 participants.\n\n2504 jcm.asm.org\n\nJournal of Clinical Microbiology\n\nAugust 2015 Volume 53 Number 8\n\nPooled Testing of Sputum with Xpert MTB/RIF\n\nTABLE 2 Individual Xpert M. tuberculosis results by smear gradea\n\nNo. (%) of participants with the following Xpert semiquantitative M. tuberculosis level:\n\nSmear grade\nNegative Scanty 1\u03e9 2\u03e9 3\u03e9\n\nNegative\n540 (98) 5 (1) 4 (1) 0 (0) 0 (0)\n\nVery low\n14 (88) 1 (6) 0 (0) 0 (0) 1 (6)\n\nLow\n15 (100) 0 (0) 0 (0) 0 (0) 0 (0)\n\nMedium\n12 (38) 3 (9) 7 (22) 8 (25) 2 (6)\n\nHigh\n3 (8) 4 (11) 8 (22) 11 (31) 10 (28)\n\nTotal\n584 (90) 13 (2) 19 (3) 19 (3) 13 (2)\n\nTotal\n\n549 (100)\n\n16 (100)\n\n15 (100)\n\n32 (100)\n\n36 (100)\n\n648 (100)\n\na Only cases that had both a valid Xpert M. tuberculosis result and a smear microscopy result are included. Speci\ufb01cally, data for only 99 of 115 individual Xpert M. tuberculosis-\npositive cases in the study are included here, as 14 did not have an available smear result, and 2 individuals with smear-positive/Xpert-positive results were missing Xpert semiquantitative M. tuberculosis level data. Smear grading was as follows: scanty, 1 to 9 acid-fast bacilli per 100 immersion \ufb01elds; 1\u03e9, 10 to 99 acid-fast bacilli per 100 immersion \ufb01elds; 2\u03e9, 1 to 10 acid-fast bacilli per immersion \ufb01eld; and 3\u03e9, \u03fe10 acid-fast bacilli per immersion \ufb01eld. Xpert semiquantitative M. tuberculosis levels were classi\ufb01ed as follows: very low, CT value of \u03fe28; low, CT value of 22 to 28; medium, CT value of 16 to 22; and high, CT value of \u03fd16.\n\nlevels frequently had lower concentrations reported in the pooled assay.\nSeven hundred eighteen sputum samples had both valid individual and pooled Xpert results, while 639 samples had a valid smear result and both valid individual and pooled Xpert results. A pooled testing strategy (pooled testing plus follow-on individual testing of each specimen from positive pools) would have detected 109 (96%) of the 114 individual M. tuberculosis-positive samples and correctly identi\ufb01ed 604 (100%) of 604 individual M. tuberculosis-negative samples. One M. tuberculosis-positive pool, however, contained only M. tuberculosis-negative samples on follow-on individual testing, thus giving an M. tuberculosis-negative result for the pooled testing strategy. Thus, the results of the pooled testing strategy agreed with those of the individual testing approach in 713 (99%) out of 718 instances (kappa value, 0.973; P \u03fd 0.001). After exclusion of samples without a smear result, the pooled testing approach would have detected 55 (98%) of 56 smear-positive samples and 42 (95%) of 44 smear-negative/M. tuberculosis-positive samples. After further exclusion of smearnegative samples pooled with smear-positive samples (which could be responsible for pool positivity), a pooled testing approach would still have detected 32 (94%) of 34 smear-negative/M. tuberculosis-positive cases.\nWe assessed the time that it took the investigator to perform the manual steps of the assay under different scenarios for the \ufb01rst 284 patients. The scenarios included processing of a single sample, simultaneous processing of a batch of 4 samples for individual testing, and processing of a pool of 4 samples. The results were used to estimate the time saved by the use of pooled testing. Test-\n\ning of samples individually required 607 h, and testing of samples individually in batches of four reduced the time by 446 h (73%), assuming that all samples were available for testing and processed simultaneously. The pooled strategy, which required testing of 71 pools followed by the individual testing of 140 samples from positive pools, reduced the testing time by 377 h (62%), assuming that individual samples from positive pools were tested simultaneously in batches of 4.\nThe cost of the cartridges for the testing of 738 samples individually was $7.365.24. Testing of 185 pools and retesting of 323 samples individually from the 81 positive pools (80 \u03eb 4 samples and 1 \u03eb 3 samples) would cost $5,069.84. Overall, a pooled testing strategy would have saved $2,295.40 (31%, equivalent to 230 cartridges). Pooled testing of the 98 community-only samples would cost $2,295.40, whereas testing of the 392 samples individually would cost $3,912.16; thus, the savings are $1,616.76 (41%). Conversely, pooled testing of the 42 district hospital-only samples would cost $1,487.02, whereas testing of the 167 samples individually would cost $1,666.66; thus, the savings are only $179.64 (11%). The theoretical cost savings for pools of different sizes using the PTB prevalence values obtained in this study are shown in Table 4.\nDISCUSSION\nNew testing platforms such as Xpert MTB/RIF have a signi\ufb01cant potential to increase the sensitivity of TB diagnostics in areas with a high prevalence of TB (4). However, their high cost relative to the cost of smear microscopy is still a limitation to their widespread use (4). This study evaluated pooled testing of sputum with the Xpert MTB/RIF\n\nTABLE 3 Cross tabulation of Xpert M. tuberculosis individual and pooled test concentrationsa\n\nXpert semiquantitative M. tuberculosis level in pooled test\n\nNo. (%) of participants with the following individual Xpert semiquantitative M. tuberculosis level:\n\nVery low\n\nLow\n\nMedium\n\nHigh\n\nTotal\n\nNegative Very low Low Medium High\n\n5 (50) 4 (40) 0 (0) 1 (10) 0 (0)\n\n0 (0) 7 (70) 3 (30) 0 (0) 0 (0)\n\n0 (0) 2 (8) 9 (38) 9 (38) 4 (17)\n\n0 (0) 0 (0) 0 (0) 8 (47) 9 (53)\n\n5 (8) 13 (21) 12 (20) 18 (30) 13 (21)\n\nTotal\n\n10 (100)\n\n10 (100)\n\n24 (100)\n\n17 (100)\n\n61 (100)\n\na Only results for pools containing one GeneXpert-positive sample are included. Xpert semiquantitative M. tuberculosis levels were classi\ufb01ed as follows: very low, CT value of \u03fe28; low, CT value of 22 to 28; medium, CT value of 16 to 22; high, CT value of \u03fd16.\n\nAugust 2015 Volume 53 Number 8\n\nJournal of Clinical Microbiology\n\njcm.asm.org 2505\n\nAbdurrahman et al.\n\nTABLE 4 Theoretical cost savings of a pooled testing strategy for 738 samples using different pool sizes by study setting\n\nStudy setting (disease prevalence)\n\nPool size\n\nNo of pooled tests \u03e9 no. of individual tests requireda\n\nCost ($) of pooled testing strategy\n\nCommunity (11%)\n\n3\n\n246 \u03e9 219\n\n4\n\n185 \u03e9 276\n\n5\n\n148 \u03e9 325\n\n4,640.70 4,600.78 4,720.54\n\nCost savings ($) with pooled testing strategyb\n2,724.54 (37) 2,764.46 (38) 2,644.70 (36)\n\nDistrict hospital (26%)\n\n3\n\n246 \u03e9 438\n\n4\n\n185 \u03e9 520\n\n5\n\n148 \u03e9 575\n\n6,826.32 7,035.90 7,215.54\n\n538.92 (7) 329.34 (4) 149.70 (2)\n\nTotal population (16%)\n\n3\n\n246 \u03e9 300\n\n4\n\n185 \u03e9 372\n\n5\n\n148 \u03e9 430\n\n5,449.08 5,558.86 5,768.44\n\n1,916.16 (26) 1,806.38 (25) 1,596.80 (22)\n\na The probability of pool testing positive is equal to 1 \u03ea (1 \u03ea P)n, where P is the prevalence of disease and n is the size of the pool. b Reduction in Xpert cartridge costs compared to individual testing at a cost of $7,365.24 for 738 samples. Values in parentheses are the percent savings compared with the cost of\ntesting of individual samples.\n\nassay as a way to increase its affordability and demonstrated substantial cost savings with only a limited loss of accuracy.\nThe overall prevalence of PTB in the study population was 16%, with 56% of con\ufb01rmed cases having smear-positive disease. Not unexpectedly, individuals recruited from the hospitals had a higher prevalence of PTB and smear-positive disease than those recruited from the community. Individuals with PTB who are identi\ufb01ed via presentation to health services tend to be more symptomatic, have more advanced disease, and have a higher rate of smear positivity than those identi\ufb01ed through active case \ufb01nding (14). Conversely, individuals with chronic cough in the community may be more likely to have other chronic respiratory problems.\nPredictably, individual Xpert testing con\ufb01rmed PTB in 44 smear-negative samples. The strong positive correlation between smear grade and Xpert semiquantitative M. tuberculosis level is in keeping with the \ufb01ndings described in previous reports (15). An Xpert \ufb01nding of a very low/low or a high M. tuberculosis level was reasonably predictive of smear-negative or -positive disease, respectively, and could be used for monitoring the quality of smear microscopy. This information may also be useful for infection control purposes. Some discrepant results were observed, such as a sample with a 3\u03e9 smear but a very low Xpert M. tuberculosis level. This could be due to sampling during sputum smear preparation, as the bacilli are not evenly distributed in the specimen. These \ufb01ndings are similar to those presented in previous reports showing that Xpert is predictive of smear status only at the extremes of cycle threshold values (16).\nThe agreement between a pooled and an individual Xpert testing strategy was 99%, with pooled testing detecting 96% of individual Xpert M. tuberculosis-positive cases overall and 94% of cases from smear-negative pools. The latter is important, as Xpert is often used as a follow-on test in smear-negative individuals. Pooling of the samples did not appear to result in PCR inhibition, as no difference in the rate of failed tests was found. The maintenance of intrinsic assay performance and diagnostic accuracy suggests that pooling of sputum for Xpert testing is a technically feasible option.\nThere were \ufb01ve false-negative pools, each containing a single sample with a very low individual M. tuberculosis level. False-negative results likely occurred because the small amount of M. tuberculosis bacilli in these positive samples was diluted below the de-\n\ntection threshold. A similar loss of detection of low-level-positive samples has been reported with pooled testing of blood for HIV (6). We also observed a dilution effect in other pools containing one M. tuberculosis-positive sample, whereby the M. tuberculosis level in the pooled sample was lower than that in the individual sample. The effects of dilution could mean that the accuracy of pooled testing may vary between populations with different sputum bacillary loads. Although the dilution effect is important, Xpert cartridges with a much higher sensitivity are expected to be released in 2016 (17), and these cartridges may be able to detect the few specimens missed by the current assay in this study.\nA further discrepant result was a positive pool containing all M. tuberculosis-negative samples on individual testing. This was an unexpected \ufb01nding, as the assay is highly speci\ufb01c (4). It may have occurred because of an uneven distribution of bacilli in the processed sample, with the portion used for individual testing not containing bacilli (sampling variability), or because of cross contamination of the pooled samples. Practically, clinical decisions would be guided by the individual test result. In these instances, repeat individual testing may be bene\ufb01cial.\nOne M. tuberculosis-positive pool provided a false RIF-sensitive result. This pool contained a mixture of RIF-resistant and RIF-sensitive isolates, which likely explains the discrepancy, as the Xpert MTB/RIF assay resistance requires 65% to 100% of the DNA present to be from the resistant isolate to produce a reliable RIF susceptibility result (18). A pooled testing strategy would still have identi\ufb01ed the RIF-resistant isolate when samples were tested individually.\nThe pooled testing of sputum samples has the potential to save time compared to the time required for the testing of individual samples, and the time required for pooled testing is comparable to that required for batched testing. However, the calculation presented assumes that samples from positive pools for individual testing are available at the time of testing, that these are tested in batches of 4, and that there are no indeterminate or failed tests. The time savings would be particularly useful in busy laboratories that receive large numbers of sputum samples or in large screening programs where large numbers of patients are tested.\nWe were able to demonstrate that in an area with a high prevalence of TB, such as Nigeria, a pooled sputum testing strategy can reduce Xpert cartridge costs by up to 31%. The savings were substantially higher when pools consisted of samples collected in the\n\n2506 jcm.asm.org\n\nJournal of Clinical Microbiology\n\nAugust 2015 Volume 53 Number 8\n\nPooled Testing of Sputum with Xpert MTB/RIF\n\ncommunity (41%) as opposed to samples collected in district hospitals (11%). This is a function of the lower disease prevalence in the community population and suggests that pooled Xpert testing may be best used to lower the costs of community-based active case \ufb01nding programs. Furthermore, the higher speci\ufb01city of Xpert than smear microscopy (99% versus 98%, respectively) would result in a lower number of false-positive results in community-based interventions. In these locations, the proportion of screened patients who have TB is lower than that in hospital settings, resulting in a lower predictive value of the test and the danger of a higher number of false-positive test results. This approach therefore would decrease the cost of active case \ufb01nding approaches, and the higher speci\ufb01city of Xpert would reduce the risk of false-positive results.\nThe predicted estimates of cost savings are comparable to, if slightly less than, what we obtained. The marginally lower values may be because the predictive model does not account for any potential clustering of positive samples that may have occurred. The estimates also support the use of a pool size of three or four in the study population. In locations with a different PTB epidemiology, the most appropriate pool size may differ, as smaller pools may be appropriate in areas of high TB prevalence. For example, in hospital patients it would be preferable to use a pool sample size of three, which would produce higher cost savings than a pool of four samples, while in the community, a pool of four or even \ufb01ve would result in higher savings.\nFrom a safety and practical point of view, pooling of sputum samples already processed with the Xpert sample reagent (SR) is superior to pooling of unprocessed samples. Processing of sputum samples with the SR virtually eliminates biohazard risks (19) and lique\ufb01es the sample, facilitating easier transfer, and if a pool tests positive, the technician simply has to add the remaining portion of the samples into individual cartridges. The extra steps involved in pooled testing heighten the potential for laboratory errors, particularly if the laboratory is dealing with large numbers of samples. Therefore, strict adherence to good laboratory practices is required, including careful handling and labeling of samples and pools and clear record keeping.\nLimitations of the study include incomplete demographic data, absent smear status, and an inability to retest failed results for some individuals. HIV coinfection status was also unavailable, although it was likely to be commonplace. Improved participant information would have aided interpretation of the \ufb01ndings, but its absence is not expected to affect the results. We were unable to use sputum culture, which would have helped resolve the results for pools with discrepant results. The simultaneous testing of pooled and individual samples was required to determine agreement, and that testing approach varies from how a pooled testing system would work in practice. Furthermore, as the same investigator performed both the pooled and individual Xpert MTB/RIF tests, the investigator was not blinded to the results of the other set of tests when performing a particular set of tests. However, as the test is fully automated and the results are objective, knowledge of the results of the other tests is not expected to bias the results. The participants consisted primarily of adults with suspected PTB; therefore, the results should not be generalized to other patient populations.\nConclusion. An Xpert MTB/RIF pooled sputum testing strategy had a high level of agreement with individual Xpert testing at a reduced cost. The \ufb01ndings suggest that a pooled testing ap-\n\nproach has the potential to increase the affordability of Xpert testing, as the cost of the test is not expected to change in the near future. This strategy would be especially suited for use in active case \ufb01nding programs and in locations where the proportion of positive cases is expected to be low. Further studies with the highsensitivity Xpert cartridges may increase the agreement between the single and pooled testing strategies and should be explored.\nACKNOWLEDGMENTS\nWe are grateful for the support of the staff of Zankli Tuberculosis Research Laboratory, the Federal Capital Territory, and the National Tuberculosis and Leprosy Control programs.\nNone of us has a con\ufb02ict of interest to declare. The study was conceived by L.E.C., L.L., S.T.A., J.O., and O.M.; data collection and sputum processing were conducted by O.M., S.T.A., O.O., and N.E.; data analysis and interpretation were conducted by L.E.C., O.M., M.B., E.R.A., and R.D.; and L.E.C., O.M., and M.B. wrote the \ufb01rst draft of the manuscript. We all contributed to the \ufb01nal manuscript. The project was funded by a Wave II TB REACH award (project number T9-370-114) and the EDCTP (SP.2011.41304.021) and its cofunders (the Medical Research Council [MRC] of the United Kingdom and the Instituto de Salud Carlos III [ISCIII] of Spain). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\nREFERENCES\n1. WHO. 2014. Global tuberculosis report 2014. WHO, Geneva, Switzerland. 2. Parsons LM, Somoskovi A, Gutierrez C, Lee E, Paramasivan CN,\nAbimiku A, Spector S, Roscigno G, Nkengasong J. 2011. Laboratory diagnosis of tuberculosis in resource-poor countries: challenges and opportunities. Clin Microbiol Rev 24:314 \u2013350. http://dx.doi.org/10.1128 /CMR.00059-10. 3. Colebunders R, Bastian I. 2000. A review of the diagnosis and treatment of smear-negative pulmonary tuberculosis. Int J Tuberc Lung Dis 4:97\u2013107. 4. Lawn SD, Mwaba P, Bates M, Piatek A, Alexander H, Marais BJ, Cuevas LE, McHugh TD, Zijenah L, Kapata N, Abubakar I, McNerney R, Hoelscher M, Memish ZA, Migliori GB, Kim P, Maeurer M, Schito M, Zumla A. 2013. Advances in tuberculosis diagnostics: the Xpert MTB/RIF assay and future prospects for a point-of-care test. Lancet Infect Dis 13: 349 \u2013361. http://dx.doi.org/10.1016/S1473-3099(13)70008-2. 5. Qin ZZ, Pai M, Van Gemert W, Sahu S, Ghiasi M, Creswell J. 2015. How is Xpert MTB/RIF being implemented in 22 high tuberculosis burden countries? Eur Respir J 45:549 \u2013554. http://dx.doi.org/10.1183/09031 936.00147714. 6. Emmanuel JC, Bassett MT, Smith HJ, Jacobs JA. 1988. Pooling of sera for human immunode\ufb01ciency virus (HIV) testing: an economical method for use in developing countries. J Clin Pathol 41:582\u2013585. http://dx.doi .org/10.1136/jcp.41.5.582. 7. Lindan C, Mathur M, Kumta S, Jerajani H, Gogate A, Schachter J, Moncada J. 2005. Utility of pooled urine specimens for detection of Chlamydia trachomatis and Neisseria gonorrhoeae in men attending public sexually transmitted infection clinics in Mumbai, India, by PCR. J Clin Microbiol 43:1674 \u20131677. http://dx.doi.org/10.1128/JCM.43.4.1674-1677 .2005. 8. Morandi PA, Schockmel GA, Yerly S, Burgisser P, Erb P, Matter L, Sitavanc R, Perrin L. 1998. Detection of human immunode\ufb01ciency virus type 1 (HIV-1) RNA in pools of sera negative for antibodies to HIV-1 and HIV-2. J Clin Microbiol 36:1534 \u20131538. 9. Mine H, Emura H, Miyamoto M, Tomono T, Minegishi K, Murokawa H, Yamanaka R, Yoshikawa A, Nishioka K, Japanese Red Cross NAT Research Group. 2003. High throughput screening of 16 million serologically negative blood donors for hepatitis B virus, hepatitis C virus and human immunode\ufb01ciency virus type-1 by nucleic acid ampli\ufb01cation testing with speci\ufb01c and sensitive multiplex reagent in Japan. J Virol Methods 112:145\u2013151. http://dx.doi.org/10.1016/S0166-0934(03)00215-5. 10. Westreich DJ, Hudgens MG, Fiscus SA, Pilcher CD. 2008. Optimizing screening for acute human immunode\ufb01ciency virus infection with pooled nucleic acid ampli\ufb01cation tests. J Clin Microbiol 46:1785\u20131792. http://dx .doi.org/10.1128/JCM.00787-07.\n\nAugust 2015 Volume 53 Number 8\n\nJournal of Clinical Microbiology\n\njcm.asm.org 2507\n\nAbdurrahman et al.\n\n11. Peeling RW, Toye B, Jessamine P, Gemmill I. 1998. Pooling of urine specimens for PCR testing: a cost saving strategy for Chlamydia trachomatis control programmes. Sex Transm Infect 74:66 \u201370. http://dx.doi.org /10.1136/sti.74.1.66.\n12. WHO. 1998. Laboratory services in tuberculosis control. Part II. Microscopy. Report WHO/TB/98.258. WHO, Geneva, Switzerland.\n13. Raboud JM, Sherlock C, Schechter MT, Lepine DG, O\u2019Shaughnessy MV. 1993. Combining pooling and alternative algorithms in seroprevalence studies. J Clin Microbiol 31:2298 \u20132302.\n14. Ward HA, Marciniuk DD, Pahwa P, Hoeppner VH. 2004. Extent of pulmonary tuberculosis in patients diagnosed by active compared to passive case \ufb01nding. Int J Tuberc Lung Dis 8:593\u2013597.\n15. Theron G, Peter J, van Zyl-Smit R, Mishra H, Streicher E, Murray S, Dawson R, Whitelaw A, Hoelscher M, Sharma S, Pai M, Warren R, Dheda K. 2011. Evaluation of the Xpert MTB/RIF assay for the diagnosis of pulmonary tuberculosis in a high HIV prevalence setting. Am J Respir Crit Care Med 184:132\u2013140. http://dx.doi.org/10.1164/rccm .201101-0056OC.\n\n16. Theron G, Pinto L, Peter J, Mishra HK, Mishra HK, van Zyl-Smit R, Sharma SK, Dheda K. 2012. The use of an automated quantitative polymerase chain reaction (Xpert MTB/RIF) to predict the sputum smear status of tuberculosis patients. Clin Infect Dis 54:384 \u2013388. http://dx.doi .org/10.1093/cid/cir824.\n17. Cepheid. 2014. Cepheid, FIND & Rutgers announce collaboration for next-generation innovations to game changing Xpert MTB/RIF test. Cepheid, Sunnyvale, CA.\n18. Blakemore R, Story E, Helb D, Kop J, Banada P, Owens MR, Chakravorty S, Jones M, Alland D. 2010. Evaluation of the analytical performance of the Xpert MTB/RIF assay. J Clin Microbiol 48:2495\u20132501. http: //dx.doi.org/10.1128/JCM.00128-10.\n19. Banada PP, Sivasubramani SK, Blakemore R, Boehme C, Perkins MD, Fennelly K, Alland D. 2010. Containment of bioaerosol infection risk by the Xpert MTB/RIF assay and its applicability to point-of-care settings. J Clin Microbiol 48:3551\u20133557. http://dx.doi.org/10.1128 /JCM.01053-10.\n\n2508 jcm.asm.org\n\nJournal of Clinical Microbiology\n\nAugust 2015 Volume 53 Number 8\n\n\n",
    "authors": [
      "Saddiq T. Abdurrahman",
      "Omezikam Mbanaso",
      "Lovett Lawson",
      "Olanrewaju Oladimeji",
      "Matthew Blakiston",
      "Joshua Obasanya",
      "Russell Dacombe",
      "Emily R. Adams",
      "Nnamdi Emenyonu",
      "Suvanand Sahu",
      "Jacob Creswell",
      "Luis E. Cuevas"
    ],
    "doi": "10.1128/JCM.00864-15",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://journals.asm.org/doi/10.1128/JCM.00864-15"
  },
  {
    "key": "66SW72HV",
    "title": "Prevalence, acceptability, and cost of routine screening for pulmonary tuberculosis among pregnant women in Cotonou, Benin",
    "abstract": "Objectives\n              We sought to evaluate the yield, cost, feasibility, and acceptability of routine tuberculosis (TB) screening of pregnant women in Cotonou, Benin.\n            \n            \n              Design\n              Mixed-methods, cross-sectional study with a cost assessment.\n            \n            \n              Setting\n              Eight participating health facilities in Cotonou, Benin.\n            \n            \n              Participants\n              Consecutive pregnant women presenting for antenatal care at any participating site who were not in labor or currently being treated for TB from April 2017 to April 2018.\n            \n            \n              Interventions\n              Screening for the presence of TB symptoms by midwives and Xpert MTB/RIF for those with cough for at least two weeks. Semi-structured interviews with 14 midwives and 16 pregnant women about experiences with TB screening.\n            \n            \n              Primary and secondary outcome measures\n              Proportion of pregnant women with cough of at least two weeks and/or microbiologically confirmed TB. The cost per pregnant woman screened and per TB case diagnosed in 2019 USD from the health system perspective.\n            \n            \n              Results\n              Out of 4,070 pregnant women enrolled in the study, 94 (2.3%) had a cough for at least two weeks at the time of screening. The average (standard deviation) age of symptomatic women was 26 \u00b1 5 years and 5 (5.3%) had HIV. Among the 94 symptomatic women, 2 (2.3%) had microbiologically confirmed TB for a TB prevalence of 49 per 100,000 (95% CI: 6 to 177 per 100,000) among pregnant women enrolled in the study. The average cost to screen one pregnant woman for TB was $1.12 USD and the cost per TB case diagnosed was $2271 USD. Thematic analysis suggested knowledge of TB complications in pregnancy was low, but that routine TB screening was acceptable to both midwives and pregnant women.\n            \n            \n              Conclusion\n              Enhanced screening for TB among pregnant women is feasible, acceptable, and inexpensive per woman screened, however in this setting has suboptimal yield even if it can contribute to enhance TB case finding.",
    "full_text": "PLOS ONE\n\na1111111111 a1111111111 a1111111111 a1111111111 a1111111111\n\nRESEARCH ARTICLE\nPrevalence, acceptability, and cost of routine screening for pulmonary tuberculosis among pregnant women in Cotonou, Benin\nM\u00eanonli Adjobimey1,2*, Serge Ade1,3, Prudence Wachinou1,2, Marius Esse1, Lydia Yaha1, Wilfried Bekou1, Jonathon R. CampbellID4, Narcisse Toundoh1, Omer Adjibode1, Geoffroy Attikpa2, Gildas Agodokpessi1,2, Dissou Affolabi1,2, Corinne S. MerleID5\n1 National Tuberculosis Program, Cotonou, Benin, 2 Faculty of Health Sciences, University of AbomeyCalavi, Cotonou, Benin, 3 Faculty of Medicine, University of Parakou, Parakou, Benin, 4 Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, Quebec, Canada, 5 Special Programme for Research & Training in Tropical Diseases (TDR), World Health Organization, Geneva, Switzerland\n* menoladjobi@yahoo.fr\nAbstract\n\nOPEN ACCESS\nCitation: Adjobimey M, Ade S, Wachinou P, Esse M, Yaha L, Bekou W, et al. (2022) Prevalence, acceptability, and cost of routine screening for pulmonary tuberculosis among pregnant women in Cotonou, Benin. PLoS ONE 17(2): e0264206. https://doi.org/10.1371/journal.pone.0264206\nEditor: Emma K. Kalk, University of Cape Town, SOUTH AFRICA\nReceived: July 15, 2021\nAccepted: February 7, 2022\n\nObjectives\nWe sought to evaluate the yield, cost, feasibility, and acceptability of routine tuberculosis (TB) screening of pregnant women in Cotonou, Benin.\nDesign\nMixed-methods, cross-sectional study with a cost assessment.\nSetting\nEight participating health facilities in Cotonou, Benin.\n\nPublished: February 22, 2022\nPeer Review History: PLOS recognizes the benefits of transparency in the peer review process; therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. The editorial history of this article is available here: https://doi.org/10.1371/journal.pone.0264206\nCopyright: \u00a9 2022 Adjobimey et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.\nData Availability Statement: All relevant data are within the paper and its Supporting information files.\n\nParticipants\nConsecutive pregnant women presenting for antenatal care at any participating site who were not in labor or currently being treated for TB from April 2017 to April 2018.\nInterventions\nScreening for the presence of TB symptoms by midwives and Xpert MTB/RIF for those with cough for at least two weeks. Semi-structured interviews with 14 midwives and 16 pregnant women about experiences with TB screening.\nPrimary and secondary outcome measures\nProportion of pregnant women with cough of at least two weeks and/or microbiologically confirmed TB. The cost per pregnant woman screened and per TB case diagnosed in 2019 USD from the health system perspective.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n1 / 14\n\nPLOS ONE\n\nIntegrating tuberculosis screening into antenatal visits\n\nFunding: TDR, the Special Programme for Research and Training in Tropical Diseases, funded the development and conduct of this study as part of the West African Regional Network for TB control (WARN-TB) implementation research training programme. TDR can conduct its work thanks to the commitment and support from a variety of funders. These include long-term core contributors from national governments and international institutions, as well as designated funding for specific projects within our current priorities. A full list of TDR donors is available at: https://www.who.int/tdr/about/funding/en/.\n\nResults\nOut of 4,070 pregnant women enrolled in the study, 94 (2.3%) had a cough for at least two weeks at the time of screening. The average (standard deviation) age of symptomatic women was 26 \u00b1 5 years and 5 (5.3%) had HIV. Among the 94 symptomatic women, 2 (2.3%) had microbiologically confirmed TB for a TB prevalence of 49 per 100,000 (95% CI: 6 to 177 per 100,000) among pregnant women enrolled in the study. The average cost to screen one pregnant woman for TB was $1.12 USD and the cost per TB case diagnosed was $2271 USD. Thematic analysis suggested knowledge of TB complications in pregnancy was low, but that routine TB screening was acceptable to both midwives and pregnant women.\n\nCompeting interests: CSM is currently a staff member of the World Health Organization; the author alone is responsible for the views expressed in this publication and they do not necessarily represent the decisions, policy or views of the WHO.\n\nConclusion\nEnhanced screening for TB among pregnant women is feasible, acceptable, and inexpensive per woman screened, however in this setting has suboptimal yield even if it can contribute to enhance TB case finding.\n\nIntroduction\nTuberculosis (TB) continues to be a major public health concern, particularly in low- and middle-income countries, despite significant efforts by the international community [1, 2]. According to the World Health Organization (WHO), there was an estimated 10 million TB cases resulting in 1.5 million deaths in 2020 [3]. Among women, TB is more common during childbearing age and is a major cause of maternal and infant mortality [4]. It is also one of the top three causes of death among women aged 15 to 49 years old [5, 6]. Pregnant women are at increased risk of TB and adverse maternal and fetal outcomes [4]. Therefore, WHO classifies pregnant women as a high risk, vulnerable population and recommends active case finding for early detection of TB [7].\nThe prevalence of symptoms consistent with TB and confirmed TB disease among pregnant women varies between epidemiologic contexts and with screening approaches. In the United Kingdom the incidence of TB among pregnant women was approximately 1.5-times higher than the general population [8], however in a similarly low TB prevalence setting of the United States the yield of screening was only 0.025% [9]. Yield of screening is also inconsistent in high-burden settings. In Pakistan, though 2.6% of pregnant women had symptoms consistent with TB, only 0.025% were diagnosed with TB [10]. Comparatively, in eSwatini, TB prevalence was 2% among HIV-negative pregnant women [11], while in South Africa, among pregnant women living with HIV, 16% had TB symptoms, and 2.5% had TB [12].\nSymptom presentation among pregnant women also varies, which affects the sensitivity of symptom screening as a triage test for microbiologic testing. Several studies report WHO recommended symptom screening among pregnant women has sensitivity that varies from 28% to 54% [4, 12] largely driven by prolonged cough. This is lower than estimates from studies informing World Health Organization symptom screening recommendations which suggest symptom screening has a sensitivity of 73% [13, 14].\nIn Benin, TB mainly affects the young adult population between 25 and 44 years of age, with nearly one-third of all people diagnosed with TB being women of childbearing age [15]. However, pregnancy data is not routinely collected on TB case report forms and there are no\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n2 / 14\n\nPLOS ONE\n\nIntegrating tuberculosis screening into antenatal visits\nnational guidelines for screening and management of TB in pregnant women, largely due to insufficient and inconsistent data [4]. In the absence of national guidelines and data on TB in pregnant women in Benin, this study was initiated. The objectives of the study were to implement enhanced case finding among pregnant women, to describe characteristics of women with TB symptoms, and to evaluate the yield, cost, feasibility, and acceptability of such a programme in Cotonou, Benin.\nMethods Study setting\nThis study took place in Benin, a country located in West Africa with an estimated TB incidence of 55 per 100,000 population in 2019. HIV seroprevalence in the general population was only 1.2% in 2012 [16], but the proportion of TB patients whose HIV status is known is 98% and HIV seroprevalence among TB patients has remained stable at around 16% from 2011 to 2017 [15]. As of 2018, there are approximately 370,000 pregnancies annually in Benin [17].\nThe city of Cotonou, with a population of approximately 685,000, was where the study was implemented. HIV prevalence in the city of Cotonou is higher than other cities at 1.9% [11] and approximately the same among pregnant women with known HIV status (1.6%) [11]. The study included eight health centers\u2014three public, three religious, and two private facilities\u2014 collaborating with the national tuberculosis programme and conducting antenatal care (ANC) activities. All health facilities were located within a radius of 15km from the National TB hospital named Centre National Hospitalier Universitaire de Pneumo-Phtisiologie de Cotonou (CNHU-PPC)\u2014which also houses the national tuberculosis program (NTP).\nStudy design\nThis was a mixed-methods cross-sectional study combined with a cost assessment that took place from April 2017 to April 2018.\nInclusion and exclusion criteria\nAll women presenting to participating health facilities for routine pregnancy assessments were assessed for inclusion during the study period. Inclusion criteria were: age between 14 and 45 years and evidence of pregnancy based on a positive pregnancy test, ultrasound or gynecological examination. Exclusion criteria were: women in labor or those with already diagnosed TB and/or under TB treatment at the time of the survey.\nSample size\nWe calculated the sample size required to estimate the prevalence of TB using a previous study from Zambia [18] which found a TB prevalence of 1.5% among pregnant women with TB symptoms. Using a Poisson regression formula, type I error rate of 5% and type II error rate of 10%, we estimated we would need to recruit 1022 pregnant women with TB symptoms to estimate this same prevalence or a total of 4444 pregnant women regardless of TB symptoms (if based on the Zambia study, where ~23% of the pregnant women were classified as presumptive TB patients [18]).\nStudy procedures\nPrior to the start of the study, all midwives performing antenatal care at participating health centres received training regarding TB, its deleterious impact on pregnancy, and TB symptoms such as cough for \ufffd2 weeks, fever, night sweats, or weight loss. Taking into account that\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n3 / 14\n\nPLOS ONE\n\nIntegrating tuberculosis screening into antenatal visits\nprolonged cough is the primary symptom indicative of TB, NTP prioritization of persons with cough for further TB evaluation, and midwife workload, after recording all symptoms of TB, only those with cough of at least two weeks received microbiologic testing with Xpert MTB/ RIF.\nPregnant women presenting to healthcare facilities for the first time gave informed consent and were screened for all TB symptoms by the attending midwife. Symptoms were selfreported. Subsequent data on women without cough for at least two weeks were not collected. For women with a cough of at least two weeks, the midwife administered a structured questionnaire to collect data on socio-demographic, occupational, household, and clinical characteristics. All symptomatic women had sputum collected by the midwife (spot sputum). Each day a representative from the NTP would call the health facilities and come to the clinic to collect sputum samples should they have been collected that day. All sputum samples were analyzed at the CNHU-PPC using Xpert MTB/RIF. Women found to have tuberculosis were managed by standard NTP protocols. All data collection and study progress were monitored by the NTP, with monthly meetings on study status.\nQuantitative analysis\nData collected during the study were entered anonymously using Epi Data V.3.1. and analyzed with Epi Data Client v.2.0.7.22; the dataset is available in S1 Dataset. We conducted descriptive analysis to detail the characteristics of pregnant women found to have cough for at least two weeks and describe in more detail those diagnosed with TB. We estimated the TB prevalence per screening visit among pregnant women and exact binomial 95% confidence interval (95% CI).\nQualitative analysis\nWe did a qualitative analysis to understand participants\u2019 experiences (both midwives and pregnant women) with TB screening during routine antenatal appointments.\nData collection. We recruited two pregnant women per site (16 total) and two midwives per site (14 total, as two sites only had one midwife) to conduct in-depth interviews and broaden the perspective and breadth of responses. Sample sizes were based on convenience and what was feasible. At each site, a number was assigned to each pregnant woman and to each midwife present on the day of the survey; random sampling was used to identify who would be interviewed. For both groups, the interviewer was experienced in qualitative data collection. After explaining the goal of the interview to the participant, the interviewer used a semi-structured questionnaire developed by the authors (Table 1). Interviews were conducted in French, Fon, or Yoruba and lasted 30\u201345 minutes for midwives and 20\u201330 minutes for pregnant women. The interviews were conducted onsite in a confidential location where the participant could not be overheard. There were no repeat interviews. The interviewer took notes and recorded the entire interview with the consent of the participants. The recordings were then transcribed.\nData analysis. The data were analyzed by topic manually (without software) and independently by a group of three people experienced in qualitative analysis who listened to the recording and read the transcript and notes beforehand. Based on their understanding, they identified themes that represented the concepts expressed in the interviews. The three-person panel reviewed the independently assessed results, and through a discussion process reached consensus on key themes.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n4 / 14\n\nPLOS ONE\n\nIntegrating tuberculosis screening into antenatal visits\n\nTable 1. Questions from in-depth interviews with midwives and pregnant women.\n\nParticipants Midwives\n\nQuestions\n1. What does it mean to you to screen for tuberculosis in pregnant women? 2. Do you enjoy routinely screening patients for tuberculosis during antenatal visits? 3. Do you think it is possible to routinely screen pregnant women for tuberculosis as part of\nevery midwife\u2019s routine? 4. What do you think might be the barriers to integrating routine tuberculosis screening into\nroutine antenatal care? 5. What do you think are the practical prerequisites for integrating tuberculosis screening into\nroutine antenatal care services? 6. The national tuberculosis programme is developing a project to integrate routine tuberculosis\nscreening into the antenatal care visit for pregnant women. Would you support this implementation? 7. In your opinion, is routine screening of pregnant women for tuberculosis during antenatal visits feasible?\n\nPregnant Women\n\n1. What does it mean to you to screen for tuberculosis in pregnant women?\n2. Does it make sense for you, your partner, and your family to screen for tuberculosis in your current pregnancy?\n3. Was it a problem for you, your partner, and your family to screen for tuberculosis in your current pregnancy?\n4. What do you think might be the barriers to integrating routine tuberculosis screening into antenatal care for pregnant women?\n5. In your opinion, is routine screening of pregnant women for tuberculosis during antenatal visits acceptable?\n\nhttps://doi.org/10.1371/journal.pone.0264206.t001\nCost analysis\nWe did a cost analysis to estimate the additional cost of the TB screening program among pregnant women during antenatal care. We employed a microcosting approach, which considered all costs associated with the TB screening program that were in excess of the current standard of care (no TB screening), excluding costs associated with research. All costs were from the health system perspective and expressed in 2019 USD. Costs were locally collected where possible. When necessary, we converted salaries using purchasing power parity and material costs using exchange rates [19, 20].\nWe considered the following costs: the costs associated with training midwives in TB symptom screening and sputum collection; the costs associated with contacting facilities about the need for sputum sample pickup; the costs associated with TB symptom screening; the costs associated with sample transport; and the costs associated with Xpert MTB/RIF analysis.\nTo estimate personnel costs, we collected average annual salaries of midwives and laboratory technicians. We used time and motion techniques to estimate the time required to perform symptom screening (in absence and presence of symptoms) and the time required to analyze samples with Xpert MTB/RIF. For midwives, a midwife from each site was randomly selected and monitored during their workday for one week to record different activities performed and the time required for each activity. For laboratory technicians, a technician was observed receiving, preparing, and analyzing a set of sputum samples with Xpert MTB/RIF. To estimate material costs, we used the Global Drug Facility for the costs of Xpert MTB/RIF cartridges and machines. The capital costs of the machine were annualized over a five-year useful lifetime at 3% per annum and averaged on a per sample basis. Costs associated with Xpert MTB/RIF maintenance came from laboratory expense logs and were averaged on a per sample basis. Laboratory overhead costs were estimated based on annual expense logs and averaged based on manpower allocation to Xpert MTB/RIF and on a per sample basis. Since TB\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n5 / 14\n\nPLOS ONE\n\nIntegrating tuberculosis screening into antenatal visits\nscreening occurred during routine antenatal care visits, we did not consider overhead costs of these visits. For costs of training and communication, we used the incurred costs of these components during the research study.\nWe calculated the overall cost of our TB screening intervention during the study period, and estimated: the cost per pregnant woman screened, the cost per pregnant woman identified with cough of at least two weeks, and the cost per pregnant woman diagnosed with TB.\nEthical considerations\nThe study was jointly authorized by Directorate of Maternal and Child Health and the NTP, two departments of the Ministry of Health of Benin, and by the managers of the health facilities. The approval of the National Committee of Ethics for Health Research (CNERS) was obtained under the number n\u02da 029 of 09/09/2016.\nWritten consent was obtained from all pregnant women and midwives. The consent forms were approved by the ethics committee. Information relating to the state of health of pregnant women has been treated with respect for confidentiality and the human person.\nResults\nBetween April 2017 and April 2018, a total of 4,070 consenting pregnant women underwent TB screening during routine antenatal care visits at eight health facilities in Cotonou, Benin (Fig 1).\nQuantitative analysis\nCharacteristics of women screened and with cough of at least two weeks at their screening visit. The 4,070 pregnant women participating in the study had an average (standard deviation) age of 27 \u00b1 5 years. Of these, 94 women (2.3%) had cough of at least two weeks and were further assessed. They ranged in age from 16 to 40 years with an average of 26 \u00b1 5 years (Table 2). The vast majority (77/94; 82%) had received BCG vaccination. On average, women were on their third pregnancy\u201475 (79%) had at least two previous pregnancies and 36 (38%) had given birth to at least two children. Among women with cough of at least two weeks, the majority (54/94; 57.4%) worked in sales and services occupations, 4 (4.2%) were exposed to second-hand smoke, and 5 (5.3%) had known previous tuberculosis contact (Table 3). The co-morbidities found among pregnant women were anemia (11/94; 11.7%), HIV (5/94; 5.3%); hepatitis B (3/94; 3.2%), and hypertension (2/94; 2.1%). Among women with a cough of at least two weeks, most had productive (i.e., with expectoration) coughs (67/94; 71.3%). The next most common symptom was fever (28/94; 29.8%), followed by weight loss/inability to gain weight (12/94; 12.8%), night sweats (9/94; 9.6%), and hemoptysis (5/94; 5.3%).\nPrevalence of TB among pregnant women with cough of at least two weeks at their screening visit. All 94 women with a cough provided sputum for Xpert MTB/RIF analysis. Of these, 2 (2.3%) were found to have bacteriologically confirmed TB. This is equivalent to a prevalence of 49 cases per 100,000 (95% CI: 6 to 177 per 100,000) pregnant women screened per screening visit (2 cases among 4070 pregnant women screened). Both women were HIVnegative and initiated TB treatment\u2014one in the first trimester and one later in pregnancy after being initially lost to follow-up\u2014and delivered their children. However, the child of the mother who initiated treatment later in pregnancy died two weeks after birth; the exact cause of death is unknown to the study team.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n6 / 14\n\nPLOS ONE\n\nIntegrating tuberculosis screening into antenatal visits\n\nFig 1. Flow diagram of participants through the study.\nhttps://doi.org/10.1371/journal.pone.0264206.g001\nQualitative analysis\nWe interviewed a total of 16 pregnant women and 14 midwives for our qualitative assessment. We found that most participants had good knowledge of TB symptoms such as cough, fever, and hemoptysis.\nOriginal Participant Response: \u201cCelui qui tousse et crache du sang doit courir pour aller \u00e0 l\u2019hopital, la tuberculose est souvent l\u00e0\u201d\nEnglish Translation: \u201cThe one who coughs, and spits blood must run to the hospital, tuberculosis is often there\u201d\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n7 / 14\n\nPLOS ONE\n\nIntegrating tuberculosis screening into antenatal visits\n\nTable 2. Characteristics of pregnant women found to have cough of at least two weeks during antenatal care visits (N = 94).\n\nAge Level of education Marital status Number of pregnancies Number of deliveries Age of pregnancy Total\n\n15\u201324 25\u201334 35\u201344 None Primary Secondary Post-Secondary Free Union Married Single First pregnancy 2\u20133 pregnancies \ufffd 4 pregnancies Zero deliveries One delivery 2\u20133 deliveries \ufffd 4 deliveries First trimester Second trimester Third trimester\n\nNumber 33 52 9 27 27 33 7 31 59 4 25 41 28 19 39 30 6 36 39 19 94\n\nPercentage (%) 35.1 55.3 9.6 28.7 28.7 35.1 7.4 33.0 62.8 4.3 26.6 43.6 29.8 20.2 41.5 31.9 6.4 38.3 41.5 20.2 100\n\nhttps://doi.org/10.1371/journal.pone.0264206.t002\n\nOriginal Participant Response:\u201cOn reconnait une personne qui a la tuberculose lorsque tousses, crache a le corps et devient blanc\u201d\nEnglish Translation: \u201cYou can recognize a person who has tuberculosis when they cough, spit on their body and turns white\u201d\nMidwives were aware of the possibility of TB complications in pregnant women, but not specific manifestations.\n\nTable 3. Profession, smoking status, medical history, and tuberculosis symptoms present among pregnant women with cough of at least two weeks during antenatal care visits (N = 94).\n\nProfessional and environmental features Medical history\nTB symptoms\n\nSales/Service Profession Second Hand Smoking Known Previous Tuberculosis Contact Presence of BCG scar Anemia HIV Hepatitis B Hypertension Previous Tuberculosis Diagnosis Productive cough \ufffd 2 weeks Fever Weight stagnation / weight loss Night sweats Hemoptysis\n\nNumber 54 4 5 77 11 5 3 2 1 67 28 12 9 5\n\nPercentage (%) 57.4 4.2 5.3 81.9 11.7 5.3 3.2 2.1 1.1 71.3 29.8 12.8 9.6 5.3\n\nhttps://doi.org/10.1371/journal.pone.0264206.t003\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n8 / 14\n\nPLOS ONE\n\nIntegrating tuberculosis screening into antenatal visits\nOriginal Midwife Response:\u201cA force de tousser la femme est fatigu\u00e9e et son b\u00e9b\u00e9 peut tomber malade\u201d\nEnglish Translation: \u201cCoughing makes the woman tired and her baby may get sick\u201d Original Midwife Response: \u201cJe ne connais pas les sp\u00e9cificit\u00e9s de la tuberculose chez une femme enceinte mais la tuberculose doit \u00eatre plus grave avec une grossesse\u201d English Translation: \u201cI don\u2019t know the specifics of tuberculosis in a pregnant woman, but tuberculosis must be more serious with a pregnancy\u201d The airborne mode of transmission was known by all respondents, transplacental transmission was known only by midwives. Transmission by digestive tract was less mentioned by both midwives and pregnant women. Original Participant Response: \u201cLa tuberculose s\u2019attrape dans l\u2019air qui est respir\u00e9\u201d English Translation: \u201cTB is caught in the air you breathe\u201d Original Midwife Response: \u201cUne maman peut donner la tuberculose \u00e0 son b\u00e9b\u00e9 dans le ventre\u201d English Translation: \u201cA mother can give tuberculosis to her baby in the womb\u201d All pregnant women interviewed supported TB screening during routine antenatal care visits. All midwives also found routine TB screening during antenatal care visits like an opportunity to detect more TB cases, but the additional workload was also mentioned as a potential barrier and therefore a factor to be taken into consideration if TB screening was included in ANC routine care. Original Participant Response: \u201cSi on peut vite d\u00e9pister la tuberculose chez la maman son b\u00e9b\u00e9 sera plus en s\u00e9curit\u00e9\u201d English Translation: \u201cIf we can quickly detect tuberculosis in the mother, her baby will be safer\u201d Original Midwife Response: \u201cLa recherche de la tuberculose pendant la grossesse est une bonne initiative mais faut pas que cela soit encore une charge additionnelle pour nous les sages -femmes, c\u2019est pourquoi il faudra penser l\u2019int\u00e9grer directement sur la carte maternelle de suivi des CPN\u201d English Translation: \u201cThe research of tuberculosis during pregnancy is a good initiative but it should not be an additional burden for us midwives, that\u2019s why we should think about integrating it directly on the maternal card of ANC monitoring\u201d\nCost analysis\nTable 4 reports the individual component costs of the TB screening intervention. Across all 4,070 pregnant women screened, the overall cost of the intervention was $4542 USD. Microbiological analysis with Xpert MTB/RIF accounted for nearly 40% of all costs and training of midwives accounted for nearly 30%. Overall, we calculated a cost of $1.12 USD per pregnant woman screened, $48.32 USD per pregnant woman identified with cough of at least 2 weeks, and $2271 USD per case of TB diagnosed.\nDiscussion\nIn this study of 4,070 pregnant women, the prevalence of cough of at least 2 weeks was 2.3%. Among the 94 pregnant women with cough, two had bacteriologically confirmed TB for an overall prevalence of 49 per 100,000 per screening visit. The cost per pregnant woman screened was inexpensive and both pregnant women and their midwives found screening to be acceptable.\nIn general, symptom screening is less sensitive in pregnancy and this can lead to an underestimation of presumptive TB cases [4, 12]. In our study, only women with cough of at least 2\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n9 / 14\n\nPLOS ONE\n\nIntegrating tuberculosis screening into antenatal visits\n\nTable 4. Cost components for the cost assessment.\n\nComponent Costs\nAdministrative Costs Training Telephone monitoring\nScreening Costs Verbal Screening for Cough\nAdditional Screening and Sputum Collection from Pregnant Women with Cough of at least 2 weeks\nTransport of sputum to lab (per sample)\nXpert MTB/RIF (per sample) Cartridge and shipping costs (per sample)\nTechnician time (per sample)\nCapital Cost of Equipment (per sample)\nEquipment Maintenance Costs (per sample)\nLaboratory Overhead Costs (per sample)\n\nUnit Cost (2019 USD)\n\nDescription\n\n$1315\n\nCost used to train midwives at each site\n\n$336\n\nPhone fees paid $28 USD per month over the\n\n12-month study for communication with each site\n\n$0.12 Midwife salary ($0.12 USD/min) multiplied by the time to ask about symptoms (1 minute per person)\n\n$1.80 Midwife salary ($0.12 USD/min) multiplied by the time for additional screening, sputum collection, and sample registering (15 minutes per person).\n\n$4.70\n\nTransportation costs per roundtrip pickup of samples from sites.\n\n$19.06\n\nSum of components below\n\n$11.26\n\nGlobal Drug Facility cost of Xpert MTB/RIF cartridge\n\n$0.59\n\nTechnician Salary ($0.09 USD/min) multiplied by time to accession, prepare, analyze, and report each sample (6.5 min per sample, when 16 prepared at once)\n\n$1.75 Annuitized cost of equipment (at a rate of 3%) for an expected useful life of 5 years\n\n$0.10\n\nAnnual maintenance costs based on national tuberculosis programme contract\n\n$5.36\ufffd\ufffd\n\nLaboratory overhead costs estimated based on laboratory manpower and overhead expenditures\nfrom the hospital\n\n\u00b6The CNHU-PPC has 2 Xpert MTB/RIF machines of 16 cartridges (144,000 USD). The lifetime of one machine is 5 years and performs 17,923 analyses per year. The $1.75 USD represents the cost adjusted over the lifetime of the machine per sample, annuitized at a 3% rate. \ufffd\ufffdThe cost of overhead associated with Xpert MTB/RIF, based on the overhead of the hospital and an overhead distribution derived from the number of people employed in the laboratory dedicated to Xpert MTB/RIF compared to the rest of the hospital (6/113).\n\nhttps://doi.org/10.1371/journal.pone.0264206.t004\n\nweeks gave sputum for Xpert MTB/RIF. The rate of cough found in our study was lower than that reported in Zambia which found 23% of pregnant women with cough [18], as well as in Kenya which found a prevalence of any symptoms of TB to be 8% among pregnant women living with HIV and 5% among HIV-negative pregnant women [14]. Differences may be explained by epidemiologic contexts and different frequencies of respiratory illness and therefore cough in these populations. However, our findings are similar to those found in BurkinaFaso (3%)\u2014a country with a similar prevalence of HIV to Benin\u2014as well as Pakistan; both studies were performed in the context of systematic screening for tuberculosis during antenatal consultation [10, 21].\nThe prevalence of pulmonary TB per visit among pregnant women was 49 per 100,000. Comparatively, TB incidence among a similarly aged female population in Benin from 2016\u2013 2018 was 28 per 100,000 [15, 22]. Other analyses of TB prevalence have found prevalence may be higher among pregnant women compared to the general population [6]. At the TB rates estimated in this study and an estimated 370,000 pregnancies occurring annually in Benin\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n10 / 14\n\nPLOS ONE\n\nIntegrating tuberculosis screening into antenatal visits\n[17], a screening program for all pregnant women would cost $414,000 USD and detect about 182 cases of TB.\nThe diagnosis of TB in both participants was made in the first trimester. Indeed, women in early pregnancy are twice as likely to develop TB as non-pregnant women [23], however risk remains throughout pregnancy and even in the postpartum period [4, 24]. Clinically, both patients had similar symptoms, which were like those found in HIV-negative pregnant women in another study [14]. In terms of treatment, only one of the two participants started antituberculosis drugs early, with good adherence. Sobhy et al [25] have shown that outcomes among pregnant women tend to improve the earlier treatment starts. In their study, women treated early in the first trimester of pregnancy had no premature births, low birth weight babies, or perinatal deaths, whereas in women treated in the second or third trimester of pregnancy, 33% of infants were premature, 61% of infants had low birth weight, and 23% had perinatal deaths [25]. Indeed, the participant in our study who started treatment later in pregnancy experienced perinatal death, though it is uncertain whether this death was a result of late initiation of effective TB therapy.\nMidwives\u2019 and pregnant women\u2019s knowledge of TB was good for routine TB symptoms but low about TB complications in pregnancy. This suggests awareness and information sessions may be required to improve knowledge on TB complications. The feasibility and acceptability of implementing TB screening was considered good among both pregnant women and their midwives, though it was stressed it should not be an additional burden. In 2018, the NTP of Benin offered free access to Xpert MTB/RIF as a first-line diagnostic for key populations, including pregnant women, which supports the feasibility and acceptability of such a screening programme.\nWe found the cost of screening pregnant women for TB was low, however within our population the yield was suboptimal. Early screening and detection of TB with prompt treatment initiation, helps minimize the adverse consequences of TB in the mother and infant [4, 5] and prevent transmission. The benefits may extend to the midwives and the health system because it may reduce strain associated with caring for pregnant women in emergency situations. The cost per TB case detected was high compared to other case-finding interventions in Benin. For example, an intervention implemented through TB REACH, in which the overall objective was to increase the detection rate of TB in populations with limited access to TB services through outreach clinics, estimated a cost of ~$610 USD per TB case diagnosed [26]. However, to formally compare different interventions, a cost-utility analysis would need to be performed to determine if any intervention is cost-effective and which should be prioritized. For pregnant women attending ANC appointments, questioning about cough can be done quickly and easily. In other populations at increased risk of TB\u2014such as people living with HIV\u2014taking advantage of routine care appointments to screen for TB is already done. Similarly integrating such screening for pregnant women could avert morbidity and mortality, with minimal additional effort.\nStrengths and limitations\nThe main strengths of this study were the diverse nature of women attending antenatal care and pragmatic implementation of the intervention reflecting integration into routine care. However, this study also has limitations. The method to identify women for further microbiologic evaluation was only cough of at least two weeks [27]. This is a less sensitive approach to TB detection than considering all possible TB symptoms for further evaluation\u2014which already has lower sensitivity in pregnant women\u2014however, was implemented because this is the policy of the NTP. We speculate our overall yield may have increased had we performed Xpert\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n11 / 14\n\nPLOS ONE\n\nIntegrating tuberculosis screening into antenatal visits\nMTB/RIF on all women with any TB symptom, but would result in an increase in the cost per pregnant woman screened. We found less symptomatic pregnant women than initially expected, which underpowered our estimates. This could be due to the lower prevalence of both cough and TB in Benin compared to other settings where this screening strategy has been evaluated. The absence of radiography in HIV-positive women is also a limitation. Indeed, on the basis of the symptomatology of productive cough, very few HIV-positive women were further evaluated for TB. Our study was cross-sectional in nature and only accounted for one screening event per pregnant woman. There is a possibility that women may have developed symptoms later and had TB or that women presented with subclinical TB that would be missed by TB symptom screening. We only used Xpert MTB/RIF to diagnose TB. While this is a rapid, molecular-based diagnostic test, it is less sensitive than culture\u2014particularly among smear-negative cases\u2014so some cases of TB may have been missed [28].\nConclusion\nIn this study, we found about 1 in 40 pregnant women had prolonged cough and the prevalence of TB per screening visit was 49 per 100,000. TB screening in ANC seems to be acceptable to both midwives and pregnant women, with a low cost per woman screened. In settings where access to care is limited, ANC visits may be a useful opportunity to perform TB screening. Although the yield of the intervention was low, it must be weighed against the increased risks of adverse maternal and fetal outcomes if a pregnant woman develops TB and against other case finding interventions.\nSupporting information\nS1 Dataset. (XLSX)\nAcknowledgments\nOur thanks go to the midwives, the pregnant women, participating sites, and the NTP of Benin.\nDisclaimer: CSM is currently a staff member of the World Health Organization; the author alone is responsible for the views expressed in this publication and they do not necessarily represent the decisions, policy or views of the WHO.\nAuthor Contributions\nConceptualization: M\u00eanonli Adjobimey, Serge Ade, Prudence Wachinou, Wilfried Bekou, Narcisse Toundoh, Omer Adjibode, Geoffroy Attikpa, Gildas Agodokpessi, Dissou Affolabi, Corinne S. Merle.\nData curation: M\u00eanonli Adjobimey, Marius Esse, Lydia Yaha, Jonathon R. Campbell, Narcisse Toundoh.\nFormal analysis: M\u00eanonli Adjobimey, Marius Esse, Wilfried Bekou, Jonathon R. Campbell. Funding acquisition: Corinne S. Merle.\nInvestigation: M\u00eanonli Adjobimey. Methodology: M\u00eanonli Adjobimey. Project administration: M\u00eanonli Adjobimey.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n12 / 14\n\nPLOS ONE\n\nIntegrating tuberculosis screening into antenatal visits\nResources: M\u00eanonli Adjobimey.\nSupervision: M\u00eanonli Adjobimey, Corinne S. Merle.\nWriting \u2013 original draft: M\u00eanonli Adjobimey, Jonathon R. Campbell.\nWriting \u2013 review & editing: M\u00eanonli Adjobimey, Serge Ade, Prudence Wachinou, Marius Esse, Lydia Yaha, Wilfried Bekou, Jonathon R. Campbell, Narcisse Toundoh, Omer Adjibode, Geoffroy Attikpa, Gildas Agodokpessi, Dissou Affolabi, Corinne S. Merle.\nReferences\n1. World Health Organization. Global tuberculosis control-epidemiology, strategy, financing. 2018. 2. Union Internationale Contre la Tuberculose et les Maladies Respiratoires (L\u2019Union). Prise en charge de\nla tuberculose, Guide des e\u00b4 le\u00b4 ments essentiels pour une bonne pratique. Sixi\u00e8me e\u00b4d. 2010. 3. WHO. Global Tuberculosis Report [Internet]. 2020. https://www.who.int/publications/i/item/\n9789240013131 4. Mathad JS, Gupta A. Tuberculosis in Pregnant and Postpartum Women: Epidemiology, Management,\nand Research Gaps. 2012; 55(11):1532\u201349. 5. Loto O, Awowole I. Tuberculosis in Pregnancy: A Review. J Pregnancy. 2012;1\u20137. https://doi.org/10.\n1155/2012/379271 PMID: 22132339 6. Mnyani CN M J. Tuberculosis in pregnancy. BJOG. 2011; 118:226\u2013231. https://doi.org/10.1111/j.1471-\n0528.2010.02771.x PMID: 21083862 7. World Health Organization. Systematic screening for active tuberculosis: principles and recommenda-\ntions. Geneva, Switzerland; 2013. 8. Zenner D, Kruijshaar ME, Andrews N, Abubakar I. Risk of Tuberculosis in Pregnancy A National, Pri-\nmary Care\u2014based Cohort and Self-controlled Case Series Study. Am J Respir Crit Care Med. 2012; 185(7):779\u201384. https://doi.org/10.1164/rccm.201106-1083OC PMID: 22161161 9. Schwartz N, Wagner SA, Keeler SM, Mierlak J, Seubert DE C A. Universal Tuberculosis Screening in Pregnancy. Am J Perinatol. 2009; 26(6):447\u201351. https://doi.org/10.1055/s-0029-1214244 PMID: 19263332 10. Feroz Ali R, Arif Siddiqi D, Malik AA, Taighoon Shah M, Javed Khan A, Hussain H, et al. Integrating tuberculosis screening into antenatal visits to improve tuberculosis diagnosis and care: Results from a pilot project in Pakistan. Int J Infect Dis. 2021; 108:391\u2013396. https://doi.org/10.1016/j.ijid.2021.05.072 PMID: 34087487 11. Pasipamire M, Broughton E, Mkhontfo M, Maphalala G, Simelane-Vilane B, Haumba S. Detecting tuberculosis in pregnant and postpartum women in Eswatini. Afr J Lab Med. 2020; 9(1):1\u20139. https://doi. org/10.4102/ajlm.v9i1.837 PMID: 32832404 12. Hoffmann CJ, Variava E, Rakgokong M, Masonoke K, Van Der Watt M. High Prevalence of Pulmonary Tuberculosis but Low Sensitivity of Symptom Screening among HIV-Infected Pregnant Women in South Africa. PLoS One. 2013; 8(4):e-62211:1\u20135. https://doi.org/10.1371/journal.pone.0062211 PMID: 23614037 13. World Health Organization. WHO consolidated guidelines on tuberculosis: tuberculosis preventive treatment ANNEX 2 GRADE summary of evidence tables (for new recommendations in 2018 & 2019 guidelines updates). 2020; http://apps.who.int/bookorders. 14. Kosgei RJ, Szkwarko D, Callens S, Gichangi P, Temmerman M, Kihara A, et al. Screening for tuberculosis in pregnancy: do we need more than a symptom screen? Experience from western Kenya. 2013; I (4):294\u20138. https://doi.org/10.5588/pha.13.0073 PMID: 26393049 15. Programme National contre la Tuberculose. Minist\u00e8re de la sante\u00b4 du Be\u00b4 nin. Direction Nationale de la Sante\u00b4 Publique. Rapports annuels sur la tuberculose au Be\u00b4nin 2016,2017,2018. 16. CNLS. Rapport de suivi de la de\u00b4claration de politique sur le Vih / Sida au Be\u00b4 nin 2016. 2016. 17. Sante\u00b4 M de la. Annuaire Statistique. Minist\u00e8re. 2008. 48\u201352 p. 18. Kancheya N, Luhanga D, Harris JB, Morse J, Kapata N, Bweupe M, et al. Integrating active tuberculosis case finding in antenatal services in Zambia. 2014; 18(January):1466\u201372. 19. World Bank. Official exchange rate\u2014United States, Benin. 2020. 20. World Bank. PPP conversion factor, GDP\u2014United States, Benin. 2020. 21. Sulis G, Pai M. Tuberculosis in Pregnancy: A Treacherous Yet Neglected Issue. J Obstet Gynaecol Canada. 2018; 40(8):1003\u20135. https://doi.org/10.1016/j.jogc.2018.04.041 PMID: 30103870\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n13 / 14\n\nPLOS ONE\n\nIntegrating tuberculosis screening into antenatal visits\n22. worldbank. https://data.worldbank.org/indicator/SP.POP.1564.TO.ZS?locations=BJ [Internet]. https:// data.worldbank.org/indicator/SP.POP.1564.MA.IN?end=2019&start=1960&view=chart\n23. Turnbull ER, Kancheya NG, Harris JB, Topp SM, Henostroza G, Reid SE. A Model of Tuberculosis Screening for Pregnant Women in Resource-Limited Settings Using Xpert MTB / RIF. J of Pregnancy Vol. 2012;(ID 565049):5 pages.\n24. Sulis G, Gnanou S, Roggi A, Konseimbo A, Giorgetti PF, Castelli F, et al. Recherche active des cas de tuberculose parmi les femmes enceintes: un projet pilote au Burkina Faso. 2016; 20(10):1306\u20138.\n25. Sobhy S; Zamora J; Kunst H. Maternal and perinatal mortality and morbidity associated with tuberculosis during pregnancy and the postpartum period: a systematic review and meta-analysis. BJOG. 2017; 124:727\u201333. https://doi.org/10.1111/1471-0528.14408 PMID: 27862893\n26. Programme National contre la Tuberculose. Rapport TB Reach. 2011. 27. WHO. Consolidated Guidelines on Tuberculosis. Module 1, Prevention: Tuberculosis Preventive Treat-\nment. Geneva; 2020. 28. Li S, Liu B, Peng M, Chen M, Yin W, Tang H, et al. Diagnostic accuracy of Xpert MTB / RIF for tuberculo-\nsis detection in different regions with different endemic burden: A systematic review and meta-analysis. 2017;1\u201313.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0264206 February 22, 2022\n\n14 / 14\n\n\n",
    "authors": [
      "M\u00eanonli Adjobimey",
      "Serge Ade",
      "Prudence Wachinou",
      "Marius Esse",
      "Lydia Yaha",
      "Wilfried Bekou",
      "Jonathon R. Campbell",
      "Narcisse Toundoh",
      "Omer Adjibode",
      "Geoffroy Attikpa",
      "Gildas Agodokpessi",
      "Dissou Affolabi",
      "Corinne S. Merle"
    ],
    "doi": "10.1371/journal.pone.0264206",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://dx.plos.org/10.1371/journal.pone.0264206"
  },
  {
    "key": "PY65REKC",
    "title": "Cost-effectiveness of GeneXpert and LED-FM for diagnosis of pulmonary tuberculosis: A systematic review",
    "abstract": "",
    "full_text": "a1111111111 a1111111111 a1111111111 a1111111111 a1111111111\n\nRESEARCH ARTICLE\nCost-effectiveness of GeneXpert and LED-FM for diagnosis of pulmonary tuberculosis: A systematic review\nKaruna D. SagiliID1*, Malaisamy Muniyandi2, Kayzad Soli Nilgiriwala3, Kalpita S. Shringarpure4, Srinath Satyanarayana1, Richard Kirubakaran5, Sarabjit S. Chadha1, Prathap Tharyan5\n1 International Union against Tuberculosis and Lung Disease, South East Asia Regional office, New Delhi, India, 2 National Institute for Research in Tuberculosis, ICMR, Chennai, India, 3 Tuberculosis Division, The Foundation for Medical Research, Mumbai, India, 4 Department of Preventive and Social Medicine, Medical College Baroda, Baroda, India, 5 Prof BV Moses Centre for Evidence- Informed Health Care, Christian Medical College, Vellore, India\n* drkarunas@gmail.com\n\nAbstract\n\nOPEN ACCESS\nCitation: Sagili KD, Muniyandi M, Nilgiriwala KS, Shringarpure KS, Satyanarayana S, Kirubakaran R, et al. (2018) Cost-effectiveness of GeneXpert and LED-FM for diagnosis of pulmonary tuberculosis: A systematic review. PLoS ONE 13(10): e0205233. https://doi.org/10.1371/journal.pone.0205233\nEditor: Vishnu Chaturvedi, Wadsworth Center, UNITED STATES\nReceived: November 3, 2017\nAccepted: September 23, 2018\nPublished: October 29, 2018\nCopyright: \u00a9 2018 Sagili et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.\nData Availability Statement: All relevant data are within the paper and its Supporting Information files.\nFunding: This work is supported by USAID (https://www.usaid.gov); Award No: AID-GHN-A00-08-00004. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\nCompeting interests: The authors have declared that no competing interests exist.\n\nBackground\nEarly and accurate diagnosis of tuberculosis is a priority for TB programs globally to initiate treatment early and improve treatment outcomes. Currently, Ziehl\u2013Neelsen (ZN) stainbased microscopy, GeneXpert and Light Emitting Diode-Fluorescence Microscopy (LEDFM) are used for diagnosing pulmonary drug sensitive tuberculosis. Published evidence synthesising the cost-effectiveness of these diagnostic tools is scarce.\nMethodology\nPubMed, EMBASE and Cost-effectiveness analysis registry were searched for studies that reported on the cost-effectiveness of GeneXpert and LED-FM, compared to ZN microscopy for diagnosing pulmonary TB. Risk of bias was assessed independently by four authors using the Consensus Health Economic Criteria (CHEC) extended checklist. The data variables included the study settings, population, type of intervention, type of comparator, year of study, duration of study, type of study design, costs for the test and the comparator and effectiveness indicators. Incremental cost-effectiveness ratio (ICER) was used for assessing the relative cost-effectiveness in this review.\nResults\nOf the 496 studies identified by the search, thirteen studies were included after removing duplicates and studies that did not fulfil inclusion criteria. Four studies compared LED-FM with ZN and nine studies compared GeneXpert with ZN. Three studies used patient cohorts and eight were modelling studies with hypothetical cohorts used to evaluate cost-effectiveness. All these studies were conducted from a health system perspective, with four studies utilising cost utility analysis. There were considerable variations in costing parameters and\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n1 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\neffectiveness indicators that precluded meta-analysis. The key findings from the included studies suggest that LED-FM and GeneXpert may be cost effective for pulmonary TB diagnosis from a health system perspective.\nConclusion\nOur review identifies a consistent trend of the cost effectiveness of LED-FM and GeneXpert for pulmonary TB diagnosis in different countries with diverse context of socio-economic condition, HIV burden and geographical distribution. However, all the studies used different parameters to estimate the impact of these tools and this underscores the need for improving the methodological issues related to the conduct and reporting of cost-effectiveness studies.\n\nIntroduction\nTuberculosis (TB) remains a leading cause of death worldwide. Globally, 10.4 million new cases were reported by WHO in 2016 [1]. India is amongst the six countries that accounted for 60% of the new cases. The Sustainable Development Goals (SDGs) and the End TB Strategy aim to end the global TB epidemic and reduce TB deaths by 90% and TB incidence by 80% in 2030 [2]. Though TB treatment averted 49 million deaths globally between 2000 and 2015, diagnostic gaps persist [1]. The WHO 2015 report estimates that about 37% of the cases were undiagnosed or not reported [3]. The potential transmission through people with undiagnosed TB to their contacts poses a serious public health problem. Hence, early and accurate diagnosis of TB is now the top priority of national TB programs globally. Delayed diagnosis contributes to continued transmission, poor health outcomes and distress to the patient and the family [4]. Early diagnosis is expected to lead to early treatment initiation and hence better outcomes. Improved diagnostic tools may facilitate early diagnosis and reduce the direct costs of the diagnostic burden on patients and family [5,6]. Currently, Ziehl\u2013Neelsen (ZN) stain-based microscopy, GeneXpert and Light Emitting Diode-Fluorescence Microscopy (LED-FM) are widely used diagnostic tools for drug-sensitive pulmonary tuberculosis by National TB programmes in high burden countries.\nCurrent diagnostic tools\nSputum microscopy has been the main tool for TB diagnosis for nearly a century; followed by sputum culture, which is considered as the gold standard. However, these two tools have their inherent limitations viz. low sensitivity for microscopy and prolonged duration to obtain culture test results. ZN stain-based smear microscopy, using Carbol-fuchsin, Ziehl-Neelsen or Kinyoun acid-fast stains with an artificial light source or reflected sunlight, is widely used to detect acid fast bacillus (AFB). However, it has variable sensitivity (78%; 95% CI 32% to 89%) though it has higher specificity (98%; 95% CI 85% to 100%) for the diagnosis of pulmonary sputum smear-positive TB [7]. Sputum smear microscopy has been relied upon as a primary diagnostic tool in resource limited settings as it is cheaper with minimal required biosafety standards [3]. Thus, it continues to be the routine diagnostic method for pulmonary TB in countries like India [8]. It is simple and inexpensive, and at the same time allows rapid detection of the most infectious cases of pulmonary TB. It can be used for TB diagnosis at the\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n2 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\nperipheral level as well [9]. Though highly specific [8], it is limited by its low sensitivity (further reduced in patients with extra-pulmonary TB, children and HIV/TB co-infected patients).\nGeneXpert (Cepheid, Sunnyvale, USA) is a newer molecular test that detects DNA of TB bacteria in sputum samples (pooled sensitivity\u2013 98%; 95% CI 85%-92% and specificity 99%; 95% CI 98%-99%) and also detects resistance to Rifampicin within two hours. This simplifies molecular testing with fully integrated and automated sample preparation, compared to the procedure and time required for amplification and detection by real-time PCR [7, 10]. The cost of GeneXpert per cartridge is US$17 universally except for some high TB burden and low income countries which receive a discounted cost of about US$10 [11]. It was reported that implementation of GeneXpert would result in a three-fold increase in the diagnosis of patients with drug-resistant TB and a two-fold increase in the number of HIV-associated TB cases [12]. It is also useful for diagnosing smear negative specimens considering the lack of accuracy of smear microscopy. While testing single sputum samples in a prospective study of people suspected to have TB, GeneXpert detected 98% to 100% of those with sputum smear-positive disease and 57% to 83% of those with smear negative disease [7]. Countries like South Africa are offering this test upfront for TB diagnosis, and India is also scaling up its GeneXpert services across the country.\nAround the same time as the introduction of GeneXpert, evidence on the efficacy of the LED-FM was provided by the WHO in 2009. Sensitivity of LED-FM is comparable to that of conventional fluorescence microscopy and it surpasses that of conventional Ziehl\u2013Neelsen microscopy by an average 10%. Conventional fluorescence microscopy replacement with LED-FM has been recommended by WHO [8, 9]. A retrospective cohort study on cost utility of LED-FM showed it to be a cost effective intervention in diagnosis of pulmonary TB in India with an Incremental Cost-effectiveness Ratio (ICER) of US$14.64 per disability-adjusted lifeyear (DALY) averted [13].\nExpenditure for TB program in India was 6398.6 million rupees (US$ 98.47 million) in 2015\u201316 [14]. Low and middle-income countries fell short of almost US$ 2 billion of the US$ 8.3 billion needed in 2016, which was required to combat the TB epidemic [1]. This amount excludes the funding required for research and development. Thus, \u201cGlobal actions and investments fall far short of those needed to end the global TB epidemic\u201d [14].\nThere are several direct and indirect costs entailed to delayed diagnosis and treatment of TB, which can be averted with early and prompt diagnosis [14, 15]. Costs are usually described in monetary units, while effects can be measured in terms of health status or another outcome of interest. The incremental cost-effectiveness ratio (ICER) summarizes the additional cost per unit of health benefit gained in switching from one medical intervention to another [16]. A common application of the ICER is in cost-utility analysis, in which case the ICER is synonymous with the cost per quality-adjusted life year (QALY) gained, where\nICER \u00bc \u00f0Cost of new diagnostic \u00c0 Cost of standard care\u00de = \u00f0Effectiveness of new diagnostic \u00c0 Effectiveness of standard care\u00de:\nConsidering the challenges in TB diagnosis and the limited resource, there is a need of a cost-effective tool as a priority that is highly sensitive and specific to be used in resource poor settings. Though there are recent systematic reviews on diagnostic accuracy of newer tools such as GeneXpert, these reviews do not report incremental costs and hence have limitation in guiding decision makers. A test having a good value doesn\u2019t always mean it is affordable or feasible [15, 17]. It is important for the national TB programs to know what additional health unit benefits would accrue, if any, by changing a diagnostic tool and what additional costs this would incur. In the absence of any systematic reviews reporting on the incremental cost-effectiveness of the newer diagnostic tools, we undertook a systematic review to evaluate the\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n3 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\nincremental cost-effectiveness of GeneXpert and LED-FM in comparison with ZN microscopy for the diagnosis of smear-positive pulmonary TB.\nMethods\nThis systematic review was conducted following the PRISMA guidelines [18] (S1 Table). The review protocol is registered at the Prospero registry (Registration No. CRD42016043333) [19]. The objective was to compare the incremental cost-effectiveness of GeneXpert and LED-FM with ZN smear microscopy in the diagnosis of smear-positive pulmonary TB. Though we had initially planned to include Chest X-ray as one of the diagnostic tests evaluated, we excluded it for this review due to the lack of studies providing data comparing costeffectiveness of Chest X-ray with ZN smear microscopy. Below is the PICO question for this review:\nP\u2014(Participants/population): Presumptive pulmonary TB patients undergoing diagnostic evaluation\nI\u2013(Interventions): GeneXpert, LED FM microscopy C\u2013(Comparator): ZN microscopy O\u2013(Outcome measures): To find out the incremental cost-effectiveness ratio (ICER) for GeneXpert and LED FM in comparison to ZN sputum microscopy from a health system perspective.\nSelection criteria\nTypes of studies. All types of studies (cross-sectional, observational, cohort, modelling, economic evaluation) that reported on cost-effectiveness of ZN microscopy, GeneXpert and LED-FM for pulmonary TB diagnosis were included.\nStudy population. Any person presumed to have pulmonary TB who was undergoing diagnostic evaluation irrespective of co-morbidities like infection with the Human Immunodeficiency Virus (HIV).\nDiagnostic tests. Studies comparing GeneXpert with ZN microscopy and LED-FM in comparison to ZN microscopy for the diagnosis of pulmonary TB, with data provided for costs as well as for effectiveness. Studies reporting cost-effectiveness of GeneXpert or LED-FM but using a comparator other than ZN microscopy were excluded. Studies reporting only costs and not reporting an effectiveness indicator were also excluded.\nOutcome measures. The primary outcome measure was incremental cost-effectiveness ratio (ICER) for GeneXpert and LED-FM compared to ZN microscopy. The secondary outcomes were additional case detection, cure rate, and time to initiate treatment post-diagnosis. The ICER [20] is an informative measure generated from economic/cost analysis and represents the ratio of the difference in cost between two health interventions to the difference in outcomes between the two interventions. Since the ICER summarizes the additional cost per unit of additional health benefit gained in switching from one health intervention to another, it serves as an important measure to guide decisions about allocating scarce resources across competing medical interventions.\nSearch strategies\nWe searched PubMed, EMBASE and Cost-effectiveness analysis registry [21] using the search strategies detailed in S2 Table. We also searched the Cochrane database [22]. The searches were conducted in April 2017, and finalised on 24th April 2017. The search has been updated till July 2018.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n4 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\nSelection of studies\nThe abstracts for all papers retrieved by the search that were considered relevant to this review were uploaded in the Rayyan software [23] and screened for duplicates. After removing duplicates, the remaining abstracts were screened independently for relevance by four authors (KDS, MM, KSN, and KSS). Conflicts were resolved through discussions among the four investigators. Full texts of articles identified as relevant were obtained. When full texts of studies mentioned the cost-effectiveness as a key objective, but did not report an effectiveness indicator, they were excluded.\nData extraction\nData from the included studies were extracted into a data extraction form independently by MM and KSN. The data variables included the study settings, population, type of intervention, type of comparator, year of study, duration of study, type of study design, costs for the test and the comparator, effectiveness indicators and others. A sample extraction form is given in the supplementary material (S1). Wherever the key data was missing, we contacted the authors; however, there was no response from the authors. In case of disagreements, it was discussed with KDS and KSS and extraction was completed after obtaining consensus.\nRisk of bias assessment. MM and KSN assessed the risk of bias for each included study using the Consensus Health Economic Criteria (CHEC) extended checklist [24]. The checklist consists of 20 items with positive responses scored 1 and negative responses scored 0. The total score for each item was summed and converted to a percentage with the range of scores ranging from zero to 100. The total CHEC score for each study was categorized into four grades: low, moderate, good and excellent using cut-off value of \ufffd50, 51\u201375, 76\u201395 and >95, respectively. Higher scores denote lower risk of bias.\nResults Study selection\nFig 1 depicts the study selection process. The search yielded 497 studies that had reportedly assessed the cost-effectiveness of GeneXpert and LED-FM. Of these 384 studies were shortlisted after excluding 112 duplicates. After the review of abstracts, 67 studies were retained for evaluation of full papers. Thirty-four studies were further excluded since they did not have a ZN smear microscopy comparator [S3 Table]. Of the remaining 33, twenty studies were excluded due to lack of effectiveness data [S4 Table]. Finally this review included 13 studies from which data were extracted; four of the included studies compared LED-FM with ZN [13, 25, 26, 27] and seven studies compared ZN with GeneXpert [28, 29, 30, 31, 32, 33, 34, 35, 36].\nCharacteristics of included studies\nOut of the 13 studies, seven were conducted in Africa [25, 28, 31, 32, 34, 35, 36] of which four were from South Africa [25, 31, 32, 36], one was a multi-centric study which included Botswana, Lesotho, Namibia, South Africa and Swaziland [34], one from Zambia [28] and one from Ethiopia [35] (Table 1). Four studies were conducted in Asia with one each from India, China, Hong Kong and Thailand [13, 26, 27, 30]. Two studies were from the Americas, one each from USA and Brazil [33, 29]. All the studies except for the one from USA were conducted in low and middle-income countries. Ten studies were conducted within the time period of 2011 to 2017 [13, 25, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36]. Seven studies were conducted in an urban or peri-urban setting [13, 25, 30, 31, 32, 33, 35], while others did not mention the study setting clearly (Table 1). Three studies used the real patient cohorts [25, 26, 27]\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n5 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\n\nFig 1. Flow diagram indicating the process of selecting the studies for this systematic review on cost-effectiveness of tolls to diagnose pulmonary TB.\nhttps://doi.org/10.1371/journal.pone.0205233.g001\nand eight used modelling studies with hypothetical cohorts to evaluate the cost-effectiveness of different diagnostic tools for pulmonary TB diagnostics.\nAll these studies were conducted from a health system perspective with seven studies utilising cost utility analysis. Four used Disability Adjusted Life Years (DALY) [13, 34, 35, 36], one [30] used Quality Adjusted Life Years (QALY) and one [31] used years of life saved (YLS) as indicators, all these being standard indicators for cost-effectiveness analysis. There were also studies that used other indicators like time duration per slide for diagnosis [25, 26, 27], additional cases diagnosed [29, 31], TB cases averted [28] and reduction in duration of\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n6 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\n\nTable 1. Characteristics of the 13 studies included in the review.\n\nSr. First No. Author,\nYear\n\nCountry Setting\n\nFunding Source\n\nType of Target Comorbidities Study Reporting Study Design Time Sensitivity\n\nEconomic Population\n\nPerspective of ICER\n\nHorizon analysis\n\nEvaluation\n\n(years)\n\nLED-FM vs ZN microscopy\n\n1 Whitelaw, South\n\n2011\n\nAfrica\n\nUrban European\n\nCEA\n\nCommission\n\n& Canadian\n\nInstitute of\n\nHealth\n\nResearch\n\nAdults\n\nHIV\n\nHealth\n\nNR\n\nsystem\n\nCrosssectional\n\n1 NR\n\n2 Kelly, 2015\n\nIndia\n\nUrban TB Reach Initiative\n\nCUA\n\nAdults\n\nNR\n\nHealth\n\nYes\n\nsystem\n\nCohort\n\n1 One-way PA\n\n3 Sohn,\n\nThailand NR USAID &\n\nCEA\n\nNR\n\nNR\n\n2009\n\nCDC\n\nHealth\n\nNR\n\nsystem\n\nCrosssectional\n\n0.25 NR\n\n4 Xia, 2013 China\n\nNR BMGF\n\nCEA\n\nNR\n\nNR\n\nHealth\n\nNo\n\nsystem\n\nCrosssectional\n\n1 NR\n\nGeneXpert vs ZN microscopy\n\n5 Mishra, Zambia NR NR 2012\n\nCEA\n\nNR\n\nHIV\n\nHealth\n\nYes\n\nsystem\n\nCohort\n\nNR NR\n\n6 Pinto,\n\nBrazil\n\nNR Bill &\n\nCEA\n\nNR\n\nHIV\n\n2016\n\nMelinda\n\nGates\n\nFoundation\n\nHealth\n\nYes\n\nsystem\n\nCohort\n\nNR Monte Carlo simulation\n\n7 You, 2015 Hongkong Urban No funding CUA\n\nAdults\n\nNR\n\nHealth\n\nYes\n\nsystem\n\nCohort\n\n10 Monte Carlo simulation\n\n8 Jha, 2016 South Africa\n\nUrban\n\nFrank &\n\nCEA\n\nKathleen Polk\n\nAssistant\n\nProfessorship\n\nin\n\nEpidemiology\n\nAdults\n\nNR\n\nHealth\n\nYes\n\nsystem\n\nModel\n\nNR One-way PA\n\n9 Andrews, South\n\nPeri- NatioNDl CUA\n\nNR\n\nHIV\n\n2012\n\nAfrica\n\nUrban Institute of\n\nGeneral\n\nMedical\n\nSciences\n\nHealth\n\nYes\n\nsystem\n\nCohort\n\nNR Two-way PA\n\n10 Millman, USA\n\nUrban American CBA\n\nNR\n\nNR\n\n2013\n\nLung\n\nAssociation,\n\nUCSF-GIVI\n\nCentre for\n\nAIDS\n\nResearch,\n\nNational\n\nInstitutes of\n\nHealth &\n\nNDtioNDl\n\nCentre for\n\nResearch\n\nResources\n\nHealth\n\nNo\n\nsystem\n\nCohort\n\n1 Monte Carlo simulation\n\n11 Menzies, Botswana, NR UNITAID & CUA\n\nNR\n\nHIV\n\n2012\n\nLesotho,\n\nMGH\n\nNamibia,\n\nProgram in\n\nSouth\n\nCancer\n\nAfrica &\n\nOutcome &\n\nSwaziland\n\nTraining\n\nHealth\n\nYes\n\nsystem\n\nModel\n\n10 Monte Carlo simulation\n\n(Continued )\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n7 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\n\nTable 1. (Continued)\n\nSr. First No. Author,\nYear\n12 Vassall, 2017\n\nCountry\nSouth Africa\n\n13 Tesfaye A, Ethiopia 2017\n\nSetting Funding Source\n\nNR Urban\n\nBill & Melinda Gates Foundation\nUSAID/TB CARE\n\nType of Target Comorbidities Study Reporting\n\nEconomic Population\n\nPerspective of ICER\n\nEvaluation\n\nCUA\n\nAdults\n\nHIV\n\nHealth\n\nYes\n\nsystem\n\nStudy Design\nclusterrandomised trial\n\nTime Horizon (years)\n1\n\nSensitivity analysis\nOne-way PA\n\nCUA\n\nNR\n\nHIV\n\nHealth\n\nYes\n\nsystem\n\nobservational 10 quantitative modeling\n\nOne-way PA\n\nNR\u2013Not reported; PA\u2013Probabilistic analysis\n\nhttps://doi.org/10.1371/journal.pone.0205233.t001\n\nhospitalisation as an effectiveness indictor [33]. Five studies mentioned the target population as adults and seven studies also included patients with HIV co-infection [13, 25, 30, 31, 35, 36]. Out of 13 studies ICER value was reported by nine studies [13, 28, 29, 30, 31, 32, 33, 34, 35, 36]. Nine studies mentioned their time horizon ranging from 3 months to ten years [13, 25, 26, 27, 30, 33, 34, 35, 36]. Eleven studies were funded by international agencies like Stop TB, USAID and DFID [13, 25, 26, 27, 29, 31, 32, 33, 34, 35, 36] and the remaining two did not mention about funding [28, 30].\n\nQuality of included studies\nTable 2 summarises the appraisal of reporting quality for each study using the Extended CHEC checklist. Of the 13 studies, seven studies were of moderate quality while five were of good quality, indicating lower risk of bias. One study was graded as low score however it was decided to include this study owing to less number of studies qualifying for review purpose. Overall, four studies fulfilled \ufffd80% of the 20 items as per the checklist [29, 31, 33, 34]. Two studies [31, 32] did not mention the time horizon over which costs and consequences were being evaluated.\nTwo studies did not clearly state the funding sources and conflict of interest [28, 30]. Out of 13 studies five studies did not include all costs components and these were not valued appropriately [13, 25, 28, 30, 32].\n\nIncremental Cost-Effectiveness of LED FM compared with ZN microscopy\nThe sample size in the four studies [13, 25, 26, 27] comparing LED FM and ZN microscopy ranged from 345 to 21450 for test and from 345 to 14,300 for comparator. One of the studies used decision tree modelling analysis [13], while the cost indicator for all the four studies was average cost per smear. The cost for LED-FM ranged from USD 0.31 to 1.97 and the cost for ZN ranged from USD 0.21 to 2.2. The effectiveness indicator used in three of the studies [25, 26, 27] was time per reading of one slide in minutes, which ranged from 1\u20132 minutes for LED-FM and 2.4\u20133.4 minutes for ZN microscopy. The ICER values for these studies were calculated in this review (Table 3). The effectiveness indicator used in one of the study [13] was DALYs, which was 27.45 for LED-FM and 40.84 for ZN microscopy and the ICER value was 14.64 (Table 3). The range of cost-effectiveness ratio observed maybe due to different study settings, populations and methodology used.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n8 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\n\nTable 2. Consensus Health Economic Criteria (CHEC) extended checklist for quality assessment of the included studies.\n\nSr.\n\nChecklist question\n\nNo.\n\n1 Is the study population clearly described?\n\n2 Are competing alternatives clearly described?\n\n3 Is a well-defined research question posed in answerable form?\n\n4 Is the economic study design appropriate to the stated objective?\n\n5 Are the structural assumptions and the validation methods of the model properly reported?\n\n6 Is the chosen time horizon appropriate in order to include relevant costs and consequences?\n\n7 Is the actual perspective chosen appropriate?\n\n8 Are all important and relevant costs for each alternative identified?\n\n9 Are all costs measured appropriately in physical units?\n\n10 Are costs valued appropriately?\n\n11 Are all important and relevant outcomes for each alternative identified?\n\n12 Are all outcomes measured appropriately?\n\n13 Are outcomes valued appropriately?\n\n14 Is an appropriate incremental analysis of costs and outcomes of alternatives performed?\n\n15 Are all future costs and outcomes discounted appropriately?\n\n16 Are all important variables, whose values are uncertain, appropriately subjected to sensitivity analysis?\n\n17 Do the conclusions follow from the data reported?\n\n18 Does the study discuss the generalizability of the results to other settings and patient/client groups?\n\n19 Does the article/report indicate that there is no potential conflict of interest of study researcher(s) and funder(s)?\n\n20 Are ethical and distributional issues discussed appropriately?\n\n% of Yes\n\nOverall Quality\n\nWhitelaw 2011 1 1 1\n\nKelly\ufffd 2015\n0 1 1\n\n1\n\n1\n\n0\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n0\n\n1\n\n0\n\n0\n\n0\n\n0\n\n0\n\n0\n\n0\n\n1\n\n1\n\n0\n\n1\n\n0\n\n0\n\n0\n\n1\n\nSohn 2009\n0 1 1\n1\n0\n1\n1 1\n1\n1 1\n0 1 0\n0\n0\n\nXia 2013\n0 1 1\n\nMishra Pinto 2012 2016\n\n0\n\n0\n\n1\n\n1\n\n1\n\n1\n\nYou 2015\n1 1 1\n\nJha 2016\n1 1 1\n\nAndrews 2012 0 1 1\n\nMillman 2013 1 1 1\n\nMenzies 2012 0 1 1\n\nVassall 2017\n1 1 1\n\nTesfaye 2017 0 1 1\n\nTotal (% of Yes)\n38 100 100\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n100\n\n0\n\n0\n\n1\n\n1\n\n0\n\n1\n\n1\n\n1\n\n1\n\n1\n\n62\n\n1\n\n0\n\n1\n\n1\n\n0\n\n0\n\n1\n\n1\n\n1\n\n1\n\n77\n\n0\n\n0\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n85\n\n1\n\n0\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n85\n\n1\n\n0\n\n1\n\n0\n\n1\n\n0\n\n1\n\n1\n\n1\n\n1\n\n69\n\n1\n\n0\n\n1\n\n0\n\n1\n\n0\n\n1\n\n1\n\n1\n\n1\n\n62\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n85\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n0\n\n69\n\n1\n\n0\n\n1\n\n0\n\n0\n\n0\n\n0\n\n0\n\n1\n\n0\n\n46\n\n0\n\n1\n\n1\n\n1\n\n1\n\n1\n\n0\n\n1\n\n1\n\n0\n\n62\n\n0\n\n0\n\n1\n\n1\n\n1\n\n1\n\n0\n\n1\n\n1\n\n1\n\n54\n\n0\n\n0\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n69\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n100\n\n0\n\n1\n\n1\n\n0\n\n0\n\n1\n\n0\n\n1\n\n1\n\n1\n\n1\n\n1\n\n1\n\n69\n\n1\n\n1\n\n0\n\n0\n\n0\n\n1\n\n1\n\n0\ufffd\n\n0\n\n0\ufffd\n\n0\ufffd\n\n1\n\n1\n\n46\n\n0\n\n1\n\n1\n\n1\n\n0\n\n1\n\n0\n\n1\n\n0\n\n1\n\n0\n\n0\n\n0\n\n46\n\n55\n\n65\n\n65\n\n60\n\n35\n\n95\n\n75\n\n80\n\n65\n\n80\n\nModerate Moderate Moderate Moderate \ufffd\ufffd Good Moderate Good Moderate Good\n\n80 Good\n\n95\n\n75\n\nGood Moderate\n\n\ufffd Conflict of interest present \ufffd\ufffd Not categorised due to lack of information\nhttps://doi.org/10.1371/journal.pone.0205233.t002\n\nIncremental cost-effectiveness of GeneXpert compared to ZN microscopy\nSample size in the seven studies comparing GeneXpert and ZN microscopy ranged from 1009 to 8,92,000 for test and comparator. Four studies [13, 28, 29, 30] used decision tree modelling analysis, one study used Cost Effectiveness of Preventing AIDS complications (CEPAC) model [32] and one study used dynamic compartmental modelling [34]. Six of the studies used average costs per sample as the cost indictor [24, 25, 26, 27, 28, 29] and one study used cost per case detected [28]. The average cost per sample for GeneXpert ranged from USD 14.45 to 218 and the cost for ZN ranged from USD 1.59 to 31. In one study, the average cost per case detected was USD 108.9 for GeneXpert and the cost for ZN was USD 75.74 [28]. These studies used different effectiveness indicators such as TB cases averted, additional case diagnosed, QALYs, DALYs, YLs and reduction in hospitalisation and ICER values were calculated accordingly (Table 2). Except in one study [28], sensitivity analysis was done using either Monte Carlo Simulation (4 studies [29, 30, 33, 34]), one way (one study, [31]) or two-way probabilistic analysis (one study) [32].\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n9 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\n\nTable 3. Description of cost-effectiveness analyses reported in the included studies.\n\nSr.\n\nFirst\n\nNo. Author,\n\nYear\n\nCountry\n\nLED-FM vs ZN microscopy\n\n1 Whitelaw, South Africa 2011\n\n2 Kelly, 2015 India\n\n3 Sohn, 2009 Thailand\n\n4 Xia, 2013 China\n\nGeneXpert vs ZN microscopy\n\n5 Mishra, 2012\n\nZambia\n\n6 Pinto, 2016 Brazil\n\n7 You, 2015 Hongkong\n\n8 Jha, 2016 South Africa\n\n9 Andrews, 2012\n\nSouth Africa\n\n10 Millman, USA 2013\n\n11 Menzies, 2012\n12 Vassall, 2017\n\nBotswana, Lesotho, Namibia, South Africa & Swaziland\nSouth Africa\n\n13 Tesfaye A, Ethiopia 2017\n\nEconomic Evaluation\nType\n\nSample size (Test)\n\nSample size (ZN)\n\nModel Type\n\nYear\n\nCost\n\nCost of Cost of Effectiveness Effectiveness- Effectiveness- ICER\n\nICER\n\nSensitivity\n\nCost Indicator Test ZN\n\nIndicator\n\nTest\n\nZN\n\nThreshold Analysis\n\nConclusion\n\nCEA CUA CEA CEA\n\n345\n\n345 NA\n\n21,450 14,300 Decision Tree\n\n30/day 30/day NA\n\n11,276 11,276 NA\n\n2009\u2013 Average cost 10 per smear\n2011\u2013 Average cost 12 per smear\n2007 Average cost per smear\n2013 Average cost per smear\n\n1.63\n\n2.1 Time per slide\n\n(min)\n\n0.31 0.21 DALYs\n\n1.03 1.16 Time per slide (min)\n\n1.97\n\n2.2 Time per slide\n\n(min)\n\n1.8 27.45\n1 2\n\n2.5 40.84 2.4 3.4\n\n0.67\ufffd 14.64 0.09\ufffd 0.16\ufffd\n\nNR\n\nNR\n\n1489 One-way PA\n\nNR\n\nNR\n\nNR\n\nNR\n\nLED-FM microscopy is cheaper\nLED-FM is cost effective at high load settings\nLED-FM is costeffective in resource limited settings\nLED-FM is costeffective in peripheral laboratories\n\nCEA CUA\nCUA\n\nNR\n\nNR Decision Tree\n\nNM Cost per case 108.9 75.74 TB cases\n\nNR\n\ndetected\n\naverted\n\nNR\n\nNR Decision Tree\n\n2014 Average cost 14.69 3.08 Additional case\n\n3.9\n\nper sample\n\ndiagnosed (%)\n\nNR\n\nNR Decision Tree\n\n2014 Average cost\n\n128\n\n7.5 QALYs\n\nNR\n\nper sample\n\nCEA CUA\n\n1,009\n\nNR Economic Model 2015 Average cost 14.45 1.59 Additional case\n\nNR\n\nper sample\n\ndiagnosed\n\nNR\n\nNR CEPAC\n\n2010 Average cost\n\n21.6\n\n4.6 Years of life\n\nNR\n\nper sample\n\nsaved (YLS)\n\nCBA\n\n1,358\n\n1,381 Decision Tree\n\n2011 Average cost\n\n218\n\n15 Reduction in\n\nNR\n\nper sample\n\nhospitalization\n\nCUA\n\n8,92,000 8,92,000 Dynamic\n\n2011 Average cost\n\n45\n\n31 DALYs\n\nNR\n\ncompartmental\n\nper sample\n\nmodel\n\nCUA\n\n2324\n\n2332 NA\n\n2012 Average cost 168.79 160.64 DALYs\n\nNR\n\nper\n\nparticipant\n\nCUA\n\n54000 113000 discrete-event\n\n2014 annualized\n\nNR\n\nNR DALYs\n\nNR\n\nsimulation\n\ncost per\n\nDALY\n\naverted\n\nNR\n\n252\n\nNR\n\nNR\n\n-\n\nNR\n\n643\n\n11,000 Monte\n\nSingle-sample\n\nCarlo\n\nGeneXpert testing\n\nsimulation can replace 2-sample\n\nsputum smear\n\nmicroscopy test\n\nNR\n\n99\n\n50,000 Monte\n\nSingle sample\n\nCarlo\n\nGeneXpert testing\n\nsimulation during initial\n\nassessment of\n\nhospitalized patients\n\nis highly cost-effective\n\nNR\n\n1,927\n\n2,000\n\nOne-way\n\nGeneXpert is likely to\n\nPA\n\nbe highly cost-\n\neffective where the\n\nlevel of empiric TB\n\ndiagnosis is low\n\nNR\n\n5,100\n\n21,300 Two-way\n\nTwo-sample\n\nPA\n\nGeneXpert testing is\n\nvery cost-effective for\n\nscreening all\n\nindividuals initiating\n\nART\n\nNR\n\n101.5\ufffd\n\nNR\n\nMonte\n\nGeneXpert provides\n\nCarlo\n\nsubstantial savings to\n\nsimulation hospitals in high\n\nincome countries by\n\nreducing overall\n\nlength of stay\n\nNR\n\n959\n\n1,000 Monte\n\nGeneXpert has the\n\nCarlo\n\npotential to produce a\n\nsimulation substantial reduction\n\nin TB morbidity and\n\nmortality\n\nNR\n\n16.37\n\nNR\n\nOne-way\n\nXpert introduction in\n\nPA\n\nSouth Africa was\n\ncost-neutral\n\nNR\n\n127\n\n690\n\nOne-way\n\nXpert is considered\n\nPA\n\ncost effective\n\n\ufffd ICER calculated; NA = Not Applicable; NR = Not Reported https://doi.org/10.1371/journal.pone.0205233.t003\n\nDifferent components of costs used for costs calculation\nFor cost calculation, broadly six components such as laboratory space, staff, training, equipment, consumables and overheads were used in the studies (Table 4). Out of 13 studies none included all the six components. Additionally, one study included waste disposal [27] and one study included transportation cost components [34]. There was variation in inclusion of different costs components. Though the reasons for this variation are not clear, individual studies perceived the importance of each component differently, and it may depend on their outcome of interest or the effectiveness indicator.\n\nDiscussion\nTo the best of our knowledge, this is the first systematic review to synthesize the evidence of cost-effectiveness of LED-FM and GeneXpert in comparison to ZN microscopy for pulmonary\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n10 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\n\nTable 4. Key Cost components reported by the studies included in the review.\n\nSr. First No. Author,\nYear 1 Whitelaw,\n2011 2 Kelly, 2015 3 Sohn, 2009 4 Xia, 2013 5 Mishra,\n2012 6 Pinto, 2016 7 You, 2015 8 Jha, 2016 9 Andrews,\n2012 10 Millman,\n2013 11 Menzies,\n2012\n12 Vassall, 2017\n13 Tesfaye, 2017\n\nCountry\nSouth Africa\nIndia Thailand China Zambia\nBrazil Hongkong South Africa South Africa\nUSA\nBotswana, Lesotho, Namibia, South Africa & Swazil& South Africa\nEthiopia\n\nLab Staff Training Equipment Consumables Overheads Disposal Transport Checkmarks Cost of Cost of\n\nspace\n\nTest ZN\n\n\u2713\u2713\n\n\u2715\n\n\u2713\n\n\u2713\n\n\u2713\n\n\u2715\n\n\u2715\n\n5\n\n1.63 2.1\n\n\u2715\u2715\n\n\u2713\n\n\u2713\n\n\u2713\n\n\u2715\n\n\u2715\n\n\u2715\n\n3\n\n0.31 0.21\n\n\u2713\u2713\n\n\u2715\n\n\u2713\n\n\u2713\n\n\u2713\n\n\u2715\n\n\u2715\n\n5\n\n1.03 1.16\n\n\u2713\u2713\n\n\u2715\n\n\u2713\n\n\u2713\n\n\u2713\n\n\u2713\n\n\u2715\n\n6\n\n1.97 2.2\n\n-\n\n-\n\n-\n\n-\n\n-\n\n-\n\n-\n\n-\n\n0\n\n108.9 75.74\n\n\u2715\u2713\n\n\u2715\n\n\u2713\n\n\u2713\n\n\u2713\n\n\u2715\n\n\u2715\n\n4\n\n14.69 3.08\n\n\u2715\u2713\n\n\u2715\n\n\u2715\n\n\u2713\n\n\u2715\n\n\u2715\n\n\u2715\n\n2\n\n128 7.5\n\n\u2713\u2713\n\n\u2715\n\n\u2713\n\n\u2713\n\n\u2713\n\n\u2715\n\n\u2715\n\n5\n\n14.45 1.59\n\n\u2715\u2713\n\n\u2715\n\n\u2715\n\n\u2713\n\n\u2715\n\n\u2715\n\n\u2715\n\n2\n\n21.6 4.6\n\n\u2715\u2713\n\n\u2715\n\n\u2713\n\n\u2713\n\n\u2713\n\n\u2715\n\n\u2715\n\n4\n\n218\n\n15\n\n\u2715\u2713\n\n\u2715\n\n\u2713\n\n\u2713\n\n\u2713\n\n\u2715\n\n\u2713\n\n5\n\n45\n\n31\n\n\u2715\u2713\n\n\u2713\n\n\u2713\n\n\u2713\n\n\u2715\n\n\u2715\n\n\u2713\n\n5\n\n168.79 160.64\n\n\u2715\u2713\n\n\u2713\n\n\u2713\n\n\u2713\n\n\u2713\n\n\u2715\n\n\u2715\n\n5\n\nNR NR\n\nAdditional health system costs per year over 10 years is used for different algorithms, to calculate ICER value, hence cost per test is not reported. NR = Not reported\n\nhttps://doi.org/10.1371/journal.pone.0205233.t004\n\nTB diagnosis. The review also appraised the reporting quality of the published evidence. The key findings from the included studies suggest that the new diagnostic tools LED-FM and GeneXpert are very cost effective for pulmonary TB diagnosis from a health system perspective, even though they are not cost saving to the health system. The evidence from 11 countries, with majority of them having high TB burden shows that these new tools are cost effective irrespective of their economic condition, HIV burden and geographical distribution.\nFor LED-FM, only one out of four studies reported ICER values and, for the remaining three studies, ICER was calculated using the data provided [13, 27\u201331, 33]. Three studies used average time per slide reading as the effectiveness indicator, while one study used DALYs. The average time taken to read one ZN stained slide is 2.8 (\u00b10.4) minutes. By using the new tool LED-FM this can be reduced to 1.6 (\u00b10.4) minutes, with an additional cost of less than one USD. This additional costs fall within the \u2018willingness to pay threshold\u2019 of each country. Hence, this tool is cost-effective to diagnose pulmonary TB. One study from India reported the long-term impact in terms of DALYs which indicated additional cost of USD 14.64 to avert one DALY. This additional cost is less than the national \u2018willingness to pay threshold\u2019 of USD 1489 for India [13]. Apart from being cost-effective, LED-FM is user-friendly and more acceptable among technicians. It can also be extended to other infectious disease diagnosis like malaria and trypanosomiasis, reducing the costs involved in providing integrated laboratory services [34]. Considering this factor, LED-FM could possibly be more cost-effective in countries with high double burden of TB and malaria.\nGeneXpert studies included in this review used different short term (additional case diagnosed, reduction in duration of hospitalisation) and long term (TB case averted, QALYs,\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n11 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\nDALYs and YLS) effectiveness indicators. There was a huge variation in terms of cost per unit of health benefit which could be due to the different effectiveness indicators, year of study and the subsidised rate of GeneXpert cartridges to high burden countries. For instance, it was observed that health system will have to pay at least USD 1927 for a short-term benefit of additional TB case diagnosed if GeneXpert is preferred in South Africa [30]. This additional cost is very close to the maximum of willingness to pay threshold USD 2000. However, another study from South Africa in 2012 [31] reported an ICER of USD 5100 to save one life-year which is a long-term benefit. This also is within the willingness-to-pay threshold of USD 21,300.\nThis review observed that the included studies analysed effectiveness in terms of different indicators. Results of these studies conclude that implementation of GeneXpert will increase case detection, reduce duration of hospitalisation, gain QALYs, reduce DALYs and save additional years of lives. Also, the investment is within the willingness to pay threshold to avert TB cases. However, most of the studies have not included the sensitivity and specificity of the test in the calculation. Additional to these benefits, GeneXpert can diagnose rifampicin resistance, contributing to early diagnosis of TB as well as rifampicin resistance TB, early treatment initiation and indirectly reduce transmission in the community. However, none of these factors have been considered in cost calculation in the included studies. Thus, the costs calculated may have been underestimated. It is possible that if these studies include the above mentioned factors, GeneXpert may prove to be even more cost-effective.\nFurthermore, the current review assessed the reporting quality of the studies using the CHEC checklist which consists of 20 items. It was observed that none of the studies included all cost components which resulted in under estimation of total costs. This indicates variability in the methods used to determine the costs involved in the diagnosis of pulmonary TB. Additionally, none of the studies are based on randomised controlled trials which provide rigorous comparison. Majority of the studies included limited cost components such as consumables and staff costs to calculate costs. Similarly, the effectiveness indicators varied in different studies due to which meta-analysis was not possible in this current review. Sensitivity analysis was performed in almost all the GeneXpert studies. None of the studies mentioned about the methods of calculations of QALYs, DALYs and YLS. This review provides the way forward to compare the ICER values and sum up the results. This review also suggests the need for improvement in several aspects of published cost effectiveness analysis [37].\nOnly five of the thirteen studies included in the review mentioned target population. Overall, majority of the studies (8/13) mention the sample size but adequate description of the characteristics of the base population is not clearly stated. Although the sample size varied considerably, the authors did not provide the value of standard deviation of average costs. However, these studies represent developed and developing nations as well as low and high TB burden countries. The conclusions of all included studies suggest the generalizability of the observation. Similarly, a systematic review on methodological issues on cost-effectiveness study has also mentioned inadequate reporting of characteristics of the target population which is important for generalizability of the results for decision making [38].\nWhile the cost-effectiveness of implementing a new tool (LED-FM or GeneXpert) is one dimension; the other dimension of clinical effectiveness is considering the sensitivity and specificity for each of the methods. A systematic review conducted on clinical effectiveness of GeneXpert showed that GeneXpert has higher sensitivity than the ZN microscopy. Test accuracy was retained; a single GeneXpert MTB/RIF test directly on sputum detected 99% of smear-positive patients and 80% of patients with smear-negative disease. Thus, GeneXpert is cost effective with increase in sensitivity [39]. It also provides additional information on drug susceptibility of rifampicin.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n12 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\nOf the included studies for GeneXpert, majority were done in South Africa (5/9) [31, 32, 34]. Since South Africa has adopted GeneXpert as an upfront diagnostic for TB, which made it possible for more studies to be conducted. One multi-centric study done in 2012 [34] including South Africa reported cost per sample was USD 45. In the same year (2012) another study was conducted only in South Africa reported cost per sample was USD 21.6 [32]. Though this study did not report the country wise costs, the higher cost may be due to the pooled estimate (due to multi-centric nature of the study). Another study conducted in South Africa in 2016 reported the cost per sample was USD 14.45; indicating that, over a period of time, implementation of GeneXpert seems to be getting more cost-effective [30].\nNone of these studies considered the patient benefits through GeneXpert to calculate costseffectiveness. It was reported that average time to detection was less than one day for GeneXpert, one day for microscopy, 17 days for liquid culture and more than 30 days for solid culture. Further, rifampicin resistance was detected in less than one day with GeneXpert compared with an average of 75 days for phenotypic drug sensitive profile. When GeneXpert results were not used to direct therapy, smear-negative TB patients were initiated with treatment in 58 days on an average, as compared to four days when GeneXpert results were used [40]. This has an impact on quality of life of TB patients and leads to increase in QALYs. Moreover, early diagnosis and initiation of treatment will also contribute in reduction of TB transmission. A study from Brazil reported that 35% reduction in TB-related mortality with less advanced disease among the smear-negative patients diagnosed by GeneXpert [41]. However, this aspect is also not considered for the calculation of cost-effectiveness. If all these parameters are taken into consideration for the cost-effectiveness estimation, GeneXpert will be more cost-effective than currently estimated for the diagnosis of pulmonary TB.\nLimitations of the review\nIn this review, we did not include unpublished studies or studies published in non-indexed journals. The heterogeneity of the included studies in terms of study design, outcome measures limited the scope for synthesising the data and interpretation.\nConclusion\nOur review identifies a consistent trend of the cost effectiveness of LED-FM and GeneXpert in different countries with diverse context of socio-economic condition, HIV burden and geographical distribution. However, all the studies used different parameters to estimate the impact of these tools and this underscores the need for improving the methodological issues related to the conduct and reporting of cost-effectiveness studies.\nSupporting information\nS1 Table. PRISMA checklist. (DOCX)\nS2 Table. Search strategy used to search various databases (PubMed/MEDLINE/EMBASE/ Cochran/CEA Registry). (DOCX)\nS3 Table. List of references excluded due to non-ZN comparator. (DOCX)\nS4 Table. List of references excluded due to non-reporting of effectiveness indicator. (DOCX)\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n13 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\nAcknowledgments\nThe review was conducted as a part of the training course aimed to build capacity on undertaking Systematic Reviews within India, especially in the context of tuberculosis. The training project was conceived and implemented by the South East Asia office of The International Union against Tuberculosis and Lung Diseases (The Union) in collaboration with Central TB Division, Ministry of Health and Family Welfare, Government of India and The Cochrane South Asia. The training was generously supported by U.S. Agency for International Development (USAID). We also thank Dr. Nandini Dendukuri, Department of Health Technology Assessment, McGill University, Canada for her inputs during protocol development.\nAuthor Contributions\nConceptualization: Karuna D. Sagili, Malaisamy Muniyandi, Kayzad Soli Nilgiriwala, Kalpita S. Shringarpure, Srinath Satyanarayana, Richard Kirubakaran, Sarabjit S. Chadha, Prathap Tharyan.\nData curation: Karuna D. Sagili, Malaisamy Muniyandi, Kayzad Soli Nilgiriwala, Kalpita S. Shringarpure.\nFormal analysis: Karuna D. Sagili, Malaisamy Muniyandi, Kayzad Soli Nilgiriwala, Kalpita S. Shringarpure.\nFunding acquisition: Sarabjit S. Chadha.\nInvestigation: Karuna D. Sagili, Malaisamy Muniyandi, Kayzad Soli Nilgiriwala, Kalpita S. Shringarpure.\nMethodology: Karuna D. Sagili, Malaisamy Muniyandi, Kayzad Soli Nilgiriwala, Kalpita S. Shringarpure, Srinath Satyanarayana, Richard Kirubakaran, Prathap Tharyan.\nProject administration: Karuna D. Sagili, Sarabjit S. Chadha.\nResources: Karuna D. Sagili, Sarabjit S. Chadha. Software: Karuna D. Sagili, Kayzad Soli Nilgiriwala, Richard Kirubakaran.\nSupervision: Karuna D. Sagili, Srinath Satyanarayana, Prathap Tharyan. Validation: Karuna D. Sagili, Malaisamy Muniyandi, Kayzad Soli Nilgiriwala, Kalpita S. Shrin-\ngarpure, Richard Kirubakaran, Prathap Tharyan.\nVisualization: Karuna D. Sagili, Malaisamy Muniyandi, Kayzad Soli Nilgiriwala, Kalpita S. Shringarpure.\nWriting \u2013 original draft: Karuna D. Sagili, Malaisamy Muniyandi, Kayzad Soli Nilgiriwala, Kalpita S. Shringarpure.\nWriting \u2013 review & editing: Karuna D. Sagili, Malaisamy Muniyandi, Kayzad Soli Nilgiriwala, Kalpita S. Shringarpure, Srinath Satyanarayana, Richard Kirubakaran, Sarabjit S. Chadha, Prathap Tharyan.\nReferences\n1. World Health Organization. Global Tuberculosis Report 2016. Geneva 2016. 2. World Health Organization. Gear up to End TB: Introducing the End TB Strategy. Geneva 2015. 3. World Health Organization. Global Tuberculosis Report 2015. Geneva 2015. 4. World Health Organization. Early detection of Tuberculosis: An overview of approaches, guidelines and\ntools. Geneva 2011.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n14 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\n5. World Health Organization, Regional Office for South East Asia. Scale up TB control initiatives to reach the missing one million cases. 2015.\n6. Laokri S, Drabo MK, Weil O, Kafando B, Dembele SM, Dujardin B. Patients Are Paying Too Much for Tuberculosis: A Direct Cost-Burden Evaluation in Burkina Faso. PLoS One. 2013; 8(2):e56752. https:// doi.org/10.1371/journal.pone.0056752 PMID: 23451079\n7. Molicotti P, Bua A, Zanetti S. Cost-effectiveness in the diagnosis of tuberculosis: choices in developing countries. J Infect Dev Ctries. 2014; 8(1):24\u201338. https://doi.org/10.3855/jidc.3295 PMID: 24423709\n8. World Health Organization. Fluorescent light-emitting diode (LED) microscopy for diagnosis of tuberculosis: policy statement. Geneva, 2011.\n9. World Health Organization. Implementing tuberculosis diagnostics: A policy framework. Geneva 2015.\n10. World Health Organization. Policy Statement: Automated Real-Time Nucleic Acid Amplification Technology for Rapid and Simultaneous Detection of Tuberculosis and Rifampicin Resistance: Xpert MTB/ RIF System. Geneva 2011.\n11. Kanabus A. Information about Tuberculosis. Global Health Education 2017. TBFacts.org\n12. World Health Organization. WHO endorses new rapid tuberculosis test: A major milestone for global TB diagnosis and care. World Health Organization, Geneva, 2010.\n13. Kelly V, Sagili KD, Satyanarayana S, Reza LW, Chadha SS, Wilson NC. Cost-utility analysis of LED fluorescence microscopy in the diagnosis of pulmonary tuberculosis in Indian settings. Int J Tuberc Lung Dis. 2015; 19(6):696\u2013701. https://doi.org/10.5588/ijtld.14.0203 PMID: 25946362\n14. Central TB Division. TB India 2017\u2014Annual Status Report. Directorate General of Health Services, Government of India, Ministry of Health and Family Welfare, New Delhi, India 2017\n15. Kaur R, Kachroo K, Sharma JK, Vatturi SM, Dang A. Diagnostic accuracy of xpert test in tuberculosis detection: A systematic review and meta-analysis. J Global Infect Dis. 2016; 8(1):32.\n16. Reynolds MR, Magnuson EA, Lei Y, Wang K, Vilain K, Li H, et al Cost-effectiveness of transcatheter aortic valve replacement compared with surgical aortic valve replacement in high-risk patients with severe aortic stenosis: results of the PARTNER (Placement of Aortic Transcatheter Valves) trial (Cohort A). J Am Coll Cardiol 2012; 60:2683\u201392. https://doi.org/10.1016/j.jacc.2012.09.018 PMID: 23122802\n17. Yan L, Xiao H, Zhang Q. Systematic review: Comparison of Xpert MTB/RIF, LAMP and SAT methods for the diagnosis of pulmonary tuberculosis. Tuberculosis. 2016; 96:75\u201386. https://doi.org/10.1016/j. tube.2015.11.005 PMID: 26786658\n18. Liberati A, Altman DG, Tetzlaff J, Mulrow C, G\u00f8tzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med. 2009; 6(7):e1000100. https://doi.org/10.1371/journal.pmed. 1000100 PMID: 19621070\n19. Sagili K, Shringarpure K, Nilgiriwala K, Muniyandi M. Cost-effectiveness of GeneXpert, LED FM and chest X-ray for diagnosis of pulmonary tuberculosis: a systematic review and meta-analysis. PROSPERO 2016:CRD42016043333 Available from http://www.crd.york.ac.uk/PROSPERO_ REBRANDING/display_record.asp?ID=CRD42016043333\n20. Pang F, Drummond M, Song F. The use of meta-analysis in economic evaluation. Centre for Health Economics, University of York, 1999.\n21. Tufts Medical Center. Cost effectiveness analysis registry. http://healtheconomics.tuftsmedicalcenter. org/cear4/\n22. Cochrane Library. Cochrane Database of Systematic Reviews. http://onlinelibrary.wiley.com/ cochranelibrary/search\n23. Ouzzani M, Hammady H, Fedorowicz Z, Elmagarmid A. Rayyan\u2014a web and mobile app for systematic reviews. Systematic reviews. 2016; 5(1):210. https://doi.org/10.1186/s13643-016-0384-4 PMID: 27919275\n24. Odnoletkova I, Goderis G, Pil L, Nobels F, Aertgeerts B, Annemans L, et al. Cost-Effectiveness of Therapeutic Education to Prevent the Development and Progression of Type 2 Diabetes. Systematic Review. J Diab Metab. 2014; 5(9).\n25. Whitelaw A, Peter J, Sohn H, Viljoen D, Theron G, Badri M, et al. Comparative cost and performance of light-emitting diode microscopy in HIV-tuberculosis-co-infected patients. Eur Res J. 2011; 38(6):1393\u2013 7.\n26. Sohn H, Sinthuwattanawibool C, Rienthong S, Varma JK. Fluorescence microscopy is less expensive than Ziehl-Neelsen microscopy in Thailand. Int J Tuberc Lung Dis 2009; 13(2):266\u20138. PMID: 19146758\n27. Xia H, Song YY, Zhao B, Kam KM, O\u2019Brien RJ, Zhang ZY, et al. Multicentre evaluation of Ziehl-Neelsen and light-emitting diode fluorescence microscopy in China. Int J Tuberc Lung Dis 2013; 17(1):107\u2013112. https://doi.org/10.5588/ijtld.12.0184 PMID: 23232010\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n15 / 16\n\nCost-effectiveness of GeneXpert and LED-FM for TB diagnosis\n28. Mishra L, Henostroza G, Harris J, Siyambango M, Krunner A, Kaunda K, et al. Evaluating the cost-effectiveness of TB diagnostic strategies in HIV-positive patients in Lusaka, Zambia. Int J of Infect Dis 2012; 16: e286.\n29. Pinto M, Steffen RE, Cobelens F, van den Hof S, Entringer A, Trajman A. Cost-effectiveness of the Xpert(R) MTB/RIF assay for tuberculosis diagnosis in Brazil. Int J Tuberc Lung Dis. 2016; 20(5):611\u20138. https://doi.org/10.5588/ijtld.15.0455 PMID: 27084814\n30. You JH, Lui G, Kam KM, Lee NL. Cost-effectiveness analysis of the Xpert MTB/RIF assay for rapid diagnosis of suspected tuberculosis in an intermediate burden area. J Infect. 2015; 70(4):409\u201314. https://doi.org/10.1016/j.jinf.2014.12.015 PMID: 25573001\n31. Jha S, Ismail N, Clark D, Lewis JJ, Omar S, Dreyer A, et al. Cost-Effectiveness of Automated Digital Microscopy for Diagnosis of Active Tuberculosis. PloS One. 2016; 11(6):e0157554. https://doi.org/10. 1371/journal.pone.0157554 PMID: 27322162\n32. Andrews JR, Lawn SD, Rusu C, Wood R, Noubary F, Bender MA, et al. The cost-effectiveness of routine tuberculosis screening with Xpert MTB/RIF prior to initiation of antiretroviral therapy: a modelbased analysis. AIDS. 2012; 26(8):987\u201395. https://doi.org/10.1097/QAD.0b013e3283522d47 PMID: 22333751\n33. Millman AJ, Dowdy DW, Miller CR, Brownell R, Metcalfe JZ, Cattamanchi A, et al. Rapid molecular testing for TB to guide respiratory isolation in the U.S.: a cost-benefit analysis. PloS One. 2013; 8(11): e79669. https://doi.org/10.1371/journal.pone.0079669 PMID: 24278155\n34. Menzies NA, Cohen T, Lin HH, Murray M, Salomon JA. Population health impact and cost-effectiveness of tuberculosis diagnosis with Xpert MTB/RIF: a dynamic simulation and economic evaluation. PLoS Med. 2012; 9(11):e1001347. https://doi.org/10.1371/journal.pmed.1001347 PMID: 23185139\n35. Tesfaye A, Fiseha D, Assefa D, Klinkenberg E, Balanco S, Langley I. Modeling the patient and health system impacts of alternative xpert(R) MTB/RIF algorithms for the diagnosis of pulmonary tuberculosis in Addis Ababa, Ethiopia. BMC Infect Dis. 2017; 17(1):318. https://doi.org/10.1186/s12879-017-2417-6 PMID: 28464797\n36. Vassall A, Siapka M, Foster N, Cunnama L, Ramma L, Fielding K, et al. Cost-effectiveness of Xpert MTB/RIF for tuberculosis diagnosis in South Africa: a real-world cost analysis and economic evaluation. Lan Glob Health. 2017; 5(7):e710\u2013e9.\n37. Wong CK, Liao Q, Guo VY, Xin Y, Lam CLK. Cost-effectiveness analysis of vaccinations and decision makings on vaccination programmes in Hong Kong: A systematic review. Vaccine 2017, 35(24):3153\u2013 3161. https://doi.org/10.1016/j.vaccine.2017.04.050 PMID: 28476628\n38. Catala\u00b4-Lo\u00b4pez F, Ridao M, Alonso-Arroyo A, Garc\u00b4\u0131a-Alte\u00b4 s A, Cameron C, Gonza\u00b4 lez-Bermejo D, et al. The quality of reporting methods and results of cost-effectiveness analyses in Spain: a methodological systematic review. Systematic reviews. 2016; 5(1):6.\n39. Dinnes J, Deeks J, Kunst H, Gibson A, Cummins E, Waugh N, et al. A systematic review of rapid diagnostic tests for the detection of tuberculosis infection. Health Technology Assessment Southampton 2007; 11(3).\n40. World Health Organization. Xpert MTB/RIF implementation manual: Technical and operational \u2018how-to\u2019: practical considerations. Geneva, Switzerland: WHO, 2014.\n41. Trajman A, Durovni B, Saraceni V, Menezes A, Cordeiro-Santos M, Cobelens F, et al. Impact on patients\u2019 treatment outcomes of XpertMTB/RIF implementation for the diagnosis of tuberculosis: followup of a stepped-wedge randomized clinical trial. PloS One. 2015; 10(4):e0123252. https://doi.org/10. 1371/journal.pone.0123252 PMID: 25915745\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0205233 October 29, 2018\n\n16 / 16\n\n\n",
    "authors": [
      "Karuna D. Sagili",
      "Malaisamy Muniyandi",
      "Kayzad Soli Nilgiriwala",
      "Kalpita S. Shringarpure",
      "Srinath Satyanarayana",
      "Richard Kirubakaran",
      "Sarabjit S. Chadha",
      "Prathap Tharyan"
    ],
    "doi": "10.1371/journal.pone.0205233",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://dx.plos.org/10.1371/journal.pone.0205233"
  },
  {
    "key": "NRBX5785",
    "title": "Comparison of laboratory costs of rapid molecular tests and conventional diagnostics for detection of tuberculosis and drug-resistant tuberculosis in South Africa",
    "abstract": "Background: The World Health Organization has endorsed the use of molecular methods for the detection of TB and drug-resistant TB as a rapid alternative to culture-based systems. In South Africa, the Xpert MTB/Rif assay and the GenoType MTBDRplus have been implemented into reference laboratories for diagnosis of TB and drug-resistance, but their costs have not been fully elucidated.\nMethods: We conducted a detailed reference laboratory cost analysis of new rapid molecular assays (Xpert and MTBDRplus) for tuberculosis testing and drug-resistance testing in South Africa, and compared with the costs of conventional approaches involving sputum microscopy, liquid mycobacterial culture, and phenotypic drug sensitivity testing.\nResults: From a laboratory perspective, Xpert MTB/RIF cost $14.93/sample and the MTBDRplus line probe assay cost $23.46/sample, compared to $16.88/sample using conventional automated liquid culture-based methods. Laboratory costs of Xpert and MTBDRplus were most influenced by cost of consumables (60-80%).\nConclusions: At current public sector pricing, Xpert MTB/RIF and MTBDRplus are comparable in cost to mycobacterial culture and conventional drug sensitivity testing. Overall, reference laboratories must balance costs with performance characteristics and the need for rapid results.",
    "full_text": "Shah et al. BMC Infectious Diseases 2013, 13:352 http://www.biomedcentral.com/1471-2334/13/352\n\nRESEARCH ARTICLE\n\nOpen Access\n\nComparison of laboratory costs of rapid molecular tests and conventional diagnostics for detection of tuberculosis and drug-resistant tuberculosis in South Africa\nMaunank Shah1*, Violet Chihota2, Gerrit Coetzee3, Gavin Churchyard2,4,5 and Susan E Dorman1\n\nAbstract\nBackground: The World Health Organization has endorsed the use of molecular methods for the detection of TB and drug-resistant TB as a rapid alternative to culture-based systems. In South Africa, the Xpert MTB/Rif assay and the GenoType MTBDRplus have been implemented into reference laboratories for diagnosis of TB and drug-resistance, but their costs have not been fully elucidated.\nMethods: We conducted a detailed reference laboratory cost analysis of new rapid molecular assays (Xpert and MTBDRplus) for tuberculosis testing and drug-resistance testing in South Africa, and compared with the costs of conventional approaches involving sputum microscopy, liquid mycobacterial culture, and phenotypic drug sensitivity testing.\nResults: From a laboratory perspective, Xpert MTB/RIF cost $14.93/sample and the MTBDRplus line probe assay cost $23.46/sample, compared to $16.88/sample using conventional automated liquid culture-based methods. Laboratory costs of Xpert and MTBDRplus were most influenced by cost of consumables (60-80%).\nConclusions: At current public sector pricing, Xpert MTB/RIF and MTBDRplus are comparable in cost to mycobacterial culture and conventional drug sensitivity testing. Overall, reference laboratories must balance costs with performance characteristics and the need for rapid results.\nKeywords: Laboratory, Costs, Tuberculosis, Diagnostics\n\nBackground Currently, less than 10% of multi-drug resistant tuberculosis (MDR-TB) cases in the world are detected [1]. Performance of drug susceptibility testing (DST) using conventional methods relies on solid or liquid media and is slow and resource intensive. Recently, the World Health Organization endorsed the use of molecular methods for the detection of TB and drug-resistvant TB as a rapid alternative to culture-based systems [2,3]. Two commercially available molecular assays using different methodologies have been implemented in South\n* Correspondence: Mshah28@JHMI.EDU 1Johns Hopkins University School of Medicine, Department of Medicine, Division of Infectious Disease, 1503 East Jefferson St, Room 118, Baltimore, MD 21231, USA Full list of author information is available at the end of the article\n\nAfrica -- the GenoType MTBDRplus (\u2018MTBDRplus,\u2019 Hain Lifescience, Nehren, Germany) and the Xpert MTB/RIF (\u2018Xpert,\u2019 Cepheid, Sunnyvale, CA).\nMTBDRplus is a line probe assay that has shown good diagnostic accuracy for the detection of M. tuberculosis in smear-positive cases and isoniazid and/or rifampin resistance in several validation studies, and test results can be available in as few as 1\u20132 days [4-6]. Xpert is an integrated specimen processing and nucleic acid amplification-based test for detection of M. tuberculosis and rifampin resistance and offers results within hours. Xpert has the advantage of high sensitivity when performed on smear microscopy-positive sputum samples and requires little laboratory processing, overhead, or labor [7].\nSouth Africa is a middle-income country that has sought to scale-up laboratory services to implement\n\n\u00a9 2013 Shah et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.\n\nShah et al. BMC Infectious Diseases 2013, 13:352 http://www.biomedcentral.com/1471-2334/13/352\n\nPage 2 of 7\n\nthese new tests. However, while Xpert and MTBDRplus offer rapid results and similar performance characteristics, each has limitations and neither may be a complete replacement for conventional culture and DST [2,3,6,7]. Both tests have reduced sensitivity for smear-negative samples, and conventional DST is needed for expanded drug resistance testing. MTBDRplus is a technically complicated assay requiring substantial laboratory resources, and both tests require expensive equipment and reagents. In August 2012, however, public sector prices for Xpert consumables were significantly reduced [8]. To date, little data is available to compare the costs and resource needs of these emerging rapid diagnostics to guide policy-makers and laboratory managers. To address this knowledge gap, we performed a detailed costanalysis from a laboratory perspective and compared the costs associated with conventional liquid culture and DST, MTBDRplus and Xpert. We further explored the costs of incorporating these assays as stand-alone tests for TB diagnosis, or alternatively in conjunction with existing diagnostics.\nMethods Costs associated with mycobacterial testing were analyzed from a laboratory perspective at the National Health Laboratory Services (NHLS) National TB Reference Laboratory in Johannesburg, South Africa. Costs were collected for testing conducted using Xpert, MTBDRplus, Ziehl-Neelsen smear microscopy using a light microscope, and florescence smear microscopy using auramine-O staining and a light emitting diode microscope [9]. Costs were also collected for sputum processing (i.e. digestion and decontamination) using Nacetyl-L-cysteine-NaOH and concentration by centrifugation [10], liquid culture using the automated BACTEC MGIT 960 system (BD Diagnostic Systems, Sparks, MD, USA), indirect phenotypic DST using the MGIT SIRE system (BD Diagnostic Systems), and anti-MPB64 monoclonal antibody-based species identification (Capilia TB-Neo, TAUNS Laboratories, Numazu, Japan) of positive cultures. For the liquid culture testing scenario, all samples were considered to require sputum processing and MGIT culture; positive cultures were tested by Ziehl Neelsen smear microscopy to assess for mycobacteria; cultures positive for mycobacteria were tested by the anti-MPB64 assay to distinguish M. tuberculosis from non-tuberculous mycobacteria, and cultures positive for M. tuberculosis were subjected to phenotypic MGIT SIRE DST. Xpert testing was conducted per manufacturer instructions and performed using a 4-module GeneXpert (Cepheid) instrument with automated readout. MTBDRplus was performed according to manufacturer instructions and consisted of DNA extraction, amplification, and hybridization steps, with\n\nhybridization performed using a GT Blot instrument (Hain Lifescience).\nCosts were analyzed using an \u201cingredients\u201d approach that involved multiplying the quantity of inputs used by their unit prices; costs and wages were gathered using detailed laboratory records. The amount of staff time, consumable supplies and equipment quantities utilized for each test were determined through direct observation of testing procedures, and included costs associated with quality assurance and quality control. Overhead laboratory costs included indirect labor costs, office and lab supplies and furniture, general operations costs, and physical infrastructure costs (i.e. building, utilities, and maintenance costs). Overhead costs were allocated based on the volume of testing and amount of physical infrastructure utilized by each diagnostic system. Equipment costs were annualized over their useful lifespans. South Africa is eligible to pay prices negotiated by the Foundation for Innovative Diagnostics (FIND), and costs for consumables and equipment reflect this pricing structure [8,11,12]. Laboratory testing capacity was estimated based on the laboratory operating for 12 hours per day. Ten percent of all sputum samples were estimated to be smear-positive and 5% of cultures were estimated to be positive for M. tuberculosis based on laboratory records. Unit costs of key equipment and consumables are shown in Table 1. We evaluated the costs of each diagnostic system individually and in combination with each other. All costs are presented in 2012 US dollars.\nResults Base-case laboratory costs for TB diagnostics are shown in Table 2. Costs per test conducted were $14.93 for Xpert, $23.46 for MTBDRplus, $3.40 for fluorescence smear microscopy, $2.25 for Ziehl-Neelson light microscopy, $12.16 for MGIT culture, and $26.19 for DST using MGIT SIRE. We calculated a total cost of $16.88 per specimen tested for the combination of fluorescence smear microscopy plus the liquid culture testing scenario (sputum processing and MGIT culture, followed by Ziehl Neelsen smear microscopy on positive cultures, anti-MPB64 assays on cultures with mycobacterial growth, and MGIT SIRE DST on cultures with growth of M. tuberculosis). This combination of fluorescence microscopy plus the liquid culture testing scenario was over $6 cheaper per sample than using MTBDRplus on all sputa, but was more expensive than Xpert (Table 3).\nThe cost of Xpert was largely attributable to consumables ($11.97 per test [80% of total]), and was driven by the cost of Xpert cartridges ($9.98 per cartridge, Tables 1 and 2). Alternatively, countries not eligible for FINDnegotiated discounts may pay up to $78,200 for purchase of the GeneXpert instrument and $71.63 per cartridge [8,13]. In the latter scenario, Xpert costs may rise to\n\nShah et al. BMC Infectious Diseases 2013, 13:352 http://www.biomedcentral.com/1471-2334/13/352\n\nPage 3 of 7\n\nTable 1 Unit costs of key consumables and equipment*\n\nConsumables\n\nUnit (quantity) Unit cost $US\n\nNALC/NAOH Kit\n\nper Sample\n\n$1.73\n\nN95 Mask\n\nper Box(100)\n\n$55.19\n\nZN stain\n\nper liter\n\n$4.87\n\nAuramine-O\n\nper liter\n\n$4.67\n\nMethylene blue\n\nper liter\n\n$3.26\n\nPotassium permanganate\n\nPer liter\n\n$3.22\n\nAFB fixative\n\nper 100 ml\n\n$8.52\n\nPANTA\n\none box(100) $89.07\n\nMGIT growth supplement\n\none box(100) $70.80\n\nAnti-MPB64 Capilia TB Neo\n\nper test\n\n$1.69\n\nMGIT Tube\n\nper Box(100)\n\n$195.00\n\nSIRE kit\n\nper kit (35)\n\n$127.50\n\nMTBDR rif kit\n\nper kit (96)\n\n$917.42\n\nXpert MTB/Rif cartridge\n\nper cartridge\n\n$9.98\n\nEquipment\n\nCentrifuge plus accessories\n\nper Instrument $22,439\n\nVortex\n\nper Instrument $304\n\nBiosafety cabinet\n\nper Instrument $3,190\n\nBiosafety cabinet filter replacement yearly\n\n$1,608\n\nBiosafety cabinet decontamination q 6months\n\n$156\n\nLight microscope\n\nper Instrument $3,409\n\nLED microscope\n\nper Instrument $4580\n\nBactec MGIT 960\n\nper Instrument $38,950.00\n\nEpicenter software\n\n\u2013\n\n$12,500.00\n\nBarcode scanner\n\nper Instrument $158.30\n\nThermocycler\n\nper Instrument $5,621.18\n\nUltrasonicator\n\nper Instrument $1,596.42\n\nGTBLOT maintenance\n\nsemi-annual\n\n$487.08\n\nGT BLOT\n\nper Instrument $17,557.40\n\nUPS (power supply)\n\nper Instrument $137.24\n\nXpert instrument\n\nper Instrument $17,000.00\n\nXpert calibration\n\nper 1800 runs $1,800.00\n\n*Not all items are shown. Additional minor consumables and equipment costs included but were not limited to costs associated with gloves, disposable gowns, pipettes and pipetters and tips, computers and supplies, soap and disinfectant, waste disposal including biohazard bags, microscopy slides, and other common microbiology supplies.\n\n$78.94 per sample and would be substantially more expensive than conventional diagnostics. Labor costs associated with performing Xpert were low ($1.13 per test) compared to MGIT culture ($2.17 per test) or MTBDRplus($3.46 per test). Overall, if the volume of testing in the laboratory were reduced by 50%, the cost of Xpert would rise to $16.50 per test.\nBy comparison, MTBDRplus costs were similarly attributable largely to consumables ($14.13 per test [60%]), but also had high labor costs ($3.46 per test) and\n\noverhead costs ($4.28 per test) due to the time involved and extensive laboratory facilities needed for test performance.\nLaboratory costs associated with alternative diagnostic algorithms incorporating multiple tests are shown in Table 3. An intensive TB diagnostic strategy involving performance of Xpert or MTBDRplus in addition to the liquid culture testing scenario on all sputum samples would nearly double laboratory costs per sample ($28.41 and $36.94 per sample for addition of Xpert and addition of MTBDRplus, respectively) compared to a strategy of using only the liquid culture testing scenario (Table 3).\nA more selective strategy considered by some laboratories to reduce costs might be to perform smear microscopy plus the liquid culture testing scenario on all sputum samples to ensure highest diagnostic sensitivity, but employ molecular testing only selectively for smear microscopy-positive samples (for MTB confirmation and rapid drug-resistance results). This strategy would lead to an incremental cost of less than $3 per sample, compared to using smear microscopy plus the liquid culture testing scenario (Table 3), while likely substantially reducing the time to diagnosis of resistance.\nAn additional strategy considered by some labs may be to utilize molecular assays primarily as a replacement for conventional DST. Such a strategy would cost $16.51 if Xpert were used in place of DST (incremental -$0.37 compared to liquid culture scenario with conventional DST) and $17.75 if MTBDRplus were used in place of DST (incremental $0.87 compared to liquid culture scenario with conventional DST).\nIn South Africa, recent guidelines suggest using GeneXpert for all high risk patients and to additionally perform conventional culture and DST for confirmation of positive molecular test results [14]. Such a strategy would cost $16.86 per TB suspect using GeneXpert as the molecular assay and $25.39 if MTBDRplus were used (Table 3).\nDiscussion Scale-up of laboratory capacity for detection of TB and drug resistance is urgently needed, but may be costly. Current reference standard approaches involving mycobacterial culture and DST are slow, and are resource intensive for laboratories to perform. The Xpert MTB/Rif and MTBDRplus are two WHO recommended platforms for rapid detection of TB and drug-resistant TB and many low and middle-income countries qualify for negotiated discounts on these assays [8]. Previously, there has been limited cost information from a laboratory perspective to guide TB control programs and laboratories in implementing these tests. Our results suggest that with recent reductions in the price of Xpert\n\nShah et al. BMC Infectious Diseases 2013, 13:352 http://www.biomedcentral.com/1471-2334/13/352\n\nTable 2 Component costs for each tuberculosis diagnostic test\n\nConsumables cost per test, in $ Equipment cost per test, in $\n\nLabor * cost per test, in $\n\nOverhead cost per test, in $ Total cost per test, in $\n\n(% of total) [uncertainty range]\u2020 (% of total) [ uncertainty range]\u2020 (% of total) [uncertainty range]\u2020 (% of total) [uncertainty range]\u2020 [uncertainty range]\u2020\n\nFluorescence smear microscopy\n\n$0.36 (10%) [$0.27\u2013$0.45]\n\n$0.12 (4%) [$0.09\u2013$0.48]\n\n$2.18 (64%) [$1.64\u2013$3.00]\n\n$0.74 (22%) [$0.18\u2013$0.93]\n\n$3.40 [$2.19\u2013$4.76]\n\nZiehl-Neelsen light smear microscopy\n\n$0.34 (15%) [$0.26\u2013$0.43]\n\n$0.11 (5%) [$0.08\u2013$0.45]\n\n$1.05 (47%) [$0.88\u2013$1.53]\n\n$0.74 (33%) [$0.18\u2013$0.93]\n\n$2.25 [$1.40\u2013$3.24]\n\nMGIT culture\n\n$8.05 (66%) [$6.04\u2013$8.65]\n\n$1.05 (9%) [$0.71\u2013$4.74]\n\n$2.17 (18%) [$2.05\u2013$2.50]\n\n$0.89 (7%) [$0.66\u2013$1.11]\n\n$12.16 [$9.46\u2013$16.99]\n\nPhenotypic DST using MGIT SIRE system\n\n$16.22 (61%) [$12.16\u2013$18.71]\n\n$2.77 (10%) [$2.08\u2013$13.42]\n\n$4.15 (16%) [$3.41\u2013$6.16]\n\n$3.26 (12%) [$2.45\u2013$4.07]\n\n$26.39 [$20.10\u2013$42.37]\n\nMTBDRplus\n\n$14.13 (60%)[$10.37\u2013$17.24]\n\n$1.60 (7%) [$4.17\u2013$7.39]\n\n$3.46 (15%) [$2.85\u2013$5.13]\n\n$4.28 (18%) [$3.21\u2013$5.35]\n\n$23.46[$20.61\u2013$35.12]\n\nXpert MTB/RIF\n\n$11.97 (80%) [$11.49\u2013$19.47]\n\n$0.93 (6%) [$0.70\u2013$3.99]\n\n$1.13 (8%) [$0.94\u2013$4.30]\n\n$0.90 (6%) [$0.22\u2013$1.12]\n\n$14.93 [$13.36\u2013$28.88]\n\nAbbreviations: MGIT Mycobacterial Growth Indicator 960 automated liquid culture system, DST Drug Susceptibility Testing using MGIT SIRE system. *Average salary for laboratory technician was $9.07 per hour based on laboratory records. Range of wages from $7.43 to $16.16 were used for sensitivity analysis based on wages of different skill levels of\nlaboratory workers. \u2020Uncertainty range is based on lowest and highest estimates of consumable and equipment components along with range of laboratory volume of testing [e.g. batch size of Xpert testing was varied from 1 sample to\n4 samples per run], along with range of salaries for laboratory technicians, and highest and lowest estimates for laboratory overhead.\n\nPage 4 of 7\n\nShah et al. BMC Infectious Diseases 2013, 13:352 http://www.biomedcentral.com/1471-2334/13/352\n\nPage 5 of 7\n\nTable 3 Expected costs of diagnostic algorithms\n\nAlgorithm\n\nCost per Incremental\n\nsample\n\ncost\n\nImplementation of conventional diagnostics versus molecular testing\n\nFluorescence microscopy plus liquid culture testing scenario* on all sputa\n\n$16.88\n\nreference\n\nXpert MTB/RIF alone on all sputa\n\n$14.93\n\n$-1.95\n\nMTBDRplus alone on all sputa\n\n$23.46\n\n$6.58\n\nIntensive implementation of molecular tests in combination with conventional diagnostics\n\nXpert MTB/RIF plus liquid culture testing scenario* on all sputa\n\n$28.41\n\n$11.53\n\nMTBDRplus plus liquid culture testing scenario* on all sputa\n\n$36.94\n\n$20.06\n\nConventional diagnostics with selective implementation of molecular tests\n\nFluorescence microscopy plus liquid culture testing scenario* on all sputa + Xpert MTB/RIF on smear-positive sputa\n\n$18.37\n\n$1.49\n\nFluorescence microscopy plus liquid culture testing scenario* on all sputa + MTBDRplus on smear-positive sputa\n\n$19.23\n\n$2.35\n\nFluorescence microscopy on all sputa plus a) Xpert MTB/RIF on smear-positive sputa and b) liquid culture on smear-negative sputa with Xpert MTB/RIF on culture positive isolates\n\n$16.51\n\n-$0.37\n\nFluorescence microscopy on all sputa plus a) MTBDRplus on smear-positive sputa and b) liquid culture on smearnegative sputa with MTBDRplus on culture positive isolates\n\n$17.75\n\n$0.87\n\nMolecular testing with selective implementation of culture and dst**\n\nXpert MTB/RIF on all sputa, with liquid culture testing scenario* on sputa with a positive molecular test\n\n$16.86\n\n$-0.02\n\nMTBDRplus on all sputa, with liquid culture testing scenario* on sputa with a positive molecular test\n\n$25.39\n\n$8.51\n\nAbbreviations: MGIT Mycobacterial Growth Indicator 960 automated liquid culture system, DST Drug Susceptibility Testing using MGIT SIRE system. * Liquid culture testing scenario consists of sputum processing and MGIT culture, followed by Ziehl Neelsen smear microscopy on positive cultures, anti-MPB64 assays on cultures with mycobacterial growth, and MGIT SIRE DST on cultures with growth of M. tuberculosis. **Utilizes current estimates of Xpert and MTBDRplus sensitivity and specificity and prevalence of TB and drug-resistant TB in South Africa.\n\ncartridges, the cost of Xpert testing is comparable to that of conventional diagnostics, making it possible to consider replacement of sputum microscopy and culture with Xpert from a laboratory cost standpoint. We found that the cost of Xpert testing ($14.93) in a reference laboratory in South Africa was similar to performing the current reference standard of smear-microscopy followed by liquid culture and conventional DST ($16.88); MTBDRplus was found to be the most costly ($23.46) but offers the benefit of rapid isoniazid resistance testing in addition to rifampin resistance testing. Costs of molecular testing were most influenced by consumable costs which accounted for 60-80% of total costs associated with Xpert and MRTBDplus. Xpert additionally offers the benefit of reduced staff time needed for testing, with potential to increase volume of testing or increased opportunities and time to perform other diagnostic tests or laboratory activities.\nOverall, laboratories and TB control programs must balance costs with performance characteristics and the need for rapid results [15]. Both molecular tests assessed in this study offered rapid detection of M. tuberculosis and drug-resistance. However, reliance on either Xpert or MTBDRplus alone for diagnosis of TB and drugresistant TB in place of liquid culture and DST has limitations. Each has suboptimal sensitivity in individuals with smear-negative TB, and neither allows testing of second line drugs; Xpert also does not allow assessment\n\nof isoniazid mono-resistance [6,7]. To maximize detection of M. tuberculosis and drug-resistance, laboratories may choose to perform conventional culture and DST in addition to newer molecular assays. We found that intensive implementation of molecular testing in conjunction with conventional diagnostics for all sputa in order to optimize both sensitivity and speed of diagnosis would lead to significant laboratory cost increases (70-120% increase), making this option potentially unaffordable for laboratories in some settings.\nTo assist laboratories in allocation of resources, we examined the costs of multiple diagnostic algorithms. Alternative diagnostic algorithms with selective application of molecular assays may be considered in some laboratories. We found that selectively performing Xpert or MTBDRplus for smear-positive samples (while additionally performing liquid culture and DST for all samples) leads to only modest increases in laboratory costs (incremental $1.49 and $2.35 per sample for Xpert and MTBDRplus, respectively). Such a strategy would allow rapid species-level assessment of M. tuberculosis and rapid identification of drug-resistant TB in those likely to be most infectious, while also allowing performance of reference standard testing on all patients. Similarly, we found only relatively small increases to laboratory costs if molecular testing were used as a rapid alternative to conventional indirect phenotypic DST.\n\nShah et al. BMC Infectious Diseases 2013, 13:352 http://www.biomedcentral.com/1471-2334/13/352\n\nPage 6 of 7\n\nOur study has several limitations. Not all countries qualify for negotiated reduced pricing, and costs associated with labor (i.e. wages) and overhead can vary geographically. Nonetheless we provide a detailed cost-breakdown to aid generalizability and allow laboratories in other settings to estimate costs associated with implementing these emerging rapid TB diagnostic tests and provide cost estimates for Xpert testing without negotiated prices. This study was a cost-analysis from a reference laboratory perspective. Cost of transportation to a reference laboratory were not included in this analysis and may represent a significant expense and vary geographically; alternatively, implementation of Xpert at a lower level of the health system may avert specimen transport costs associated with mycobacterial culture and conventional DST that must be performed in more established laboratories. TB control programs making decisions on diagnostic algorithms must additionally consider costs associated with clinical evaluation and TB treatment in addition to laboratory costs, as well as consider the local prevalence of TB and drug-resistance and the need for rapid diagnosis. Nevertheless, our study provides important information regarding the likely diagnostic costs associated with incorporating Xpert and MTBDRplus as part of future diagnostic strategies.\nOn the other hand, our study has several strengths. We incorporated recent negotiated price reductions to aid generalizability to other low and middleincome settings, and report the component of costs attributable to consumables, equipment, and labor for each TB diagnostic system. We found that the Xpert and MTBDRplus laboratory costs are comparable to those of conventional diagnostics and should be considered as part of TB diagnostic algorithms; we additionally offer insight into the costs of alternative algorithms that some laboratories may be considering. Our results provide important information to aid future studies evaluating cost-effectiveness and implementation of emerging TB diagnostic algorithms and TB case-finding strategies.\nConclusions The cost of newer molecular diagnostic tests are comparable to conventional diagnostic methods, when paying reduced negotiated pricing for Xpert and MTBDRplus. We present detailed cost information related to implementation of these rapid molecular assays to guide laboratories seeking to scale up TB diagnostics. Overall, laboratories and TB programs must balance costs with performance characteristics and the need for rapid results. Intensive implementation of molecular assays as an addition to conventional automated liquid culture and DST may lead to significant laboratory cost\n\nincreases; selective implementation of molecular assays could be considered for some settings.\nCompeting interests The authors declare that they have no competing interests.\nAuthors\u2019 contributions MS conducted data collection, data analysis, and led manuscript writing and study design. VC, GC, GC assisted with data collection and manuscript writing. SD assisted with manuscript writing, study design, and conceived the study. All authors read and approved the final manuscript.\nAcknowledgements This work was supported by grants from the U.S. National Institutes of Health (RO1 AI51528 to S.E.D. and K23AI089259 to MS). We gratefully acknowledge the assistance of Minty van der Meulen at the National Health Laboratory Services National TB Reference Laboratory.\nAuthor details 1Johns Hopkins University School of Medicine, Department of Medicine, Division of Infectious Disease, 1503 East Jefferson St, Room 118, Baltimore, MD 21231, USA. 2The Aurum Institute for Health Research, Johannesburg, South Africa. 3National Health Laboratory Services, Johannesburg, South Africa. 4London School of Hygiene and Tropical Medicine, London, UK. 5School of Public Health, University of Witwatersrand, Johannesburg, South Africa.\nReceived: 8 February 2013 Accepted: 25 July 2013 Published: 29 July 2013\nReferences 1. WHO: Multidrug and Extensively drug-resistant TB: 2010 Global Report on\nSurveillance and Response; 2010. http://whqlibdoc.who.int/publications/2010/ 9789241599191_eng.pdf 2. Molecular Line Probe Assays for Rapid Screening of Patients at Risk of Multidrug-Resistant Tuberculosis. [http://www.who.int/tb/features_archive/ policy_statement.pdf] 3. Automated Real-Time Nucleic Acid Amplification Technology for Rapid and Simultaneous Detection of Tuberculosis and Drug Resistant Tuberculosis: Policy Statement. [http://whqlibdoc.who.int/publications/ 2011/9789241501545_eng.pdf] 4. Shah NS, Lan NT, Huyen MN, Laserson K, Iademarco MF, Binkin N, Wells C, Varma JK: Validation of the line-probe assay for rapid detection of rifampicin-resistant Mycobacterium tuberculosis in Vietnam. Int J Tuberc Lung Dis 2009, 13(2):247\u2013252. 5. Huyen MN, Tiemersma EW, Lan NT, Cobelens FG, Dung NH, Sy DN, Buu TN, Kremer K, Hang PT, Caws M, et al: Validation of the GenoType MTBDRplus assay for diagnosis of multidrug resistant tuberculosis in South Vietnam. BMC Infect Dis 2010, 10:149. 6. Barnard M, Albert H, Coetzee G, O'Brien R, Bosman ME: Rapid molecular screening for multidrug-resistant tuberculosis in a high-volume public health laboratory in South Africa. Am J Respir Crit Care Med 2008, 177(7):787\u2013792. 7. Boehme CC, Nabeta P, Hillemann D, Nicol MP, Shenai S, Krapp F, Allen J, Tahirli R, Blakemore R, Rustomjee R, et al: Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med 2010, 363(11):1005\u20131015. 8. Negotiated Prices for Xpert MTB/RIF. [http://www.finddiagnostics.org/ about/what_we_do/successes/find-negotiated-prices/xpert_mtb_rif. html] 9. Global TB Programme, World Health Organization. Laboratory services in TB control, Part II: Microscopy. WHO/TB/98.258. [http://whqlibdoc.who.int/ hq/1998/WHO_TB_98.258_(part2).pdf] 10. Kent P T KGP: Public Health mycobacteriology: a fuide for the level III laboratory. Atlanta, GA, USA: Centers for Disease Control; 1985. 11. Negotiated Prices and Country List for Line Probe Assay and associated instrumentation. [http://www.finddiagnostics.org/about/what_we_do/ successes/find-negotiated-prices/mtbdrplus.html] 12. FIND-Negotiated Prices for BACTEC and MGIT and Country List. [http:// www.finddiagnostics.org/about/what_we_do/successes/find-negotiatedprices/bactec-mgit.html]\n\nShah et al. BMC Infectious Diseases 2013, 13:352 http://www.biomedcentral.com/1471-2334/13/352\n13. Pricing to the FIND Target Market of 145 Countries. [http://www. cepheidcares.com/tb/cepheid-vision.html]\n14. Management of Drug-Resistant Tuberculosis: http://www.tbonline.info/ media/uploads/documents/mdr-tb_sa_2010.pdf\n15. Jacobson KR, Theron D, Kendall EA, Franke MF, Barnard M, van Helden PD, Victor TC, Streicher EM, Murray MB, Warren RM: Implementation of genotype MTBDRplus reduces time to multidrug-resistant tuberculosis therapy initiation in South Africa. Clin Infect Dis 2013, 56(4):503\u2013508.\ndoi:10.1186/1471-2334-13-352 Cite this article as: Shah et al.: Comparison of laboratory costs of rapid molecular tests and conventional diagnostics for detection of tuberculosis and drug-resistant tuberculosis in South Africa. BMC Infectious Diseases 2013 13:352.\n\nPage 7 of 7\n\nSubmit your next manuscript to BioMed Central and take full advantage of:\n\u2022 Convenient online submission \u2022 Thorough peer review \u2022 No space constraints or color \ufb01gure charges \u2022 Immediate publication on acceptance \u2022 Inclusion in PubMed, CAS, Scopus and Google Scholar \u2022 Research which is freely available for redistribution\nSubmit your manuscript at www.biomedcentral.com/submit\n\n\n",
    "authors": [
      "Maunank Shah",
      "Violet Chihota",
      "Gerrit Coetzee",
      "Gavin Churchyard",
      "Susan E Dorman"
    ],
    "doi": "10.1186/1471-2334-13-352",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://bmcinfectdis.biomedcentral.com/articles/10.1186/1471-2334-13-352"
  },
  {
    "key": "CGETEGP8",
    "title": "Rapid Diagnosis of Tuberculosis with the Xpert MTB/RIF Assay in High Burden Countries: A Cost-Effectiveness Analysis",
    "abstract": "Background: Xpert MTB/RIF (Xpert) is a promising new rapid diagnostic technology for tuberculosis (TB) that has characteristics that suggest large-scale roll-out. However, because the test is expensive, there are concerns among TB program managers and policy makers regarding its affordability for low- and middle-income settings.\nMethods and Findings: We estimate the impact of the introduction of Xpert on the costs and cost-effectiveness of TB care using decision analytic modelling, comparing the introduction of Xpert to a base case of smear microscopy and clinical diagnosis in India, South Africa, and Uganda. The introduction of Xpert increases TB case finding in all three settings; from 72%\u201385% to 95%\u201399% of the cohort of individuals with suspected TB, compared to the base case. Diagnostic costs (including the costs of testing all individuals with suspected TB) also increase: from US$28\u2013US$49 to US$133\u2013US$146 and US$137\u2013US$151 per TB case detected when Xpert is used \u2018\u2018in addition to\u2019\u2019 and \u2018\u2018as a replacement of\u2019\u2019 smear microscopy, respectively. The incremental cost effectiveness ratios (ICERs) for using Xpert \u2018\u2018in addition to\u2019\u2019 smear microscopy, compared to the base case, range from US$41\u2013$110 per disability adjusted life year (DALY) averted. Likewise the ICERS for using Xpert \u2018\u2018as a replacement of\u2019\u2019 smear microscopy range from US$52\u2013$138 per DALY averted. These ICERs are below the World Health Organization (WHO) willingness to pay threshold.\nConclusions: Our results suggest that Xpert is a cost-effective method of TB diagnosis, compared to a base case of smear microscopy and clinical diagnosis of smear-negative TB in low- and middle-income settings where, with its ability to substantially increase case finding, it has important potential for improving TB diagnosis and control. The extent of costeffectiveness gain to TB programmes from deploying Xpert is primarily dependent on current TB diagnostic practices. Further work is required during scale-up to validate these findings.",
    "full_text": "Rapid Diagnosis of Tuberculosis with the Xpert MTB/RIF Assay in High Burden Countries: A Cost-Effectiveness Analysis\nAnna Vassall1,2, Sanne van Kampen1, Hojoon Sohn3, Joy S. Michael4, K. R. John5, Saskia den Boon6, J. Lucian Davis7, Andrew Whitelaw8,9, Mark P. Nicol8,9, Maria Tarcela Gler10, Anar Khaliqov11, Carlos Zamudio12, Mark D. Perkins13, Catharina C. Boehme13, Frank Cobelens1*\n1 Department of Global Health, and Amsterdam Institute of Global Health and Development, Academic Medical Center, Amsterdam, The Netherlands, 2 Department of Global Health and Development, London School of Hygiene & Tropical Medicine, London, United Kingdom, 3 Department of Epidemiology and Biostatistics, McGill University, Canada, 4 Christian Medical College, Vellore, India, 5 National TB Program, Vellore, India, 6 Makerere University - University of California, San Francisco (MUUCSF) Research Collaboration, Kampala, Uganda, 7 Division of Pulmonary and Critical Care Medicine, San Francisco General Hospital, University of California, San Francisco, United States of America, 8 National Health Laboratory Service, Groote Schuur Hospital, Cape Town, South Africa, 9 Division of Medical Microbiology and Institute for Infectious Diseases and Molecular Medicine, University of Cape Town, South Africa, 10 Tropical Disease Foundation, Manila, Philippines, 11 Special Treatment Institution, Baku, Azerbaijan, 12 Instituto de Medicina Tropical Alexander von Humboldt, Universidad Peruana Cayetano Heredia, Lima, Peru, 13 Foundation for Innovative New Diagnostics (FIND), Geneva, Switzerland\nAbstract\nBackground: Xpert MTB/RIF (Xpert) is a promising new rapid diagnostic technology for tuberculosis (TB) that has characteristics that suggest large-scale roll-out. However, because the test is expensive, there are concerns among TB program managers and policy makers regarding its affordability for low- and middle-income settings.\nMethods and Findings: We estimate the impact of the introduction of Xpert on the costs and cost-effectiveness of TB care using decision analytic modelling, comparing the introduction of Xpert to a base case of smear microscopy and clinical diagnosis in India, South Africa, and Uganda. The introduction of Xpert increases TB case finding in all three settings; from 72%\u201385% to 95%\u201399% of the cohort of individuals with suspected TB, compared to the base case. Diagnostic costs (including the costs of testing all individuals with suspected TB) also increase: from US$28\u2013US$49 to US$133\u2013US$146 and US$137\u2013US$151 per TB case detected when Xpert is used \u2018\u2018in addition to\u2019\u2019 and \u2018\u2018as a replacement of\u2019\u2019 smear microscopy, respectively. The incremental cost effectiveness ratios (ICERs) for using Xpert \u2018\u2018in addition to\u2019\u2019 smear microscopy, compared to the base case, range from US$41\u2013$110 per disability adjusted life year (DALY) averted. Likewise the ICERS for using Xpert \u2018\u2018as a replacement of\u2019\u2019 smear microscopy range from US$52\u2013$138 per DALY averted. These ICERs are below the World Health Organization (WHO) willingness to pay threshold.\nConclusions: Our results suggest that Xpert is a cost-effective method of TB diagnosis, compared to a base case of smear microscopy and clinical diagnosis of smear-negative TB in low- and middle-income settings where, with its ability to substantially increase case finding, it has important potential for improving TB diagnosis and control. The extent of costeffectiveness gain to TB programmes from deploying Xpert is primarily dependent on current TB diagnostic practices. Further work is required during scale-up to validate these findings.\nPlease see later in the article for the Editors\u2019 Summary.\nCitation: Vassall A, van Kampen S, Sohn H, Michael JS, John KR, et al. (2011) Rapid Diagnosis of Tuberculosis with the Xpert MTB/RIF Assay in High Burden Countries: A Cost-Effectiveness Analysis. PLoS Med 8(11): e1001120. doi:10.1371/journal.pmed.1001120\nAcademic Editor: Douglas Wilson, Edendale Hospital, South Africa\nReceived April 7, 2011; Accepted September 30, 2011; Published November 8, 2011\nCopyright: \u00df 2011 Vassall et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.\nFunding: The Foundation for Innovative New Diagnostics (FIND) supported the study financially and had a role in data interpretation and writing the manuscript. FIND authors (MDP and CCB) had a role in data interpretation and writing the manuscript (commented on the results and the draft manuscript). The first and corresponding authors (AV and FC) had full access to all study data and, in consultation with the other authors, made a final decision to submit this work for publication.\nCompeting Interests: The authors have declared that no competing interests exist.\nAbbreviations: DALY, disability adjusted life year; DST, drug susceptibility testing; ICER, incremental cost effectiveness ratio; LJ, Lowenstein\u2013Jensen; LPA, line probe assay; MDR, multidrug resistant; MGIT, mycobacteria growth indicator tube; TB, tuberculosis; WTP, willingness to pay; Xpert, Xpert MTB/RIF.\n* E-mail: f.cobelens@aighd.org\n\nPLoS Medicine | www.plosmedicine.org\n\n1\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\nCost-Effectiveness of Xpert MTB/RIF\n\nIntroduction\nTuberculosis (TB) control in developing countries is hampered by the inadequate care that can be delivered on the basis of diagnosis by microscopy alone\u2014an issue that is acute in populations with a high prevalence of HIV and/or multidrug resistant (MDR)-TB. It is estimated that 1.7 million people died from TB in 2009, many of them remaining undiagnosed [1]. The Xpert MTB/RIF assay (referred to as Xpert in this article), is a real-time PCR assay that is a design-locked, all-within-cartridge test, and that has demonstrated high performance and could be deployed in a range of low- and middle-income settings [2,3]. It has recently been endorsed by the World Health Organization (WHO) as a promising new rapid diagnostic technology that has the potential for large-scale roll-out (www.who.int/tb/laboratory). Xpert is commercially produced and sold at concessional prices. However, because the price is considerably higher than that of smear microscopy, there is a concern among TB program managers and policy makers that Xpert may not be cost-effective in low- and middle-income settings.\nThere is little previous research into the cost-effectiveness of TB diagnostics. A study considering a hypothetical TB diagnostic found that cost-effectiveness would be maximized by highspecificity, low-cost tests deployed in regions with poor infrastructure [4]. Other studies have examined the cost-effectiveness of culture, PCR, and novel methods for drug susceptibility testing such as line-probe assays (LPA). These studies all found that these diagnostic tests are potentially cost-effective [5\u20137]. However, because of their technical requirements, mycobacterial culture, PCR, and LPA can only be deployed in specialised laboratories. We present the first (to our knowledge) economic evaluation of the Xpert rapid test for TB. [2].\nMethods\nThe aim of this study was to assess whether Xpert is likely to result in an improvement of the cost-effectiveness of TB care in low- and middle-income settings. We did this by estimating the impact of Xpert on the costs and cost-effectiveness of TB care in three countries, using decision analytic modelling, comparing the introduction of Xpert to a base case of sputum microscopy and clinical diagnosis. The model\u2019s primary outcome measure is the cost per disability adjusted life year (DALY) averted.\nOur model followed a cohort of 10,000 individuals suspected of having TB through the diagnostic and treatment pathway, estimating costs and health gains. In the diagnostic pathway, the TB cases among the individuals with suspected TB were either diagnosed as having TB or not, depending on the test sensitivities in the pathway. Similarly, individuals with suspected TB who were not TB cases may have been diagnosed as having TB, depending on the pathway\u2019s test specificities. A diagnosis of TB was followed by treatment. Individuals with suspected TB completed the pathway when they were either cured, failed treatment, died, or, if they had no TB from the start, remained without TB.\nThree different diagnostic scenarios are compared (Figure 1). The base case is defined as two sputum microscopy examinations followed, in smear-negative individuals with suspected TB, by clinical diagnosis that might include chest X-ray and antibiotic trial [8]. The inclusion of an antibiotic trial (empirical treatment with one or more broad-spectrum antibiotics to exclude other infectious causes of pulmonary disease) is no longer part of the WHO diagnostic strategy for HIV-infected patients. However, in the clinics participating in the demonstration study from which the diagnostic performance parameters were sourced [2], an antibiotic\n\ntrial was still commonly provided during the diagnostic process as an adjunct to the treatment of smear-negative individuals with suspected TB. Antibiotic trial was therefore included in the base case; the model assumed that for each country the use of antibiotic trial and chest X-ray was proportional to the observed use in the demonstration study clinics. In comparison, two alternative pathways involving Xpert were considered: (1) two smear examinations, if negative followed by Xpert on a single sputum specimen (\u2018\u2018in addition to\u2019\u2019); (2) Xpert instead of smear examination: single sputum specimen tested by Xpert for all individuals with suspected TB (\u2018\u2018replacement of\u2019\u2019).\nEach scenario included drug resistance testing of previously treated patients [9], either by conventional drug susceptibility testing (DST) or Xpert. All patients diagnosed with TB were treated using the standard WHO-recommended regimens. Patients awaiting DST results were started on first-line treatment (isoniazid [H], rifampicin [R], pyrazimamide [Z], and ethambutol [E] for 2 mo followed by HR for 4 mo for new patients, and HRZE for 3 mo with streptomycin added during the first 2 mo followed by HRE for 5 mo for patients with a history of previous TB treatment) and switched to second-line treatment when a DST result of rifampicin resistance became available. The second-line treatment regimens differed between the countries but commonly included a fluoroquinolone and an aminoglycoside (kanamycin, amikacin) or capreomycin in addition to one or more first-line drugs and ethionamode, cycloserine, and/or 4-aminosalicylic acid (PAS). If Xpert identified rifampicin resistance, this was confirmed by conventional DST or LPA as practice in the respective countries. LPA, used as a screening test on smear-positive sputum samples in South Africa, detects rifampicin resistance within 24 h by molecular methods. While awaiting this result, the patient was started on second-line treatment, but then switched to first-line treatment if resistance to rifampicin was not confirmed. TB cases that remained undiagnosed were assumed to return to the clinic after 3 mo, with 10% of undiagnosed cases becoming smearpositive within that time.\nKey model input parameters are shown in Table 1 and further details can be found in Text S1. The model was parameterised for three settings: India (low HIV prevalence, low MDR prevalence), Uganda (high HIV prevalence, low MDR prevalence), and South Africa (high HIV prevalence, high MDR prevalence). In each cohort, TB cases were characterized as: (1) new or previously treated, (2) HIV-negative or HIV-positive, and (3) MDR or drug susceptible. These proportions were sourced from country reports [1,10,11].\nDiagnostic test performance data were sourced from a demonstration study of Xpert use in nine facilities [2]. Sensitivity and specificity parameters for all diagnostic tests and procedures were calculated taking sputum culture as the reference standard. The sensitivity and specificity of Xpert and sputum microscopy (light-emitting diode [LED]) fluorescence microscopy) was based on weighted averages across the sites. Since clinical diagnostic practice of smear negatives in the base case varied considerably between sites, site-specific data were used to estimate performance of the clinical TB diagnosis. A patient was defined as having clinically diagnosed TB if microscopy was negative, but the onset of treatment preceded the availability of the culture result.\nEstimates of the economic costs of each pathway were made from a health service perspective. All costs were estimated using the ingredient costing approach. This approach identifies all the inputs (and their quantities) required to perform a test or deliver treatment and then values them to arrive at a cost per test/person treated.\n\nPLoS Medicine | www.plosmedicine.org\n\n2\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\nCost-Effectiveness of Xpert MTB/RIF\n\nFigure 1. Simplified schematic of model. doi:10.1371/journal.pmed.1001120.g001\n\nDiagnostic costs were collected at the demonstration sites. These costs included all building, overhead, staff, equipment and consumables, quality control and maintenance, and calibration inputs. The resource use associated with each test was measured through observations of practice, a review of financial reporting, and interviews with staff in the Xpert demonstration sites. Resource use measurement took into account the allocation of fixed resources between Xpert and any other uses. Estimates of device and test prices, calibration, and training costs were obtained from suppliers. Costs for treatment were estimated using drugs costs from the Global Drug Facility and the MSH International Price Tracker, and unit costs for outpatient visits and hospitalisation sourced from WHO-CHOICE [12]. A review of previous costing studies was used to validate these estimates [13\u201318]. As our constructed estimates were higher than those found in our review, we took the mid-point between our estimate and the lowest estimate found in the literature. All local costs were converted using the\n\naverage exchange rate for 2010 (imf.statex.imf.org). Where relevant, costs were annualised using a standard discount rate of 3% [19]. All costs are reported in 2010 US$. Treatment outcome probabilities were taken from published meta-analyses of clinical trials, cohort studies, and systematic reviews [20\u201328]. DALYs averted from patients being cured were estimated using the standard formula [19]. Further details can be found in Text S1.\nSince the Xpert scenarios are most likely to be more costly and more effective than the base case, an incremental cost effectiveness ratio (ICER) was calculated to describe the additional cost for any additional DALYs averted by Xpert over the base case. This ICER was then compared to WHO\u2019s suggested country-specific willingness to pay (WTP) threshold, defined as the cost per DALY averted of each country\u2019s per capita GDP (US$1,134 for India, US$5,786 for South Africa, and US$490 for Uganda in 2010). If the ICER is below this threshold the intervention is considered cost-effective.\n\nPLoS Medicine | www.plosmedicine.org\n\n3\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\nCost-Effectiveness of Xpert MTB/RIF\n\nTable 1. Model inputs: cohort composition and diagnostic parameters, by country.\n\nCohort Proportions and Diagnostic Parameters\nCohort proportions Smear-positive TB Smear-positive TB among pulmonary TB cases, HIV-negative Smear-positive TB among pulmonary TB cases, HIV-positive Previous TB treatment among pulmonary TB cases Multidrug resistance, among new TB cases Multidrug resistance, among previously treated TB cases HIV infection, among pulmonary TB cases Diagnostic parameters Sensitivity for diagnosing pulmonary TB (SEM) Xpert MTB RIF, smear-positive TB cases Xpert MTB RIF, smear-negative TB cases, HIV-negative Xpert MTB RIF, smear-negative cases, HIV-positive Smear microscopy (two slides), HIV-positive Smear microscopy (two slides), HIV-negative Mycobacterial culture Clinical diagnosis Proportion culture-positive individuals with suspected TB who had chest X-ray Proportion culture-positive individuals with suspected TB who had antibiotic trial Specificity for diagnosing pulmonary TB (SEM) Xpert MTB RIF Smear microscopy (two slides) Mycobacterial culture Clinical diagnosis Proportion culture-negative individuals with suspected TB who had chest X-ray Proportion culture-negative individuals with suspected TB who had antibiotic trial Sensitivity for detecting rifampicin-resistance (SEM) Xpert MTB RIF Conventional drug susceptibility testing Line-probe assay Specificity for detecting rifampicin-resistance (SEM) Xpert MTB RIF Drug susceptibility testing Line-probe assay Cost parameters US$ 2010 (min, max) First-line category 1 treatment: total\nFirst-line category 2 treatment: total\nCotrimoxazol preventive treatment: 1 mo Treatment of bacterial infection Chest X-ray\nSecond-line treatment total\nDALY parameters: DALYs averted (min, max) HIV positive, sputum smear-negative\nHIV negative, sputum smear-negative\n\nIndia\n0.1 0.723 0.446 0.192 0.023 0.172 0.006\n0.983 (0.005) 0.793 (0.025) 0.718 (0.040) 0.723 (0.015) 0.446 (0.036) 1 (\u2014) 0.160 (0.073) 0.032\n1\n0.990 (0.002) 1 (\u2014) 1 (\u2014) 0.942 (0.009) 0.037\n1\n0.944 (0.015) 1 (\u2014) \u2014\n0.983 (0.005) 1 (\u2014) \u2014\n227 (103, 352)\n352 (159, 546)\n4, 50 3, 66 11 (9, 13)\n2,256 (1,463, 3,050)\n9.38 (8.62, 10.39) 13.18 (12.32, 13.96)\n\nSouth Africa Uganda\n\n0.1 0.723 0.446 0.168 0.066 0.245 0.588\n\n0.1 0.723 0.446 0.073 0.011 0.117 0.593\n\n0.983 (0.005) 0.793 (0.025) 0.718 (0.040) 0.723 (0.015) 0.446 (0.036) 1 (\u2014) 0.209 (0.039) 0.262\n\n0.983 (0.005) 0.793 (0.025) 0.718 (0.040) 0.723 (0.015) 0.446 (0.036) 1 (\u2014) 0.444 (0.096) 0.867\n\n0.051\n\n0.241\n\n0.990 (0.002) 1 (\u2014) 1 (\u2014) 0.953 (0.007) 0.059\n\n0.990 (0.002) 1 (\u2014) 1 (\u2014) 0.869 (0.030) 0.790\n\n0.009\n\n0.887\n\n0.944 (0.015) \u2014 1 (\u2014)\n\n0.944 (0.015) 1 (\u2014) \u2014\n\n0.983 (0.005) \u2014 1 (\u2014)\n\n0.983 (0.005) 1 (\u2014) \u2014\n\n454(306, 602) 185 (146, 224)\n\n998 (451, 1546) 287 (130, 445)\n\n10, 53 9, 70 16 (14, 18)\n\n3, 25 2, 41 3 (2.6, 3.7)\n\n3,492 (2,068, 4,917)\n\n1,759 (1,285, 2,233)\n\n10.71 (9.85, 11.90)\n13.83 (12.83, 14.72)\n\n11.58 (10.63, 12.90)\n18.65 (17.56, 19.61)\n\nDistribution\nBeta Beta Beta Beta Beta Beta Beta\nBeta Beta Beta Beta Beta\nBeta Beta Beta\nBeta\nBeta Beta Beta\nBeta \u2014 \u2014\nBeta \u2014 \u2014\nTriangular Triangular Triangular Triangular Triangular Triangular\nTriangular Triangular\n\nSource\nModel assumption Demonstration study, all sites [2] Demonstration study, all sites [2] WHO [1] WHO [10] WHO [10], survey [11] WHO [1]\nDemonstration study, all sites [2] Demonstration study, all sites [2] Demonstration study, all sites [2] Demonstration study, all sites [2] Demonstration study, all sites [2] Model assumption Demonstration study [2] Demonstration study [2]\nDemonstration study [2]\nDemonstration study, all sites [2] Model assumption Model assumption Demonstration study [2] Demonstration study [2]\nDemonstration study [2]\nDemonstration study, all sites [2] Model assumption Model assumption\nDemonstration study, all sites [2] Model assumption Model assumption\nWHO-CHOICE [13], literature review [14\u201319] WHO-CHOICE [13], literature review [14\u201319] WHO-CHOICE [13] WHO-CHOICE [13] WHO-CHOICE [13], literature review [14\u201319] WHO-CHOICE [13], literature review [14\u201319]\nSee Text S1\nSee Text S1\n\nPLoS Medicine | www.plosmedicine.org\n\n4\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\nTable 1. Cont.\n\nCost-Effectiveness of Xpert MTB/RIF\n\nCohort Proportions and Diagnostic Parameters HIV positive, sputum smear-positive\nHIV negative, sputum smear-positive\n\nIndia\n9.67 (8.62, 10.39)\n16.43 (16.02, 16.79)\n\nSouth Africa\n11.03 (9.85, 11.90)\n17.52 (17.05, 17.93)\n\nUganda\n11.92 (10.63, 12.90)\n22.63 (22.13, 23.07)\n\nDistribution Source\n\nTriangular\n\nSee Text S1\n\nTriangular\n\nSee Text S1\n\nThe distribution column indicates which probability distribution was specified for each parameter in the Monte Carlo simulations. For triangular distributions the mode, upper and lower limit are given. All beta distributions have boundaries (0, 1). SEM, standard error of the mean. doi:10.1371/journal.pmed.1001120.t001\n\nIn the demonstration study from which our parameter estimates were sourced [2], the probability that an individual with suspected TB was a true TB case varied considerably by location; the proportion with smear-positive TB being 8.9% in India, 14.3% in South Africa, and 32.4% in Uganda. This variation probably reflects the local patterns of (self-) referral, in particular for the extremely high proportion of TB cases among the individuals with suspected TB in Uganda. Therefore to enable generalizability, we assumed a 10% proportion of smear-positive TB in individuals with suspected TB for all three countries as our point estimate with a range of 2.5% to 25% in our uncertainty and sensitivity analyses [29].\nA large number of one- and two-way sensitivity analyses were conducted to assess the robustness of our model. These analyses examine the robustness of our results when one or two parameters are varied between the outer limits of their confidence intervals. We examined the sensitivity of our results to the probability that a suspect has TB or MDR-TB or has been previously treated. We examined the impact of varying treatment costs on our results. We tested for different prices of Xpert cartridge. We examined the impact of varying the proportion of individuals with suspected TB who get chest X-ray in addition to Xpert, as physicians may continue clinical diagnosis for smear-negative TB. Similarly we examined the impact of assuming that all HIV-infected individuals with suspected TB who have negative Xpert undergo the clinical diagnosis procedure, with costs based on site-specific use of chest X-rays and antibiotics, and sensitivity and specificity based on site-specific diagnostic performance of clinical diagnosis. We assessed the sensitivity of our results to the performance of the base case in three ways: (1) assuming one instead of two smears; (2) by varying the sensitivity of smear examination; and (3) by replacing the site-specific performance estimates for clinical diagnosis with estimates averaged across the three sites. Recognising that the performance of clinical diagnosis is a trade-off between sensitivity and specificity, we varied the sensitivity and specificity in opposite directions across a plausible range of values. As physicians in the demonstration study were aware that they would receive the results of sputum culture of all individuals with suspected TB, we tested for the effect of deferring treatment decisions until the availability of culture results. For each site culture was costed and assessed on the basis of current practice. We did not include a sensitivity analysis of the use of alternatives to culture such as microscopic observation drug susceptibility test (MODS) [30], as this was not practiced on site, and we found no good source of costing data. We examined the effect of reprogramming Xpert so that no resistance result is obtained.\nIn addition, we conducted a probabilistic sensitivity analysis (Monte Carlo simulation) to explore the effect of uncertainty across our model parameters. This analysis randomly sampled each parameter in our model simultaneously from their probability\n\ndistribution (Table 1; Text S1), and repeated this 10,000 times to generate confidence intervals around our estimates of incremental cost per DALY averted.\nThe model and the analyses were constructed using TreeAge software. Percentage ranges in the text reflect ranges across countries unless stated otherwise.\nThe demonstration study was endorsed by national TB programmes of participating countries and approved by nine governing institutional review boards (IRBs). The requirement to obtain individual informed consent was waived. The costing and costeffectiveness assessments were outlined in the study protocol reviewed by the IRBs.\nResults\nThe cost for the Xpert test (including all costs, such as the cartridge, equipment, salaries) ranges from US$22.63 in India to US$27.55 in Uganda, at an Xpert cartridge price of US$19.40 (including a 25% mark-up for transportation) and US$17,000 per four-module instrument (Tables 2 and 3) [2]. This cost falls to as low as US$14.93 with volume-driven price reductions. As FIND has negotiated a fixed price for Xpert, the difference in costs between sites is primarily determined by the intensity of use of the four-module instrument. Other factors also influence costs, but to a lesser extent; these include local wage levels and the room space used. A single sputum smear examination costs between US$1.13 and US$1.63. Unit costs for culture (Lo\u00a8wenstein\u2013Jensen [LJ] or mycobacteria growth indicator tube [MGIT]) range from US$13.56 to US$18.95. Unit costs for tests that diagnose MDRTB (where relevant for all first-line drugs) range from US$20.23 for LPA only to US$44.88 for MGIT and LPA.\nThe use of Xpert substantially increases TB case finding in all three settings; from 72%\u201385% to 95%\u201399% of the TB suspect cohort (Table 4). When Xpert is deployed \u2018\u2018as a replacement of\u2019\u2019 instead of \u2018\u2018in addition to\u2019\u2019 smear microscopy, the number of TB cases detected is similar\u2014while the number of MDR-TB cases detected increases substantially. When undiagnosed TB patients are assumed not to return for diagnosis, TB case detection increases from 62%\u201376% in the base case to 86%\u201394% in the Xpert scenarios.\nThe diagnostic cost (including the costs of testing all individuals with suspected TB) per TB case detected is US$28\u2013US$49 for the base case and increases significantly to US$133\u2013US$146 and US$137\u2013US$151 when Xpert is used \u2018\u2018in addition to\u2019\u2019 and \u2018\u2018as a replacement of\u2019\u2019 smear microscopy, respectively, depending on the setting (Table 4). The resulting change in treatment costs is more moderate, due to a reduction in the numbers of false positives in the base case from clinical diagnosis. For example, in India, the percentage of treatment costs spent on false-positive diagnoses falls from 22% to 4% when Xpert is used \u2018\u2018as a replacement of\u2019\u2019 smear microscopy in comparison to the base case.\n\nPLoS Medicine | www.plosmedicine.org\n\n5\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\nCost-Effectiveness of Xpert MTB/RIF\n\nTable 2. Cost of diagnostic tests at the study sites (2010 US$).\n\nDiagnostic Test\nAFB Smear (one smear) Xpert (current pricing) US$19.4 including transport Xpert (volume.1.5 million/y) US$15.5 including transport Xpert (volume.3.0 million/y) US$11.7 including transport Culture (LJ) Culture (MGIT) Culture + DST (LJ) Culture + DST (MGIT) DST (MGIT + LPA) DST (LPA), on sputum\ndoi:10.1371/journal.pmed.1001120.t002\n\nType of Laboratory\nPeripheral/hospital Peripheral/hospital Peripheral/hospital Peripheral/hospital Reference Reference Reference Reference Reference Reference\n\nCosts per Test (2010 US$)\n\nIndia\n\nSouth Africa\n\n1.13 22.63 18.73 14.93 13.56 \u2014 22.33 \u2014 \u2014 \u2014\n\n1.58 25.90 22.00 18.20 \u2014 15.24 \u2014 41.17 33.01 20.23\n\nUganda\n1.63 27.55 23.61 19.85 15.45 18.95 23.98 44.88 38.82 21.84\n\nICERs for each Xpert scenario are presented in Table 5. The mean ICER for using Xpert \u2018\u2018in addition to\u2019\u2019 smear microscopy compared to the base case ranges from US$41 to US$110 per DALY averted depending on the setting. The mean ICER for using Xpert \u2018\u2018as a replacement of\u2019\u2019 smear microscopy ranges from US$52 to US$138 per DALY averted. The mean ICER for using Xpert as \u2018\u2018a replacement of\u2019\u2019 smear microscopy compared to using Xpert \u2018\u2018in addition to\u2019\u2019 smear microscopy ranges between US$343 and US$650. This higher ICER is due to the fact that the effectiveness gain from using Xpert as \u2018\u2018replacement of smear microscopy\u2019\u2019 is derived from additional MDR-TB cases detected, and the cost-effectiveness of treating MDR-TB is lower than that for drug-susceptible TB. All the ICERs found are well below the WTP threshold.\nThe results of the probabilistic sensitivity analysis (Monte Carlo simulation) are also shown in Table 5. Aside from the replacement of smear microscopy in Uganda all estimates remain cost-effective. Figure 2 provides an illustration of the cost-effectiveness of Xpert deployed as \u2018\u2018a replacement of\u2019\u2019 smear microscopy in comparison to the \u2018\u2018in addition to\u2019\u2019 scenario for a range of WTP thresholds. This graph, known as an acceptability curve, shows that if the WTP is US$490 in Uganda, there is around a 75% probability that Xpert as a replacement of smear is cost-effective when compared to the \u2018\u2018in addition to\u2019\u2019 scenario.\n\nTable 3. Cost of Xpert (current pricing) by input type (2010 US$).\n\nInput Type\nOverhead Building space Equipment Staff Reagents and chemicals Consumables Total\n\nCosts per Test (2010 US$)\n\nIndia\n\nSouth Africa\n\nUganda\n\n0.18 0.02 2.84 0.11 19.40 0.07 22.63\n\n0.88 0.08 3.50 1.82 19.40 0.22 25.90\n\n0.40 0.12 7.00 0.24 19.40 0.38 27.55\n\ndoi:10.1371/journal.pmed.1001120.t003\n\nNearly all of our one- and two-way sensitivity analyses did not increase the ICER compared to the base case of either Xpert scenario above the WTP threshold (Table 6). Figure 3 shows ICER variation when parameters for the suspect population and the performance of the base case change. Varying the true proportion of those with TB and MDR-TB in the cohort has little effect on our results, although Xpert ICERs substantially worsen when the proportion of smear-positive TB cases becomes 5% or less (translating into 7%\u20139% with any type of TB). Varying assumptions on the performance of the base case alters ICERs substantially. Increasing the sensitivity of smear examination reduces the cost-effectiveness of Xpert, but not below the WTP threshold. If clinical diagnosis has a higher specificity and lower sensitivity than in our study sites, Xpert ICERs worsen, but also remain below the WTP threshold. But, if clinical diagnosis has a lower specificity and higher sensitivity than in our study sites, ICERs for Xpert substantially improve. Adding chest X-ray for 50% of the individuals with suspected TB tested by Xpert has limited impact on the cost-effectiveness of Xpert. Adding clinical diagnosis for all HIV-positive individuals with suspected TB with a negative Xpert result has no or limited effect in India and South Africa, but doubles ICERs for Xpert in Uganda (although not above the WTP threshold). This reflects differences in HIV prevalence as well as relatively high cost and low specificity of clinical diagnosis in Uganda owing to more extensive use of X-ray. Incorporating the cost of culture and increasing the proportion of TB diagnosis based on the culture result, has a mixed effect. Xpert remains cost-effective up until the point where 40%\u201370% of patients receive a culture-based diagnosis. Above proportions of 50%\u201390%, the base case becomes more effective. If however, culture performance is less than 100%, the base case does not become more effective than the Xpert-based scenarios until nearly 100% of patients receive a culture-based diagnosis (unpublished data).\nDiscussion\nOur results suggest that Xpert is likely to be more cost-effective than a base case of smear microscopy and clinical diagnosis of smear-negative TB. The extent and type of cost-effectiveness gain from deploying Xpert is dependent on a number of different setting-specific factors. First and foremost of these factors is the performance of current TB diagnostic practice. Where the sensitivity of current practice is low, but specificity high, Xpert\n\nPLoS Medicine | www.plosmedicine.org\n\n6\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\nTable 4. Cohort, cases detected, total cohort costs, and costs per case detected.\n\nPLoS Medicine | www.plosmedicine.org\n\n7\n\nCountry\n\nScenario\n\nCohort\n\nIndia South Africa Uganda\n\nBase case In addition to smear Replacement of smear Base case In addition to smear Replacement of smear Base case In addition to smear\n\nTuberculosis (MDR) Tuberculosis (no MDR) No tuberculosis Total Tuberculosis (MDR) Tuberculosis (no MDR) No tuberculosis Total Tuberculosis (MDR) Tuberculosis (no MDR) No tuberculosis Total Tuberculosis (MDR) Tuberculosis (no MDR) No tuberculosis Total Tuberculosis (MDR) Tuberculosis (no MDR) No tuberculosis Total Tuberculosis (MDR) Tuberculosis (no MDR) No tuberculosis Total Tuberculosis (MDR) Tuberculosis (no MDR) No tuberculosis Total Tuberculosis (MDR) Tuberculosis (no MDR) No tuberculosis Total\n\nn Individuals among the Cohort Who Have TB\n\nPercent Total TB of TB Total MDR Cases Cases Cases Detected Detected Detected\n\nPercent of MDR Cases Detected\n\n72 1,318 8,611 10,000 72 1,318 8,611 10,000 72 1,318 8,611 10,000 184 1,729 8,087 10,000 184 1,729 8,087 10,000 184 1,729 8,087 10,000 36 1,882 8,082 10,000 36 1,882 8,082 10,000\n\n59\n\n82\n\n38\n\n52\n\n1,079 82\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n1,138 82\n\n38\n\n\u2014\n\n71\n\n99\n\n49\n\n68\n\n1,300 99\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n1,371 99\n\n49\n\n71\n\n99\n\n67\n\n93\n\n1,298 99\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n1,369 99\n\n67\n\n131\n\n72\n\n56\n\n31\n\n1,237 72\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n1,368 72\n\n56\n\n175\n\n95\n\n112\n\n61\n\n1,649 95\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n1,824 95\n\n112\n\n175\n\n95\n\n165\n\n90\n\n1,645 95\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n1,820 95\n\n165\n\n30\n\n85\n\n14\n\n38\n\n1,594 85\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n1,625 85\n\n14\n\n34\n\n95\n\n22\n\n63\n\n1,794 95\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n1,828 95\n\n22\n\nTotal Diagnostic Costs (2010 US$)\n1,077 8,412 46,106 55,595 2,335 13,831 184,298 200,464 3,038 28,986 174,538 206,562 2,345 13,772 22,014 38,131 7,131 30,341 205,858 243,331 9,504 46,866 193,053 249,423 499 11,282 51,565 63,345 1,392 34,694 230,369 266,455\n\nDiagnostic Cost per TB Case Detected, Excluding MDR (US$ 2010)\n\u2014 \u2014 \u2014 49 \u2014 \u2014 \u2014 146 \u2014 \u2014 \u2014 151 \u2014 \u2014 \u2014 28 \u2014 \u2014 \u2014 133 \u2014 \u2014 \u2014 137 \u2014 \u2014 \u2014 39 \u2014 \u2014 \u2014 146\n\nAdditional Diagnostic Cost per MDR Case Detected (2010 US$)\n\nTreatment Costs (2010 US$)\n\n\u2014\n\n89,223\n\n\u2014\n\n268,122\n\n\u2014\n\n100,759\n\n165\n\n458,103\n\n\u2014\n\n115,932\n\n\u2014\n\n325,381\n\n\u2014\n\n22,414\n\n116\n\n463,727\n\n\u2014\n\n151,603\n\n\u2014\n\n328,669\n\n\u2014\n\n22,414\n\n24\n\n502,687\n\n\u2014\n\n230,989\n\n\u2014\n\n659,365\n\n\u2014\n\n156,213\n\n86\n\n1,046,567\n\n\u2014\n\n423,146\n\n\u2014\n\n882,010\n\n\u2014\n\n45,788\n\n57\n\n1,350,945\n\n\u2014\n\n583,064\n\n\u2014\n\n880,190\n\n\u2014\n\n45,788\n\n30\n\n1,509,043\n\n\u2014\n\n26,422\n\n\u2014\n\n282,928\n\n\u2014\n\n171,803\n\n163\n\n481,154\n\n\u2014\n\n41,123\n\n\u2014\n\n320,685\n\n\u2014\n\n14,908\n\n124\n\n376,717\n\nTreatment Costs Percent of Total Cohort\n19 59 22 100 25 70 5 100 30 65 4 100 22 63 15 100 31 65 3 100 39 58 3 100 5 59 36 100 11 85 4 100\n\nCost-Effectiveness of Xpert MTB/RIF\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\nTable 4. Cont.\n\nn Individuals among the Cohort Who Have TB\n\nPercent Total TB of TB Total MDR Cases Cases Cases Detected Detected Detected\n\nPercent of MDR Cases Detected\n\nTotal Diagnostic Costs (2010 US$)\n\nDiagnostic Cost per TB Case Detected, Excluding MDR (US$ 2010)\n\nAdditional Diagnostic Cost per MDR Case Detected (2010 US$)\n\nPLoS Medicine | www.plosmedicine.org\n\nTreatment Costs (2010 US$)\n\nTreatment Costs Percent of Total Cohort\n\nCountry\n\nScenario\n\nCohort\n\nReplacement of smear\n\nTuberculosis (MDR) Tuberculosis (no MDR) No tuberculosis Total\n\n36\n\n34\n\n95\n\n32\n\n90\n\n1,882\n\n1,790 95\n\n\u2014\n\n\u2014\n\n1,849 57,204 217,185 276,238\n\n\u2014\n\n\u2014\n\n56,488\n\n14\n\n\u2014\n\n\u2014\n\n322,502\n\n82\n\n8,082\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n14,908\n\n4\n\n10,000\n\n1,824 95\n\n32\n\n151\n\n27\n\n393,899\n\n100\n\ndoi:10.1371/journal.pmed.1001120.t004\n\nCost-Effectiveness of Xpert MTB/RIF\n\nhas a substantial impact on effectiveness. Where the sensitivity of current practice is high, but specificity low, Xpert will lower treatment costs by reducing the number of false positives. This latter effect is illustrated by the case of Uganda, where the model predicts a reduction in the treatment costs of false positives from US$171,803 to US$14,908, contributing to the overall reduction in treatment costs.\nOther factors that are likely to determine the extent of costeffectiveness gain include the proportion of those co-infected with HIV and the proportion of those with MDR-TB, and the numbers of true TB cases in the suspect population. However, our results show that increasing proportions of HIV in the suspect population will not always reduce the ICER of Xpert (Figure 3). This finding is counter-intuitive. One would expect the cost-effectiveness of a diagnostic test that diagnoses smear-negative TB to improve with increases in HIV prevalence. However, as the proportion of individuals co-infected with HIV in the suspect population increases, so the sensitivity of Xpert decreases. Depending on the relative costs and performance of the base case, this countereffect means that the relationship between HIV prevalence and Xpert\u2019s cost-effectiveness is weaker than anticipated.\nNor can we conclude on the direction of the relationship between cost-effectiveness gain and the level of prevalence of MDR-TB in the suspect population at this time. Our model demonstrates that when transmission effects are excluded, the ICER of Xpert increases as the MDR-TB prevalence increases (Figure 3). This result occurs because although the effectiveness of Xpert increases with a higher MDR-TB prevalence, the ICER of treating MDR-TB is higher than that of drug susceptible TB, thus countering the gain from increased effectiveness.\nUnsurprisingly, we also find that higher proportions of TB cases in the suspect population improve the cost-effectiveness of Xpert. The cost per TB case detected will also decrease with increases in TB prevalence. As TB programmes already fund elements of the base case, cost-effectiveness may therefore be initially improved by using existing diagnostic tools, such as X-ray and clinical scores, to screen the TB suspect population prior to Xpert. In the longer run, however, the expansion of X-ray as a permanent approach for suspect screening is unlikely to be cost-effective, and further work examining alternative screening approaches may be required. Moreover, different approaches are likely to be adopted for specific suspect populations, most notably those already known to be HIV infected, those who have already failed treatment, and those at a high risk of MDR-TB. We therefore recommend that further work is conducted to explore the impact on costeffectiveness of different algorithms when Xpert is applied to more limited suspect groups.\nA number of factors limit our analysis. Firstly, the assumption of no transmission effects or additional mortality benefit from early diagnosis is a conservative approach and will underestimate the cost-effectiveness of Xpert\u2014particularly where the introduction of Xpert is likely to increase the numbers of drug-resistant patients who are appropriately and rapidly treated. Likewise, we do not factor in patient costs. A full societal evaluation would make all options less cost-effective, but Xpert is likely to fare better than alternatives, as it requires fewer patient visits. In addition, if Xpert can achieve earlier diagnosis, substantial reductions in patient costs prior to treatment may be achieved [31]. The reference standard for the test performance parameters in our model did not include culture-negative TB based on response to treatment, because this diagnostic category will include cases with no TB or extrapulmonary TB that cannot be diagnosed by sputum-based tests. This situation may have lead to overestimation of the sensitivity\n\n8\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\nCost-Effectiveness of Xpert MTB/RIF\n\nTable 5. Cost per DALY (US$ 2010).\n\nCountry India\nSouth Africa\nUganda\n\nScenario\nBase case\nIn addition to smear\nReplacement of smear\nBase case\nIn addition to smear\nReplacement of smear\nBase case\nIn addition to smear\nReplacement of smear\n\nTotal Cost 513,698 664,191\n709,248\n1,084,698 1,594,276\n1,758,467\n544,499 643,172\n670,137\n\nTotal DALYS 17,133 19,887\n20,019\n15,805 20,420\n20,702\n22,182 24,570\n24,611\n\nCost per DALY 30 33\n35\n69 78\n85\n25 26\n27\n\nICER Compared Monte Carlo\n\nto Base Case, Simulation ICER,\n\nMean\n\nMedian (2.5, 97.5)\n\n\u2014\n\n\u2014\n\n55\n\n54 (40, 70)\n\n68\n\n68 (51, 87)\n\n\u2014\n\n\u2014\n\n110\n\n109 (88, 133)\n\n138\n\n136 (105, 172)\n\n\u2014\n\n\u2014\n\n41\n\n34 (23, 69)\n\n52\n\n37 (0, 73)\n\ndoi:10.1371/journal.pmed.1001120.t005\n\nICER Compared to in Addition to, Mean\n\u2014\n\u2014\n\nMonte Carlo Simulation ICER, Median (2.5, 97.5)\n\u2014\n\u2014\n\n343\n\n361 (239, 578)\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n582\n\n594 (353, 956)\n\n\u2014\n\n\u2014\n\n\u2014\n\n\u2014\n\n650\n\n276 (21895, 2,406)\n\nand underestimation of the specificity of Xpert. Owing to lack of evidence, we only included one repeat visit for false negatives in our model, to capture those who quickly progress to smearpositive TB. This number may be insufficient and miss both the additional costs and effectiveness of further repeated visits. On the other hand, our assumption that 100% of false negatives still alive and with TB after 3 mo have a repeat visit may be an overestimation, thereby inflating ICERs for the Xpert scenarios. We assumed that a negative Xpert result does not lead to further TB diagnostic procedures. This assumption may not be true in practice, in particular not for HIV-infected individuals with suspected TB [32]. Our sensitivity analyses show that adding clinical diagnostic procedures for this group can substantially reduce cost-effectiveness of Xpert when HIV prevalence is high and X-ray is used extensively. Also because of the lack of data, we have not included a high MDR-TB, but low HIV-prevalence setting. This lack of data restricts our ability to generalise findings to all low- and middle-income settings, particularly the former Soviet states, where this epidemiological pattern is common in suspect populations. Finally, our sensitivity analysis demonstrates that Xpert may not be cost-effective when compared to a base case in which a high proportion of smear-negative TB cases are diagnosed by culture. However, this result is based on our assumption that culture performs at 100% sensitivity and specificity. In addition, we did not include costs of specimen transport, increased risk of false-negative cultures or contamination, reduced sensitivity when only one specimen is cultured, and possible delay effects on mortality and patient drop out. All these simplifications will inflate the cost-effectiveness of a base case that includes culture.\nAs is standard practice, we determine cost-effectiveness in comparison to the WHO WTP threshold. Unfortunately, achieving this threshold does not mean that the resources are available in low- and middle-income countries, merely that Xpert should be afforded [33]. In reality, resourcing for tuberculosis services in low- and middle-income countries is extremely constrained. Countries may therefore need to prioritise. In terms of priorities, suspect populations with a high likelihood of TB, particularly in settings with high HIV and MDR-TB prevalence,\n\nare an obvious choice. However, our findings illustrate that it is also important to balance these factors with the current performance of the existing diagnostic pathway. Countries or areas that have the weakest performance in terms of diagnosing smear-negative cases may benefit the most, even when they have relatively low levels of MDR-TB and HIV; although additional investment may be required to strengthen aspects of the health system to ensure that Xpert can be implemented successfully. Funding Xpert may also mean that scarce resources are not made available to other equally deserving areas. Care must therefore be exercised to take into account the broader tuberculosis control and health system needs of any particular setting when funding Xpert.\nOur model is robust given the current evidence and data available. However, key data in this area\u2014particularly on the characteristics of TB suspect populations, the feasibility of implementing Xpert at scale, and the extent to which clinicians allow diagnostic test results to influence treatment decisions\u2014 remain scarce. Moreover, it is likely that there will be costs associated with Xpert scale-up that we cannot predict at this point. Although our model strongly suggests that Xpert will be costeffective in a wide variety of settings, Xpert scale-up will substantially increase TB diagnostic costs. Given this increase, and the current data paucity, we recommend careful monitoring and evaluation of initial roll-out before full scale-up. Funding should be provided for implementation studies, including pragmatic trials, in a number of countries to accelerate this process. As we did not assess cost-effectiveness in a setting with high MDR but low HIV prevalence, we also recommend additional economic modelling studies before embarking on rollout in these settings, taking into consideration operational factors that may affect outcomes such as patient drop-out and physician behavior [34]. Finally, although Xpert is a highly promising technology, there is still room for improvement in TB diagnostics. Xpert has incomplete sensitivity for smear-negative TB and rifampicin resistance and does not detect resistance to isoniazid and other drugs. Other promising tests, such as microscopic observation drug susceptibility test (MODS) [35], should be evaluated for their cost-effectiveness, including comparisons with\n\nPLoS Medicine | www.plosmedicine.org\n\n9\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\nCost-Effectiveness of Xpert MTB/RIF\n\nFigure 2. Cost-effectiveness acceptability curves. ICER \u2018\u2018replacement of smear\u2019\u2019 compared with \u2018\u2018in addition to smear.\u2019\u2019 doi:10.1371/journal.pmed.1001120.g002\n\nXpert. Our finding should not discourage investment in other promising new TB diagnostic technologies, particularly those that further improve the diagnostic sensitivity and detection of wider forms of drug resistance and can be implemented at peripheral health care level at low cost.\nConclusion Despite the fact that there is considerable concern from policy\nmakers about the costs and affordability of new diagnostic technologies in low- and middle-income countries, our results\n\nsuggest that Xpert is likely to be a highly cost-effective investment. If demonstrated test performance is maintained at scale, Xpert has the potential to substantially increase TB case detection. Moreover, in the settings modelled, TB treatment costs are not predicted to substantially increase with the introduction of Xpert; instead, treatment is likely to be switched from those who do not benefit from treatment, to those who do. Our results suggest that funding should be provided to initiate the roll-out of Xpert in lowand middle-income countries, as a promising means of enabling access to effective treatment for all those with the disease. We\n\nPLoS Medicine | www.plosmedicine.org\n\n10\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\nTable 6. Costs per DALY 2010 US$: sensitivity analyses.\n\nPLoS Medicine | www.plosmedicine.org\n\n11\n\nAssumption\n\nICER Compared to:\n\nPrimary estimate\nReprogrammed test: no signal MDR\nClinical diagnosis performance pooled across countries\nProportion retreatment doubles\nCartridge cost reduces to US$11.70\n50% of individuals with suspected TB have X-ray added to Xpert\nHIV-infected individuals with suspected TB have clinical diagnosis when Xpert is negative\nUndiagnosed patients with TB do not return for diagnosis\nSingle smear examination in base case\n60% of case receive culture diagnosis\nSensitivity of smear examination increase by 15%\nNA, not available. doi:10.1371/journal.pmed.1001120.t006\n\nBase case In addition to smear Base case In addition to smear Base case\nIn addition to smear Base case In addition to smear Base case In addition to smear Base case\nIn addition to smear Base case\nIn addition to smear Base case\nIn addition to smear\nBase case In addition to smear Base case\nIn addition to smear Base case\nIn addition to smear\n\nBase Case \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014\n\u2014 \u2014 \u2014 \u2014 \u2014 \u2014\n\nIndia\n\nIn Addition to Smear\n\nReplacement of Smear\n\n55\n\n68\n\n343\n\n50\n\n51\n\n107\n\n62\n\n78\n\n342\n\n115\n\n119\n\n170\n\n42\n\n54\n\n\u2014\n\n318\n\n73\n\n87\n\n\u2014\n\n378\n\n55\n\n68\n\n343\n\n50\n\n67\n\n\u2014\n\nDominated by in\n\naddition to scenario\n\n48\n\n58\n\n\u2014\n\n343\n\nDominates base case\n\u2014\n\nDominates base case\n343\n\n106\n\n130\n\n\u2014\n\n343\n\nBase Case \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014 \u2014\n\u2014 \u2014 \u2014 \u2014 \u2014 \u2014\n\nSouth Africa\n\nIn Addition Replacement to Smear of Smear\n\n110\n\n138\n\n\u2014\n\n582\n\nBase Case\n\u2014 \u2014\n\n87\n\n86\n\n\u2014\n\n\u2014\n\nNA\n\n\u2014\n\n89\n\n121\n\n\u2014\n\n\u2014\n\n582\n\n\u2014\n\n209\n\n220\n\n\u2014\n\n\u2014\n\n334\n\n\u2014\n\n102\n\n129\n\n\u2014\n\n570\n\n\u2014\n\n126\n\n154\n\n\u2014\n\n\u2014\n\n606\n\n\u2014\n\n132\n\n157\n\n\u2014\n\n610\n\n\u2014\n\n109\n\n138\n\n\u2014\n\n\u2014\n\n1,442\n\n\u2014\n\n105\n\n130\n\n\u2014\n\n\u2014\n\n582\n\n\u2014\n\n67\n\n311\n\n\u2014\n\n\u2014\n\n582\n\n\u2014\n\n131\n\n165\n\n\u2014\n\n\u2014\n\n582\n\n\u2014\n\nUganda\n\nIn Addition to Smear\n\nReplacement of Smear\n\n41\n\n52\n\n\u2014\n\n650\n\n37\n\n40\n\n\u2014\n\n289\n\n53\n\n58\n\n\u2014\n\n650\n\n67\n\n73\n\n\u2014\n\n200\n\n26\n\n36\n\n\u2014\n\n561\n\n47\n\n58\n\n\u2014\n\n686\n\n82\n\n90\n\n706\n\n33\n\n43\n\n\u2014\n\nDominated\n\nby in addition\n\nto scenario\n\n40\n\n48\n\n\u2014\n\n650\n\nBase case more cost-effective\n\u2014\n\nBase case more cost-effective\n650\n\n59\n\n74\n\n\u2014\n\n650\n\nCost-Effectiveness of Xpert MTB/RIF\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\nCost-Effectiveness of Xpert MTB/RIF\n\nFigure 3. Selected sensitivity analyses. Sensitivity of the model for the prevalence of tuberculosis, for the prevalence of multidrug-resistant tuberculosis, and for the accuracy of clinical diagnosis. Patterns of ICERs in 2010 US$ for varying the proportion of individuals with suspected TB in the cohort who have smear-positive TB (A, D, G); for varying the proportion of new patients with TB who have multidrug-resistant TB (MDR-TB, B, E, H); and for varying the specificity of the clinical diagnosis of TB in the base case (C, F, I). (A, B, and C), South Africa; (E,D, and F), India; (G, H, and I), Uganda. Black lines, Xpert assay in addition to sputum smear examination; grey lines, Xpert assay as replacement of sputum smear examination. The proportion of individuals with suspected TB in the cohort who have smear-negative TB varies along with the proportion of individuals with suspected TB in the cohort who have smear-positive TB in a linear manner, depending on the HIV-infection prevalence (A, D, G; see Table 1 and Text S1). Similarly, the proportion of previously treated patients with TB who have MDR-TB varies linearly with the proportion of new patients with TB who have MDR-TB (B, E, H; see Table 1 and Text S1). The sensitivity of clinical diagnosis in the base case varies inversely with the specificity (range, 8%\u2013 80%; C, F, I). doi:10.1371/journal.pmed.1001120.g003\n\nrecommend, however, that this roll-out is carefully evaluated to validate our results before full scale-up\u2014to ensure that Xpert implementation is done in a way that does not negatively impact TB programmes, their funding, and the health systems that support them.\nSupporting Information\nText S1 Details of model assumptions, test turnaround times (Table S[A]), treatment outcome probabilities (Table S[B]), probabilities of death and spontaneous recovery with falsenegative tuberculosis diagnosis (Table S[C]), and variables used in the DALY calculations (Table S[D]). (DOC)\n\nAuthor Contributions\nConceived and designed the experiments: AV Svk FC. Analyzed the data: AV Svk FC. Wrote the first draft of the manuscript: AV Svk FC. Contributed to the writing of the manuscript: AV SvK HS JSM KRJ SdB JLD AW MPN MTG AK CZ MDP CCB FC. ICMJE criteria for authorship read and met: AV SvK HS JSM KRJ SdB JLD AW MPN MTG AK CZ MDP CCB FC. Agree with manuscript results and conclusions: AV SvK HS JSM KRJ SdB JLD AW MPN MTG AK CZ MDP CCB FC. AV, Svk, and FC designed the overall study and the decision analytic model. SvK constructed the model. AV, SvK, and HS designed and carried out cost data collection. AV, SvK, and FC analyzed the final data and developed the first manuscript draft. JSM, KRJ, SdB, JLD, AW, MPN, MTG, AK, CZ, CCB, and MDP contributed to collecting and analyzing the diagnostic field data on which the decision models were based. All authors contributed to data collection, interpretation of data and revision of the article.\n\nReferences\n1. WHO | Global tuberculosis control 2010 (z.d.). Available: http://www.who.int/ tb/publications/global_report/2010/en/index.html. Accessed 24 January 2011.\n\n2. Boehme CC, Nicol M, Nabeta P, Michael JS, Gotuzzo E, et al. (2011) Feasibility, diagnostic accuracy, and effectiveness of decentralised use of the\n\nPLoS Medicine | www.plosmedicine.org\n\n12\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\nCost-Effectiveness of Xpert MTB/RIF\n\nXpert MTB/RIF test for diagnosis of tuberculosis and multidrug resistance: a multicentre implementation study. Lancet 377: 1495\u20131505. 3. Boehme CC, Nabeta P, Hillemann D, Nicol MP, Shenai S, et al. (2010) Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med 363: 1005\u20131015. 4. Dowdy DW, O\u2019Brien MA, Bishai D (2008) Cost-effectiveness of novel diagnostic tools for the diagnosis of tuberculosis. Int J Tuberc Lung Dis 12: 1021\u20131029. 5. Dowdy DW, Lourenc\u00b8o MC, Cavalcante SC, Saraceni V, King B, et al. (2008) Impact and cost-effectiveness of culture for diagnosis of tuberculosis in HIV-infected Brazilian adults. PLoS ONE 3: e4057. doi:10.1371/journal. pone.0004057. 6. Acuna-Villaorduna C, Vassall A, Henostroza G, Seas C, Guerra H, et al. (2008) Cost-effectiveness analysis of introduction of rapid, alternative methods to identify multidrug-resistant tuberculosis in middle-income countries. Clin Infect Dis 47: 487\u2013495. 7. Roos BR, van Cleeff MR, Githui WA, Kivihya-Ndugga L, Odhiambo JA, et al. (1998) Cost-effectiveness of the polymerase chain reaction versus smear examination for the diagnosis of tuberculosis in Kenya: a theoretical model. Int J Tuberc Lung Dis 2: 235\u2013241. 8. WHO (2007) Improving the diagnosis and treatment of smear-negative pulmonary and extrapulmonary tuberculosis among adults and adolescents. Recommendations for HIV-prevalent and resource-constrained settings. Geneva: World Health Organization. 9. WHO | Guidelines for surveillance of drug resistance in tuberculosis (z.d.). Available: http://www.who.int/tb/publications/2009/surveillance_guidelines/ en/index.html. Accessed 24 January 2011. 10. WHO | Surveillance of drug resistance in tuberculosis (z.d.). Available: http:// www.who.int/tb/publications/mdr_surveillance/en/index.html. Accessed 24 January 2011. 11. Lukoye D, Cobelens FGJ, Ezati N, Kirimunda S, Adatu FE, et al. (2011) Rates of anti-tuberculosis drug resistance in Kampala-Uganda are low and not associated with HIV infection. PLoS ONE 6: e16130. doi:10.1371/journal. pone.0016130. 12. WHO | WHO-CHOICE (z.d.). Available: http://www.who.int/choice/en/. Accessed 18 April 2011. 13. van Cleeff MRA, Kivihya-Ndugga LE, Meme H, Odhiambo JA, Klatser PR (2005) The role and performance of chest X-ray for the diagnosis of tuberculosis: a cost-effectiveness analysis in Nairobi, Kenya. BMC Infect Dis 5: 111. 14. Floyd K, Arora VK, Murthy KJR, Lonnroth K, Singla N, et al. (2006) Cost and cost-effectiveness of PPM-DOTS for tuberculosis control: evidence from India. Bull World Health Organ 84: 437\u2013445. 15. Okello D, Floyd K, Adatu F, Odeke R, Gargioni G (2003) Cost and costeffectiveness of community-based care for tuberculosis patients in rural Uganda. Int J Tuberc Lung Dis 7: S72\u2013S79. 16. Pantoja A, Lo\u00a8nnroth K, Lal SS, Chauhan LS, Uplekar M, et al. (2009) Economic evaluation of public-private mix for tuberculosis care and control, India. Part II. Cost and cost-effectiveness. Int J Tuberc Lung Dis 13: 705\u2013712. 17. Sinanovic E, Floyd K, Dudley L, Azevedo V, Grant R, et al. (2003) Cost and cost-effectiveness of community-based care for tuberculosis in Cape Town, South Africa. Int J Tuberc Lung Dis 7: S56\u201362. 18. Sinanovic E, Kumaranayake L (2006) Financing and cost-effectiveness analysis of public-private partnerships: provision of tuberculosis treatment in South Africa. Cost Eff Resour Alloc 4: 11. doi:10.1186/1478-7547-4-11.\n\n19. Tan-Torres Edejer TBR, Adam T, Hutubessy R, Acharya A, Evans DB, et al. (2003) WHO Guide to Cost-effectiveness Analysis. Geneva: World Health Organization.\n20. Lew W, Pai M, Oxlade O, Martin D, Menzies D (2008) Initial drug resistance and tuberculosis treatment outcomes: systematic review and meta-analysis. Ann Intern Med 149: 123\u2013134.\n21. Espinal MA, Kim SJ, Suarez PG, Kam KM, Khomenko AG, et al. (2000) Standard short-course chemotherapy for drug-resistant tuberculosis: treatment outcomes in 6 countries. JAMA 283: 2537\u20132545.\n22. Menzies D, Benedetti A, Paydar A, Royce S, Madhukar P, et al. (2009) Standardized treatment of active tuberculosis in patients with previous treatment and/or with mono-resistance to isoniazid: a systematic review and meta-analysis. PLoS Med 6: e1000150. doi:10.1371/journal.pmed.1000150.\n23. Nathanson E, Lambregts-van Weezenbeek C, Rich ML, Gupta R, Bayona J, et al. (2006) Multidrug-resistant tuberculosis management in resource-limited settings. Emerging Infect Dis 12: 1389\u20131397.\n24. Akksilp S, Karnkawinpong O, Wattanaamornkiat W, Viriyakitja D, Monkongdee P, et al. (2007) Antiretroviral therapy during tuberculosis treatment and marked reduction in death rate of HIV-infected patients, Thailand. Emerging Infect Dis 13: 1001\u20131007.\n25. Varma JK, Nateniyom S, Akksilp S, Mankatittham W, Sirinak C, et al. (2009) HIV care and treatment factors associated with improved survival during TB treatment in Thailand: an observational study. BMC Infect Dis 9: 42.\n26. Wells CD, Cegielski JP, Nelson LJ, Laserson KF, Holtz TH, et al. (2007) HIV infection and multidrug-resistant tuberculosis: the perfect storm. J Infect Dis 196 Suppl 1: S86\u2013S107.\n27. Seung KJ, Omatayo DB, Keshavjee S, Furin JJ, Farmer PE, et al. (2009) Early outcomes of MDR-TB treatment in a high HIV-prevalence setting in Southern Africa. PLoS ONE 4: e7186. doi:10.1371/journal.pone.0007186.\n28. Abdool Karim SS, Naidoo K, Grobler A, Padayatchi N, Baxter C, et al. (2010) Timing of initiation of antiretroviral drugs during tuberculosis therapy. N Engl J Med 362: 697\u2013706.\n29. Rieder HL, Enarson DA (1995) A computer-based ordering system for supplies in national tuberculosis programs. Tuber Lung Dis 76: 450\u2013454.\n30. Moore DAJ, Evans CAW, Gilman RH, Caviedes L, Coronel J, et al. (2006) Microscopic-observation drug-susceptibility assay for the diagnosis of TB. N Engl J Med 355: 1539\u20131550.\n31. Vassall A, Seme A, Compernolle P, Meheus F (2010) Patient costs of accessing collaborative tuberculosis and human immunodeficiency virus interventions in Ethiopia. Int J Tuberc Lung Dis 14: 604\u2013610.\n32. Holtz TH, Kabera G, Mthiyane T, Zingoni T, Nadesan S, et al. (2011) Use of a WHO-recommended algorithm to reduce mortality in seriously ill patients with HIV infection and smear-negative pulmonary tuberculosis in South Africa: an observational cohort study. Lancet Infect Dis 11: 533\u2013540.\n33. Shillcutt SD, Walker DG, Goodman CA, Mills AJ (2009) Cost effectiveness in low- and middle-income countries: a review of the debates surrounding decision rules. Pharmacoeconomics 27: 903\u2013917.\n34. Dowdy DW, Cattamanchi A, Steingart KR, Pai M (2011) Is scale-up worth it? Challenges in economic analysis of diagnostic tests for tuberculosis. PLoS Med 8: e1001063. doi:10.1371/journal.pmed.1001063.\n35. Oberhelman RA, Soto-Castellares G, Gilman RH, Caviedes L, Castillo ME, et al. (2010) Diagnostic approaches for paediatric tuberculosis by use of different specimen types, culture methods, and PCR: a prospective case-control study. Lancet Infect Dis 10: 612\u2013620.\n\nPLoS Medicine | www.plosmedicine.org\n\n13\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\nCost-Effectiveness of Xpert MTB/RIF\n\nEditors\u2019 Summary\nBackground. Tuberculosis (TB) is a bacterial disease that infects one-third of the world\u2019s population. The disease is caused by Mycobacterium tuberculosis, a bacterium that most commonly infects the lungs (known as pulmonary TB) and is transmitted from person to person when an infected individual coughs, sneezes, or talks. The symptoms of TB include chest pain, weight loss, fever, and a persistent cough that sometimes contains blood. Only 5%\u201310% of people who are infected with TB become sick or infectious, but people with weakened immune systems, such as individuals who are HIV-positive, are more likely to develop the disease. TB is estimated to have killed 1.7 million people in 2009 and is currently the leading cause of death among people infected with HIV.\nWhy Was This Study Done? Although TB can be treated with a six-month course of antibiotics, effectively diagnosing TB is not always straightforward and drug resistance is becoming an increasing problem. One of the most common and simple methods to diagnose TB is a technique called sputum smear microscopy, which involves examining matter from the lungs under a microscope for the presence of TB-causing bacteria. However, despite being cheap and relatively simple, the test does not always detect active TB (smear-negative) and cannot determine whether the TBcausing bacteria are resistant to antibiotics. The World Health Organization has recently endorsed a new rapid test, called Xpert MTB/RIF (referred to as Xpert), for the initial diagnosis of TB. The test uses DNA amplification methods to reliably and quickly detect TB and whether infecting bacteria are resistant to the antibiotic rifampicin. The new test is expensive so there are concerns that the test might not be cost-effective in low- and middle-income countries.\nWhat Did the Researchers Do and Find? The researchers used a technique called modeling to simulate the outcome of 10,000 individuals with suspected TB as they went through a hypothetical diagnostic and treatment pathway. The model compared the costs associated with the introduction of Xpert to a base case for two different scenarios. In the base case all individuals with suspected TB had two sputum smear microscopy examinations followed by clinical diagnosis if they were smear-negative. For the different scenarios Xpert was either used in addition to the two sputum smear microscopy examinations (if the patient was smear-negative) or Xpert was used as a replacement for sputum smear microscopy for all patients. Different input parameters, based on country-specific estimates, were applied so that the model reflected the implementation of Xpert in India, South Africa, and Uganda. In the researcher\u2019s model the introduction of Xpert increased the proportion of TB-infected patients who were correctly\n\ndiagnosed with TB in any of the settings. However, the cost per TB case detected increased by approximately US$100 in both scenarios. Although the cost of detection increased significantly, the cost of treatment increased only moderately because the number of false-positive cases was reduced. For example, the percentage of treatment costs spent on falsepositive diagnoses in India was predicted to fall from 22% to 4% when Xpert was used to replace sputum smear microscopy. The model was used to calculate incremental cost effectiveness ratios (ICERs\u2014the additional cost of each disability-adjusted life year [DALY] averted) for the different scenarios of Xpert implementation in the different settings. In comparison to the base case, introducing Xpert in addition to sputum smear microscopy produced ICERs ranging from US$41 to US$110 per DALY averted, while introducing Xpert instead of sputum smear microscopy yielded ICERs ranging from US$52 to US$138 per DALY averted.\nWhat Do These Findings Mean? The findings suggest that the implementation of Xpert in addition to, or instead of, sputum smear microscopy will be cost-effective in lowand middle-income countries. The calculated ICERs are below the World Health Organization\u2019s \u2018\u2018willingness to pay threshold\u2019\u2019 for all settings. That is the incremental cost of each DALY averted by introduction of Xpert is below the gross domestic product per capita for each country ($1,134 for India, $5,786 South Africa, and $490 for Uganda in 2010). However, the authors note that achieving ICERs below the \u2018\u2018willingness to pay threshold\u2019\u2019 does not necessarily mean that countries have the resources to implement the test. The researchers also note that there are limitations to their study; additional unknown costs associated with the scale-up of Xpert and some parameters, such as patient costs, were not included in the model. Although the model strongly suggests that Xpert will be cost-effective, the researchers caution that initial roll-out of Xpert should be carefully monitored and evaluated before full scale-up.\nAdditional Information. Please access these Web sites via the online version of this summary at http://dx.doi.org/10. 1371/journal.pmed.1001120.\nN The World Health Organization provides information on all\naspects of tuberculosis, including tuberculosis diagnostics and the Stop TB Partnership (some information is in several languages)\nN The US Centers for Disease Control and Prevention has\ninformation about tuberculosis, including information on the diagnosis of tuberculosis disease\nN MedlinePlus has links to further information about\ntuberculosis (in English and Spanish)\n\nPLoS Medicine | www.plosmedicine.org\n\n14\n\nNovember 2011 | Volume 8 | Issue 11 | e1001120\n\n\n",
    "authors": [
      "Anna Vassall",
      "Sanne Van Kampen",
      "Hojoon Sohn",
      "Joy S. Michael",
      "K. R. John",
      "Saskia Den Boon",
      "J. Lucian Davis",
      "Andrew Whitelaw",
      "Mark P. Nicol",
      "Maria Tarcela Gler",
      "Anar Khaliqov",
      "Carlos Zamudio",
      "Mark D. Perkins",
      "Catharina C. Boehme",
      "Frank Cobelens"
    ],
    "doi": "10.1371/journal.pmed.1001120",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://dx.plos.org/10.1371/journal.pmed.1001120"
  },
  {
    "key": "ZC7KRQU4",
    "title": "Optimizing Tuberculosis Case Detection through a Novel Diagnostic Device Placement Model: The Case of Uganda",
    "abstract": "",
    "full_text": "RESEARCH ARTICLE\nOptimizing Tuberculosis Case Detection through a Novel Diagnostic Device Placement Model: The Case of Uganda\nMai T. Pho1*, Sarang Deo2, Kara M. Palamountain3, Moses Lutaakome Joloba4, Francis Bajunirwe5, Achilles Katamba6\n1 Department of Medicine, Sections of Hospital Medicine and Infectious Diseases & Global Health, University of Chicago, Chicago, United States of America, 2 Indian School of Business, Hyderabad, India, 3 Kellogg School of Management, Northwestern University, Evanston, United States of America, 4 Department of Medical Microbiology, Makerere University College of Health Sciences, Kampala, Uganda, 5 Department of Community Health, Mbarara University of Science and Technology, Mbarara, Uganda, 6 Department of Medicine, Makerere University College of Health Sciences, Kampala, Uganda\n* mpho@bsd.uchicago.edu\n\nAbstract\n\nOPEN ACCESS\nCitation: Pho MT, Deo S, Palamountain KM, Joloba ML, Bajunirwe F, Katamba A (2015) Optimizing Tuberculosis Case Detection through a Novel Diagnostic Device Placement Model: The Case of Uganda. PLoS ONE 10(4): e0122574. doi:10.1371/ journal.pone.0122574\nAcademic Editor: Anil Kumar Tyagi, University of Delhi, INDIA\nReceived: August 5, 2014\nAccepted: February 20, 2015\nPublished: April 1, 2015\nCopyright: \u00a9 2015 Pho et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.\nData Availability Statement: All relevant data are within the paper and its Supporting Information files.\nFunding: Grand Challenges Canada #0004-02, grandchallenges.ca (MTP SD KMP FB AK). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\nCompeting Interests: The authors have declared that no competing interests exist.\n\nBackground\nXpert MTB/RIF (Xpert) is being widely adopted in high TB burden countries. Analysis is needed to guide the placement of devices within health systems to optimize the tuberculosis (TB) case detection rate (CDR).\nMethods\nWe used epidemiological and operational data from Uganda (139 sites serving 87,600 individuals tested for TB) to perform a model-based comparison of the following placement strategies for Xpert devices: 1) Health center level (sites ranked by size from national referral hospitals to health care level III centers), 2) Smear volume (sites ranked from highest to lowest volume of smear microscopy testing), 3) Antiretroviral therapy (ART) volume (sites ranked from greatest to least patients on ART), 4) External equality assessment (EQA) performance (sites ranked from worst to best smear microscopy performance) and 5) TB prevalence (sites ranked from highest to lowest). We compared two clinical algorithms, one where Xpert was used only for smear microscopy negative samples versus another replacing smear microscopy. The primary outcome was TB CDR; secondary outcomes were detection of multi-drug resistant TB, number of sites requiring device placement to achieve specified rollout coverage, and cost.\nResults\nPlacement strategies that prioritized sites with higher TB prevalence maximized CDR, with an incremental rate of 6.2\u201312.6% compared to status quo (microscopy alone). Diagnosis of MDR-TB was greatest in the TB Prevalence strategy when Xpert was used in place of\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n1 / 14\n\nOptimizing TB Diagnostic Device Placement\nsmear microscopy. While initial implementation costs were lowest in the Smear Volume strategy, cost per additional TB case detected was lowest in the TB prevalence strategy.\nConclusion\nIn Uganda, placement of Xpert devices in sites with high TB prevalence yielded the highest TB CDR at the lowest cost per additional case diagnosed. These results represent novel use of program level data to inform the optimal placement of new technology in resourceconstrained settings.\n\nIntroduction\nTuberculosis (TB) remains a major global public health challenge, causing substantial morbidity and mortality [1]. One of the most important risk factors for the increasing TB burden is HIV/AIDS, which contributes to the difficulty in diagnosing TB in co-infected patients [1]. The introduction of the Xpert MTB/RIF (Cepheid, Sunnyvale, CA, USA) represents the first true \u201cgame-changer\u201d in the field of TB diagnostics in decades due to improved sensitivity, ease of use, and rapid turn-around-time of results made possible by the molecular platform, and in 2010, the World Health Organization endorsed the use of the Xpert device [2, 3].\nThe feasibility and cost-effectiveness of the Xpert have been well studied and support the implementation of the device in resource-limited settings [4\u20138]. When allocating new diagnostics, it has been recommended that public health decision-makers utilize not only data from such technical studies and registration trials, but also examine the existing epidemiology, health care infrastructure and clinical practice to optimize implementation and scale up [9]. However, few evidence-based guidelines exist on how the new technology should be integrated into a country\u2019s existing laboratory infrastructure, or how placement of the new technology should be prioritized [10].\nIn Uganda, a country with high TB and HIV prevalence and the focus of our analysis, Xpert testing is provided mainly to a limited number of individuals with known HIV infection. To guide the placement of the Xpert device within the existing national tuberculosis laboratory network, priority is given to centers with high smear microscopy volumes, centers providing HIV/AIDS care, and sites serving areas with traditionally poor health care access, including islands and prisons [11]. These criteria have led to placement of Xpert devices in regional referral hospitals, district hospitals and some large volume HIV/AIDS care centers.\nWhile prioritization of Xpert placement in sites with high smear volume and HIV and TB prevalence is intuitive given the burden of TB disease at these locations, it remains unclear as to which of these criteria will maximize case detection rates for TB. Further, operational factors such as the quality of smear microscopy may also impact optimal integration of the Xpert device, as sites with poorer microscopy performance are likely to benefit more from this device. In this analysis we used national epidemiologic and program level data to perform a modelbased comparison of the impact of different placement strategies for the Xpert device on the TB case detection rate (CDR) and cost in Uganda. To our knowledge this is the first attempt to use health facility level operational data to evaluate the placement of the Xpert device.\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n2 / 14\n\nOptimizing TB Diagnostic Device Placement\nMethods\nWe developed a decision-analytic model using TreeAge Pro 2012 (Williamstown, MA) and Microsoft Excel 2010 (Redmond, WA). The analytic schema is presented in Fig. 1. The primary outcome was TB CDR. Secondary outcomes were cases of MDR-TB identified, number of sites requiring Xpert devices and cost.\nEthics\nThe study was approved by the Makerere University School of Medicine Research and Ethics Committee and the Ugandan National Council of Science and Technology. All records were de-identified of protected health information prior to analysis.\nEstimation of Model Parameters\nWe created a model of all 1,089 sites in the Uganda healthcare system that provided smear microscopy services for the diagnosis of tuberculosis from January 1 to December 31, 2011. Site level data were obtained from the National TB Reference Laboratory (NTRL), the National TB Control Program, and the National AIDS Control Program and are provided in S1 Dataset (supporting information) [12\u201314]. Individual sites were characterized by level of health care provided at the facility, TB prevalence, annual smear microscopy volume, cumulative number of patients enrolled in antiretroviral therapy, and quality of microscopy performance (sensitivity and specificity) based on external quality assessment (EQA) performed by the NTRL through expert review of a sample of smears sent by each site. Exclusion criteria included sites processing less than two sputum samples for smear microscopy daily (Fig. 1). Number of individuals tested per site was estimated by dividing the total annual smear samples reported to the NTRL in 2011 at each site by 1.5 to reflect the average number of samples provided per individual based on expert opinion. Prevalence of TB at smear microscopy centers was estimated by obtaining smear microscopy results as reported to the NTRL and adjusting them by the EQAderived sensitivity and specificity for each facility as described in further detail below. Rates of multi-drug resistant disease and HIV co-infection were based on a national drug resistance survey of TB which examined a nationally representative sample of new and previously treated sputum smear-positive TB patients utilizing Lowenstein-Jensen (L-J) culture methods, PCR confirmation of Mycobacterium tuberculosis, and drug susceptibility testing using the L-J proportional method [1, 15]. The HIV infection rate among individuals evaluated at sites that were known to be ART centers was assumed to be 100% HIV.\nDiagnostic model structure and validation\nThe diagnostic algorithm in the model varied depending on whether Xpert was placed at a site or not. In sites where Xpert was not placed, individuals were evaluated using two sputum samples for smear microscopy using the Ziehl-Neelsen staining procedure. Smear negative individuals underwent clinical evaluation, with sensitivity and specificity for TB diagnosis based on published literature. In sites where Xpert was placed, individuals were first examined using two sputum samples for microscopy. Sputum samples for smear negative individuals were examined using the Xpert device including rifampicin resistance testing (integrated diagnostic algorithm). AFB culture and drug susceptibility testing was not considered in the model as access to these tests is limited to only 0.6 per 5 million individuals in the general population via line probe assay in Uganda.\nThe model was externally validated to WHO estimations of TB case detection in 2011 for Uganda, prior to the introduction of Xpert into the national laboratory.\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n3 / 14\n\nOptimizing TB Diagnostic Device Placement\n\nFig 1. Analytical schema. Of the total 1089 health care sites with smear available in Uganda, 139 met inclusion criteria (performed on average at least two samples daily) for analysis. These 139 sites were variably ranked based on placement schema as indicated in pentagonal boxes. Sensitivity and specificity of smear for all sites was adjusted by EQA data. Xpert was rolled out over the patient population. The decision tree was used to calculate estimates for TB case detection rates, number of MDR-TB cases, and number of health care sites that would require Xpert device placement to achieve Xpert rollout by any given placement schema.\ndoi:10.1371/journal.pone.0122574.g001\nXpert Placement Strategies\nWe examined five different Xpert placement strategies based on the current recommendations of the Ugandan National TB Reference Laboratory, as well as novel strategies based on the microscopy performance of individual sites and site-specific TB prevalence (Fig. 1).\n1. Health Center Level. In the Health Center Level strategy, Xpert was allocated to health care sites based on the status of the site in the hierarchy of facilities as follows. Xpert was placed first in national referral hospitals, followed by regional referral hospitals, general hospitals, health care level IV sites (considered mini-hospitals, capable of admitting patients, with laboratory and staffed by physicians) and health care level III sites (typically outpatient clinic and maternity ward, with laboratory and staffed by clinical officer). Within each level, sites were prioritized by highest to lowest smear microscopy volume.\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n4 / 14\n\nOptimizing TB Diagnostic Device Placement\n2. Smear Volume. In the Smear Volume strategy, Xpert was placed in sites ranked by highest to lowest annual number of smear samples evaluated by microscopy.\n3. ART Volume. In the Antiretroviral (ART) Volume strategy, Xpert was allocated to health care sites ranked from highest to lowest cumulative number of patients ever enrolled in ART and subsequently by smear volume, if additional devices were available after allocation to ART sites. Based on current expectations of stakeholders in the Ugandan Ministry of Health, Xpert was assumed to be placed in the site laboratories, as opposed to patient care areas, thus allowing access to for both HIV-infected and non-infected patients.\n4. EQA Performance. In the External Quality Assessment (EQA) Performance strategy, Xpert was placed in sites ranked by lowest to highest sensitivity and specificity of smear microscopy by each site according to EQA. EQA was conducted by the NTRL on random samples submitted by each site for retesting and comparison of observed results by the site compared to rechecked results by the reference laboratory.\n5. TB Prevalence. In the TB Prevalence strategy, Xpert was placed in sites ranked from highest to lowest TB prevalence. As noted above, the prevalence of TB at each site was estimated by obtaining smear microscopy results as reported to the NTRL and adjusting them by the EQA performance of the site.\nFor each strategy, increasing Xpert rollout (0%, 25%, 50%, 75%, 100%) to the total patient population tested for TB was considered, starting with the highest ranked sites and proceeding to the lowest ranked sites. In any given strategy, individuals without access to Xpert due to incomplete rollout were evaluated as in the status quo, for which only smear microscopy and clinical evaluation were available.\nTest characteristics\nBaseline sensitivity and specificity of smear microscopy, clinical examination, and Xpert MTB/ RIF were obtained from demonstration studies in the literature [5, 8]. Site-specific smear microscopy performance was calculated by multiplying the sensitivity and specificity derived via routine EQA and the published sensitivity and specificity.\nSensitivity Analysis\nSeveral sensitivity analyses were performed to assess the robustness of the results. We assessed the impact of switching the diagnostic algorithm to reflect the current algorithm adopted in South Africa, where Xpert is performed instead of smear microscopy as opposed to only used on smear microscopy negative samples [16]. We also varied EQA-derived sensitivity and specificity at each site within +10% to -10% of the base case to reflect the limitations of the EQA process in assessing the quality of smear microscopy at the sites.\nCost analysis\nThe implementation cost of providing Xpert in the first year of rollout was estimated based on the five year amortization of a four-bay Xpert device, assuming a capital outlay for device of $17,000 2014 USD, a device life of 5 years, and discount rate of 3% [17]. Annual costs attributed to each strategy were calculated adding test costs (the number of tests performed multiplied by the cost per test ($9.98 2014 USD)) and device costs (the number of devices required based on sites with Xpert available multiplied by the cost of the device as above) [17]. Cost per additional TB cases detected was calculated by dividing incremental case detection for each strategy\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n5 / 14\n\nOptimizing TB Diagnostic Device Placement\nover one year compared to the use of smear microscopy alone. Costs of personnel, training, and infrastructural requirement were not available and hence excluded in the analysis.\nResults\nModel Cohort and base case results\nOf the total of 1089 sites performing smear microscopy in 2011, 139 sites met the inclusion criteria of processing an average of at least two sputum samples daily (Table 1). 87,600 individuals evaluated for TB at these sites were simulated in the diagnostic model. This represented 52% of all estimated individuals tested in 2011. The mean TB prevalence was 25.3%. After adjusting for EQA, the average sensitivity of smear microscopy in HIV-negative and HIV-positive individuals was 0.654 and 0.404, respectively (Table 2). Table 1 describes the characteristics of individuals served at each health center level, including mean number, number enrolled on ART, TB prevalence, as well as the sensitivity and specificity of smear microscopy performed at these sites.\nTB case detection rate in the status quo (diagnosis by smear microscopy only) was estimated to be 72.3% (Table 3). This rate lay within the range of WHO estimates of 51\u201376% [1]. Given the lack of AFB culture and drug susceptibility testing, no cases of MDR-TB were identified in the status quo. Xpert placement in sites prioritized by highest to lowest TB prevalence led to a greatest number of TB diagnoses among all strategies. Compared to the status quo, incremental case detection rates increased from 6.2\u201312.6% in the TB Prevalence strategy as Xpert was made available to 25\u201375% of the patient population (Table 3). ART volume strategy was second best for coverage of less than 30% but its performance became inferior to that of EQA performance strategy for coverage greater than 30% (Fig. 2). Case detection rates were similar in the Health Center Level and Smear Volume strategies.\nThe number of Xpert devices required for the ART Volume strategy was greater at any given level of Xpert rollout as compared to other strategies (Table 3). For example, to achieve testing of 25% of the patient population with Xpert, 39 sites required Xpert placement in the ART Volume strategy as compared to 17 sites in the Health Center Level strategy and 14 sites in the Smear Volume strategy (Table 3, Fig. 3). This variation in the number of sites reflects the differences in the smear volume of sites prioritized based on placement strategies. In the above example, the Smear Volume strategy prioritized the highest volume smear microscopy sites, resulting in fewer numbers of sites requiring Xpert placement at assigned rollout, whereas the ART Volume strategy prioritized sites based on cumulative ART enrollment, which performed fewer smears per site in comparison. Diagnosis of MDR-TB was greatest in the TB Prevalence strategy (Table 3), with an increase of 0.5\u20132.5% greater MDR-TB detection compared to the next best strategy.\nSensitivity Analysis\nWhen the diagnostics algorithm was adjusted to completely replace smear microscopy with the Xpert device (versus integrating the Xpert with smear microscopy for smear-negative individuals only), the ART Volume strategy was superior in both total CDR and detection of MDR-TB to the TB prevalence strategy when Xpert availability was limited to less than 39% of the patient population. At higher levels of device availability the TB prevalence strategy became superior (Table 4, Fig. 4).\nWhen EQA-derived sensitivity and specificity was varied to underestimate the observed microscopy sensitivity and specificity by 10%, base case results remained stable. However, the superiority of placement by the EQA Performance strategy compared to placement by ART\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n6 / 14\n\nOptimizing TB Diagnostic Device Placement\n\nTable 1. Characteristics of sites included in analysis.\nTotal sites selected in analysis Smear volume for selected sites, 2011 Estimated number individuals tested for TB Cumulative patients enrolled in ART Health center level, number of sites (%)\nNational referral hospital Individuals tested for TB, mean ART Volume, mean Microscopy sensitivity by EQA, mean Microscopy speci\ufb01city by EQA, mean TB prevalence, mean\nRegional referral hospital Individuals tested for TB, mean ART Volume, mean Microscopy sensitivity by EQA, mean Microscopy speci\ufb01city by EQA, mean TB prevalence, mean\nHigh volume HIV center Individuals tested for TB, mean ART Volume, mean Microscopy sensitivity by EQA, mean Microscopy speci\ufb01city by EQA, mean TB prevalence, mean\nHospital Individuals tested for TB, mean ART Volume, mean Microscopy sensitivity by EQA, mean Microscopy speci\ufb01city by EQA, mean TB prevalence, mean\nHealth center level IV Individuals tested for TB, mean ART Volume, mean Microscopy sensitivity by EQA, mean Microscopy speci\ufb01city by EQA, mean TB prevalence, mean\nHealth center level III Individuals tested for TB, mean ART Volume, mean Microscopy sensitivity by EQA, mean Microscopy speci\ufb01city by EQA, mean TB prevalence, mean\n\nValue\n139 131,400 87,600 149,633\n4 (2.1%) 1928 1699 0.89 0.95 0.36 11 (8.0%) 924 2925 0.85 0.99 0.29 12 (8.6%) 577 2360 0.85 0.98 0.24 52 (37.4%) 652 1180 0.92 0.98 0.24 37 (26.6%) 504 481 0.93 0.98 0.22 23 (16.5%) 445 447 0.88 0.98 0.24\n\nTB: Tuberculosis, ART: Antiretroviral therapy, EQA: External quality assurance\n\ndoi:10.1371/journal.pone.0122574.t001\n\nSource\n[12] [12] [12] [14] [12] [12] [12] [14] [13] [13] [12] [12] [12] [14] [13] [13] [12] [12] [12] [14] [13] [13] [12] [12] [12] [14] [13] [13] [12] [12] [12] [14] [13] [13] [12] [12] [12] [14] [13] [13] [12]\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n7 / 14\n\nOptimizing TB Diagnostic Device Placement\n\nVolume was lost when EQA\u2014derived sensitivity and specificity was varied to overestimate observed microscopy performance by greater than 2%. (Results not shown)\n\nCosts\nImplementation costs in the first year of Xpert device rollout were greatest in the ART volume strategy and were lowest in the Smear Volume strategy (Table 5). This finding was due to the greater number of sites requiring Xpert device placement at each stage of rollout in the ART volume strategy compared to the remaining strategies. Based on the measure of incremental cost per additional tuberculosis case diagnosed (compared to smear microscopy alone), the ART volume strategy remained most costly, whereas TB prevalence strategy was least costly (Table 5).\n\nDiscussion\nWhile the Xpert device has changed the landscape of TB diagnosis in resource-limited settings, the full benefit of the test can be realized only if rational evidence-based approaches towards placement decisions are adopted. Our analysis demonstrates that in Uganda, optimal placement of Xpert is dependent upon diagnostic algorithm, degree of device availability, and key health care site characteristics. When Xpert use is integrated with smear microscopy,\n\nTable 2. Demographics of patient population included in analysis and model input parameters.\n\nParameter\nCohort Characteristics Prevalence of TB amongst tested individuals. mean Smear-positive TB among HIV-negative TB cases Smear-positive TB among HIV-positive TB cases Treatment-experience amongst TB cases MDR prevalence amongst treatment-na\u00efve, TB cases MDR prevalence amongst treatment-experienced TB cases HIV prevalence amongst TB cases Diagnostic Parameters Sensitivity Smear microscopy, HIV-negative Smear microscopy, HIV-positive Xpert MTB/RIF, smear-positive TB cases Xpert MTB/RIF, smear-negative, HIV-negative cases Xpert MTB/RIF, smear-negative, HIV-positive cases Xpert MTB/RIF rifampin testing Clinical diagnosis of TB Speci\ufb01city Smear microscopy Xpert MTB/RIF Xpert MTB/RIF rifampin testing Clinical diagnosis of TB Return for results, probability (range) While awaiting smear microscopy While awaiting for Xpert MTB/RIF\n\nBase Case Value\n0.253 0.723 0.446 0.1 .011 0.12 0.53\n0.654 0.404 0.983 0.793 0.718 0.983 0.444\n0.982 0.990 0.983 0.869\n1 (.87\u20131) 1 (.74\u20131)\n\nReference\n[12] [5, 8] [5, 8] [1] [1, 8] [1, 8] [1]\n[5, 8, 13] [5, 8, 13] [5, 8] [5, 8] [5, 8] [5, 8] [5, 8]\n[5, 8, 13] [5, 8] [5, 8] [5, 8]\n[18] [5, 19]\n\nTB: Tuberculosis, MDR: Multi-drug resistant, ART: Antiretroviral therapy. EQA: External quality assurance\n\ndoi:10.1371/journal.pone.0122574.t002\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n8 / 14\n\nOptimizing TB Diagnostic Device Placement\n\nTable 3. Base case results.\nPlacement Strategy\nStatus Quo, smear only Case Detection\nHealth Center Level Case Detection % MDR TB detected Number of sites with Xpert\nSmear Volume Case Detection % MDR TB detected Number of sites with Xpert\nART Volume Case Detection % MDR TB detected Number of sites with Xpert\nEQA Performance Case Detection % MDR TB detected Number of sites with Xpert\nTB Prevalence Case Detection % MDR TB detected Number of sites with Xpert\n\n**% of individuals with access to Xpert\n\n**0% 72.3% **25% 76.2% 11.4% 17 **25% 76.2% 11.4% 14 **25% 78.0% 13.5% 39 **25% 77.7% 9.3% 34 **25% 78.5% 14.0% 28\n\n**0% 72.3% **50% 80.7% 19.9% 44 **50% 80.2% 19.7% 39 **50% 79.9% 20.9% 72 **50% 82.3% 18.9% 59 **50% 82.3% 23.4% 57\n\n**0% 72.3% **75% 83.6% 27.2% 87 **75% 82.6% 26.2% 79 **75% 82.5% 28.0% 104 **75% 84.7% 24.4% 86 **75% 84.9% 29.9% 96\n\nTB case detection by strategy and increasing access to Xpert. ** Indicates percent of the patient population with access to the Xpert device\n\ndoi:10.1371/journal.pone.0122574.t003\n\nplacement of the device based on TB prevalence is superior to all other strategies regardless of the degree of Xpert rollout. This strategy is successful as higher TB prevalence improves the positive predictive value of the test. Placement of Xpert based on ART volume is second best at lower degrees of Xpert rollout, whereas placement based on poor EQA performance is equivalent to placement by TB prevalence at greater degrees of Xpert rollout. This may be explained by improved CDR with Xpert in HIV-infected individuals concentrated in high volume ART clinics, with loss of this advantage after early saturation of these clinics. At higher degrees of rollout, placement based on EQA performance capitalizes on the greater incremental sensitivity that Xpert holds over microscopy at these sites. Placement of Xpert devices at sites with poor microscopy performance addresses the known limitations of smear microscopy such as dependence on operator skill and poor diagnostic sensitivity in the setting of high HIV incidence [20\u201322]. When Xpert completely replaced smear microscopy in the diagnostic algorithm, the slightly lower overall CDR could be explained by the combination of test sensitivities of both smear microscopy and Xpert in the integrated diagnostic algorithm.\nIn our analysis the costs of initial implementation of Xpert was lowest based on prioritization of sites with greatest smear volume, which can be considered intuitive. However the cost per additional TB case diagnosed compared to smear microscopy alone was lowest when sites were prioritized by TB prevalence, a finding that highlights the need to consider the long-term implications of placement strategies.\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n9 / 14\n\nOptimizing TB Diagnostic Device Placement\n\nFig 2. TB CDR by placement strategy using integrated diagnostic algorithm. This figure demonstrates the TB case detection rates by Xpert placement strategy as % of individuals with access to Xpert increases. Xpert is used for smear microscopy negative individuals only. Case detection rate is defined by # TB cases diagnosed / estimated total TB cases.\ndoi:10.1371/journal.pone.0122574.g002\n\nFig 3. Number of health care sites with Xpert by placement strategy. This figure demonstrates the number of health care sites with Xpert device placement as the % of individuals with access to Xpert increases. Variation by strategy reflects different volumes of individuals tested at each site, and different sites selected based on placement strategy.\ndoi:10.1371/journal.pone.0122574.g003\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n10 / 14\n\nOptimizing TB Diagnostic Device Placement\n\nWhile our study is novel in its use of multiple site level data to inform placement decisions of a new diagnostic test, there are several limitations to consider. Relative differences in CDR rates between the best and worst performing placement strategies were small (2.3%). When considering high TB burden settings such as Uganda, however, where 65,000 incident cases are estimated per year, these differences could be considered clinically significant (1500 additional cases per year). Several sites had very few observations of the EQA data which could have introduced noise in the estimate of their EQA performance. In sensitivity analysis we found that the superiority of the TB prevalence strategy at higher degrees of Xpert rollout remained stable, however the results for prioritization of sites by EQA performance did not. This underscores the need for strengthening EQA data collection and further investigation of this approach.\nWe did not allow for Xpert performance to vary across sites. However, given the automated platform and minimal operator hands-on time required to use the device we assumed that this variation would be minimal. We did not have site level data on HIV prevalence, which may have impacted rate of MDR-TB detection in the ART volume strategy. We did not have sitespecific data to quantify the impact of reduced turnaround time for results on reduced loss to follow-up of patients. While anecdotal evidence supports such a relationship, very few studies have actually quantified it [23].\nWe only analyzed discrete placement strategies that have been most widely discussed in the policy and implementation literature and that are each based on a single criterion. It may be possible further elaborate upon this placement model using mathematical optimization of\n\nTable 4. Sensitivity analysis based on use of Xpert only (no smear microscopy) where Xpert device is available.\n\nPlacement Schema\n\n**% of individuals with access to Xpert\n\nStatus Quo, smear only Case Detection\nHealth Center Level Case Detection % MDR TB detected Number of sites with Xpert\nSmear Volume Case Detection % MDR TB detected Number of sites with Xpert\nART Volume Case Detection % MDR TB detected Number of sites with Xpert\nEQA Performance Case Detection % MDR TB detected Number of sites with Xpert\nTB Prevalence Case Detection % MDR TB detected Number of sites with Xpert\n\n0% 72.3% **25% 76.0% 26.7% 17 **25% 76.0% 25.8% 14 **25% 78.3% 39.3% 39 **25% 75.5% 25.3% 34 **25% 78.2% 35.3% 28\n\n0% 72.3% **50% 80.3% 50.5% 44 **50% 79.9% 49.3% 39 **50% 80.6% 55.5% 72 **50% 81.8% 52.0% 59 **50% 81.8% 58.1% 57\n\n0% 72.3% **75% 83.0% 67.5% 87 **75% 82.1% 63.6% 79 **75% 82.6% 68.1% 104 **75% 84.2% 68.3% 86 **75% 84.3% 75.5% 96\n\n** Indicates percent of the patient population with access to the Xpert device\n\ndoi:10.1371/journal.pone.0122574.t004\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n11 / 14\n\nOptimizing TB Diagnostic Device Placement\n\nFig 4. TB CDR by placement strategy when Xpert replaces smear microscopy. This figure demonstrates the TB case detection rates by Xpert placement strategy as % of individuals with access to Xpert increases. Xpert completely replaces smear microscopy in the diagnostic algorithm. Case detection rate is defined by # TB cases diagnosed / estimated total TB cases\ndoi:10.1371/journal.pone.0122574.g004\ncombined strategies as demonstrated in the case of early infant HIV diagnosis (Deo and Sohoni, forthcoming in the journal \u201cManufacturing and Service Operations Management\u201d). Finally we did not include personnel, training, and infrastructural costs such as electricity necessary for Xpert implementation as these data were not readily available.\nIn conclusion, we found that for Uganda, placement of the Xpert device in sites prioritized by high TB prevalence was superior to other strategies based on optimization of case detection rate and implementation cost. More broadly, our work makes two important contributions. First, our analytical framework demonstrates the value of combining operational decisions regarding site selection with clinical decisions regarding diagnostic algorithm. Second, it underlines the value of using program level data (e.g. smear volume, EQA performance) to inform critical decisions and strengthen regional capacity surrounding the placement of new\n\nTable 5. Implementation cost of Xpert in the first year of Xpert rollout and cost per additional TB case diagnosed by strategy.\n\nImplementation cost in the \ufb01rst year\n\nCost per additional TB case diagnosed compared to status quo\n\nPlacement Schema\nHealth Center Level Smear Volume ART Volume EQA Performance TB Prevalence\n\n**25%\n$235,392 $229,984 $370,689 $299,932 $269,492\n\n**50%\n$523,916 $504,829 $672,008 $577,209 $558,494\n\n**75%\n$874,923 $855,255 $950,370 $867,219 $897,277\n\n**25%\n$193 $189 $230 $193 $155\n\n**50%\n$236 $239 $330 $225 $218\n\n**75%\n$307 $324 $363 $279 $285\n\n** Indicates percent of the patient population with access to the Xpert device\n\ndoi:10.1371/journal.pone.0122574.t005\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n12 / 14\n\nOptimizing TB Diagnostic Device Placement\ntechnology. Future areas of research should include the assessment of placement decisions on outcomes such as treatment initiation or life expectancy, the impact of point-of-care versus central laboratory placement of Xpert, and the impact of device placement on geographical patient-initiated provider-directed referral patterns.\nSupporting Information\nS1 Dataset. This table contains data of sputum smear microscopy workload, performance by External Quality Assessment, and cumulative ART enrollment from individual sites as obtained from the National TB Reference Laboratory (NTRL), the National TB Control Program, and the National AIDS Control Program for 2011. (PDF)\nAcknowledgments\nWe would like to acknowledge Li Chen for her assistance in data analysis, and Diana Nadunga for organizing the NTRL dataset\nAuthor Contributions\nConceived and designed the experiments: MTP SD KMP FB AK. Performed the experiments: MTP SD KMP. Analyzed the data: MTP SD KMP FB AK. Contributed reagents/materials/analysis tools: MTP SD MLJ AK. Wrote the paper: MTP SD KMP FB MLJ AK.\nReferences\n1. WHO. Global Tuberculosis Report. Geneva: 2013. Available: http://apps.who.int/iris/bitstream/10665/ 91355/1/9789241564656_eng.pdf.\n2. Evans CA. GeneXpert\u2014a game-changer for tuberculosis control? PLoS Med. 2011; 8(7):e1001064. doi: 10.1371/journal.pmed.1001064 PMID: 21814497\n3. WHO. WHO endorses new rapid tuberculosis test. London. News release. 2010 December 8, 2010. Available: http://www.who.int/mediacentre/news/releases/2010/tb_test_20101208/en/.\n4. Abimbola TO, Marston BJ, Date AA, Blandford JM, Sangrujee N, Wiktor SZ. Cost-effectiveness of tuberculosis diagnostic strategies to reduce early mortality among persons with advanced HIV infection initiating antiretroviral therapy. J Acquir Immune Defic Syndr. 2012; 60(1):e1\u20137. doi: 10.1097/QAI. 0b013e318246538f PMID: 22240465\n5. Boehme CC, Nicol MP, Nabeta P, Michael JS, Gotuzzo E, Tahirli R, et al. Feasibility, diagnostic accuracy, and effectiveness of decentralised use of the Xpert MTB/RIF test for diagnosis of tuberculosis and multidrug resistance: a multicentre implementation study. Lancet. 2011; 377(9776):1495\u2013505. doi: 10. 1016/S0140-6736(11)60438-8 PMID: 21507477\n6. Dorman SE, Chihota VN, Lewis JJ, Shah M, Clark D, Grant AD, et al. Performance characteristics of the Cepheid Xpert MTB/RIF test in a tuberculosis prevalence survey. PLoS One. 2012; 7(8):e43307. doi: 10.1371/journal.pone.0043307 PMID: 22905254\n7. Lawn SD, Brooks SV, Kranzer K, Nicol MP, Whitelaw A, Vogt M, et al. Screening for HIV-associated tuberculosis and rifampicin resistance before antiretroviral therapy using the Xpert MTB/RIF assay: a prospective study. PLoS Med. 2011; 8(7):e1001067. doi: 10.1371/journal.pmed.1001067 PMID: 21818180\n8. Vassall A, van Kampen S, Sohn H, Michael JS, John KR, den Boon S, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: a cost-effectiveness analysis. PLoS Med. 2011; 8(11):e1001120. doi: 10.1371/journal.pmed.1001120 PMID: 22087078\n9. Pai NP, Vadnais C, Denkinger C, Engel N, Pai M. Point-of-care testing for infectious diseases: diversity, complexity, and barriers in low- and middle-income countries. PLoS Med. 2012; 9(9):e1001306. doi: 10.1371/journal.pmed.1001306 PMID: 22973183\n10. Pai M, Palamountain KM. New tuberculosis technologies: challenges for retooling and scale-up. Int J Tuberc Lung Dis. 2012; 16(10):1281\u201390. doi: 10.5588/ijtld.12.0391 PMID: 23107630\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n13 / 14\n\nOptimizing TB Diagnostic Device Placement\n11. Programme NTaLC. Draft: Planning & implementation of Xpert MTB/RIF test for detection of tuberculosis. Uganda: 2012.\n12. Laboratory NTR. Uganda TB laboratory nework sputum smear microscopy workload data. Kampala, Uganda: 2011.\n13. Laboratory NTR. Uganda TB laboratory network external quality assurance data. Kampala, Uganda: 2011.\n14. STD/AIDS Control Programme MoH. Status of antiretroviral therapy services in Uganda for 2011. Kampala, Uganda: 2011.\n15. Lukoye D, Adatu F, Musisi K, Kasule GW, Were W, Odeke R, et al. Anti-tuberculosis drug resistance among new and previously treated sputum smear-positive tuberculosis patients in Uganda: results of the first national survey. PLoS One. 2013; 8(8):e70763. doi: 10.1371/journal.pone.0070763 PMID: 23936467\n16. Stevens W. GeneXpert implementation in South Africa public sector: One year later. lessons learnt 4th Annual WHO/GLI Meeting2012 [cited 2013 September 26]. Available: http://www.stoptb.org/wg/gli/ assets/html/day%203/Stevens%20-%20South%20Africa.pdf.\n17. Gane EJ SC, Hyland RH, et. al. ELECTRON: once daily PSI-7977 plus RBV in HCV GT1/2/3. 47th Annual Meeting of the European Association for the Study of the Liver2012.\n18. Khan MS, Khan S, Godfrey-Faussett P. Default during TB diagnosis: quantifying the problem. Trop Med Int Health. 2009; 14(12):1437\u201341. doi: 10.1111/j.1365-3156.2009.02406.x PMID: 19843298\n19. Cohen G DP, Noubary F, Cloete C, Nixon K, Parker G, Bassett I. Diagnostic delays associated with Xpert MTB/RIF assay in a centralized laboratory for pulmonary TB among HIV+ adults: South Africa. 2013.\n20. Foulds J, O'Brien R. New tools for the diagnosis of tuberculosis: the perspective of developing countries. Int J Tuberc Lung Dis. 1998; 2(10):778\u201383. PMID: 9783521\n21. Pai M, Kalantri S, Dheda K. New tools and emerging technologies for the diagnosis of tuberculosis: part II. Active tuberculosis and drug resistance. Expert Rev Mol Diagn. 2006; 6(3):423\u201332. PMID: 16706744\n22. Perkins MD. New diagnostic tools for tuberculosis. Int J Tuberc Lung Dis. 2000; 4(12 Suppl 2):S182\u20138. PMID: 11144551\n23. Mugambi ML, Deo S, Kekitiinwa A, Kiyaga C, Singer ME. Do diagnosis delays impact receipt of test results? Evidence from the HIV early infant diagnosis program in Uganda. PLoS One. 2013; 8(11): e78891. doi: 10.1371/journal.pone.0078891 PMID: 24282502\n\nPLOS ONE | DOI:10.1371/journal.pone.0122574 April 1, 2015\n\n14 / 14\n\n\n",
    "authors": [
      "Mai T. Pho",
      "Sarang Deo",
      "Kara M. Palamountain",
      "Moses Lutaakome Joloba",
      "Francis Bajunirwe",
      "Achilles Katamba"
    ],
    "doi": "10.1371/journal.pone.0122574",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://dx.plos.org/10.1371/journal.pone.0122574"
  },
  {
    "key": "LV7UGHEI",
    "title": "Cost Analysis of Tuberculosis Diagnosis in Cambodia with and without Xpert\u00ae MTB/RIF for People Living with HIV/AIDS and People with Presumptive Multidrug-resistant Tuberculosis",
    "abstract": "Background: The Xpert\u00ae MTB/RIF (Xpert) test has been shown to be effective and costeffective for diagnosing tuberculosis (TB) under conditions with high HIV prevalence and HIV-TB co-infection but less is known about Xpert\u2019s cost in low HIV prevalence settings. Cambodia, a country with low HIV prevalence (0.7%), high TB burden, and low multidrug-resistant (MDR) TB burden (1.4% of new TB cases, 11% of retreatment cases) introduced Xpert into its TB diagnostic algorithms for people living with HIV (PLHIV) and people with presumptive MDR TB in 2012. The study objective was to estimate these algorithms\u2019 costs pre- and post-Xpert introduction in four provinces of Cambodia.",
    "full_text": "Author Manuscript\n\nAuthor Manuscript\n\nHHS Public Access\nAuthor manuscript\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\nPublished in final edited form as: Appl Health Econ Health Policy. 2018 August ; 16(4): 537\u2013548. doi:10.1007/s40258-018-0397-3.\nCost analysis of tuberculosis diagnosis in Cambodia with and without Xpert\u00ae MTB/RIF for people living with HIV/AIDS and people with presumptive multidrug-resistant tuberculosis\nSarah Wood Pallas, PhDa, Marissa Courey, PhDa, Chhaily Hy, BAb, Wm. Perry Killam, MDb, Dora Warren, PhDb, and Brittany Moore, MPHc a. Division of Global HIV/AIDS, Center for Global Health, U.S. Centers for Disease Control and Prevention (CDC), 1600 Clifton Road NE, Atlanta, GA 30329-4027, USA\nb. Division of Global HIV/AIDS, Center for Global Health, U.S. Centers for Disease Control and Prevention (CDC), National Institute of Public Health, #80, 289 Samdach Penn Nouth St. (289), Phnom Penh, Cambodia\nc. Division of Tuberculosis Elimination, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, U.S. Centers for Disease Control and Prevention (CDC), 1600 Clifton Road NE, Atlanta, GA 30329-4027, USA\nAbstract\nBackground: The Xpert\u00ae MTB/RIF (Xpert) test has been shown to be effective and costeffective for diagnosing tuberculosis (TB) under conditions with high HIV prevalence and HIV-TB co-infection but less is known about Xpert\u2019s cost in low HIV prevalence settings. Cambodia, a country with low HIV prevalence (0.7%), high TB burden, and low multidrug-resistant (MDR) TB burden (1.4% of new TB cases, 11% of retreatment cases) introduced Xpert into its TB diagnostic algorithms for people living with HIV (PLHIV) and people with presumptive MDR TB in 2012. The study objective was to estimate these algorithms\u2019 costs pre- and post-Xpert introduction in four provinces of Cambodia.\nCorresponding Author: Sarah Pallas, U.S. Centers for Disease Control and Prevention, 1600 Clifton Road NE, MS A-04, Atlanta, GA 30329-4027, USA; spallas@cdc.gov; Tel: +1-404-718-8759. Ethical review: The U.S. Centers for Disease Control and Prevention (CDC), Atlanta, GA, USA, and the Cambodia National Tuberculosis Control Program (CENAT) determined the study to be a program evaluation and not research involving human subjects, and therefore no IRB approval was required. Conflicts of interest: Sarah Wood Pallas, Marissa Courey, Chhaily Hy, William Perry Killam, Dora Warren, and Brittany Moore declare that they have no conflicts of interest. Data Availability Statement The data that support the findings of this study are not publicly available due to them containing information considered procurement sensitive by the Cambodia National TB Program (CENAT). The data are however available on reasonable request from the corresponding author (SWP) and with permission of CENAT. Author Contributions Conceptualization and design of the study: SWP, MC, WPK, DW, BM Acquisition of data: SWP, CH, BM Analysis of data: SWP, MC Interpretation of data and analysis: SWP, MC, CH, WPK, DW, BM Drafting of the manuscript: SWP Revisions of the manuscript for important intellectual content: SWP, MC, CH, WPK, DW, BM Final approval of the manuscript for publication: SWP, MC, CH, WPK, DW, BM\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 2\n\nMethods: Using a retrospective, ingredients-based microcosting approach, primary cost data on personnel, equipment, maintenance, supplies, and specimen transport were collected at four sites through observation, records review, and key informant consultations.\nResults: Across the sample facilities, the cost per Xpert test was US$33.88-US$37.11,clinical exam cost US$1.22-US$1.84, chest x-ray cost US$2.02-US$2.14, fluorescent microscopy (FM) smear cost US$1.56-US$1.93, Ziehl-Neelsen (ZN) smear cost US$1.26, liquid culture test cost US $11.63-US$22.83, follow-on work-up for positive culture results and Mycobacterium tuberculosis complex (MTB) identification cost US$11.50-US$14.72, and drug susceptibility testing (DST) cost US$44.26. Specimen transport added US$1.39-US$5.21 per sample. Assuming clinician adherence to the algorithms and perfect test accuracy, the normative cost per patient correctly diagnosed under the post-Xpert algorithms would be US$25-US$29 more per PLHIV and US$34US$37 more per person with presumptive MDR TB (US$41 more per PLHIV when accounting for variable test sensitivity and specificity).\nConclusions: Xpert test unit costs could be reduced through lower cartridge prices, longer usable life of GeneXpert\u00ae (Cepheid, USA) instruments, and increased test volumes; however, epidemiological and test eligibility conditions in Cambodia limit the number of specimens received at laboratories, leading to sub-optimal utilization of current instruments. Improvements to patient referral and specimen transport could increase test volumes and reduce Xpert test unit costs in this setting.\n\n1. Introduction\nGlobally, an estimated 10.4 million people developed tuberculosis (TB) in 2015, of whom an estimated 11% were co-infected with HIV (1). There were 480,000 estimated new cases of multidrug-resistant (MDR) TB in 2015 (1). Traditional diagnostic procedures of clinical screening, chest x-ray, and sputum smear microscopy often fail to accurately diagnose TB among people living with HIV (PLHIV) and do not diagnose drug resistance, while more accurate diagnostics of culture and drug susceptibility testing (DST) take multiple weeks to complete before results are available for clinicians and patients (2\u20135). Delayed diagnosis and treatment of TB among PLHIV and of MDR TB increases opportunities for disease transmission and can accelerate disease progression and mortality (1, 6\u20139).\nThe Xpert\u00ae MTB/RIF (Xpert) rapid test conducted on the GeneXpert\u00ae platform (GeneXpert; Cepheid, USA) was developed in response to these challenges. Xpert is a polymerase chain reaction-based test that delivers a result in roughly two hours for both Mycobacterium tuberculosis complex (MTB) detection and resistance to one of the first-line anti-TB drugs, rifampicin (RIF) (3, 10). As the test is largely automated, it also addresses concerns about shortages of trained laboratory personnel in low- and middle-income country settings with the highest burdens of TB among PLHIV and MDR TB (10). In 2010, the World Health Organization recommended Xpert use for TB diagnosis in PLHIV and for people with presumptive MDR TB, and in 2013 for pediatric and some types of extrapulmonary TB (3, 11). Previous literature has shown Xpert to have high sensitivity and specificity for the diagnosis of TB and detection of RIF resistance (12\u201317), and to be costeffective (18\u201323). Prior cost-effectiveness studies, however, have focused on settings with high HIV prevalence such as southern Africa using model-based approaches. Less is known\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 3\nabout the costs of Xpert under routine implementation conditions in settings with low HIV prevalence.\nAccordingly, this paper presents results from a cost analysis of Xpert integration into TB diagnostic algorithms for PLHIV and people with presumptive MDR TB in Cambodia, which has an estimated adult HIV prevalence of 0.7% (0.3%\u22121.5%) but is one of 30 WHOdesignated high TB burden countries (1, 24). Cambodia\u2019s estimated TB incidence rate is 380 per 100,000 (246\u2013543), and an estimated 2.4% of incident TB cases are co-infected with HIV.(1) MDR TB cases in Cambodia are estimated at 1.8% (0.77%\u22122.8%) of new TB cases and 11% (1.4%\u221220%) of retreatment TB cases (1). In 2012, the Cambodian Ministry of Health introduced Xpert testing into its diagnostic algorithms for PLHIV and people with presumptive MDR TB (Supplemental Figures 1\u20134) and placed 4-module GeneXpert instruments in selected laboratories throughout the country. For PLHIV, the primary change to the algorithm was that PLHIV with a positive symptom screen for TB would initially receive an Xpert test rather than initially receiving clinical exam, chest x-ray, and sputum smear microscopy for TB diagnosis. For patients with presumptive MDR TB, the primary change to the algorithm was to first receive an Xpert test and be started on empirical treatment based on Xpert results while awaiting culture and DST results, rather than waiting for culture and DST results before beginning treatment.\nThis rollout of Xpert has been the subject of an ongoing program evaluation by the Cambodian National TB Program (CENAT), and the U.S. Centers for Disease Control and Prevention (CDC) (25), which includes assessing the fidelity of implementation of the algorithms and the costs of the diagnostic algorithms before and after Xpert introduction. The objective of the cost analysis was to identify the unit costs of each procedure in the diagnostic algorithms for PLHIV and people with presumptive MDR TB, as well as to estimate the cost per patient diagnosed and to understand the factors influencing TB diagnosis costs in Cambodia. This study \u2013 the first cost analysis of TB diagnostics in Cambodia \u2013 offers insights into the cost drivers of Xpert introduction and use that are also relevant to a broader range of countries seeking to optimize their use of the GeneXpert platform.\n\n2. Methods\n2.1. Study Setting and Site Sample\nThe study setting was four provinces in northwest Cambodia \u2013 Battambang, Banteay Meanchey, Pursat, and Pailin \u2013 selected based on the CDC\u2019s existing presence there. In 2012, 4-module GeneXpert instruments were placed in two provincial referral hospitals (Battambang Provincial Referral Hospital in Battambang and Mongkol Borei Provincial Referral Hospital in Banteay Meanchey) to provide Xpert testing for the four-province catchment area. Across the four provinces, there are a total of 12 outpatient departments located at provincial or lower-level operational district hospitals, each of which has a TB clinic that screens TB patients and patients with presumptive MDR TB and 10 of which have co-located HIV clinics that screen PLHIV for TB symptoms. PLHIV who screen positive for TB in HIV clinics are referred to TB clinics for TB testing and treatment. Two of the provincial referral hospitals (Battambang and Mongkol Borei) provide MDR TB treatment\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 4\n\nto any diagnosed MDR TB patient within the four provinces; MDR TB treatment at CENAT in the capital city of Phnom Penh is also an option. Ten of the hospitals in the catchment area currently have functioning x-ray equipment that could be used for chest x-rays and all have laboratories that are equipped to perform sputum smear microscopy. Public sector TB culture testing is available at the laboratory of the Battambang Provincial Referral Hospital or the CENAT national laboratory in Phnom Penh.\nFour sites were included in the sample for the cost analysis: the two provincial referral hospitals in the four-province catchment area in which a GeneXpert instrument had been placed in 2012 (Battambang (BTB) and Mongkol Borei (MKB)), one operational district (OD) hospital in Mong Russey (MR), and the CENAT national laboratory. The BTB and MKB provincial referral hospitals were selected as these were the only two sites in the fourprovince study setting in which public sector Xpert testing was conducted. Mong Russey OD hospital was selected for convenience based on proximity as a feeder facility to the BTB provincial referral hospital. The CENAT national laboratory in Phnom Penh was selected as the only facility in Cambodia that performs DST and because hospitals in the study setting may choose to send samples for culture to the CENAT laboratory rather than to BTB. Selected sites were not intended to be a representative sample for Cambodia as a whole.\n\n2.2. Study Design\nThe cost analysis was conducted from a public health sector perspective, considering costs to the publicly-funded health care provider, which for this evaluation includes CENAT, the National HIV/AIDS Program, and other Ministry of Health costs, including costs within the public health sector that may be funded by international donor agencies or nongovernmental organizations. This perspective excluded patients\u2019 out-of-pocket and opportunity costs, as well as the costs of private sector health care providers, government services outside of the health sector, and health services not directly related to the study populations and diagnostic algorithms being evaluated.\nThe cost analysis used an ingredients-based, micro-costing approach to obtain the empirical economic costs of each procedure in the TB diagnostic algorithms for PLHIV and people with presumptive MDR TB as implemented in the study sites. These procedures were clinical exam, chest x-ray, fluorescent microscopy (FM) smear, Ziehl-Neelsen (ZN) smear, Xpert, Mycobacteria growth indicator tube (MGIT) liquid culture (MGIT 960, Becton Dickinson, USA), follow-on work-up for a positive liquid culture result and MTB identification, and DST.\nThe cost categories included were personnel, equipment, equipment maintenance, supplies, and specimen transport directly related to the diagnostic procedures. Costs for general hospital and laboratory administration, facility space, and utilities were not included as these were not expected to differ due to the introduction of Xpert testing in the study sites. Personnel time was valued at average monthly wages and benefits for each position as paid by the Cambodian Ministry of Health, including the costs for personnel hired as full-time government employees and as consultants using international donor funds. Donated equipment and supplies were valued at their replacement cost using current market prices quoted by vendors that regularly supply CENAT and the Ministry of Health. Costs of\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 5\n\ntransporting centrally procured drugs, supplies, and equipment from the CENAT warehouse in Phnom Penh to sites were not included. Equipment maintenance was defined to include all services and investments required to keep diagnostic procedure equipment in working order (e.g., personnel time, spare parts, calibration services). Specimen transport included the costs of personnel time, courier services, transportation, and per diem.\nThe time frame for the cost analysis was from the date of GeneXpert placement in 2012 (February for BTB, June for MKB) through the end of the program evaluation in June 2013. When historical cost data were not available, cost data that were current as of the date of data collection (May-July 2014) were used. For each site, a single set of procedure unit costs were calculated, which were then used to calculate the total cost per TB case diagnosed under each algorithm: the pre-Xpert algorithms in use before GeneXpert placement in the evaluation sites in 2012, and the post-Xpert algorithms introduced after 2012. The differences in pre- and post-Xpert algorithm costs at each site therefore only reflect the differences in the type and number of procedures used under each algorithm (i.e., differences in the combination of procedures) and not differences in the unit costs of the individual procedures across algorithms.\n\n2.3. Data Collection Procedures\nPrimary cost data were collected by CDC staff at the study sites in May-July 2014 using a standard set of Excel-based tools for TB clinic costs, laboratory costs, and CENAT national lab and program costs. Not all procedures were performed at all sites. Cost data were collected only for procedures performed at each site per the diagnostic algorithms for PLHIV and people with presumptive MDR TB in the four-province catchment area. Data were collected through consultation with clinical and laboratory staff at the study sites, procurement and financial staff at CENAT, and private sector vendors that regularly supply CENAT and the Ministry of Health, as well as through observation of laboratory procedures and review of laboratory records. Personnel time was collected through direct observation of laboratory procedures and from clinical and laboratory staff members\u2019 self-reports of their time or percent effort spent on each activity. The costs of equipment and supplies that were shared across procedures were allocated on the basis of the reported percent of space or percent of time used for each procedure as reported by laboratory staff at each site. For each site, the data collected included the prices and quantities of inputs used for each procedure, allocation of inputs across procedures, and monthly procedure volumes. Cost data were recorded manually during site visits and then entered into electronic versions of the tools.\n\n2.4. Cost Analysis\nThe cost analysis had three components: (i) calculation of the unit cost for each procedure included in either the pre- or post-Xpert algorithm for PLHIV and patients with presumptive MDR TB, (ii) calculation of the total costs per TB case diagnosed under the pre- and postXpert algorithms for PLHIV and patients with presumptive MDR TB under the hypothetical assumptions of perfect (i.e., 100%) test accuracy and that the algorithms were followed as written, and (iii) calculation of the total costs per TB case diagnosed and treated under the pre- and post-Xpert algorithms for PLHIV accounting for differential sensitivity and specificity of FM smear versus Xpert.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 6\nIn the first component of calculating the unit cost for each procedure, the purchase price, annual cost, or monthly cost of each input (i.e., each type of personnel, supplies, equipment, and equipment maintenance) was converted to a per-procedure cost (Supplemental Box 1). For personnel, the average monthly salary for a given cadre was multiplied by the share of time that that cadre of personnel spent on the procedure in the evaluation site divided by the number of procedures performed per month in that site (Supplemental Table 2). For singleuse supplies (e.g., glass slide, Xpert cartridge), the purchase price was divided by the quantity purchased and multiplied by the quantity used in the procedure. For supplies used across more than one test (e.g., gloves, pens, markers, waste bags), the purchase price was divided by the quantity purchased and the number of tests performed during the time period before the supply needed to be replaced (e.g., per month). For equipment, the purchase price was converted to an annual equivalent cost based on the number of years of useful life of the equipment, the resale value of the equipment at the end of its useful life (assumed to be US $0 in all cases), and an assumed 3% annual discount rate. The annual equivalent cost of each piece of equipment was then divided by the actual average number of procedures performed per year in the evaluation site to obtain the per-procedure cost. For maintenance costs, the reported annual cost of maintenance for each piece of equipment was divided by the average number of procedures performed per year in the evaluation site to obtain the per-procedure cost. The unit cost for each procedure was then calculated as the sum of the per-procedure costs for each input. Costs collected in Cambodian riel were converted to 2014 U.S. dollars at the rate of 4000 riel per dollar, the average current exchange rate at the time of data collection, and subsequently adjusted to 2017 U.S. dollars using the GDP implicit price deflator.(26)\nSensitivity analyses were performed to examine how the unit cost of an Xpert test would change under different scenarios for the useful life of the GeneXpert instrument, the cartridge cost, and the average monthly test volume. These factors were selected for sensitivity analysis because of the Ministry of Health\u2019s uncertainty around the performance of GeneXpert under conditions of routine use in Cambodia (as the instruments were new technology placed in 2012, it was not known how long they would last beyond the date of data collection) and the availability of future concessional pricing for cartridges. The effect of varying monthly test volume was explored as a factor that directly impacts the unit cost of an Xpert test and that may be amenable to policy or programmatic intervention. We also examined how the unit costs of other procedures would change if monthly procedure volumes were held constant across sites to examine the degree to which factors besides volume influenced costs.\nIn the second component of the cost analysis, the normative cost to diagnose TB in a single patient under each of the four diagnostic algorithms (pre- and post-Xpert for PLHIV and pre- and post-Xpert for patients with presumptive MDR TB) was calculated as the sum of each procedure\u2019s empirical unit cost multiplied by the normative frequency of that procedure as written in the algorithm, under the hypothetical assumptions that the algorithm was followed as written and that each procedure was perfectly (i.e., 100%) accurate. Although empirical procedure unit costs were used, this is termed a normative cost analysis of the algorithm because it calculates the cost for the algorithm as it should have been implemented (i.e., the algorithm \u201con paper\u201d), not the actual practice of clinicians in the evaluation sites,\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 7\nwhich may have differed from the algorithm. In the third cost analysis component, this normative analysis was repeated for the PLHIV pre- and post-Xpert algorithms replacing the assumption that each procedure is perfectly accurate with sensitivity and specificity values from the literature for FM smear and Xpert to examine the potential implications for treatment costs under the post-Xpert algorithm.\n\n3. Results\n3.1. Empirical Unit Costs per Procedure\nUnit costs per procedure and average monthly procedure volumes varied by site (reported in Table 1). Unit costs were lower in the facilities with higher test volumes, as the costs of equipment, maintenance, and personnel were amortized over a larger number of tests.\nThe inputs representing the largest share of the unit cost per procedure at each site were fairly consistent across sites (Table 2), though the exact shares varied based on procedure volume and some differences in available equipment and personnel. Consumables and equipment represented the largest input cost shares for all procedures, with the exception of clinical diagnostic assessment, FM smear at BTB, and DST at the CENAT national lab for which personnel time was among the two most costly inputs.\nSpecimen transport costs for Xpert, culture, and DST varied by site location and patient type (PLHIV versus people with presumptive MDR TB) and were therefore calculated separately from the unit costs per procedure. In the study\u2019s four-province catchment area, for both PLHIV and people with presumptive MDR TB, the cost to transport a set of three sputum samples from lower-level health clinics and operational district hospitals to the two provincial referral hospitals for Xpert testing ranged from US$4.17-US$14.59 (US$1.39-US $4.87 per sample). Samples sent from MKB to BTB for culture testing incurred an additional US$10.42 transport cost (US$3.47 per sample) for both PLHIV and people with presumptive MDR TB. If the lower-level facility referred a patient with presumptive MDR TB to the provincial referral hospital to provide sputum samples there, the patient could receive US$15.63 to cover transport and overnight stay costs (US$5.21 per sample). Specimen transport from the provincial referral hospitals to the CENAT national laboratory for DST for both PLHIV and people with presumptive MDR TB cost US$109.43 per month on average (US$4.72 per sample based on average monthly sample volumes). There were no specimen transport costs for sputum smear microscopy, which can be performed at lowerlevel health facilities.\nCosts for community directly-observed therapy short course (C-DOTS) were calculated as US$11.21 per month for the BTB provincial referral hospital catchment area and US$8.60 per month for the MKB provincial referral hospital catchment area. Differences in C-DOTS costs were due to differences in the personnel and funding models used during the study period in each catchment area. First and second line drug costs were assumed to be the same across all study sites as drugs are centrally procured by CENAT. On average, a category 1 first line regimen of RIF and isoniazid (each for six months) plus pyrazinamide and ethambutol (each for the first two months only) cost US$26.05 per patient, whereas a five- to eight-month category 2 regimen for treatment failure, relapse, and retreatment cases that\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 8\n\nadded streptomycin cost US$104.22 per patient, and a 24-month regimen of second line drugs (various formulations) for patients with multidrug resistance cost US$3,126.50 per patient. Patients receiving treatment also receive an average of three ZN smears performed at lower-level health clinics or operational district hospitals to monitor their response to treatment.\n\n3.2. Normative Cost Estimates per Patient Diagnosed and Treated\nAssuming hypothetical perfect test accuracy, for PLHIV, the cost per patient diagnosed with TB under the post-Xpert algorithm is US$34-US$37 compared to US$9 under the base case pre-Xpert algorithm, whereas for people with presumptive MDR TB, the cost per patient diagnosed with MDR TB under the post-Xpert algorithm is US$169-US$201 compared to US$135-US$164 under the base case pre-Xpert algorithm (Table 3). In this case of perfect test accuracy, the cost per patient diagnosed by these multi-step algorithms is equivalent to the cost per patient tested, assuming no dropout at each stage of testing. Ranges reflect the variation in unit costs per procedures across the study site sample. For comparison, the treatment cost implications for PLHIV were estimated based on the estimated sensitivity and specificity of FM smear (Se: 73%, Sp: 100%) versus Xpert (Se: 88%, Sp: 99%).(11, 27) Under these parameters, the post-Xpert algorithm results in an estimated 1% of patients receiving a false positive diagnosis and 12% receiving a false negative diagnosis following Xpert testing, compared to no false positives and 27% false negatives under the pre-Xpert algorithm. The reduction in false negatives and slight increase in false positives under the post-Xpert algorithm means that more of the PLHIV tested would be initiated on TB treatment under this algorithm than the pre-Xpert algorithm; this increase in treatment costs (US$13-US$16 per PLHIV diagnosed with TB) plus the increase in diagnostic costs (US $25-US$28 per PLHIV diagnosed with TB) increases the total diagnosis and treatment cost of the post-Xpert algorithm byUS$41 per PLHIV diagnosed with TB relative to the preXpert algorithm using FM smear.\n\n3.3. Sensitivity Analyses\nChanges to the useful life of the GeneXpert instrument, the price per Xpert test cartridge, and the average monthly test volume change the unit cost per Xpert test, with alternative unit costs ranging from US$13.23-US$42.32 (Table 4). The actual unit cost was calculated based on an expected 5-year useful life for the 4-module GeneXpert instrument and the reported bulk cartridge procurement price at the time of data collection of US$12.51 per cartridge. The greatest reductions in unit costs were achieved by increases in average monthly test volumes from current levels (40 at BTB and 25 at MKB) to maximum daily utilization of GeneXpert (three runs per day of a 4-module instrument, or 240 tests per month), even when the test cartridge price was increased to commercial levels for non-bulk purchases. Changes to the test cartridge price change the unit cost per Xpert test directly (e.g., if the Cambodian Ministry of Health had to procure cartridges in smaller batches at the commercially available price of US$17.72 per cartridge, the unit cost would increase by US$5.21 per test). Reducing the test cartridge price to US$10.42 per cartridge under concessional pricing that Cambodia currently receives reduces the unit cost per test more than would doubling the useful life of the GeneXpert instrument from five to 10 years.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 9\nIn a supplemental analysis holding constant average monthly procedure volumes across sites (Supplemental Table 1), unit costs increased when the monthly procedure volumes of the lower volume site were applied to the higher volume site. For chest x-ray, FM smear, and Xpert, the hypothetical unit costs at BTB provincial referral hospital (US$2.28, US$2.54, and US$40.40, respectively) under the lower procedure volumes from MKB provincial referral hospital were higher than the empirical unit costs at either site, reflecting the presence of more expensive personnel and equipment at the BTB laboratory. For liquid culture using MGIT, the hypothetical unit cost at the CENAT national laboratory (US$20.50) under the lower procedure volumes from BTB provincial referral hospital was higher than the empirical unit costs for the CENAT national laboratory but still lower than the empirical unit costs for BTB provincial referral hospital, suggesting that the national laboratory may still have some process efficiencies in performing liquid culture compared to BTB. For follow-on work up for positive MGIT results and MTB identification, the hypothetical unit cost at the CENAT national laboratory (US$16.60) under the lower procedure volumes from BTB provincial referral hospital were higher than the empirical unit costs for both sites.\n\n4. Discussion\nThis study presents the first cost analysis of tuberculosis diagnostic procedures in Cambodia. Our unit costs are within the range of estimates for these procedures from other countries, falling towards the upper end of previously published estimates (18, 20, 28\u201332). Prior cost studies do not always report the average volume of tests performed in their settings, however, making it difficult to determine the extent to which test throughput (rather than input prices) explains variation in unit costs. Procedure volume was an important driver of unit costs across the sites sampled for this study as many equipment and supply prices were constant across sites due to centralized government procurement; however, the supplemental sensitivity analysis holding volume constant indicates that this is not the only cost driver. Variations in the number and cadre of personnel and in the allocation of tasks across personnel also partially explained differences in unit costs across study sites. For example, BTB had additional tuberculosis staff at the time of the study as a legacy of other donorfunded projects related to TB-HIV co-infection, whereas such additional clinical staff were not funded at MKB. As another example, sites that used higher-paid temporary contractors (BTB and CENAT) paid by international donor funds to fill in staffing gaps rather than hiring lower-paid full-time government employees had higher unit costs for procedures performed by contractors. Our analysis used the salary rates of these internationally-funded contractors as there was no indication of planned transition of these roles to governmentfunded staff, and it was not clear if similarly skilled personnel could be recruited and retained at government salary rates. As a hypothetical counterfactual, if the same quality and quantity of tests could have been performed by laboratory staff paid at lower government salary rates, the unit costs for the procedures performed by these staff would have been reduced (BTB FM smear: from US$1.56 to US$1.43; CENAT DST: from US$44.26 to US $36.57) and personnel time would not have been among the two most costly inputs per procedure at these sites; instead, the two inputs contributing the largest shares of the unit cost per test would have been the monetized value of the share of the equipment used for FM smear at BTB (biosafety cabinet, 19% (or US$0.27 per test), and binocular fluorescent\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 10\nmicroscope, 15% (or US$0.22 per test)) and supplies for DST at CENAT (sterile filter tips 100\u20131000 ul, 18% (or US$6.59 per test); sterile filter tips 20\u2013200 ul, 15% (or US$5.48 per test)).\nThe sensitivity analysis indicates that Xpert unit costs in Cambodia could be reduced through access to further discounted concessional pricing per Xpert cartridge, as had been negotiated through UNITAID (1), or through extending the life of current GeneXpert instruments beyond the five years assumed by program planners in Cambodia. The greatest reductions in Xpert unit cost would come from increasing utilization of current GeneXpert instruments. A primary strategy to increase GeneXpert instrument utilization would be to ensure that all individuals currently eligible for Xpert testing (i.e., PLHIV with a positive TB symptom screen and people with presumptive MDR TB) receive it through strengthened case finding, patient referral, and specimen transport. Given Cambodia\u2019s epidemiological context, however, there is a ceiling on the maximum number of Xpert tests that could be performed if Xpert testing continues to be limited to PLHIV with a positive TB symptom screen and people with presumptive MDR TB. Under Cambodia\u2019s current diagnostic algorithms for these populations, there are not sufficient numbers of individuals in the four provinces of our study setting who are eligible for Xpert testing to permit the current GeneXpert instruments to be used at their maximum run capacity of 240 tests per month.\nUnder these conditions, placing additional GeneXpert instruments at lower-level health facilities may not be efficient, as doing so would only further reduce the testing volume at the current provincial referral hospital laboratories and increase per test costs. In addition, TB control program officials in Cambodia and other countries should consider carefully whether high-volume procurement of Xpert cartridges to secure more favorable pricing is warranted when the population eligible for Xpert testing is limited, so as to avoid oversupply and expiration of cartridges. Exploring additional options for pooled procurement, including at the regional level, may be a way to obtain lower prices for smaller volumes appropriate to the country\u2019s epidemiological profile and Xpert testing eligibility criteria. Improvements in sample transport and patient referral may also facilitate increased utilization of current GeneXpert instruments to ensure that all patients who are currently eligible for Xpert testing receive it. As the WHO has recommended expansion of Xpert testing to pediatric patients and patients presumed to have some types of extrapulmonary TB, there may be scope to increase the utilization of current GeneXpert instruments in Cambodia through expansions in the eligibility guidelines to cover children with presumptive TB, HIV-associated TB, and drug-resistant TB. WHO also has a conditional recommendation for the use of Xpert as a follow-up test to microscopy in settings in which HIV and MDR TB are of lesser concern, which may also suggest options for broader use of existing GeneXpert instruments in Cambodia.\nIn addition, our illustrative normative cost estimates do not account for health system cost savings due to reduced TB transmission in the community or for the societal benefits of improved quality of life and survival for PLHIV with TB. For people with presumptive MDR TB, the post-Xpert algorithm does not offer any cost savings per patient diagnosed or treated as Xpert testing is added on to the existing diagnostic tests of culture and DST. In theory, it is possible that the post-Xpert algorithm could still be more cost-effective than the\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 11\n\npre-Xpert algorithm for this population when the benefits of reduced MDR TB transmission and earlier treatment initiation on quality of life and survival are taken into account; however, this cost analysis does not assess the algorithms\u2019 cost-effectiveness.\n\n4.1. Limitations\nThe cost analysis results should be interpreted in light of several limitations. First, the site sample is not a representative sample of all health facilities in Cambodia, limiting the generalizability of the results. Second, historical cost and procedure volume data were not available in all cases; our results therefore do not capture changes in unit costs before and after Xpert placement that might reflect changes in laboratory workflow or efficiencies from learning the new Xpert testing procedure. Third, the value of personnel time was based on self-reported time or percent effort by clinical and laboratory staff. Some validation of these reports was conducted through observation of staff performing diagnostic procedures at the CENAT and BTB labs; however, it is possible that what was reported and what was observed by data collectors does not reflect how time is spent in routine practice without observation. Fourth, the normative analysis of cost per patient diagnosed under each algorithm, while based on empirical unit costs per procedure, relied on assumptions of test accuracy and clinician guideline adherence that are not expected to represent actual programmatic implementation conditions in Cambodia; this normative analysis helps isolate the variation in algorithm costs due to the combination of procedures used in each algorithm but does not necessarily predict what the actual empirical cost per patient diagnosed will be in this setting. Fifth, our estimates excluded overhead costs, which may be relevant in other settings. Finally, the perspective of the analysis did not include patient costs, and therefore does not capture changes in patient\u2019s out-of-pocket costs to provide sputum specimens or obtain test results and health care due to changes in the diagnostic algorithms.\n\n5. Conclusions and Implications for Future Research\nAs a high-burden TB country with low HIV prevalence, Cambodia offers a distinct epidemiological context in which to examine the costs and cost-effectiveness of Xpert implementation, contributing to the existing literature on Xpert costs in settings with higher prevalence of HIV and TB among PLHIV. Our initial normative cost estimate shows that the use of Xpert for diagnosing TB and RIF resistance in PLHIV and people with presumptive MDR TB is more expensive than previous diagnostic algorithms without Xpert; however, a full cost-effectiveness analysis is needed to weigh these increased costs against Xpert\u2019s benefits of more timely and accurate diagnoses given that a major rationale for Xpert adoption is the reduced time to diagnosis and treatment and, by extension, the reduced number of new TB infections. This cost analysis did not consider the differential effectiveness of the pre- and post-Xpert algorithms in terms of disease transmission, morbidity, or mortality; these cost data are intended as inputs to cost-effectiveness analyses and should not be interpreted as a recommendation in favor or against Xpert use without consideration of Xpert\u2019s potential public health impacts. The empirical cost per patient diagnosed will depend on the observed sensitivity and specificity of each procedure as well as the proportion of patients receiving each procedure in the study sites, as actual clinical practice may deviate from the algorithms; these data are forthcoming from the broader\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 12\nprogram evaluation. Nevertheless, our cost analysis results can inform future planning to optimize GeneXpert placement and use within Cambodia and to project the costs that the Ministry of Health may need to assume in the future if the sources of external aid that supported the initial GeneXpert rollout decrease. Our study also contributes to the literature on the costs of Xpert in routine implementation settings. Future research should examine the relative costs and benefits of placing GeneXpert instruments at lower-level versus more centralized health facilities and how improvements in patient referral and specimen transport may be able to increase test volumes with existing instruments. Future research should also examine the costs and cost-effectiveness of Xpert testing using the Xpert\u00ae MTB/RIF Ultra test with higher sensitivity and specificity for smear-negative TB and if other tests using the GeneXpert platform (such as HIV viral load) are available, which could increase the population health benefits derived from existing GeneXpert investments.\n\nSupplementary Material\nRefer to Web version on PubMed Central for supplementary material.\n\nAcknowledgements:\nThe authors would like to acknowledge the input and feedback received from staff at the Cambodia National TB Program (CENAT), Dr. Kanara Nong and Mr. Huot Uong of CDC/Cambodia, and respondents at cost data collection sites in Battambang Provincial Referral Hospital, Mongkol Borei Provincial Referral Hospital, Mong Russey Operational District Hospital, and the CENAT National Laboratory.\nCompliance with Ethical Standards\nFunding: This work was supported by funding provided by the U.S. Agency for International Development and the U.S. Centers for Disease Control and Prevention, in addition to funding from the President\u2019s Emergency Plan for AIDS Relief (PEPFAR) through the U.S. Centers for Disease Control and Prevention. The findings and conclusions in this presentation are those of the authors and do not necessarily represent the official position of the U.S. Centers for Disease Control and Prevention.\nReferences\n1. World Health Organization. Global Tuberculosis Report 2014 . Geneva: World Health Organization; 2014.\n2. World Health Organization. Rapid implementation of the Xpert MTB/RIF diagnostic test: Technical and operational \u2018How to\u2019 practical considerations . Geneva: World Health Organization; 2011.\n3. World Health Organization. Automated Real-time Nucleic Acid Amplification Technology for Rapid and Simultaneous Detection of Tuberculosis and Rifampicin Resistance: Xpert MTB/RIF System: Policy Statement . Geneva: World Health Organization; 2011.\n4. Padmapriyadarsini C, Narendran G, Swaminathan S. Diagnosis and treatment of tuberculosis in HIV co-infected patients Indian J Med Research. 2011; 134(6):16.\n5. World Health Organization. Tuberculosis Diagnostics Xpert MTB/RIF Test Fact Sheet . Geneva: World Health Organization; 2013.\n6. Jones BE, Young SM, Antoniskis D, Davidson PT, Kramer F, Barnes PF. Relationship of the manifestations of tuberculosis to CD4 cell counts in patients with human immunodeficiency virus infection Am Rev Respir Dis. 1993; 148(5):6. [PubMed: 8317816]\n7. Perlman DC, el-Sadr WM, Nelson ET. Variation of chest radiographic patterns in pulmonary tuberculosis by degree of human immunodeficiency virus-related immunosuppression Clin Infect Dis. 1997; 25(2):5.\n8. Whalen C, Horsburgh CR, Hom D, Lahart C, Simberkoff M, Ellner J. Accelerated course of human immunodeficiency virus infection after tuberculosis Am J Respir Crit Care Med. 1995; 151(1):7.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 13\n9. Gandhi NR, Shah NS, Andrews JR. HIV coinfection in multidrug- and extensively drug-resistant tuberculosis results in high early mortality Am J Respir Crit Care Med. 2010; 181(1):7. [PubMed: 19815810]\n10. Piatek A, van Cleef M, Alexander H, Coggin W, Rehr M, van Kampen S. GeneXpert for TB diagnosis: planned and purposeful implementation Glob Health Sci Pract. 2013; 1(1):6.\n11. World Health Organization. Automated real-time nucleic acid amplification technology for rapid and simultaneous detection of tuberculosis and rifampicin resistance: Xpert MTB/RIF assay for the diagnosis of pulmonary and extrapulmonary TB in adults and children. Policy update. Geneva : World Health Organization ; 2013.\n12. Boehme CC, Nicol MP, Nabeta P, Michael JS, Gotuzzo E, Tahirli R. Feasibility, diagnostic accuracy, and effectiveness of decentralised use of the Xpert\u00ae MTB/RIF test for diagnosis of tuberculosis and multidrug resistance: a multicentre implementation study Lancet. 2011; 377(9776):11. [PubMed: 20797780]\n13. Boehme CC, Nabeta P, Hillemann D, Nicol MP, Shenai S, Krapp F. Rapid Molecular Detection of Tuberculosis and Rifampin Resistance N Engl J Med. 2010; 363(11):11. [PubMed: 20505173]\n14. Lawn SD, Nicol MP. Xpert\u00ae MTB/RIF assay: development, evaluation and implementation of a new rapid molecular diagnostic for tuberculosis and rifampicin resistance Future Microbiology. 2011; 6(9):16.\n15. Carriquiry G, Otero L, Gonzalez-Lagos E, Zamudio C, Sanchez E, Nabeta P. A Diagnostic Accuracy Study of Xpert\u00ae (R) MTB/RIF in HIV-Positive Patients with High Clinical Suspicion of Pulmonary Tuberculosis in Lima, Peru PLoS ONE. 2012; 7(9)\n16. Chang K, Lu W, Wang J, Zhang K, Jia S, Li F. Rapid and effective diagnosis of tuberculosis and rifampicin resistance with Xpert\u00ae MTB/RIF assay: A meta-analysis Journal of Infection. 2012; 64(6):9.\n17. Yoon C, Cattamanchi A, Davis JL, Worodria W, den Boon S, Kalema N. Impact of Xpert\u00ae MTB/RIF Testing on Tuberculosis Management and Outcomes in Hospitalized Patients in Uganda PLoS ONE. 2012; 7(11)\n18. Vassall A, van Kampen S, Hojoon S, Michael JS, John KR, den Boon S. Rapid Diagnosis of Tuberculosis with the Xpert MTB/RIF Assay in High Burden Countries: A Cost-Effectiveness Analysis PLoS Medicine. 2011; 8(11):e1001120. [PubMed: 22087078]\n19. Andrews JR, Lawn SD, Rusu C, Wood R, Noubary F, Bender MA. The cost-effectiveness of routine tuberculosis screening with Xpert MTB/RIF prior to initiation of antiretroviral therapy: a model-based analysis AIDS. 2012; 26(8):987\u201395. [PubMed: 22333751]\n20. Menzies N, Cohen T, Lin H- H, Murray M, Salomon J. Population Health Impact and CostEffectiveness of Tuberculosis Diagnosis with Xpert MTB/RIF: A Dynamic Simulation and Economic Evaluation PLoS Medicine. 2012; 9(11):e1001347. [PubMed: 23185139]\n21. Choi HW, Miele K, Dowdy D, Shah M. Cost-effectiveness of Xpert\u00ae MTB/RIF for diagnosing pulmonary tuberculosis in the United States Int J Tuberc Lung Dis. 2013; 17(10):1328\u201335. [PubMed: 24025386]\n22. Pantoja A, Fitzpatrick C, Vassall A, Weyer K, Floyd K. Xpert MTB/RIF for diagnosis of tuberculosis and drug-resistant tuberculosis: a cost and affordability analysis Eur Respir J. 2012; 42(3):708\u201320. [PubMed: 23258774]\n23. Abimbola TO, Marston BJ, Date AA, Blandford JM, Sangrujee N, Wiktor SZ. Cost-effectiveness of tuberculosis diagnostic strategies to reduce early mortality among persons with advanced HIV infection initiating antiretroviral therapy J Acquir Immune Defic Syndr. 2012; 60(1):e1\u2013e7. [PubMed: 22240465]\n24. UNAIDS. The Gap Report. Geneva : Joint United Nations Programme on HIV/AIDS ; 2014.\n25. Auld SC, Moore BK, Killam WP, Eng B, Nong K, Pevzner EC. Rollout of Xpert\u00ae MTB/RIF in Northwest Cambodia for the diagnosis of tuberculosis among PLHA Public Health Action. 2014; 4(4):216\u201321. [PubMed: 26400699]\n26. Gross Domestic Product Implicit Price Deflator [GDPDEF] [Internet]. 2018 [cited April 26, 2018]. Available from: https://fred.stlouisfed.org/series/GDPDEF.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 14\n27. Steingart KR, Henry M, Ng V, Hopewell PC, Ramsay A, Cunningham J. Fluorescence versus conventional sputum smear microscopy for tuberculosis: a systematic review Lancet Infectious Diseases. 2006; 6:570\u201381. [PubMed: 16931408]\n28. Smit P, Sollis K, Fiscus S, Ford N, Vitoria M, Essajee S. Systematic Review of the Use of Dried Blood Spots for Monitoring HIV Viral Load and for Early Infant Diagnosis PLoS ONE. 2014; 9(3):e86461. [PubMed: 24603442]\n29. Van Rie A, Page-Shipp L, Hanrahan C, Schnippel K, Dansey H, Bassett J. Point-of-care Xpert\u00ae MTB/RIF for smear-negative tuberculosis suspects at a primary care clinic in South Africa International Journal of Tuberculosis and Lung Disease : the official journal of the International Union against Tuberculosis and Lung Disease. 2013; 17(3):368\u201372.\n30. Schnippel K, Meyer-Rath G, Long L, MacLeod W, Sanne I, Stevens W. Scaling up Xpert MTB/RIF technology: the costs of laboratory- vs. clinic-based roll-out in South Africa Tropical Medicine & International Health. 2012; 17:1142\u201351. [PubMed: 22686606]\n31. Atif M, Sulaiman SA, Shafie AA, Saleem F, Ahmad N. Determination of chest x-ray cost using activity based costing approach at Penang General Hospital, Malaysia The Pan African Medical Journal. 2012; 12:40. [PubMed: 22891098]\n32. Pang Y, Li Q, Ou X, Sohn H, Zhang Z, Li J. Cost-Effectiveness Comparison of Genechip and Conventional Drug Susceptibility Test for Detecting Multidrug-Resistant Tuberculosis in China PLoS ONE. 2013; 8(7)\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nPallas et al.\n\nPage 15\nKey Points for Decision Makers \u2022 Across 4 provinces in Cambodia, Xpert MTB/RIF costs were US$33.88-US\n$37.11 per test. \u2022 Tuberculosis diagnostic algorithms with Xpert were more expensive than pre-\nXpert. \u2022 Increasing use of existing GeneXpert instruments would decrease Xpert test\ncosts. \u2022 Test eligibility and specimen referral practices may limit optimal GeneXpert\nuse.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nTable 1. Unit costs per procedure and average monthly procedure volumes by site (costs in 2017 U.S. dollars)\n\nPallas et al.\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nProcedure\n\nBTB\n\nMKB\n\nMR OD CENAT\n\nClinical diagnostic assessment\n\nUnit Cost\n\nUS$1.84\n\nUS$1.22\n\nProcedure Volume\n\n500\n\n500\n\nChest x-ray, automatic development (*manual development)\n\nUnit Cost\n\nUS$2.02\n\nUS$2.14 (US$2.06*)\n\nProcedure Volume\n\n275 (chest)\n\n200 (chest)\n\n531 (all x-rays) 350 (all x-rays)\n\nSmear Microscopy \u2013 FM\n\nUnit Cost\n\nUS$1.56\n\nUS$1.93\n\nProcedure Volume\n\n500\n\n300\n\nSmear Microscopy \u2013 ZN\n\nUnit Cost\n\nUS$1.26\n\nProcedure Volume\n\n140\n\nXpert\n\nUnit Cost\n\nUS$33.88\n\nUS$37.11\n\nProcedure Volume\n\n40\n\n25\n\nLiquid Culture \u2013 MGIT\n\nUnit Cost\n\nUS$22.83\n\nUS$11.63\n\nProcedure Volume\n\n112\n\n350\n\nFollow-on work up for positive MGIT results and MTB identification Unit Cost\n\nUS$14.72\n\nUS$11.63\n\nProcedure Volume (Follow-on work-up)\n\n23\n\n85\n\nProcedure Volume (MTB identification)\n\n28\n\n45\n\nDST\n\nUnit Cost\n\nUS$44.26\n\nProcedure Volume\n\n23\n\n\u2022 BTB: Battambang Provincial Referral Hospital; MKB: Mongkol Borei Provincial Referral Hospital; MR OD: Mong Russey Operational District Hospital; CENAT: National TB Control Program Lab \u2022 At the time of the study, the GeneXpert instruments did not operate at maximum possible capacity every day at either the BTB or MKB site due to low specimen volumes received.\n\nPage 16\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nInputs representing largest share of unit costs per procedure by site\n\nTable 2.\n\nPallas et al.\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nProcedure Clinical diagnostic assessment Chest x-ray, automatic development Smear Microscopy \u2013 FM Smear Microscopy \u2013 ZN\nXpert Liquid Culture \u2013 MGIT Follow-on work up for positive MGIT results and MTB identification DST\n\nBTB\nClinician time (49%), Nurse time (35%)\nX-ray film (52%), X-ray machine (14%)\nLab technician time (17%), Biosafety cabinet (16%)\n\nMKB\nClinician time (74%), Nurse time (18%)\nX-ray film (49%), X-ray machine (21%)\nN95 mask (24%), Microscope (18%)\n\nXpert cartridge (37%),\n\nGeneXpert instrument (36%),\n\nGeneXpert instrument (24%) Xpert cartridge (34%)\n\nMGIT 960 instrument (32%), MGIT 7mL tube (10%)\n\nMTB identification test (22%), Autoclave (12%)\n\nMR OD\nReagent kit (41%), Sputum cup (15%)\n\nCENAT\nMGIT 7mL tube (20%), MGIT 960 instrument (14%) MTB identification test (29%), Autoclave (15%) Lab specialist (contractor) time (21%), Sterile filter tips (15%)\n\n\u2022 BTB: Battambang Provincial Referral Hospital; MKB: Mongkol Borei Provincial Referral Hospital; MR OD: Mong Russey Operational District Hospital; CENAT: National TB Control Program Lab\n\nPage 17\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nTable 3. Procedure volume and estimated normative cost per patient diagnosed and treated under Cambodia\u2019s TB diagnostic algorithms for PLHIV and people with presumptive MDR TB (costs in 2017 U.S. dollars)\n\nPallas et al.\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nProcedure\n\nClinical diagnostic assessment\n\nProcedure Frequency\n\nUnit Cost\n\nChest x-ray, automatic development\n\nProcedure Frequency\n\nUnit Cost\n\nSmear Microscopy \u2013 FM\n\nProcedure Frequency\n\nUnit Cost\n\nXpert\u00ae MTB/RIF\n\nProcedure Frequency\n\nUnit Cost\n\nLiquid Culture \u2013 MGIT\n\nProcedure Frequency\n\nUnit Cost\n\nFollow-on work up for positive MGIT results and MTB identification\n\nProcedure Frequency\n\nUnit Cost\n\nDrug Susceptibility Testing (DST)\n\nProcedure Frequency\n\nUnit Cost\n\n* DIAGNOSTIC TEST COSTS per TB+ patient (assuming perfect accuracy)\n\nCommunity Directly Observed Therapy Short Course (C-DOTS)\n\nProcedure Frequency\n\nUnit Cost (per month)\n\nFirst Line Drugs (6 months)\n\nProcedure Frequency\n\nUnit Cost (per month)\n\nTreatment Monitoring Smear Microscopy \u2013 ZN\n\nProcedure Frequency\n\nUnit Cost\n\nTREATMENT COSTS per TB+ patient (assuming perfect test accuracy)\n\nScenario without perfect test accuracy\n\n1\nFalse positive rate (1-specificity)\n\n1\nFalse negative rate (1-sensitivity)\n\nNecessary treatment costs per patient due to true positives\n\nUnnecessary treatment costs per patient due to false positives\n\nTreatment cost increase per patient under post-Xpert algorithm compared to pre-Xpert algorithm (without perfect test accuracy, using assumed sensitivity and specificity)\n\nDiagnostic cost increase per patient under post-Xpert algorithm compared to pre-Xpert algorithm (without perfect test accuracy, using assumed sensitivity and specificity)\n\nTotal increase in diagnosis and treatment costs per patient under post-Xpert algorithm compared to pre-Xpert algorithm (without perfect test accuracy, using assumed sensitivity and specificity)\n\nPLHIV Pre-Xpert Algorithm (no drug resistance) 1 time US$1.22-US$1.84 1 time US$2.02-US$2.14 3 tests US$1.56-US$1.93\nUS$8.56-US$9.14 6 months US$8.60-US$11.21 6 months US$4.35 3 tests US$1.26 US$81.42-US$97.12\n0% 27% US$59.43-US$70.90 US$0.00\n\nPLHIV Post-Xpert Algorithm (no drug resistance)\n1 test US$33.88-US$37.11\nUS$33.88-US$37.03 6 months US$8.60-US$11.21 6 months US$4.35 3 tests US$1.26 US$81.42-US$97.12 1% 12% US$71.65-US$85.47 US$0.81-US$0.97 US$13.03-US$15.54 US$25.32-US$27.89 US$40.86-US$40.92\n\nMDR TB Pre-Xpert Algorithm\n2 tests US$11.63-US$22.83 2 tests US$11.50-US$14.72 2 tests US$44.26 US$134.77-US$163.62\n\nMDR TB Post-Xpert Algorithm\n1 test US$33.88-US$37.11 2 tests US$11.63-US$22.83 2 tests US$11.50-US$14.72 2 tests US$44.26 US$168.65-US$200.65\n\n* Range in costs is due to the variation in unit costs per procedure across study sites. 1. Based on assumed sensitivity of 73% and specificity of 100% for FM smear and sensitivity of 88% and specificity of 99% for Xpert (11, 27).\n\nPage 18\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nAuthor Manuscript\n\nTable 4. Unit cost of Xpert test under different scenarios for GeneXpert useful life, cartridge prices, and monthly test volume (costs in 2017 U.S. dollars)\n\nPallas et al.\n\nAppl Health Econ Health Policy. Author manuscript; available in PMC 2019 August 01.\n\nScenario Alternative scenario 5 Alternative scenario 11 Alternative scenario 2 Alternative scenario 4 Alternative scenario 10 Alternative scenario 8 Alternative scenario 1 Alternative scenario 7 Alternative scenario 9 Alternative scenario 3 Actual study unit costs Alternative scenario 6\n\nUseful life of GeneXpert instrument 5 years 10 years 5 years 5 years 10 years 5 years 5 years 5 years 10 years 5 years 5 years 5 years\n\nPrice per Cartridge\n\nAverage Monthly Xpert Test Volume BTB Xpert Unit Cost MKB Xpert Unit Cost\n\nUS$10.42 (concessional pricing)\n\nMax utilization (BTB:240; MKB:240)\n\nUS$12.51 (current bulk purchase pricing) Max utilization (BTB:240; MKB:240)\n\nUS$12.51 (current bulk purchase pricing) Max utilization (BTB:240; MKB:240)\n\nUS$10.42 (concessional pricing)\n\nDouble current levels (BTB:80; MKB:50)\n\nUS$12.51 (current bulk purchase pricing) Double current levels (BTB:80; MKB:50)\n\nUS$17.72 (commercial pricing)\n\nMax utilization (BTB:240; MKB:240)\n\nUS$12.51 (current bulk purchase pricing) Double current levels (BTB:80; MKB:50)\n\nUS$17.72 (commercial pricing)\n\nDouble current levels (BTB:80; MKB:50)\n\nUS$12.51 (current bulk purchase pricing) Current levels (BTB:40; MKB:25)\n\nUS$10.42 (concessional pricing)\n\nCurrent levels (BTB:40; MKB:25)\n\nUS$12.51 (current bulk purchase pricing) Current levels (BTB:40; MKB:25)\n\nUS$17.72 (commercial pricing)\n\nCurrent levels (BTB:40; MKB:25)\n\nUS$14.22 US$15.65 US$16.30 US$21.25 US$21.41 US$21.51 US$23.33 US$28.54 US$30.05 US$31.80 US$33.88 US$39.09\n\nUS$13.23 US$14.67 US$15.31 US$22.85 US$21.86 US$20.52 US$24.94 US$30.15 US$30.96 US$35.03 US$37.11 US$42.32\n\nNotes: Bold cells indicate dimensions that are varied compared to the actual study unit costs. Max utilization: maximum utilization of three runs per day of a 4-module GeneXpert instrument (12 tests per working day). All unit costs include costs of annual calibration kit (1 kit per instrument, US$521/kit), but not for replacement cartridges or sending instruments abroad for servicing as these were not reported during the study period.\n\nPage 19\n\n\n",
    "authors": [
      "Sarah Wood Pallas",
      "Marissa Courey",
      "Chhaily Hy",
      "Wm. Perry Killam",
      "Dora Warren",
      "Brittany Moore"
    ],
    "doi": "10.1007/s40258-018-0397-3",
    "year": null,
    "item_type": "journalArticle",
    "url": "http://link.springer.com/10.1007/s40258-018-0397-3"
  },
  {
    "key": "B5UNS3PG",
    "title": "Cost-Effectiveness Analysis of Xpert MTB/RIF for Multi-Outcomes of Patients With Presumptive Pulmonary Tuberculosis in Thailand",
    "abstract": "Objectives: The cost-effectiveness of screening adult patients for pulmonary tuberculosis is not clear. As such, this study aims to identify the cost-effectiveness between the Xpert MTB/RIF assay and the sputum acid-fast bacilli (AFB) smear. Multioutcomes were correct diagnosis, time to achieve correct diagnosis, and gain in quality-adjusted life-years (QALYs).\nMethods: A decision tree model was constructed to reveal a possible clinical pathway of tuberculosis diagnosis. The researchers used a clinical study to establish the probability of all clinical pathways for input into this model. The sample size was calculated following the correct diagnosis. Participants were randomly divided into 2 groups. A structural questionnaire and the Thai version of quality of life (EQ-5D-5L) were used for interviewing.\nResults: The results showed that the time to achieve the correct diagnosis for the group using Xpert MTB/RIF was shorter than that for the group using the sputum AFB smear. Both the correct diagnosis and QALYs of the base case analysis presented the Xpert MTB/RIF method as dominant. A Monte Carlo model, which analyzed the Xpert MTB/RIF method, revealed that the average number of patients who were correctly diagnosed was 673, the QALYs were 945.85 years, and the total cost was $143 110.64. For the sputum AFB smear method, the average number who received a correct diagnosis was 592, the QALYs were 940.40 years, and the total cost was $196 666.84. Probabilistic and one-way sensitivity analysis con\ufb01rmed that the Xpert MTB/RIF remained dominant.\nConclusions: These results provide useful information for the National Strategic Plan to screen all adult patients for pulmonary tuberculosis.",
    "full_text": "Contents lists available at sciencedirect.com Journal homepage: www.elsevier.com/locate/vhri\nEconomic Evaluation\nCost-Effectiveness Analysis of Xpert MTB/RIF for Multi-Outcomes of Patients With Presumptive Pulmonary Tuberculosis in Thailand\nJiraporn Khumsri, PhD,1,2 Piya Hanvoravongchai, MD, ScD,2,3 Narin Hiransuthikul, MD, PhD,2,* Charoen Chuchottaworn, MD4\n1 Department of Medical Services, Nopparat Rajathanee Hospital, Ministry of Public Health, Bangkok, Thailand; 2Department of Preventive and Social Medicine, Chulalongkorn University, Bangkok, Thailand; 3Thailand Research Center for Health Services System, Chulalongkorn University, Bangkok, Thailand; 4Department of Medical Services, Central Chest Institute of Thailand, Ministry of Public Health, Nonthaburi, Thailand\nABSTRACT\nObjectives: The cost-effectiveness of screening adult patients for pulmonary tuberculosis is not clear. As such, this study aims to identify the cost-effectiveness between the Xpert MTB/RIF assay and the sputum acid-fast bacilli (AFB) smear. Multioutcomes were correct diagnosis, time to achieve correct diagnosis, and gain in quality-adjusted life-years (QALYs). Methods: A decision tree model was constructed to reveal a possible clinical pathway of tuberculosis diagnosis. The researchers used a clinical study to establish the probability of all clinical pathways for input into this model. The sample size was calculated following the correct diagnosis. Participants were randomly divided into 2 groups. A structural questionnaire and the Thai version of quality of life (EQ-5D-5L) were used for interviewing. Results: The results showed that the time to achieve the correct diagnosis for the group using Xpert MTB/RIF was shorter than that for the group using the sputum AFB smear. Both the correct diagnosis and QALYs of the base case analysis presented the Xpert MTB/RIF method as dominant. A Monte Carlo model, which analyzed the Xpert MTB/RIF method, revealed that the average number of patients who were correctly diagnosed was 673, the QALYs were 945.85 years, and the total cost was $143 110.64. For the sputum AFB smear method, the average number who received a correct diagnosis was 592, the QALYs were 940.40 years, and the total cost was $196 666.84. Probabilistic and one-way sensitivity analysis con\ufb01rmed that the Xpert MTB/RIF remained dominant. Conclusions: These results provide useful information for the National Strategic Plan to screen all adult patients for pulmonary tuberculosis.\nKeywords: cost-effectiveness analysis, multi-outcomes, pulmonary tuberculosis, sputum AFB smear, Thailand, Xpert MTB/RIF.\nVALUE IN HEALTH REGIONAL ISSUES. 2020; 21(C):264\u2013271\n\nIntroduction\nPulmonary tuberculosis (PTB) was one of the top 10 causes of mortality worldwide in 2016.1 The World Health Organization, which has been seeking an optimal tool to help early treatment, recommended the use of Xpert MTB/RIF as the initial diagnostic test for PTB in adult patients by 2035. The World Health Organization\u2019s goal is to reduce new patients with TB to less than 10:1 000 000 of the population.1 In low- or middle-income countries with a high prevalence of TB, diagnosis is usually achieved from a sputum acid-fast bacilli (AFB) smear,2-4 although this smear has a relative low sensitivity for detecting MTB (the detection rate is between 20%-70%).5 It is the most widely used test for diagnosing TB, together with chest x-rays and patient symptoms.\n\nA few studies have explored the cost-effectiveness of Xpert MTB/RIF.6-13 Most of these have investigated the costeffectiveness from healthcare or provider perspectives.6,12,13 Although it is dif\ufb01cult to conduct useful assessments of the cost and cost-effectiveness of tuberculosis services, policymakers increasingly need to know the cost and costeffectiveness of their services to justify the continued allocation of resources.\nThe aim of this study was to evaluate the costs of tuberculosis services and multi-outcomes, which were correct diagnosis, time to achieve the correct diagnosis, and quality-adjusted life-years (QALYs) between the Xpert MTB/ RIF and sputum AFB smear among persons suspected of having TB.\n\nCon\ufb02ict of interest: The authors have indicated that they have no con\ufb02icts of interest with regard to the content of this article.\n* Address correspondence to: Narin Hiransuthikul, MD, Department of Preventive and Social Medicine, Chulalongkorn University, Rama 4 Rd, Bangkok 10330, Thailand. Email: nhiransu@yahoo.com\n2212-1099/$36.00 - see front matter \u00aa 2020 ISPOR\u2013The professional society for health economics and outcomes research. Published by Elsevier Inc. https://doi.org/10.1016/j.vhri.2019.09.010\n\nECONOMIC EVALUATION\n\n265\n\nMethods\nTarget Population\nParticipants were men or women aged 18 years or older with presumptive TB disease. The researchers constructed a decision tree diagram to represent a possible clinical pathway for TB diagnosis and treatment (see Appendix Fig. 1 in Supplemental Materials found at https://doi.org/10.1016/j.vhri.2019.09.010). There were insuf\ufb01cient data to input for establishing a clinical pathway for TB diagnosis; therefore, the researchers used a clinical study to establish the probability of all clinical pathways.\nSetting and Location\nIn the clinical study, data were collected from the outpatient department (OPD) of a tertiary care hospital in Thailand on patients aged $18 years with presumptive TB disease during the period of October 2016 to March 2017.\nThe clinical study calculated sample size using the PS Power and Sample Size Calculator by William D. Dupont and Walton D. Plummer. The level of power in the sample size calculation was 90% at a .05 level of statistical signi\ufb01cance. Prior data indicated that the accuracy of the diagnostic test of sputum AFB smear was 0.65, and that of Xpert MTB/RIF was 0.95.14 The calculated sample size was 35 patients per group. To prevent possible data error during data collection, 30% of the estimated sample number was added; thus, the required sample size was 45 patients per group.\nAll the participants were given diagnostic tests, both the sputum AFB smear and the Xpert MTB/RIF, to calculate prevalence, sensitivity, and speci\ufb01city. The physicians and medical technologists in this study were blinded to the randomization process of assigning diagnostic tests to the participants. All participants voluntarily gave written informed consent before the start of the structured questionnaire interview. The questionnaire also asked for the participants to mention the total time between their \ufb01rst visit to OPD with suspected PTB and when they received a correct diagnosis from the doctor. The questionnaire also included questions related to the Thai version of quality of life (EQ-5D-5L) and the patient\u2019s costs. Participants were randomly divided into 2 groups by simple random sampling. The \ufb01rst group was examined using Xpert MTB/ RIF, and the second group used the sputum AFB smear. All samples were tested further with sputum cultures including a liquid culture (BACTEC MGIT) and a solid culture (2% Ogawa medium). Moreover, drug susceptibility was also tested to determine to which anti-TB drugs the participants were sensitive to.\nStudy Perspective\nThe cost of implementing each tool was derived from a societal perspective. Costs were divided into 2 categories: medical costs and patient costs. Data were gathered by reviewing the hospital\u2019s database report and interviewing healthcare staff, patients, and their relatives to calculate total costs. The medical costs included capital costs, material costs, and labor costs. Capital costs included the costs of the buildings and machines comprising each tool. Material costs covered the costs of equipment used in each test and were assessed by direct observation. Labor costs consisted of the incomes of all healthcare staff who diagnosed the presumptive PTB patients in the laboratory, radiology, and OPD department. Patient\u2019s costs related to the costs of traveling, food, drink, and lost income as a result of absence from work. The treatment costs were excluded because this study aimed to explore only the costs of diagnosis. Moreover, the tool that could detect patients\u2019 positive to MTB or rifampicin resistance is a good outcome, but it is very\n\ncostly and other forms of treatment were cheaper. Accordingly, it was not appropriate to calculate the treatment costs in this model.\nThe costs were reported in US dollars (USD). The costs in Thai Baht (THB) were converted to USD at a rate of 31.5 THB per 1 USD (in February 2019).15 The commercial prices of the Xpert MTB/RIF cartridges ($22.22 per cartridge) and the 4-Cartridge Xpert instrument ($31 811.27) were indicated by sales agencies in Thailand, whereas the process of the solid culture ($8.25 per test) and DST ($6.35 per test) were equal to the charge rates indicated by the Department of Central Chest Institute of Thailand.\nTime Horizon\nThe researchers compared the progress of the QALYs over a time span of 1 year. The timing for utility calculation was classi\ufb01ed in 3 periods: before diagnosis, 1 month later, and 3 months later. The counting time for each period, which provided a de\ufb01nition for management, included TBD (the average time it took to get the correct diagnosis from each tool); T1 MO (the time of 1 month, which was computed from 90 days \u2013 TBD), and T3 MO (the period of three months, which was computed from 365 days \u2013 [TBD 1 T1 MO]).\nDiscount Rate\nCosts and health outcomes were not discounted because of the short measure of timing.\nOutcomes\nThe health outcomes of both tools had 3 outcomes: correct diagnosis, time to get the correct diagnosis, and gain in QALYs.\nMeasurement of Effectiveness\nFour categories of patients were correctly diagnosed: (1) patients who were true positive (TP) for MTB, who were treated with anti-TB drugs; (2) patients with TP for MTB with RR (TPrif), who were treated with multidrug-resistant (MDR) drugs; (3) patients with false negative (FN) for MTB, who were treated with anti-TB drugs; and (4) patients with true negative (TN), who were not treated. The time to achieve the correct diagnosis was recorded from the \ufb01rst outpatient department visit until the correct diagnosis was made.\nThe QALYs gain was calculated after 1 year by computation among all patients, who fell under 6 categories: (1) patients with TP for MTB, who were treated with anti-TB drugs; (2) patients with TP for MTB, who were not treated with anti-TB drugs; (3) patients with TP for RR (TPrif), who were not treated with MDR drugs; (4) patients with FN for MTB, who were treated with antiTB drugs; (5) patients with TN for MTB, who were not treated with anti-TB drugs; and (6) patients with TP for RR (TPrif), who were treated with MDR drugs. In the measurement of QALYs, death is valued as 0 and perfect health as 1.\nSynthesis-Based Estimates\nThe researcher calculated the probability or the branch or pathway of patients with presumptive TB since their \ufb01rst visit to the OPD until they received the correct diagnosis and then, put the probability of each branch into the decision tree model for 2 simulated outcomes: correct diagnosis and gain in QALYs. One thousand cohorts were calculated simultaneously to estimate the correct diagnosis, QALYs gain, and the costs of both tools. The time to get the correct diagnosis used the result of this outcome derived from the implementation of clinical operations of the participant in clinical study.\n\n266\n\nVALUE IN HEALTH REGIONAL ISSUES\n\nMAY 2020\n\nClinical Inputs\nAfter \ufb01nishing the clinical study, the researchers aimed to derive the probability of all clinical pathways for input into decision tree model. For example, the probability of patients with TP for MTB were treated with anti-TB drugs or not treated; the probability of patients with TP were tested further with sputum culture or not tested.\nUtility Inputs\nFrom the clinical study, all participants were interviewed with the Thai version of quality of life (EQ-5D-5L). Participants were interviewed 3 times: during the \ufb01rst OPD visit before diagnosis, at 1 month, and 3 months after that.\nThe process of calculating utility provided a de\ufb01nition for management over 3 periods including: (1) utility before diagnosis (UBD), which was the average utility before diagnosis; (2) utility at 1 month (U1 MO), which was UBD plus the mean difference of utility before diagnosis \u2013 utility at 1 month UBD and U1 MO; and (3) utility at 3 months (U3 MO) was UBD plus the mean difference of utility before diagnosis 2 utility at 3 months UBD and U3 MO.\nCost Inputs\nThe researchers used the primary data from the clinical study. The total costs of the TB diagnostic tools were computed according to costs per unit. The unit costs of the Xpert MTB/RIF, the sputum AFB smear, solid culture, and DST were input into the decision tree model to calculate the total costs.\nCost-Effectiveness Analysis\nThe incremental cost-effectiveness ratio (ICER) was calculated by dividing the differences in total costs between Xpert MTB/RIF and the sputum AFB smear by the differences of the numbers of correct diagnosis and QALY gained. The formula below demonstrates how the ICER was calculated.\nTotal cost Xpert MTB/RIF \u2013 Total cost Sputum AFB smear / Total outcome Xpert MTB/RIF \u2013 Total outcome Sputum AFB smear.\nThis study used a Thai ceiling threshold of 160 000 THB/QALY ($5079.36/QALY).16 The concept of threshold indicates that if the cost-effectiveness (CE) ratio of the new method was not greater than the threshold, it was cost-effective and appropriate for adoption. However, if the CE ratio of the new method was greater than the threshold, the policymakers would limit its adoption.17\nSensitivity Analysis\nSensitivity analysis involves assessment of whether changes in some of the key costs or estimates of effectiveness would affect the conclusions to be drawn from the baseline analysis. To determine the robustness of these \ufb01ndings, uncertainty analysis using the Monte-Carlo simulation method was performed to estimate the relationships between total costs and total correct diagnosis, and total costs and QALYs gain. One-way sensitivity analysis was performed to determine the effect of costs, diagnostic tests, and the prevalence of PTB patients. The results are displayed as a Tornado diagram.\nEthical Considerations\nApproval for this study was obtained from the Ethics Committee of the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand (certi\ufb01cate of approval No. 606/2016).\n\nResults\nClinical Part of This Study\nNinety patients were enrolled, and the data of 87 patients were analyzed. Forty-four patients were assigned to the Xpert MTB/RIF group, and 43 patients into the sputum AFB smear group. The clinical characteristics of both groups were tested by Fisher exact tests. There were no statistically signi\ufb01cant differences between the Xpert MTB/RIF group and the sputum AFB smear group in relation to sex, age, history of illness, and history of sputum AFB smear examination (see Appendix Table 1 in Supplemental Materials found at https://doi.org/10.1016/j. vhri.2019.09.010).\nThe utility of patients with TP for MTB, patients with TN for MTB, and patients with TP for RR (TPrif) was 0.648, 0.715, and 0.915, respectively. For the results of the utility in one year for the calculation of QALYs see Appendix Tables 2 and 3 in Supplemental Materials found at https://doi.org/10.1016/j.vhri.2019. 09.010).\nBase-Case Analysis\nThe value of the probability that was used in the clinical inputs was extracted from the clinical study (Table 1). Unit costs of the sputum AFB smear were $195.73, and the Xpert MTB/RIF were $139.14. In the cohort of 1000 presumptive PTB patients, the Xpert MTB/RIF method had a greater number of correct treatments (907 patients) than that of the sputum AFB smear (724 patients). Next, the Xpert MTB/RIF method achieved signi\ufb01cantly shorter times (1.88 days) to get the correct diagnosis than the sputum AFB smear method (4.11 days; P , .001). In comparison with the sputum AFB smear, the Xpert MTB/RIF took 2.23 days shorter to correctly detect TB (95% con\ufb01dence interval [CI]; 23.047 to 21.425; see Appendix Table 1 in Supplemental Materials found at https://doi.org/10.1016/j.vhri.2019.09.010). Finally, QALY gained per 1000 presumptive PTB patients in 1 year was equal to 947.14 scores for the Xpert MTB/RIF and 939.84 scores for the sputum AFB smear (Table 2).\nThe results of base-case analysis are presented in Table 2. The Xpert MTB/RIF was more effective and less costly than the sputum AFB smear. The ICER results indicated that the Xpert MTB/RIF was more cost-effective than the sputum AFB smear (Table 3), and it was dominated by the Xpert MTB/RIF method.\nSensitivity Analysis\nOne-way sensitivity analysis was performed to determine the effect of costs, diagnostic tests, and prevalence of PTB patients. The parameter inputs of the one-way sensitivity analysis can be seen in Appendix Tables 4 and 5 in Supplemental Materials found at https://doi.org/10.1016/j.vhri.2019.09.010. The results are displayed as a Tornado diagram. The results of the base case were robust to one-way sensitivity analysis (Figs. 1 and 2). One thousand cohort simulations of the Xpert MTB/RIF method revealed the average of correct diagnosis was 673 (95% con\ufb01dence interval [CI], 655.21-691.22), the QALYs gained were 945.85 (95% CI, 945.7945.98), and the average costs were $143 110.64 (95% CI, $143 009-$143 212). For the sputum AFB smear method, the average of correct diagnoses was 592 (95% CI, 577.34-605.84), the QALYs gained were 940.40 (95% CI, 940.27-940.53), and the average costs were $196 666.84 (95% CI, $196 552-$196.781). This study found the Xpert MTB/RIF method increased the correct diagnosis by more than 81 patients, the health gained increased by 5.45 years,\n\nECONOMIC EVALUATION\n\n267\n\nTable 1. Model inputs (in USD).\n\nParameter\n\nDistribution\n\nMean\n\nSE\n\na\n\nb\n\nReference\n\nPrevalence\n\nBeta\n\n0.333\n\n0.050\n\n29\n\n58\n\nClinical study\n\nSensitivity of Xpert MTB/RIF\n\nBeta\n\n0.828\n\n0.069\n\n24\n\n5\n\nClinical study\n\nSensitivity of sputum AFB smear\n\nBeta\n\n0.483\n\n0.091\n\n14\n\n15\n\nClinical study\n\nProportion of test Proportion for SM2 Proportion for SM1 Proportion for XP2 Proportion for XP1 Proportion culture for SM1 Proportion culture for XP1\n\nDistribution Beta Beta Beta Beta Beta Beta\n\nMean 0.194 1.000 0.065 1.000 1.000 1.000\n\nSE 0.015 0.077 0.005 0.077 0.077 0.077\n\na 0.165 0.850 0.055 0.850 0.850 0.850\n\nb\n0.220 1.150 0.070 1.150 1.150 1.150\n\nReference Clinical study Clinical study Clinical study Clinical study Clinical study Clinical study\n\nParameter\n\nDistribution\n\nMean\n\nSE\n\n215%\n\n15%\n\nReference\n\nLabor costs of Xpert MTB/RIF\n\nGamma\n\n41.25\n\n0.58\n\n162.29\n\n0.0081\n\nClinical study\n\nLabor costs of sputum AFB smear\n\nGamma\n\n79.27\n\n1.60\n\n78.02\n\n0.0323\n\nClinical study\n\nBuilding costs of Xpert MTB/RIF\n\nGamma\n\n6.99\n\n0.04\n\n965.38\n\n0.0002\n\nClinical study\n\nBuilding costs of sputum AFB smear\n\nGamma\n\n11.09\n\n0.04\n\n2,432.02\n\n0.0001\n\nClinical study\n\nCapital costs of Xpert MTB/RIF\n\nGamma\n\n1.14\n\n0.03\n\n41.55\n\n0.0009\n\nClinical study\n\nCapital costs of sputum AFB smear\n\nGamma\n\n0.77\n\n0.00\n\n1,099.77\n\n0.0000\n\nClinical study\n\nMaterial costs of Xpert MTB/RIF\n\nGamma\n\n24.74\n\n1.22\n\n13.13\n\n0.0598\n\nClinical study\n\nMaterial costs of sputum AFB smear\n\nGamma\n\n12.96\n\n0.37\n\n38.68\n\n0.0106\n\nClinical study\n\nElectricity costs of Xpert MTB/RIF\n\nGamma\n\n8.32\n\n0.37\n\n16.40\n\n0.0161\n\nClinical study\n\nElectricity costs of sputum AFB smear\n\nGamma\n\n1.51\n\n0.00\n\n3,231.90\n\n0.0000\n\nClinical study\n\nWater supply costs of Xpert MTB/RIF\n\nGamma\n\n0.18\n\n0.01\n\n29.58\n\n0.0002\n\nClinical study\n\nWater supply costs of sputum AFB smear\n\nGamma\n\n0.49\n\n0.00\n\n508.71\n\n0.0000\n\nClinical study\n\nPatient's costs of Xpert MTB/RIF\n\nGamma\n\n56.47\n\n0.19\n\n2,843.67\n\n0.0006\n\nClinical study\n\nPatient's costs of sputum AFB smear\n\nGamma\n\n89.62\n\n0.30\n\n2,843.67\n\n0.0010\n\nClinical study\n\nNote. Patent's costs related to the costs of traveling, food, drink, and lost income as a result of absence from work. AFB indicates acid-fast bacilli; SM, sputum AFB smear; SM2, negative to MTB; SM1, positive to MTB; XP, XPERT MTB/RIF; XP2, no found MTB or MTB not detected; XP1, found MTB or MTB detected.\n\nTable 2. The results of base-case analysis (in 2018 USD).\n\nMethod Costs USD per 1000\n\nTime to get correct\n\npresumptive PTB patients diagnosis Mean (SD)\n\nXpert MTB/ RIF\n\n143 119.53\n\n1.88 (1.07)\n\nSputum AFB smear\n\n198 030.97\n\n4.11 (2.22)\n\nCorrect diagnosis per 1000 presumptive PTB patients\n907\n724\n\nAFB indicates acid-fast bacilli; PTB, pulmonary tuberculosis; QALY, quality-adjusted life-years; SD, standard deviation.\n\nQALYs gained per 1000 presumptive PTB patients\n947.14\n939.84\n\nTable 3. Base-case analysis of the average costs per correct diagnosis and costs per QALYs (in 2018 USD).\n\nMethod\n\nCosts (USD)\n\nCorrect diagnosis (n)\n\nCE ratio\n\nXpert MTB/RIF\n\n143 119.53\n\n907\n\n158.31\n\nSputum AFB smear\n\n198 030.97\n\n724\n\n273.52\n\nICER\n\n254 911.45\n\n81\n\nDominant\n\nCE indicates cost-effectiveness; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life-years; USD, US dollar.\n\nQALYs 947.14 939.84\n7.29\n\nCE ratio 151.11 208.95\nDominant\n\n268\n\nVALUE IN HEALTH REGIONAL ISSUES\n\nMAY 2020\n\nFigure 1. Tornado diagram reveals the percentage change of factors that were sensitive to ICER of the correct diagnosis.\n\nICER indicates incremental cost-effectiveness ratio; PT, patients.\nFigure 2. Tornado diagram reveals the percentage change of factors that were sensitive to ICER of the quality-adjusted life-years.\n\nICER indicates incremental cost-effectiveness ratio.\n\nand the costs were less than the sputum AFB smear ($53 556.20). The probabilistic sensitivity analysis was presented by CE plane, in which the Xpert MTB/RIF method remained dominant (Figs. 3 and 4).\n\nDiscussion\nThe values of probability used in clinical inputs in this study were credible because the general characteristics of the Xpert MTB/\n\nFigure 3. Scatter plots for the estimated ICER of the correct diagnosis.\n\nECONOMIC EVALUATION\n\n269\n\nICER indicates incremental cost-effectiveness ratio; PT, patients; USD, US dollar.\n\nRIF group and the sputum AFB smear group were not different (Appendix Table 6 in Supplemental Materials found at https://doi. org/10.1016/j.vhri.2019.09.010). Consequently, the value of the prevalence, sensitivity, and speci\ufb01city of both tools resemble that of previous studies.11,18-20 The speci\ufb01city of both tools was 100% (95% CI, 94.0%-100%), which appeared to be similar to studies in other countries, where the speci\ufb01city ranged from 95.1% to 99%. In our study, a liquid culture and a solid culture were used, which may have\n\nled to the difference in the level of speci\ufb01city in this study. For this reason, the outcome from the decision tree model might be credible as well.\nOne-third of the outcomes were the time to get a correct diagnosis. The Xpert MTB/RIF lasted 1.88 days, whereas the sputum AFB smear lasted 4.11 days. This can be explained by the fact that the sputum AFB smear took approximately 15 to 20 minutes per test, whereas the Xpert MTB/RIF took about 2 hours. The\n\nFigure 4. Scatter plots for the estimated ICER of the quality-adjusted life-years (QALY).\n\nICER indicates incremental cost-effectiveness ratio; USD, US dollar.\n\n270\n\nVALUE IN HEALTH REGIONAL ISSUES\n\nMAY 2020\n\nsputum AFB smear needs 3 samples and takes 2 to 3 days to interpret the results, but the Xpert MTB/RIF required only 1 sample. Consequently, the Xpert MTB/RIF method was signi\ufb01cantly shorter in achieving a correct diagnosis. Unfortunately, it was not possible to compare this \ufb01nding of the present article on the time taken to get a correct diagnosis with other recent studies due to the fact that there is irrelevant research on this issue. The time taken to achieve a correct diagnosis in this study was consistent with the \ufb01nding of a previous systematic literature review on the delays in diagnosis and treatment of PTB in India. Treatment delay was de\ufb01ned as \u201cthe time interval between diagnosis and initiation of anti-TB treatment.\u201d The health system delay, HD, was classi\ufb01ed into diagnosis delay, in which the median (days) was 27.2 (2-87 days).21\nThe number of correct diagnoses and QALYs gained, as multioutcomes, revealed slight differences between the Xpert MTB/ RIF and the sputum AFB smear method.\nThe total diagnostic cost of the Xpert MTB/RIF was $139.14 per patient, which is consistent with a previous research carried out in a high-burden TB country. This previous research indicated that the diagnostic cost of the Xpert MTB/RIF was $137-151 per patient.12 Furthermore, the total diagnostic cost of the sputum AFB smear was $195.73 per patient, which differed from the previous \ufb01ndings ($28-$49 per patient).7,12 In our study, the sputum AFB smear method had a higher diagnostic cost than it did in other studies.\nHowever, the costs covered all the activities and equipment related to TB diagnosis. The higher cost of the AFB smear might be due to the longer duration of diagnosis. We also found that the Xpert MTB/RIF was more cost-effective than the sputum AFB smear. The \ufb01ndings in this study are similar to \ufb01ndings of previous studies.6,7,12,13 Moreover, one-way sensitivity analysis and probabilistic sensitivity analysis, ICER of Xpert MTB/RIF, were dominant. The Xpert MTB/RIF method increases the chances of a correct diagnosis and the QALY gained over 1 year for presumptive PTB patients. Overall, we can conclude that using the Xpert MTB/RIF as the initial tool for TB diagnosis is the most costeffective.\nThis study had several limitations. First, the structure of the model involved a real-life situation of PTB diagnosis. Various algorithms were not created, so it was not possible to compare the results with the base case analysis results. Second, the clinical study was conducted in a single hospital, which contributed to a high degree of internal validity but a lack of external validity.\nThis study had several important strengths. A liquid culture and a solid culture were used to con\ufb01rm MTB in the clinical study. This provided greater precision in the probability of clinical inputs. The \ufb01ndings are able to verify information related to adults with presumptive TB. Therefore, the results of this study could be extrapolated for new presumptive TB cases. Furthermore, we are the \ufb01rst research group to examine and compare multi-outcomes with cost-effectiveness analysis.\nEvaluation of the economic factors involved is essential for policymakers in making decisions about the planning and administration of the national TB service system.\nConclusions\nUsing Xpert MTB/RIF in the initial TB diagnosis in adult patients presumed to have TB can shorten the time to achieve a correct diagnosis. Moreover, Xpert MTB/RIF is considerably costeffective for multi-outcomes. These study results are essential\n\nempirical evidence for policy makers who wish to evaluate the National Strategic Plan for TB in Thailand in setting up and choosing TB diagnostic tools that are signi\ufb01cantly cost-effective and worthy.\nAcknowledgments\nWe are grateful to the staff of the Nopparat Rajathanee Hospital, and the Bureau of Tuberculosis, Department of Medical Services, Ministry of Public Health, Thailand, for their collaboration.\nSupported by the National Research Council of Thailand and the 90th Anniversary of Chulalongkorn University Ratchadaphiseksomphot Fund.\nSupplemental Materials\nSupplementary data associated with this article can be found in the online version at https://doi.org/10.1016/j.vhri.2019.09.010.\nREFERENCES\n1. World Health Organization. Global tuberculosis report 2016. Geneva, Switzerland. 2016. http://apps.who.int/iris/bitstream/10665/250441/1/978 9241565394-eng.pdf?ua=1. Accessed April 22, 2020.\n2. Mao TE, Okada K, Yamada N, et al. Cross-sectional studies of tuberculosis prevalence in Cambodia between 2002 and 2011. Bull World Health Organ. 2014;92:573\u2013581.\n3. Mekonnen A. Smear-positive pulmonary tuberculosis and AFB examination practices according to the standard checklist of WHO\u2019s tuberculosis laboratory assessment tool in three governmental hospitals, Eastern Ethiopia. BMC Res Notes. 2014;7(1):1\u20138.\n4. Priyakanta N, Ajay KMV, Mareli C, et al. Comparing same day sputum microscopy with conventional sputum microscopy for the diagnosis of tuberculosis\u2013Chhattisgarh, India. PLoS ONE. 2013;8(9):e74964.\n5. World Health Organization. Toman\u2019s Tuberculosis Case Detection, Treatment, and Monitoring. 2 ed. Geneva, Switzerland: WHO; 2004.\n6. Choi HW, Miele K, Dowdy D, Shah M. Cost-effectiveness of Xpert\u00ae MTB/RIF for diagnosing pulmonary tuberculosis in the United States. Int J Tuberc Lung Dis. 2013;17(10):1328\u20131335.\n7. da Silva Antunes R, Pinto M, Trajman A. Patient costs for the diagnosis of tuberculosis in Brazil: comparison of Xpert MTB/RIF and smear microscopy. Int J Tuberc Lung Dis. 2014;18(5):547\u2013551.\n8. Dowdy DW, van't Hoog A, Shah M, Cobelens F. Cost-effectiveness of rapid susceptibility testing against second-line drugs for tuberculosis. Int J Tuberc Lung Dis. 2014;18(6):647\u2013654.\n9. Langley I, Lin HH, Squire SB. Cost-effectiveness of Xpert MTB/RIF and investing in health care in Africa. Lancet Glob Health. 2015;3(2):e83\u2013e84.\n10. Shah M, Dowdy D, Joloba M, et al. Cost-effectiveness of novel algorithms for rapid diagnosis of tuberculosis in HIV-infected individuals in Uganda. AIDS (London, England). 2013;27(18):2883\u20132892.\n11. Theron G, Zijenah L, Chanda D, et al. Feasibility, accuracy, and clinical effect of point-of-care Xpert MTB/RIF testing for tuberculosis in primary-care settings in Africa: a multicentre, randomised, controlled trial. The Lancet. 2014;383(9915):424\u2013435.\n12. Vassall A, van Kampen S, Sohn H, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: a cost-effectiveness analysis. PLoS Med. 2011;8(11):e1001120.\n13. You JH, Lui G, Kam KM, Lee NL. Cost-effectiveness analysis of the Xpert MTB/ RIF assay for rapid diagnosis of suspected tuberculosis in an intermediate burden area. J Infect. 2015;70(4):409\u2013414.\n14. Boehme CC, Nabeta P, Hillemann D, et al. Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med. 2010;363(11):1005\u20131015.\n15. Bank of Thailand. Foreign exchange rates. 2018. https://www.bot.or.th/Thai/ Statistics/FinancialMarkets/ExchangeRate/ExchangeRate_EN_PDF/ER_PDF_28 092018.PDF. Accessed April 22, 2020.\n16. Teerawattananon Y, Tritasavit N, Suchonwanich N, Kingkaew P. The use of economic evaluation for guiding the pharmaceutical reimbursement list in Thailand. Z Evid Fortbild Qual Gesundhwes. 2014;108(7):397\u2013404.\n17. Marseille E, Larson B, Kazi SD, Kahnd GJ, Rosenb S. Thresholds for the cost\u2013 effectiveness of interventions: alternative approaches. Bull World Health Organ. 2015;93:118\u2013124.\n18. Steingart KR, Sohn H, Schiller I, et al. Xpert(R) MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev. 2013;1:Cd009593.\n19. Pinyopornpanish K, Chaiwarith R, Pantip C, et al. Comparison of Xpert MTB/ RIF assay and the conventional sputum microscopy in detecting mycobacterium tuberculosis in Northern Thailand. Tuberc Res Treat. 2015;2015:1\u20136.\n\nECONOMIC EVALUATION\n\n271\n\n20. Bajrami R, Mulliqi G, Kurti A, Lila G, Raka L. Comparison of GeneXpert MTB/ RIF and conventional methods for the diagnosis of tuberculosis in Kosovo. J Infect Dev Ctries. 2016;10(4):418\u2013422.\n\n21. Sreeramareddy CT, Panduru KV, Menten J, Van den Ende J. Time delays in diagnosis of pulmonary tuberculosis: a systematic review of literature. BMC Infect Dis. 2009;9:91.\n\n\n",
    "authors": [
      "Jiraporn Khumsri",
      "Piya Hanvoravongchai",
      "Narin Hiransuthikul",
      "Charoen Chuchottaworn"
    ],
    "doi": "10.1016/j.vhri.2019.09.010",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://linkinghub.elsevier.com/retrieve/pii/S221210992030008X"
  },
  {
    "key": "A59ATLY5",
    "title": "Xpert\u00aeMTB/RIF for the Diagnosis of Tuberculosis in a Remote Arctic Setting: Impact on Cost and Time to Treatment Initiation",
    "abstract": "",
    "full_text": "RESEARCH ARTICLE\nXpert1MTB/RIF for the Diagnosis of Tuberculosis in a Remote Arctic Setting: Impact on Cost and Time to Treatment Initiation\nOlivia Oxlade1,2, Jordan Sugarman1, Gonzalo G. Alvarez3,4, Madhukar Pai1,2, Kevin Schwartzman1,2*\n1 Respiratory Epidemiology and Clinical Research Unit, Department of Epidemiology, McGill University, Montreal, QC, Canada, 2 McGill International Tuberculosis Centre, McGill University, Montreal, QC, Canada, 3 Clinical Epidemiology, The Ottawa Hospital Research Institute, Ottawa, ON, Canada, 4 Division of Respirology, Department of Medicine, University of Ottawa, Ottawa, ON, Canada\n* kevin.schwartzman@mcgill.ca\n\nOPEN ACCESS\n\nAbstract\n\nCitation: Oxlade O, Sugarman J, Alvarez GG, Pai M, Schwartzman K (2016) Xpert1MTB/RIF for the Diagnosis of Tuberculosis in a Remote Arctic Setting: Impact on Cost and Time to Treatment Initiation. PLoS ONE 11(3): e0150119. doi:10.1371/journal. pone.0150119\nEditor: Ulla Kou Griffiths, London School of Hygiene and Tropical Medicine, UNITED KINGDOM\nReceived: September 5, 2015\nAccepted: February 9, 2016\n\nBackground\nTuberculosis (TB) remains a significant health problem in the Canadian Arctic. Substantial health system delays in TB diagnosis can occur, in part due to the lack of capacity for onsite microbiologic testing. A study recently evaluated the yield and impact of a rapid automated PCR test (Xpert1MTB/RIF) for the diagnosis of TB in Iqaluit (Nunavut). We conducted an economic analysis to evaluate the expected cost relative to the expected reduction in time to treatment initiation, with the addition of Xpert1MTB/RIF to the current diagnostic and treatment algorithms used in this setting.\n\nPublished: March 18, 2016\nCopyright: \u00a9 2016 Oxlade et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.\nData Availability Statement: All relevant data are within the paper and its Supporting Information files.\n\nMethods\nA decision analysis model compared current microbiologic testing to a scenario where Xpert1MTB/RIF was added to the current diagnostic algorithm for active TB, and incorporated costs and clinical endpoints from the Iqaluit study. Several sensitivity analyses that considered alternative use were also considered. We estimated days to TB diagnosis and treatment initiation, health system costs, and the incremental cost per treatment day gained for each individual evaluated for possible TB.\n\nFunding: Support was provided by the Canadian Institutes of Health Research MOP 89918 to MP GA & AOH 126659 to KS OO JS [www.cihr.irsc.gc.ca]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\nCompeting Interests: The authors have declared that no competing interests exist.\n\nResults\nWith the addition of Xpert1MTB/RIF, costs increased while days to TB treatment initiation were reduced. The incremental cost per treatment day gained (per individual investigated for TB) was $164 (95% uncertainty range $85, $452). In a sensitivity analysis that considered hospital discharge after a single negative Xpert1MTB/RIF, the Xpert1MTB/RIF scenario was cost saving.\n\nPLOS ONE | DOI:10.1371/journal.pone.0150119 March 18, 2016\n\n1 / 13\n\nEconomic Analysis of Xpert1MTB/RIF for Diagnosis of TB\nInterpretation\nAdding Xpert1MTB/RIF to the current diagnostic algorithm for TB in Nunavut appears to reduce time to diagnosis and treatment at reasonable cost. It may be especially well suited to overcome some of the other logistical barriers that are unique to this and other remote communities.\n\nIntroduction\nTuberculosis (TB) remains a significant health problem in the Canadian Arctic; in 2012, incidence rates for all forms of TB in Nunavut were 234/100,000, nearly 60 times higher than rates in southern Canada [1]. The reasons for this high TB burden are multi-faceted, reflecting many overlapping risk factors. One factor is health system delays in diagnosis that can occur due to limited capacity for on-site microbiologic testing at the hospital in Iqaluit, the only hospital in the Qikiqtaaluk (Baffin) region of Nunavut. All microbiologic specimens for investigation of possible TB from Qikiqtaaluk (Baffin) are prepared at the hospital and flown South to Ottawa, Ontario, for smears, cultures, and drug susceptibility testing. This contributes to delays in diagnosis and treatment initiation, as well as high costs.\nIn 2012, the World Health Organization issued an updated strong recommendation for using Xpert1MTB/RIF (Cepheid Inc, Sunnyvale, CA), a fully automated TB diagnostic molecular test, as the initial test for diagnosis of pulmonary TB in certain populations [2]. The Xpert1MTB/RIF (approved by Health Canada in 2012 and the United States Food and Drug Administration in 2013), has high specificity and sensitivity [3], a rapid turnaround time (which ideally translates to faster time to treatment initiation), and minimal bio-safety requirements and training needs. Several studies have considered programmatic and patient benefits from management decisions that reflected Xpert1MTB/RIF test results. One suggested that use of Xpert1MTB/RIF reduced potentially unnecessary empiric therapy [4]; another reported that one negative Xpert1MTB/RIF result can reduce unnecessary and expensive hospital isolation for inpatients initially labelled as possibly having active TB [5].\nBased on the potential merits of Xpert1MTB/RIF, a study to evaluate its yield and impact in Nunavut was recently completed [6]. The study showed test sensitivity and specificity were high at 85% and 99% respectively, and time to treatment initiation was significantly shortened, particularly for smear-negative, culture-positive cases, which comprise 2/3 of cases in Nunavut [6]. Given these findings, there are suggestions to incorporate the Xpert1MTB/RIF test into routine practice in Nunavut, consistent with the Canadian TB Standards [6,7]. Xpert1MTB/RIF implementation will require substantial resources for equipment, supplies, and laboratory technician time. In view of the unique epidemiologic, sociodemographic and geographic features of Nunavut, we conducted an economic analysis to evaluate expected costs and changes in times to diagnosis and treatment initiation attributable to adding Xpert1MTB/RIF to diagnostic and TB management algorithms in Nunavut. We also considered costs and the impact on time to treatment initiation of potential changes to TB program policies arising from Xpert1MTB/RIF implementation.\nMethods\nGeneral Description of Model\nA deterministic decision analysis model was developed using Tree Age software (Tree Age Pro 2012, Williamstown, MA, USA). We considered a hypothetical cohort of individuals evaluated\n\nPLOS ONE | DOI:10.1371/journal.pone.0150119 March 18, 2016\n\n2 / 13\n\nEconomic Analysis of Xpert1MTB/RIF for Diagnosis of TB\nfor possible TB within the Qikiqtaaluk (Baffin) region. We simulated expected events and costs, with and without the integration of Xpert1MTB/RIF into the routine testing algorithm for diagnosis and treatment of active TB. Simulation began with initial contact at a local clinic, and ended at completion of treatment for active TB, or at the end of investigation for those found not to have active TB. Model outputs for each scenario were: 1) days from first presentation to health services to the start of TB treatment and 2) health system costs for all investigations and treatment for suspected or confirmed active TB. Use of Xpert1MTB/RIF to detect rifampin-resistant TB (marker of multi-drug resistance) was not considered, since all TB isolates tested in Nunavut were susceptible to all first line drugs in 2013, and no MDR TB cases were identified in Nunavut from 2003 to 2013 [8].\nThe mean interval (in days) from initial evaluation to treatment initiation were calculated for: 1) the full cohort of individuals evaluated for TB (calculation includes contribution from both confirmed TB cases, and those who do not have TB and therefore contribute 0 days to the time between initial evaluation and treatment); and 2) for microbiologically confirmed cases only. The incremental cost per treatment day gained (per individual evaluated for TB), was used to summarize the cost relative to the clinical impact of the strategy with Xpert1MTB/RIF added, versus the status quo where only smear and culture are used. The analysis was conducted from the perspective of the health system. No discounting was used, due to the short timeframe.\nStudy Population\nA hypothetical cohort of persons evaluated for active TB was considered; the prevalence of TB in this group was determined from the study of Xpert1MTB/RIF in Nunavut [6]. Two types of settings were considered in order to capture differences in cost and clinical management of persons evaluated for TB in Qikiqtaaluk (Baffin): 1) Iqaluit (i.e., territorial capital with a hospital) and 2) remote communities (nursing station- no hospital).\nDiagnostic Strategies\nTwo diagnostic strategies were considered. The testing algorithms for each strategy are illustrated in Fig 1. With Strategy 1, routine tests for TB diagnosis included smear and culture, both performed in Ottawa. With Strategy 2, Xpert1MTB/RIF performed in Iqaluit was added to the testing algorithm. With both strategies, persons in Iqaluit with a high clinical suspicion for active TB and a suggestive chest X-ray, were hospitalized for empiric treatment initiation and microbiologic testing. Similarly, with both strategies, persons living in remote communities with a high clinical suspicion for active TB, and a suggestive chest X-ray, began empiric treatment in their communities. Hence for such individuals, Xpert1MTB/RIF results could not advance treatment initiation. However, positive Xpert1MTB/RIF results led to earlier treatment for others whose clinical and radiographic findings did not warrant immediate empiric treatment. Details about these diagnostic strategies and the yield of the various tests are provided in an on-line appendix (See S1 File). Relevant clinical and epidemiolgoic parameters, as well as intervals (in days) between milestones in the diagnostic algorithms are listed in Tables 1 and 2.\nCosting\nTB-related health system costs included materials, shipping, patient transport, fourteen days hospitalization at Qikiqtani General Hospital (based on information provided by the Nunavut Health Department) and professional wages. Nunavut physician fees were obtained from the Government of Nunavut Physician Services Department; Ontario physician fees were obtained\n\nPLOS ONE | DOI:10.1371/journal.pone.0150119 March 18, 2016\n\n3 / 13\n\nEconomic Analysis of Xpert1MTB/RIF for Diagnosis of TB\n\nFig 1. Testing algorithms with the two strategies: Status Quo and Xpert1MTB/RIF.\ndoi:10.1371/journal.pone.0150119.g001\nfrom the 2013 Ontario Health Insurance Plan Fee Schedule. Component costs and total costs for the status quo strategy (Strategy 1) are listed in Table 3. Additional costs for the Xpert1MTB/RIF strategy are listed in Tables 4 and 5. Machine costs (calibration, cartridges, shipping) and salary for the laboratory technologist working with specimens were included. Costs associated with unreadable or faulty Xpert1MTB/RIF modules were not included, as none were reported in the Iqaluit pilot study. In the base case, the cost per specimen tested using Xpert1MTB/RIF was $133.03, assuming 1651 specimens analyzed per year (the number reported for Nunavut in 2013 by the Nunavut Health Department). All costs were reported in 2014 Canadian dollars. Costs from 2013 were inflated using inflation rates from the Bank of Canada [9].\nSensitivity Analysis\nTwo additional scenarios for use of Xpert1MTB/RIF were considered as part of our sensitivity analysis:\n1) Three Sputum Samples per Individual. Extracts from 3 sputum samples were analyzed using Xpert1MTB/RIF (vs 1 in the base case). Cost per sample analyzed was reduced, but test\n\nPLOS ONE | DOI:10.1371/journal.pone.0150119 March 18, 2016\n\n4 / 13\n\nEconomic Analysis of Xpert1MTB/RIF for Diagnosis of TB\n\nTable 1. Clinical and Epidemiologic Probabilities Associated with TB Investigation and Care in Nunavut, Canada.\n\nDescription\n\nValue\n\nRange Source\n\nProportion of Nunavut population that lives in Iqaluit Prevalence of active TB among those who present to clinic for TB evaluation\nSmear result among persons with active TB smear positive\nsmear negative Sensitivity of relying on clinical symptoms for initial TB evaluation\nSpeci\ufb01city of relying on clinical symptoms for initial TB evaluation\nSensitivity of chest x-ray\nSpeci\ufb01city of chest x-ray\nProbability of producing suitable sputum samples given CXR abnormality\nProbability of producing suitable sputum samples given no CXR abnormality Probability of an individual being evaluated for TB in a remote community being sent to Iqaluit for further TB related clinical evaluation Anti-tuberculosis treatment outcomes\nCure\nFailure Death Sensitivity of Xpert1MTB/RIF for smear positive TB\nSensitivity of Xpert1MTB/RIF for smear negative TB\nSpeci\ufb01city of Xpert1MTB/RIF\n\n21.0% 27/ 344 = 7.8%\n32.1% 67.9% 61.0% 83.0% 90.5% 60.0% 81.2% 50% 4.0%\n94.6% 1.2% 4.2% 95.0% 57.0% 99.0%\n\n\u2014\n\n[17]\n\n5.0\u201310.6 [6]\n\n23.5\u2013 40.8 0.51\u2013 0.71 0.82\u2013 0.84 90.0\u2013 100 60.0\u2013 70.0 74.7\u2013 89.6 0\u2013100 0\u20135.0\n94.2\u2013 96.8\n\n[7]\n[18] [18] [1, 19] [1, 19] [20] [21] Personal communication Van Dyk [22]\n\n2.3\u20136.2\n\n85.0\u2013\n\n[6]\n\n1.00\n\n20.4\u2013\n\n[6]\n\n93.6\n\n98.0\u2013\n\n[6]\n\n1.00\n\ndoi:10.1371/journal.pone.0150119.t001\n\ncost per individual was increased (Table 5). Based on 3 samples, test sensitivity increased to 99.8% for smear-positive cases and 90.2% for smear-negatives [10].\n2) Hospital Discharge with Negative Tests. Individuals hospitalized for empiric treatment initiation without microbiologic confirmation, were discharged from hospital 24 hours after receipt of a negative Xpert1MTB/RIF result from either, a) one Xpert1MTB/RIF test or, b) 3 consecutive Xpert1MTB/RIF tests. However, outpatient treatment and associated costs continued until a negative culture results were obtained (ie. after two months).\n\nUnivariate Sensitivity Analysis\nUnivariate sensitivity analysis was conducted to understand the influence of key variables on projected costs and days to treatment initiation. Given uncertainty in point estimates, the following were varied extensively: 1) Costs per sample analyzed with Xpert1MTB/RIF (total cost including cartridge, machine, labor from $50-$350); 2) Proportion of individuals under investigation in remote communities moved to Iqaluit hospital for further TB work up (from 0%10%); 3) Underlying prevalence of TB disease in those evaluated for possible TB (from 0%20%), as this likely varies within the region; and 4) Cost for medical evacuation (from $8,000$12,000).\n\nPLOS ONE | DOI:10.1371/journal.pone.0150119 March 18, 2016\n\n5 / 13\n\nEconomic Analysis of Xpert1MTB/RIF for Diagnosis of TB\n\nTable 2. Intervals for Patients Ultimately Diagnosed with TB disease, From Entry to Health System to Eventual TB Diagnosis and Treatment Initiation in Nunavut, Canada.\n\nInterval prior to diagnosis\nIQALUIT Interval between ordering CXR and result provided to doctor Interval between obtaining 3 spontaneous sputum samples and lab stamp in Iqaluit Interval between obtaining 3 sputum samples via induction in Iqaluit and lab stamp in Iqaluit Interval between sending out all 3 sputum samples from Iqaluit to smear result and \ufb01rst dose of meds when required (regardless of setting) Interval between sending out 3 sputum samples from Iqaluit to culture result and \ufb01rst dose of meds when required (regardless of setting) REMOTE COMMUNITY Interval between ordering CXR and result provided to doctor Interval between obtaining 3 spontaneous sputum samples and lab stamp in Iqaluit (2 extra days added for shipment of sample to Iqaluit) Interval in remote communities if spontaneous sputum cannot initially be obtained (2nd attempt after 2 weeks is assumed to be successful) XPERT1MTB/RIF Interval between obtaining single sputum sample (either spontaneous or induced) for Xpert1MTB/RIF and lab stamp in Iqaluit Interval between obtaining single sputum sample for Xpert1MTB/RIF and lab stamp in remote community Interval between lab stamp of receipt of single sputum sample to Xpert result and \ufb01rst treatment doses (if needed) in I Iqaluit Interval between lab stamp of receipt of single sputum sample to Xpert result and \ufb01rst treatment dose (if needed) in remote community (2 extra days added for shipment of sample to Iqaluit) MISSED CASE Interval associated with missing a case\n\nAverage Days\n1.5 7 3 7.7 37.1\n4 9 14\n1 3 1.8 3.8\n60\n\nRange Reference\n\n1.1, 1.9 Personal communication Van Dyk 5.3, 8.8 Personal communication, Alvarez and DeMaio\n\n2.3, 3.8 Personal communication, Alvarez and DeMaio\n\n5.8, 9.6 [6]\n\n27.8,\n\n[6]\n\n46.4\n\n3.0, 5.0\n6.8, 11.3\n10.5, 17.5\n\nPersonal communication Van Dyk Personal communication, Alvarez and DeMaio\nAssumption\n\n0.8, 1.3 2.3, 3.8 1.4, 2.3\n\nAssumes one day for collection of single sample to use for Xpert1MTB/RIF\nAssumes one day for collection of single sample to use for Xpert1MTB/RIF, plus 2 days travel\n[6]\n\n2.9, 4.8 Assumption\n\n45.0, 75.0\n\nAssumption\n\nFOOTNOTE: Table shows the average number of days taken to complete each activity, according to information obtained from Nunavut sources, or assumption when no estimate available.\n\ndoi:10.1371/journal.pone.0150119.t002\n\nProbabilistic Sensitivity Analysis\nFinally, probabilistic sensitivity analysis was conducted using 1,000 Monte Carlo trials, to obtain a 95% uncertainty range (2.5th and 97.5th percentiles) around the point estimates for projected outcomes for each strategy. Distributions were defined for probabilities, and times to diagnosis and treatment milestones. Costs were not varied as most values were provided from reliable sources. For most probabilities used in the model, beta distributions were fitted to 95% confidence intervals obtained from empiric data. For times to diagnosis and treatment inputs, a range of +/- 25% of the point estimate was used, as empiric data were not always available. Triangular distributions were fitted using these maximum and minimum estimates. More detail on distributions is provided in appendix (Tables A and B in S1 File).\n\nResults\nIn the base case scenario, with the status quo strategy, total cost per individual evaluated for TB for the full population (Iqaluit and remote communities combined) was $2278 (95% uncertainty range (UR): $1668, $2649). The mean days to treatment initiation was 1.7 (0.7, 3.0) per\n\nPLOS ONE | DOI:10.1371/journal.pone.0150119 March 18, 2016\n\n6 / 13\n\nEconomic Analysis of Xpert1MTB/RIF for Diagnosis of TB\n\nTable 3. Costs Associated with TB investigation and Care in Nunavut, Canada. (2014 $CAD).\n\nDescription\n\nValue (2014 Source $CAD)\n\nInitial TB assessment Iqaluit (1 MD visit) Chest x-ray in Nunavut in a hospital or remote community\n-Technician Fee: 15-min appointment -Two \ufb01lms (two views) -Radiologic Interpretation in Ottawa Spontaneous sputum collection Nunavut in a hospital or remote community -Sending South -Cups Sputum induction in Iqaluit Analysis of sputum If result is negative -Base sputum analysis fee for 3 samples If result is positive -Base sputum analysis fee for 3 samples -1 PCR probe if positive Hospitalization per day in Iqaluit, Nunavut -Hospitalization Qikiqtani General Hospital Mean Days spent in hospital for TB treatment initiation in Iqaluit, Nunavut Initial TB assessment in a remote community in Nunavut -Nurse fee for TB assessment (20 mins / assessment at 70.01/hr\u2014range for hourly wage 52.74\u201387.30) Administering DOT in Nunavut -Nurse fee for 104 DOT administrations (10 mins/visit) Drugs for TB therapy Follow up following treatment initiation -5 visits with Nurse (20 mins/visit) -5 Liver Function Tests (5.31 each) From remote community to Iqaluit, Medical Evacuation From Iqaluit to remote community, \ufb02ight home\n\n64.49 67.32 10.14\n6.13 51.06\n3.49\n\nNunavut Health Department\nQikiqtani General Hospital Qikiqtani General Hospital [23]\n\n3.02 0.47 96.00\n\nFirst Air Cargo Division Fischer Scienti\ufb01c [21]\n\n29.04 29.04 80.09 29.04 51.05 2463.23 2463.23 14\n\nGamma Dynacare Ottawa\nGamma Dynacare Ottawa Gamma Dynacare Ottawa\nNunavut Health Department Nunavut Health Department\n\n23.83 23.83 [24]\n\n1410.04 1410.04\n432.15 146.27 119.16\n27.11 10000 1428.83\n\n[24] Qikiqtani General Hospital\n[24] Nunavut Health Department Assumption First Air (Average common return fare to/from Iqaluit)\n\ndoi:10.1371/journal.pone.0150119.t003\n\nperson evaluated for possible TB, and 21.9 (7.9, 36.8) per confirmed active TB case (Table 6). Costs were highest in the Iqaluit hospital community $5261 ($3397, $6815) mainly due to hospitalization, however, time to treatment initiation was shorter because the testing process was faster.\nWith the addition of Xpert1MTB/RIF, overall costs increased by roughly $100 per individual investigated, while days to treatment initiation were reduced by 0.6 days per person evaluated for possible TB, and by 7.8 days per confirmed TB case. The estimated incremental cost per treatment day gained (per individual evaluated for TB) was $164 ($85, $452) (Table 6). In the Iqaluit hospital community, the estimated incremental cost per treatment day gained (per individual evaluated for TB) was higher at $340 ($137, $1141), while in the remote community it was $145 ($72, $392).\n\nSensitivity Analyses\nWhen 3 sputum specimens were analyzed with Xpert1MTB/RIF (vs 1 in the baseline scenario) (Table 7), costs were higher, but time to treatment initiation was further reduced\n\nPLOS ONE | DOI:10.1371/journal.pone.0150119 March 18, 2016\n\n7 / 13\n\nEconomic Analysis of Xpert1MTB/RIF for Diagnosis of TB\n\nTable 4. Costs Associated with Xpert1MTB/RIF.\n\nCost Component\n\nBase Case Cost (2014 $CAD)\n\nRange\n\nReference\n\nGene Xpert1MTB/RIF 4 Module, machine with laptop and printer Life expectancy of machine Depreciated annual machine cost assuming 10 year life expectancy Gene Xpert1MTB/RIF Cartridges Calibration Module Laboratory Technologist dedicated to Xpert1MTB/RIF\n\n$60,000 10 years $6000 $60/cartridge + $2.50 shipping $450/year $110,000 annual salary (includes bene\ufb01ts)\n\n$30,000-$90,000 $35-$60 -\n\nInter Medico QC* Inter Medico QC Calculated Inter Medico QC Inter Medico QC [25, 26]\n\n* Inter Medico is the of\ufb01cial distributor for Cepheid/ Xpert1MTB/RIF in Canada\n\ndoi:10.1371/journal.pone.0150119.t004\n\nbecause of increased sensitivity with 3 samples. The estimated incremental cost per treatment day gained (per individual evaluated for TB) was $197, a small change compared to $164 for a single sample.\nIn the scenario with hospital discharge after one negative Xpert result (Table 7), costs for the Xpert1MTB/RIF scenario were substantially reduced, and it became cost saving in all settings. For the entire cohort, costs per individual investigated were reduced by almost $1000. In Iqaluit, costs were reduced by $2495 per person, because all individuals initiated treatment in hospital. In the remote communities, hospital release following one negative Xpert1MTB/RIF had much less impact as only 4% of individuals under investigation were transferred to the Iqaluit hospital. Days to treatment initiation remained the same as in the base case scenario, as they were unrelated to hospital discharge.\nIf three negative Xpert1MTB/RIF results were required for hospital release (Table 7), savings were still projected relative to the status quo.\n\nUnivariate Sensitivity Analysis\nWhen the total \"all in\" costs per sample analyzed using Xpert1MTB/RIF were varied from $50 to as high as $350, costs per day of earlier treatment initiation per individual investigated for TB increased proportionately. With a high-end cost of $350 per sample analyzed, the incremental cost per day gained was approximately $400. When the proportion of individuals\n\nTable 5. Per Sample Component Costs for Xpert1MTB/RIF, by Number of Individuals and Samples Evaluated.\n\nCost Component (per sample)\nMachine Cost (Calculated using base case depreciated machine cost shown above) Gene Xpert1MTB/RIF Cartridges Calibration Module Laboratory Technologist Total per Sample Total per Individual Investigated\n\nCost for 1651 Individuals 1 Cost for 1651 Individuals 3\n\nsample per individual*\n\nsamples per individual**\n\n$3.63\n\n$1.21\n\n$62.50 $0.27 $66.63 $133.03 $133.03\n\n$62.50 $0.09 $22.21 $86.01 $258.03\n\n* Cost used in base case analysis as well as Sensitivity Analysis 2a (Hospital discharge with 1 negative test) ** Cost used in Sensitivity Analysis 1 (3 sputum samples per individual) and Sensitivity Analysis 2b (Hospital discharge with 3 negative tests)\n\ndoi:10.1371/journal.pone.0150119.t005\n\nPLOS ONE | DOI:10.1371/journal.pone.0150119 March 18, 2016\n\n8 / 13\n\nEconomic Analysis of Xpert1MTB/RIF for Diagnosis of TB\n\nTable 6. Base Case Analysis: Projected Cost and Time to Treatment Initiation, Incremental Outcomes and Incremental Cost per Treatment Day Gained in Nunavut, Canada.\n\nStatus Quo\n\nGene Xpert1MTB/RIF Added\n\nGene Xpert1MTB/RIF vs Status Quo\n\n$ per Individual evaluated for TB\n\nTime to treatment initiation per individual evaluated for TB\n\nIqaluit\n\n5261\n\n1.3\n\nHospital\n\nCommunity\n\nRemote\n\n1486\n\n1.8\n\nNunavut\n\nCommunity\n\nFull\n\n2278\n\n1.7\n\nCohort\n\ndoi:10.1371/journal.pone.0150119.t006\n\nTime to treatment initiation per con\ufb01rmed TB case\n16.5\n23.4\n21.9\n\n$ per Individual Evaluated for TB\n5455\n157\n2390\n\nTime to treatment initiation per individual evaluated for TB\n0.7\n1.2\n1.1\n\nTime to treatment initiation per con\ufb01rmed TB case\n9.2\n15.4\n14.1\n\nIncremental cst\n\nTreatment day gained per individual evaluated for TB\n\nTreatment day gained per con\ufb01rmed TB case\n\nIncremental cost per treatment day gained (per individual evaluated for TB)\n\n194\n\n0.6\n\n7.29\n\n340\n\n90\n\n0.6\n\n7.99\n\n145\n\n100\n\n0.6\n\n7.83\n\n164\n\ntransported to Iqaluit hospital for further investigation was varied from 0\u201310%, the effect on the incremental cost per treatment day gained for the Xpert1MTB/RIF vs status quo was minimal, increasing from $144 to $187. The effect of varying the underlying prevalence of TB among those referred for investigation was more substantial. Increasing the prevalence from 1% to 20% decreased the incremental cost per treatment day gained from over $1000 to approximately $60 (Fig 2). Increasing the cost for medical evacuation from $8,000-$12,000 increased projected costs only by a small amount because of the small number of people who required evacuation.\nAdditional results from the probabilistic sensitivity analysis are shown in Table C in S1 File.\n\nInterpretation\nAdding Xpert1MTB/RIF to the current Nunavut diagnostic algorithm for TB is likely to be cost effective. The estimated reduction in time to treatment initiation was substantial, 7.8 days per confirmed active case, and the incremental cost per treatment day gained was only $164, versus the status quo scenario where only smear and culture are used for microbiologic diagnosis of active TB.\nXpert1MTB/RIF has been judged cost effective in other settings, both high income [11] and low income [12\u201314]. The health care infrastructure and clinical practices in place prior to Xpert1MTB/RIF implementation were important determinants of study findings in such settings. The clinical care landscape in the Canadian Arctic is unique. It is a remote environment with enormous distances between communities, where healthcare and laboratory infrastructure is sparse, and clinical samples must be shipped to southern Canada for microbiological confirmation. Consequently, TB patients are often only diagnosed and treated after some delay, and may therefore have more advanced disease, in contrast to high income urban settings where Xpert1MTB/RIF use has been studied, such as San Francisco [4] and Montreal [15]. In Montreal, where Xpert1MTB/RIF appeared to have particularly limited utility, most individuals investigated for TB were asymptomatic and referred via immigration screening [15], which differs substantially from the Arctic setting. Hence, because of the differences in infrastructure and patient characteristics, Xpert1MTB/RIF may be particularly valuable in the Canadian Arctic. The Canadian TB Standards recommend that where there is no on-site\n\nPLOS ONE | DOI:10.1371/journal.pone.0150119 March 18, 2016\n\n9 / 13\n\nEconomic Analysis of Xpert1MTB/RIF for Diagnosis of TB\n\nTable 7. Sensitivity Analysis: Projected Cost and Time to Treatment Initiation, Incremental Cost (savings) and Treatment Days Gained, and Incremental Cost per Treatment Days Gained, per Individual Evaluated for TB, in Nunavut, Canada.\n\nStatus Quo\n\nGene Xpert1MTB/RIF Added\n\nGene Xpert1MTB/RIF vs Status Quo\n\n$ per Individual Time to\n\nevaluated for treatment\n\nTB\n\ninitiation\n\n$ per Individual evaluated for TB\n\nTime to treatment initiation\n\n*Incremental cost\n\nTreatment days gained\n\nIncremental cost per treatment day gained\n\nBase Case\n\nIqaluit Hospital\n\n5261\n\n1.3\n\n5455\n\n0.7\n\n194\n\n0.6\n\n340\n\nCommunity\n\nRemote\n\n1486\n\n1.8\n\n1576\n\n1.2\n\n90\n\n0.6\n\n145\n\nNunavut\n\nCommunity\n\nFull Cohort\n\n2278\n\n1.7\n\n2390\n\n1.1\n\n100\n\n0.6\n\n164\n\nSensitivity Analysis 1: 3 specimens analyzed with Xpert1MTB/RIF per individual\n\nIqaluit Hospital\n\n5261\n\n1.3\n\n5510\n\n0.5\n\n249\n\n0.8\n\n304\n\nCommunity\n\nRemote\n\n1486\n\n1.8\n\n1633\n\n1.0\n\n147\n\n0.9\n\n169\n\nNunavut\n\nCommunity\n\nFull Cohort\n\n2278\n\n1.7\n\n2447\n\n0.9\n\n169\n\n0.9\n\n197\n\nSensitivity Analysis 2a: Hospital discharge for those started on empiric therapy who are Xpert1MTB/RIF negative- 1 specimen analyzed with Xpert1MTB/RIF\n\nIqaluit Hospital\n\n5261\n\n1.3\n\n2960\n\n0.7\n\n-2301\n\n0.6\n\nCommunity\n\nSaving\n\nRemote\n\n1486\n\n1.8\n\n982\n\nNunavut\n\nCommunity\n\n1.2\n\n-504\n\n0.6\n\nSaving\n\nFull Cohort\n\n2278\n\n1.7\n\n1397\n\n1.1\n\n-881\n\n0.6\n\nSaving\n\nSensitivity Analysis 2b: Hospital discharge for those started on empiric therapy who are Xpert1MTB/RIF negative- 3 specimens analyzed with Xpert1MTB/RIF\n\nIqaluit Hospital\n\n5261\n\n1.3\n\n3349\n\n0.5\n\n-1912\n\n0.8\n\nCommunity\n\nSaving\n\nRemote\n\n1486\n\n1.8\n\n1061\n\n1.0\n\n-425\n\n0.9\n\nNunavut\n\nCommunity\n\nSaving\n\nFull Cohort\n\n2278\n\n1.7)\n\n1541\n\n0.9\n\n-727\n\n0.9\n\nSaving\n\n* Negative incremental cost indicates \"Savings\" with Xpert1MTB/RIF strategy relative to the Status Quo strategy\n\ndoi:10.1371/journal.pone.0150119.t007\n\ncapacity for smear microscopy and culture, an automated test can be used to make rapid decisions about TB treatment and isolation, pending routine smear and culture results [7].\nOur analysis focused on the Qikiqtaaluk (Baffin) region, where the great majority of Nunavut TB cases are found. We had to make some generalizations about clinical decision making in our study. Wherever possible, the likelihood of clinical events was informed by Nunavut policy recommendations, or by input from clinicians with extensive experience in the investigation of possible TB in Nunavut. Study inputs were relevant and accurate; most, including test performance, epidemiologic and costing data came from the study of Xpert1MTB/RIF in Nunavut [6]. We did not account for secondary transmission from infectious TB patients. The greatest reduction in diagnostic delays involved smear-negative patients, for whom the probability of onward transmission is substantially lower than with smear-positive disease [16]; their effect would likely be relatively small at a population level [13].\nCosts were considered only from a health system perspective, we did not address those borne by patients and families. Costs were limited to TB diagnosis and treatment; downstream\n\nPLOS ONE | DOI:10.1371/journal.pone.0150119 March 18, 2016\n\n10 / 13\n\nEconomic Analysis of Xpert1MTB/RIF for Diagnosis of TB\n\nFig 2. Underlying Prevalence of TB vs. Incremental Cost per Treatment Day Gained.\ndoi:10.1371/journal.pone.0150119.g002\ncosts such as contact investigation were not considered. There was some uncertainty as to the number of sputum samples to be analyzed using Xpert1 RIF/MTB, therefore we considered scenarios where either 1 or 3 samples per individual were analyzed. There was also uncertainty about the placement of the Xpert machine; placing the machine in more remote communities could further reduce delays and costs by eliminating the transfer of specimens to Iqaluit. We did not evaluate the logistics and costs of on-site Xpert1MTB/RIF testing in remote communities.\nA major strength of our study is the integration of clinical and cost data from a field evaluation of Xpert1MTB/RIF in Iqaluit. Wherever possible, Nunavut data and experience informed our modelling. We focused on current practice and examined potential practice changes as a consequence of Xpert1MTB/RIF adoption.\nIn conclusion, Xpert1MTB/RIF appears potentially cost effective as an addition to the diagnostic algorithm in Nunavut. Although we focused on the Qikiqtaaluk (Baffin) region, our study provides an in-depth analysis of the cost and potential yield of Xpert1MTB/RIF in the Canadian Arctic, and considers possible scenarios for its use in guiding clinical management. The assessment of cost and clinical impact, and of scenarios for optimal use, may be relevant for other remote communities where there is ongoing TB-related morbidity, and where it is proposed for on-site TB diagnosis. Xpert1MTB/RIF may be especially well suited to overcome some of the logistical barriers unique to this remote Arctic setting.\nSupporting Information\nS1 File. Supplemental Methods -Details Of Diagnostic Strategies & Supplemental Results. Table A. Range and Distribution Around Point Estimate of Clinical and Epidemiologic Model Probabilities. Table B. Range and Distribution Around Point Estimate of Average Intervals to Diagnosis Associated with Tuberculosis Care. Table C. Base Case and Sensitivity Analysis with 95% Uncertainty ranges: Projected Cost and Time to treatment initiation, Incremental Cost (savings) and treatment days gained, and Cost per Treatment days gained, per Individual Evaluated for TB in Nunavut, Canada. (DOCX)\n\nPLOS ONE | DOI:10.1371/journal.pone.0150119 March 18, 2016\n\n11 / 13\n\nEconomic Analysis of Xpert1MTB/RIF for Diagnosis of TB\nAcknowledgments\nThe study investigators would like to acknowledge Sonia Marchand and all of the staff at the Qikiqtani General Hospital Laboratory, Dr. Maureen Baikie and the staff at Iqaluit Public Health, Natan Obed from Nunavut Tunngavik Inc., and the members of the Taima TB steering committee. We would also like to thank Dr. Patricia DeMaio and Ms. Deborah Van Dyk for their assistance and guidance in providing study data.\nAuthor Contributions\nConceived and designed the experiments: OO GA MP KS. Performed the experiments: OO GA JS KS. Analyzed the data: OO GA JS KS. Contributed reagents/materials/analysis tools: GA MP. Wrote the paper: OO GA JS MP KS.\nReferences\n1. Public Health Agency of Canada. Tuberculosis in Canada\u2013 2012 Pre-Release. Ottawa, ON, Canada. 2012. Available: http://www.phac-aspc.gc.ca/tbpc-latb/pubs/tbcan12pre/tab-eng.php#tab1.\n2. World Health Organization. Tuberculosis Diagnostics Xpert MTB/RIF system Test: Updated Recommendations Geneva: World Health Organization; 2013.\n3. Steingart KR, Sohn H, Schiller I, Kloda L, Boehme CC, Pai M, et al. Xpert1 MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev. 2014; 1.\n4. Davis JL, Kawamura LM, Chaisson LH, Grinsdale J, Benhammou J, Ho C, et al (2014) Impact of GeneXpert MTB/RIF1 on Patients and Tuberculosis Programs in a Low-Burden Setting: A Hypothetical Trial. Am J Respir Crit Care Med. 2014 Jun 15; 189(12):1551\u20139. doi: 10.1164/rccm.201311-1974OC PMID: 24869625\n5. Chaisson LH, Roemer M, Cantu D, Haller B, Millman AJ, Cattamanchi A, et al. Impact of GeneXpert MTB/RIF Assay on Triage of Respiratory Isolation Rooms for Inpatients With Presumed Tuberculosis: A Hypothetical Trial. Clin Infect Dis. 2014; 59:1353\u201360. doi: 10.1093/cid/ciu620 PMID: 25091300\n6. Alvarez GG, Van Dyk DD, Desjardins M, Yasseen AS III, Aaron SD, Cameron DW, et al. The feasibility, accuracy and impact of Xpert1 MTB/RIF testing in a remote Aboriginal community in Canada. Chest. 2015; 2015 Mar 19. doi: 10.1378/chest.14-2948\n7. Menzies D, editor. Canadian Tuberculosis Standards. Public Health Agency of Canada, Ottawa, ON, Canada; 2014.\n8. McGuire M GV, Bourgeois A-C, Ogunnaike-Cooke S. A summary of tuberculosis drug resistance in Canada, 2003\u22122013. Canada Communicable Disease Report 2015; 41 S-2:8\u201315.\n9. Bank of Canada. Inflation Calculator (cited 2015 April 21). Available: http://www.bankofcanada.ca/ rates/related/inflation-calculator/.\n10. Boehme CC, Nabeta P, Hillemann D, Nicol MP, Shenai S, Krapp F, et al. Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med. 2010; 363(11):1005\u201315. doi: 10.1056/ NEJMoa0907847 PMID: 20825313\n11. Choi H, Miele K, Dowdy D, Shah M. Cost-effectiveness of Xpert1 MTB/RIF for diagnosing pulmonary tuberculosis in the United States. Int J Tuberc Lung Dis. 2013; 17(10):1328. doi: 10.5588/ijtld.13.0095 PMID: 24025386\n12. Langley I, Lin H-H, Egwaga S, Doulla B, Ku C-C, Murray M, et al. Assessment of the patient, health system, and population effects of Xpert MTB/RIF and alternative diagnostics for tuberculosis in Tanzania: an integrated modelling approach. Lancet Glob Health. 2014; 2(10):e581\u2013e91. doi: 10.1016/S2214109X(14)70291-8 PMID: 25304634\n13. Menzies NA, Cohen T, Lin HH, Murray M, Salomon JA. Population health impact and cost-effectiveness of tuberculosis diagnosis with Xpert MTB/RIF: a dynamic simulation and economic evaluation. PLoS Med. 2012; 9(11):e1001347. doi: 10.1371/journal.pmed.1001347 PMID: 23185139\n14. Vassall A, van Kampen S, Sohn H, Michael JS, John K, den Boon S, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: a cost-effectiveness analysis. PLoS medicine. 2011; 8(11):e1001120. doi: 10.1371/journal.pmed.1001120 PMID: 22087078\n15. Sohn H, Aero AD, Menzies D, Behr M, Schwartzman K, Alvarez GG, et al. Xpert MTB/RIF testing in a low TB incidence, high-resource setting: limitations in accuracy and clinical impact. Clin Infect Dis. 2014:ciu022.\n\nPLOS ONE | DOI:10.1371/journal.pone.0150119 March 18, 2016\n\n12 / 13\n\nEconomic Analysis of Xpert1MTB/RIF for Diagnosis of TB\n16. Behr M, Warren S, Salamon H, Hopewell P, de Leon AP, Daley C, et al. Transmission of Mycobacterium tuberculosis from patients smear-negative for acid-fast bacilli. Lancet. 1999; 353(9151):444\u20139. PMID: 9989714\n17. Goverment of Canada. Canadian Census. 2011.\n18. van't Hoog AH, Meme HK, Laserson KF, Agaya JA, Muchiri BG, Githui WA, et al. Screening strategies for tuberculosis prevalence surveys: the value of chest radiography and symptoms. PloS one. 2012; 7 (7):e38691. doi: 10.1371/journal.pone.0038691 PMID: 22792158\n19. Koppaka R, Bock N. How reliable is chest radiography? In: Toman\u2019s tuberculosis: case detection, treatment and monitoring 2nd ed Geneva, Switzerland World Health Organization; 2004. p. 51\u201360.\n20. Schoch OD, Rieder P, Tueller C, Altpeter E, Zellweger J-P, Rieder HL, et al. Diagnostic yield of sputum, induced sputum, and bronchoscopy after radiologic tuberculosis screening. Am J Respir Crit Care Med. 2007; 175(1):80\u20136. PMID: 17053204\n21. Sugarman J, Alvarez G, Schwartzman K, Oxlade O. Sputum induction for tuberculosis diagnosis in an Arctic setting: a cost comparison. Int J Tuberc Lung Dis. 2014; 18(10):1223\u201330. doi: 10.5588/ijtld.14. 0163 PMID: 25216837\n22. Public Health Agency of Canada. 2002\u20132010 Tuberculosis in Canada Reports. Ottawa, ON, Canada: 2002\u20132010.\n23. Schedule of benefits for physician services under the Health Insurance Act. Toronto, ON, Canada: Government of Ontario, 2013. Available: http://www.health.gov.on.ca/english/providers/program/ohip/sob/ physserv/physser. Accessed 2015 Apr 21.\n24. Nunavut nurses: salary and bonuses. Iqaluit, NU, Canada: Government of Nunavut, 2008. Available: http://www.nunavutnurses.ca/english/jobs/salary_bonuses.shtml. Accessed 2015 Apr 21.\n25. Patient cost estimator. Ottawa, ON, Canada: CIHI, 2013. Available: http://www.cihi.ca/CIHI-ext-portal/ internet/en/document full/spending\u00feand\u00fehealth\u00feworkforce/spending/pce_application. Accessed 2015 Apr 21.\n26. Nunavut Employees Union. Collective Agreement between the Nunavut Employees Union and the minister responsible for the Public Service Act. Available: http://www.neu.ca/Collective_Agreements. Accessed 2015 Apr 21.\n\nPLOS ONE | DOI:10.1371/journal.pone.0150119 March 18, 2016\n\n13 / 13\n\n\n",
    "authors": [
      "Olivia Oxlade",
      "Jordan Sugarman",
      "Gonzalo G. Alvarez",
      "Madhukar Pai",
      "Kevin Schwartzman"
    ],
    "doi": "10.1371/journal.pone.0150119",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://dx.plos.org/10.1371/journal.pone.0150119"
  },
  {
    "key": "WT5DLV98",
    "title": "Costs and cost-effectiveness of Gene Xpert compared to smear microscopy for the diagnosis of pulmonary tuberculosis using real-world data from Arsi zone, Ethiopia",
    "abstract": "Background\n              Early diagnosis and treatment are one of the key strategies of tuberculosis control globally, and there are strong efforts in detecting and treating tuberculosis cases in Ethiopia. Smear microscopy examination has been a routine diagnostic test for pulmonary tuberculosis diagnosis in resource-constrained settings for decades. Recently, many countries, including Ethiopia, are scaling up the use of Gene Xpert without the evaluation of the cost and cost-effectiveness implications of this strategy. Therefore, this study evaluated the cost and cost-effectiveness of Gene Xpert (MTB/RIF) and smear microscopy tests to diagnosis tuberculosis patients in Ethiopia.\n            \n            \n              Methods\n              We compared the costs and cost-effectiveness of tuberculosis diagnosis using smear microscopy and Gene Xpert among 1332 patients per intervention in the Arsi zone. We applied combinations of top-down and bottom-up costing approaches. The costs were estimated from the health providers\u2019 perspective within one year (2017\u20132018). We employed \u201ccases detected\u201d as an effectiveness measure, and the incremental cost-effectiveness ratio was calculated by dividing the changes in cost and change in effectiveness. All costs and incremental cost-effectiveness ratio were reported in 2018 US$.\n            \n            \n              Results\n              The unit cost per test for Gene Xpert was $12.9 whereas it is $3.1 for AFB smear microscopy testing. The cost per TB case detected was $77.9 for Gene Xpert while it was $55.8 for the smear microscopy method. The cartridge kit cost accounted for 42% of the overall Gene Xpert\u2019s costs and the cost of the reagents and consumables accounted for 41.3% ($1.3) of the unit cost for the smear microscopy method. The ICER for the Gene Xpert strategy was $20.0 per tuberculosis case detected.\n            \n            \n              Conclusion\n              Using Gene Xpert as a routine test instead of standard care (smear microscopy) can be potentially cost-effective. In the cost scenario analysis, the price of the cartridge, the number of tests performed per day, and the life span of the capital equipment were the drivers of the unit cost of the Gene Xpert method. Therefore, Gene Xpert can be a part of the routine TB diagnostic testing strategy in Ethiopia.",
    "full_text": "PLOS ONE\n\na1111111111 a1111111111 a1111111111 a1111111111 a1111111111\n\nRESEARCH ARTICLE\nCosts and cost-effectiveness of Gene Xpert compared to smear microscopy for the diagnosis of pulmonary tuberculosis using real-world data from Arsi zone, Ethiopia\nAbdene Weya KasoID1\u262f*, Alemayehu HailuID2\u262f\n1 School of Public Health, College of Medicine and Health Science, Dilla University, Dilla, Ethiopia, 2 Department of Global Public Health and Primary Care, Bergen Centre for Ethics and Priority Setting, University of Bergen, Bergen, Norway\n\u262f These authors contributed equally to this work. * abdannekaso@gmail.com\n\nAbstract\n\nOPEN ACCESS\nCitation: Kaso AW, Hailu A (2021) Costs and costeffectiveness of Gene Xpert compared to smear microscopy for the diagnosis of pulmonary tuberculosis using real-world data from Arsi zone, Ethiopia. PLoS ONE 16(10): e0259056. https://doi. org/10.1371/journal.pone.0259056\nEditor: Frederick Quinn, The University of Georgia, UNITED STATES\nReceived: July 22, 2021\nAccepted: October 11, 2021\nPublished: October 25, 2021\nCopyright: \u00a9 2021 Kaso, Hailu. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.\nData Availability Statement: All relevant data are within the manuscript and its Supporting Information files.\nFunding: AK was supported by an Addis Ababa University School of Public Health Grant. AH was supported by a Bill & Melinda Gates Foundation Grant (OPP1162384). The funders had no role in study design, data collection, analysis, decision to publish, or manuscript preparation.\nCompeting interests: The authors have declared that no competing interests exist.\n\nBackground\nEarly diagnosis and treatment are one of the key strategies of tuberculosis control globally, and there are strong efforts in detecting and treating tuberculosis cases in Ethiopia. Smear microscopy examination has been a routine diagnostic test for pulmonary tuberculosis diagnosis in resource-constrained settings for decades. Recently, many countries, including Ethiopia, are scaling up the use of Gene Xpert without the evaluation of the cost and costeffectiveness implications of this strategy. Therefore, this study evaluated the cost and costeffectiveness of Gene Xpert (MTB/RIF) and smear microscopy tests to diagnosis tuberculosis patients in Ethiopia.\nMethods\nWe compared the costs and cost-effectiveness of tuberculosis diagnosis using smear microscopy and Gene Xpert among 1332 patients per intervention in the Arsi zone. We applied combinations of top-down and bottom-up costing approaches. The costs were estimated from the health providers\u2019 perspective within one year (2017\u20132018). We employed \u201ccases detected\u201d as an effectiveness measure, and the incremental cost-effectiveness ratio was calculated by dividing the changes in cost and change in effectiveness. All costs and incremental cost-effectiveness ratio were reported in 2018 US$.\nResults\nThe unit cost per test for Gene Xpert was $12.9 whereas it is $3.1 for AFB smear microscopy testing. The cost per TB case detected was $77.9 for Gene Xpert while it was $55.8 for the smear microscopy method. The cartridge kit cost accounted for 42% of the overall Gene Xpert\u2019s costs and the cost of the reagents and consumables accounted for 41.3% ($1.3) of\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0259056 October 25, 2021\n\n1 / 10\n\nPLOS ONE\n\nCost and cost-effectiveness of Gene Xpert and smear microscopy for the diagnosis of tuberculosis in Arsi zone\n\nAbbreviations: AFB, Acid Fast Bacilli; ACER, average cost-effectiveness ratio; DALY, DisabilityAdjusted Life Year; HR, human resources; HIV, human immunodeficiency virus; ICER, incremental cost-effectiveness ratio; MDR-TB, multi-drug resistant TB; MDR/RIF, multi-drug resistant TB/ rifampicin; MTB, mycobacterium TB; TB/HIV, TB/ human immune virus; PSA, probabilistic sensitivity analysis; TB, tuberculosis; US$, United States dollar; WTP, Willingness-to-pay; WHO, World Health Organization.\n\nthe unit cost for the smear microscopy method. The ICER for the Gene Xpert strategy was $20.0 per tuberculosis case detected.\nConclusion\nUsing Gene Xpert as a routine test instead of standard care (smear microscopy) can be potentially cost-effective. In the cost scenario analysis, the price of the cartridge, the number of tests performed per day, and the life span of the capital equipment were the drivers of the unit cost of the Gene Xpert method. Therefore, Gene Xpert can be a part of the routine TB diagnostic testing strategy in Ethiopia.\n\nIntroduction\nTuberculosis (TB) remains a considerable public health threat in Africa [1]. It is also the leading cause of mortality in Ethiopia with a TB incidence of 163 cases per 100,000 population reported in 2016. In 2015, the prevalence of multiple drug-resistant TB (MDR-TB) reached 2.7% in new TB cases and 14.0% in retreated TB cases in Ethiopia [2,3].\nEarly case detection and treatment of TB cases were among the key strategies of the National TB and Leprosy Control Program (NTLCP). Thus, an advanced diagnostic tool is required to detect and treat TB in Ethiopia [4]. In Ethiopia, smear microscopy was a routine TB diagnostic tool for decades [5]. It has low sensitivity, doesn\u2019t enable the detection of MDR-TB, and has little value in extrapulmonary TB and children [6,7]. Therefore, the detection of MDR-TB strains requires the introduction of more sensitive and highly advanced diagnostic tools. Gene Xpert is one of these advanced technologies and has the potential for rapid diagnosis of TB and MDR-TB [8,9]. In 2010, the World Health Organization (WHO) endorsed this rapid and advanced molecular tool for the diagnosis of tuberculosis [10]. According to the recent systematic review, Gene Xpert has 85% pooled sensitivity and 98% pooled specificity in TB case detection whereas it has 96% pooled sensitivity and 98% pooled specificity in Rifampicin resistance detection [11]. Furthermore, Gene Xpert has certain benefits over the routine Acid-fast bacilli (AFB) smear examination method. It has low human resource requirements and it doesn\u2019t need biosafety during diagnosis [12]. This has generated a new hope in populations with high burdens of TB such as Ethiopia [13].\nHowever, the scale-up of Gene Xpert as a routine diagnostic test in the health care delivery system has substantial economic implications. A previous study shows that using Gene Xpert for diagnosis of TB in routine services among suspected patients increases the health system testing costs compared with using smear microscopy [14]. Recently, in Ethiopia, Gene Xpert was used for routine TB diagnosis without evaluating the cost and cost-effectiveness implications of the method [15]. Therefore, this study aimed to evaluate the cost and cost-effectiveness of Gene Xpert (MTB/RIF) and smear microscopy tests to diagnose TB in Ethiopia.\nMethods\nStudy setting\nA cost and cost-effectiveness analysis of TB diagnostic strategies was conducted among 1332 patients per intervention in public health facilities in the Arsi zone. The zone comprises 28 Woreda and two town administrations. It also consists of 7 hospitals and 104 health centers. The current Ethiopian health tier system classified health facilities as a primary, secondary,\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0259056 October 25, 2021\n\n2 / 10\n\nPLOS ONE\n\nCost and cost-effectiveness of Gene Xpert and smear microscopy for the diagnosis of tuberculosis in Arsi zone\nand tertiary healthcare systems. In this study, one tertiary level healthcare (i.e. Regional referral hospital), one secondary level healthcare (i.e. General hospital), and six primary healthcare facilities (three hospitals and three health centers) were included. Among these facilities, one referral and general hospital perform TB testing using the Gene Xpert technique whereas the three hospitals and health centers use the AFB smear microscopy method. In the hospitals and health centers, the laboratory operates under NTLCP of Ethiopia.\nStudy design and description of interventions\nThis study compared the cost and cost/effectiveness of two TB diagnostic methods: Gene Xpert MTB/RIF assay and smear microscopy method. A single spot sputum specimen was tested for all suspected TB cases in the Gene Xpert methods (MTB/RIF assay). A fresh sputum sample was taken from the patients presented with signs and symptoms of pulmonary TB based on WHO TB and Ethiopia NTLCP screening criteria [10,16]. The sputum sample was combined with the reagent in a 2:1 ratio for sputum liquefaction and inactivation. Then, it was incubated at room temperature for 15\u201320 minutes. After incubation, a total of 2 ml of the mixture was introduced into the Gene Xpert cartridge and loaded into the Gene Xpert instrument for analysis. The instrument generates a test report automatically within 2 hours. In the smear microscopy method, two spot-spot sputum samples (i.e. a first spot sputum sample at the first arrival of patients to laboratory and another one spot sample after 30 minute) were collected from all suspected TB patients and examined for acid-fast bacilli using Ziehl\u2013Nielsen or fluorescent staining technique [13,17].\nData collection\nThe cost estimation was based on service data collected from three health centers and five hospitals in the Arsi Zone. This costing was conducted from the health providers\u2019 perspective, with all costs to the health system for diagnosing TB among suspected individuals with smear microscopy and the Gene Xpert technique were included. Various cost components, such as building space, equipment, consumables, overhead, and human resources (HR), were included. Additionally, weekly quality control, external quality control, and annual calibration costs for Gene Xpert were included. However, the maintenance cost for both methods was not included due to a lack of data.\nThe cost data were collected from hospital procurement invoices, expert interviews, managers\u2019 opinions, and other administrative reports available in the health facilities. The persontime elapsed per test was estimated based on observation and consultation with laboratory technicians. The cost-of-training data were obtained from Oromia Public Health Research Capacity Building & Quality Assurance Laboratory. The annual calibration cost was obtained from the Stop TB Partnership source [18]. The cost types and quantity of each resource used in each diagnostic technique were recorded using a Microsoft Excel 2010 spreadsheet. The data collection tools were developed by modifying the WHO guidelines for the cost and costeffectiveness analysis of TB control. After the pretest was conducted, necessary corrections were made to the data collection tools.\nData analysis\nThe cost estimation was conducted with a one-year time frame (2017\u20132018) using an ingredients-based and top-down approach. In ingredient-based approach, all inputs required to perform diagnostic testing was identified, quantified and multiplied with their prices to arrive at a unit cost per patient tested. Moreover, in top-down method, we divided the gross cost data expenditure on building, overhead, reagents and equipment allocated specifically to each\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0259056 October 25, 2021\n\n3 / 10\n\nPLOS ONE\n\nCost and cost-effectiveness of Gene Xpert and smear microscopy for the diagnosis of tuberculosis in Arsi zone\ndiagnostic method by the numbers of tests conducted. We valued personnel cost based on an estimated proportion of working time spent on TB diagnosis by each method. The smear microscopy diagnostic procedure took 35 minutes per test, and the Gene Xpert method took 25 minutes per test. The number of training days for conducting tests by AFB smear microscopy and Gene Xpert was assumed to be five days for smear microscopy and three days for Gene Xpert technique. The cost of reagents and consumables for smear microscopy was calculated using the gross cost of each reagent over the number of patients tested with that amount. Moreover, the cost of the Xpert cartridge kit was obtained from a published source [19].\nThe cost of equipment for each diagnostic method was obtained by dividing the equipment\u2019s annualized cost by the number of tests performed. The useful life of microscopy was assumed to be 10 years, running five tests per day. A Gene Xpert machine was assumed to be operating four simultaneous tests and running for eight hours per day. The instrument was assumed to have a useful life of 10 years and to process, on average, eight sputum samples per day. For buildings, an expected lifetime of 30 years was used. Laboratory space cost was allocated based on the proportional size required for conducting TB diagnosis.\nThe cost of quality control for smear microscopy was estimated by identifying all the resources needed to perform the quality control procedure based on WHO Stop TB guidelines. The cost of overhead for each technique was calculated by taking 5% of the total health overhead cost based on laboratory heads\u2019 expert opinions. Capital costs were annualized using a 3% discount rate per year [20,21]. Local costs were collected in Ethiopian birr and converted to United States dollars (US$) using average exchange rates from the National Bank of Ethiopia (US$1 = 27.18 birr) [22]. All the costs were adjusted for inflation using the consumer price index of the year 2018 as a base year cost.\nAn incremental cost-effectiveness ratio (ICER) was calculated using the ratio of the change in unit cost per case detected. As there is no widely accepted willingness-to-pay (WTP) threshold for this intermediate outcome (i.e., ICER in terms of cost per case detected), we did not use any threshold, and therefore we presented only the estimated ICER in this study.\nEthics approval and consent to participate\nEthical approval was obtained from the Addis Ababa University School of Public Health Institutional Review Board. The ethical review board provided a waiver for consent to participate, as the data were collected from patient records.\nResults Socio-demographic characteristics of the participants\nAmong TB suspected patients diagnosed by AFB smear microscopy method, 692 (52%) were male, and 639(48%) were in the age group 5\u201325 years old. Around 1,258 (94.4%) patients had a smear-negative result, whereas 74 (5.6%) had a smear-positive result. Among 1332 patients enrolled in the Gene Xpert algorithm, more than half (54.4%) were males and around 221 (16.6%) patients have TB infection whereas the remaining 1111(83.4%) were negative for TB. The majority of patients diagnosed using the Gene Xpert method were in the age group 5\u201325 years with the mean age of 34\u00b1 17 standard deviation (SD) (Table 1).\nCost per patient tested\nThe unit cost of testing suspected TB patients using smear microscopy and the Gene Xpert algorithm varies with the volume of testing and level of health facilities. The average cost of the testing using the smear microscopy technique was $3.1 (ranging from $2.40 to $4.96) whereas\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0259056 October 25, 2021\n\n4 / 10\n\nPLOS ONE\n\nCost and cost-effectiveness of Gene Xpert and smear microscopy for the diagnosis of tuberculosis in Arsi zone\n\nTable 1. Characteristics of TB suspected patients by testing technique.\n\nCategory Age 5\u201325 years 26\u201334 years 35\u201344 years 45\u201354 years 55 and above years Sex\nMale Female Test results Test positive Test negative\n\nSmear microscopy Frequency (%)\n639(48%) 348(26.1%) 174(13.1%) 124(9.3%) 47(3.5%)\n692 (52.0%) 640 (48.0%)\n74 (5.6%) 1,258 (94.4%)\n\nhttps://doi.org/10.1371/journal.pone.0259056.t001\n\nGene Xpert Frequency (%)\n504(37.8%) 279(21%) 199(14.9%) 144(10.8%) 206(15.5%)\n724 (54.4%) 608 (45.6%)\n221 (16.6%) 1,111(83.4%)\n\nit was $12.9 for the Gene Xpert method (ranging from $12.69 to $13.22). Consumables accounted for 41.3% of AFB smear examination costs while the Xpert cartridge cost was the major determinants of Gene Xpert\u2019s unit cost. The staff salary costs accounted for 29% of the Gene Xpert technique whereas it was 5.4% ($0.9 per test) for AFB smear microscopy (Table 2).\n\nCost per TB case detected\nIn AFB smear microscopy techniques, the unit cost per TB case detected was $55.8. The TB staining supplies accounted for 41.1% of the AFB smear examination costs whereas around $16.6 (29.7%) was attributable to overhead and equipment costs. Besides, the cost per TB case detected for the Gene Xpert testing technique was $77.9. The medical supplies (i.e., cartridge kit) cost accounted for 82.5% of the Gene Xpert costs. The ICER for the Gene Xpert techniques compared to the AFB smear microscopy method was $20.0 per TB case detected (Tables 3 and 4).\n\nScenario analysis of the costs\nThe reduction of the capital equipment\u2019s shelf life time from 10 to five years drives the unit cost of AFB smear examination to $3.60. Likewise, the unit cost would reduce by 16.13% as the\n\nTable 2. Unit cost per patient tested of smear microscopy and the Gene Xpert diagnostic method (2018 US$).\n\nHealth facilities Smear microscopy Bokoji HSP Bokoji HC Kersa HSP Sagure HC Gobesa HSP Meraro HC Gene Xpert Asella HSP Didea HSP\n\nConsumable 1.3\n375.1 241.0 214.1 312.3 272.0 285.6 10.7 8,003.0 6,204.2\n\nOverhead and space 0.4\n106.7 59.3 103.3 59.4 103.3 58.2 0.2 130.2 102.1\n\nEquipment 0.7\n210.7 64.6 151.9 67.9 188.0 66.8 1.3 882.2 879.1\n\nHR 0.9 265.5 190.3 170.2 189.9 192.3 183.6 0.7 543.4 466.1\n\nNote: HR = human resources; HC = health center; HSP = hospital.\n\nhttps://doi.org/10.1371/journal.pone.0259056.t002\n\nTested (annual)\n343 167 129 262 209 222\n753 579\n\nUnit cost 3.1 2.8 3.3 5.0 2.4 3.6 2.7 12.9 12.7 13.2\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0259056 October 25, 2021\n\n5 / 10\n\nPLOS ONE\n\nCost and cost-effectiveness of Gene Xpert and smear microscopy for the diagnosis of tuberculosis in Arsi zone\n\nTable 3. Cost per case detected of smear microscopy and Gene Xpert diagnostic methods by health facilities in Arsi Zone, Ethiopia, 2019 (2018 US$).\n\nHealth facilities Smear microscopy Bokoji HSP Bokoji HC Kersa HSP Sagure HC Gobesa HSP Meraro HC GeneXpert Asella HSP Didea HSP\n\nSupplies\n\nOverhead and space\n\nEquipment\n\nHR\n\nNumber of cases detected\n\nCost per case detected\n\n23.0\n\n6.6\n\n10.0\n\n16.1\n\n74\n\n55.8\n\n18.7\n\n5.3\n\n10.5\n\n13.3\n\n20\n\n47.9\n\n24.1\n\n5.9\n\n6.5\n\n19.0\n\n10\n\n55.5\n\n21.4\n\n10.3\n\n15.2\n\n17.0\n\n10\n\n63.9\n\n26.0\n\n4.9\n\n5.6\n\n15.8\n\n12\n\n52.5\n\n17.0\n\n6.4\n\n11.7\n\n12.0\n\n16\n\n47.2\n\n47.6\n\n9.7\n\n11.1\n\n30.6\n\n6\n\n99.0\n\n64.3\n\n1.0\n\n8.0\n\n4.6\n\n221\n\n77.9\n\n68.4\n\n1.1\n\n7.5\n\n4.6\n\n117\n\n81.7\n\n59.7\n\n1.0\n\n8.4\n\n4.5\n\n104\n\n73.6\n\nNote: HR = human resources; HC = health center; HSP = hospital.\n\nhttps://doi.org/10.1371/journal.pone.0259056.t003\n\nvolume of tests performed increased from 5 to 10 per day. Besides this, the full utilization of microscopy used to the AFB smear examination at health centers would increase the test\u2019s cost by 7%.\nIn the Gene Xpert technique, the reduction of the cartridge price by 10% reduces the test\u2019s cost to $11.7. In addition, the increment of the number of tests executed per day from eight to 16 reduces the test\u2019s costs by 8.1%. However, the cost per test would increase to $13.8 if the life span of the four-module Gene Xpert machine reduced from ten to five years (Table 5).\n\nDiscussion\nTuberculosis is a global public health threat especially in sub-Saharan African countries [1]. The introduction of Gene Xpert technology has offered the potential to increases the case detection rate of TB and MDR-TB in the health service delivery of low and middle-income countries [23]. This study evaluated the cost and cost-effectiveness of Gene Xpert and AFB smear microscopy in TB diagnosis. In our study, the use of the Gene Xpert algorithm is a costeffective intervention from the health provider\u2019s perspective with an ACER of $78. This is consistent with a recent systematic review which suggests that using Gene Xpert in a routine health care delivery system is cost-effective in various settings [11]. Our findings are also in line with a study from South African that indicated using Gene Xpert for routine TB diagnosis was a cost-saving method [24].\nIn this study, the cost per test for Gene Xpert diagnostic method ($12.9) was substantially lower than AFB smear microscopy techniques ($3.1). In our cost-scenario analysis, the volume of tests is an essential factor influencing the unit cost of the diagnostic methods. As the number of tests per day increases, the unit cost per test would decrease for both smear microscopy and Gene Xpert. Our finding is consistent with studies from Sub-Saharan African countries that\n\nTable 4. Incremental cost-effectiveness ratios of GeneXpert compared to smear microscopy.\n\nDiagnostic methods Smear microscopy GeneXpert\n\nCost per case detected 55.8 77.8\n\nIncremental cost per case detected Ref. 20.0\n\nRef: Reference strategy (smear microscopy was the reference strategy). Effectiveness was measured in terms of TB cases detected.\n\nhttps://doi.org/10.1371/journal.pone.0259056.t004\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0259056 October 25, 2021\n\n6 / 10\n\nPLOS ONE\n\nCost and cost-effectiveness of Gene Xpert and smear microscopy for the diagnosis of tuberculosis in Arsi zone\n\nTable 5. Scenario analysis of cost assumptions for smear microscopy and Gene Xpert diagnostic methods (2018US$).\n\nParameter Smear microscopy Base-case unit cost Reduce useful life of capital equipment from 10 to 5 years Increase the number of tests to 10 per day Allocate 100% of microscopy to smear microscopy at the health center GeneXpert Base-case unit cost Reduce the useful life of capital equipment from 10 to 5 years Increase the number of tests to 16 per day Reduce the price of the cartridge by 10%\n\nCost per test\n3.1 3.6 2.6 3.3\n12.9 13.8 11.9 11.7\n\nChange (%)\n+15.8 -16.1 +7.1\n+7.0 -8.1 -9.8\n\nhttps://doi.org/10.1371/journal.pone.0259056.t005\n\nCost per case detected\n55.8 64.6 46.7 59.7\n77.9 83.4 71.6 70.3\n\nChange (%)\n+15.7 -16.4 +6.9\n+7.1 -8.1 -9.7\n\nreported a similar trend, and the primary explanation for the discrepancy in unit cost across health facilities by test volume was that facilities incurred the fixed costs (i.e., equipment, HR, etc.) regardless of the number of tests performed [4,25\u201329].\nEven though we found that Gene Xpert is a cost-effective diagnostic method, its unit cost per test is higher than the AFB smear examination technique. The relatively high cost of Gene Xpert is attributable to the high prices of the equipment and cartridge. Moreover, in our study, the highest share of the unit cost of the Gene Xpert method is attributable to supplies (42%) especially cartridge costs. This estimate was in line with reports from South African and Ugandan studies\u2019 which indicated that the cartridge\u2019s cost accounted for the largest share of the Gene Xpert technique\u2019s unit cost [20,30]. Other studies\u2019 findings from settings in low- and middle-income countries consistently reported the same trends that Xpert cartridge costs are critical drivers of the overall cost [31\u201333]. Thus, the high cost of cartridges can be a barrier in the scale-up of Gene Xpert as a routine TB testing intervention. Therefore, to ensure this technology\u2019s financial sustainability in low-income settings, reducing the cartridge and equipment price is essential [34].\nTo the best of our knowledge, this is the first study that evaluated the cost and cost-effectiveness of Gene Xpert compared to smear microscopy in Ethiopian settings. However, this study has few drawbacks. First, the maintenance cost for the diagnostic machines was not included in this analysis. Excluding the cost of maintenance might significantly affect the total unit cost of the diagnostic methods [35,36]. Furthermore, although the information on the diagnosis accuracy of the techniques in TB-HIV co-infection cases would be relevant, because of the data limitation, TB-HIV co-infection was not included. Additionally, we used an intermediate outcome in this study (i.e., the ICER is presented in cost per TB case detected). Thus, the use of 1-times or 3-times gross domestic product per capita per DALY averted as a WTP threshold to determine the strategies\u2019 cost-effectiveness may not be directly applicable to our study.\n\nConclusion\nIn conclusion, using the Gene Xpert diagnostic method in routine TB management compared to smear microscopy was cost-effective. In the cost scenario analysis, the price of the cartridge, the number of tests performed per day, and the life span of the capital equipment were the drivers of the unit cost of Gene Xpert techniques. Therefore, Gene Xpert can be considered as a part of the routine TB diagnostic testing strategy in Ethiopia.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0259056 October 25, 2021\n\n7 / 10\n\nPLOS ONE\n\nCost and cost-effectiveness of Gene Xpert and smear microscopy for the diagnosis of tuberculosis in Arsi zone\nSupporting information\nS1 Dataset. (RAR)\nAuthor Contributions\nConceptualization: Abdene Weya Kaso, Alemayehu Hailu.\nFormal analysis: Abdene Weya Kaso, Alemayehu Hailu.\nInvestigation: Abdene Weya Kaso, Alemayehu Hailu.\nMethodology: Abdene Weya Kaso, Alemayehu Hailu.\nSoftware: Abdene Weya Kaso, Alemayehu Hailu.\nSupervision: Abdene Weya Kaso.\nValidation: Abdene Weya Kaso, Alemayehu Hailu.\nVisualization: Abdene Weya Kaso, Alemayehu Hailu.\nWriting \u2013 original draft: Abdene Weya Kaso, Alemayehu Hailu.\nWriting \u2013 review & editing: Abdene Weya Kaso, Alemayehu Hailu.\nReferences\n1. WHO. Global tuberculosis report. 2018. 2. Kyu HH, Maddison ER, Henry NJ, Mumford JE, Barber R, Shields C, et al. The global burden of tubercu-\nlosis: results from the Global Burden of Disease Study 2015. The Lancet Infectious Diseases. 2018; 18 (3):261\u201384. https://doi.org/10.1016/S1473-3099(17)30703-X PMID: 29223583 3. FMOH. Report on National TB/HIV Sentinel Surveillance. Addis Ababa, Ethiopia: Ethiopian Public Health Institute, December 2015. 4. Dobler CC. Screening strategies for active tuberculosis: focus on cost-effectiveness. Clinicoecon Outcomes Res. 2016; 8:335\u201347. Epub 2016/07/16. https://doi.org/10.2147/CEOR.S92244 PMID: 27418848; PubMed Central PMCID: PMC4934456. 5. Brodie D, Schluger NW. The diagnosis of tuberculosis. Clin Chest Med. 2005; 26(2):247\u201371, vi. Epub 2005/04/20. https://doi.org/10.1016/j.ccm.2005.02.012 PMID: 15837109. 6. Detjen AK, DiNardo AR, Leyden J, Steingart KR, Menzies D, Schiller I, et al. Xpert MTB/RIF assay for the diagnosis of pulmonary tuberculosis in children: a systematic review and meta-analysis. Lancet Respiratory Medicine. 2015. 7. Churchyard GJ, Stevens WS, Mametja LD, McCarthy KM, Chihota V, Nicol MP, et al. Xpert MTB/RIF versus sputum microscopy as the initial diagnostic test for tuberculosis: a cluster-randomized trial embedded in South African roll-out of Xpert MTB/RIF. Lancet Global Health. 2015;3. 8. Piatek AS, Cleeff MV, Alexander H, Coggin WL, Rehr M, Kampen SV, et al. GeneXpert for TB diagnosis: planned and purposeful implementation. Global Health Science and Practice 2013; 1(1). https://doi. org/10.9745/GHSP-D-12-00004 PMID: 25276513 9. Al-Darraji HA, Abd Razak H, Ng KP, Altice FL, Kamarulzaman A. The diagnostic performance of a single GeneXpert MTB/RIF assay in an intensified tuberculosis case finding survey among HIV-infected prisoners in Malaysia. PLoS One. 2013; 8(9):e73717. Epub 2013/09/17. https://doi.org/10.1371/journal. pone.0073717 PMID: 24040038; PubMed Central PMCID: PMC3767617. 10. WHO. Xpert MTB/RIF assay for the diagnosis of TB. Geneva, Switzerland: 2016. 11. Sagili KD, Muniyandi M, Nilgiriwala KS, Shringarpure KS, Satyanarayana S, Kirubakaran R, et al. Costeffectiveness of GeneXpert and LED-FM for diagnosis of pulmonary tuberculosis: A systematic review. PLoS One. 2018; 13(10):e0205233. Epub 2018/10/30. https://doi.org/10.1371/journal.pone.0205233 PMID: 30372436; PubMed Central PMCID: PMC6205591. 12. Tre\u00b4 bucq A, Enarson DA, Chiang CY, Deun AV, Harries AD, Boillot F, et al. Xpert\u00ae MTB/RIF for national tuberculosis programmes in low-income countries: when, where and how? International journal of tuberculosis and lung disease. 2011; 15(12):1567\u201371.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0259056 October 25, 2021\n\n8 / 10\n\nPLOS ONE\n\nCost and cost-effectiveness of Gene Xpert and smear microscopy for the diagnosis of tuberculosis in Arsi zone\n13. WHO. Automated real-time nucleic acid amplification technology for rapid and simultaneous detection of tuberculosis and rifampicin resistance: Xpert MTB/RIF assay for the diagnosis of pulmonary and extrapulmonary TB in adults and children. 2013.\n14. Vassall A, van Kampen S, Sohn H, Michael JS, John KR, den Boon S, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: a cost-effectiveness analysis. PLoS Med. 2011; 8(11):e1001120. Epub 2011/11/17. https://doi.org/10.1371/journal.pmed.1001120 PMID: 22087078; PubMed Central PMCID: PMC3210757.\n15. Ethiopia FMoHo. National strategic plan for Tuberclosis and Leprosy prevention and control (2013/14\u2013 2020/21). Addis Ababa: Ministry of Health of Ethiopia, 2017.\n16. FMOH. National policy updates summary on six edition of National TB, TB/HIV and Leprosy guidelines. Addis Ababa2016.\n17. FMOH. AFB Smear Microscopy External Quality Assessment Guideline. second ed. Addis Ababa, Ethiopia: Ethiopian Public Health Institute; January 2018.\n18. partnership/GDF ST. Ensuring an interrupted supply of quality assured, affordable tuberculosis medicines and diagnostic to the world. Geneva, Switzerland. July, 2018.\n19. Diagnostics F. Negotiated Product Pricing 21, March, 2019. Available from: https://www.finddx.org/findnegotiated-product-pricing/.\n20. Van Rie A, Page-Shipp L, Hanrahan CF, Schnippel K, Dansey H, Bassett J, et al. Point-of-care Xpert (R) MTB/RIF for smear-negative tuberculosis suspects at a primary care clinic in South Africa. Int J Tuberc Lung Dis. 2013; 17(3):368\u201372. Epub 2013/02/15. https://doi.org/10.5588/ijtld.12.0392 PMID: 23407225; PubMed Central PMCID: PMC4445423.\n21. Yakhelef Nadia, Audibert Martine, Varaine Francis, Chakaya Jeremiah, Sitienei J. Is introducing rapid culture in the diagnostic algorithm of smear-negative tuberculosis cost-effective? halshs-00866530. 2013;17.\n22. Ethiopia: Macroeconomic and Social Indicators (2018\u20132019) [Internet]. 2020. Available from: https:// nbebank.com/wp-content/uploads/pdf/annualbulletin/report-2018-2019.pdf.\n23. Uddin MKM, Chowdhury MR, Ahmed S, Rahman MT, Khatun R, Leth FV, et al. Comparison of direct versus concentrated smear microscopy in the detection of pulmonary tuberculosis. BMC Research Notes 2013; 6(291). https://doi.org/10.1186/1756-0500-6-291 PMID: 23885922\n24. Jha S, Ismail N, Clark D, Lewis JJ, Omar S, Dreyer A, et al. Cost-Effectiveness of Automated Digital Microscopy for Diagnosis of Active Tuberculosis. PLoS One. 2016; 11(6):e0157554. Epub 2016/06/21. https://doi.org/10.1371/journal.pone.0157554 PMID: 27322162; PubMed Central PMCID: PMC4913947.\n25. Adelman MW, McFarland DA, Tsegaye M, Aseffa A, Kempker RR, HM B. Cost-effectiveness of WHORecommended Algorithms for TB Case Finding at Ethiopian HIV Clinics. Infectious disease society of America. 2017. https://doi.org/10.1093/ofid/ofx269 PMID: 29399596\n26. Andrews JR, Lawn SD, Rusu C, Wood R, Noubary F, Bender MA, et al. The cost-effectiveness of routine tuberculosis screening with Xpert MTB/RIF prior to initiation of antiretroviral therapy: a modelbased analysis. AIDS. 2012; 26(8):987\u201395. Epub 2012/02/16. https://doi.org/10.1097/QAD. 0b013e3283522d47 PMID: 22333751; PubMed Central PMCID: PMC3517815.\n27. Choi HW, Miele K, Dowdy D, M S. Cost-effectiveness of Xpert\u00ae MTB/RIF for diagnosing pulmonary tuberculosis in the United States. International Journal of Tuberculosis and Lung Disease. 2013; 17 (10):1328\u201335. https://doi.org/10.5588/ijtld.13.0095 PMID: 24025386\n28. Zwerling AA, Sahu M, Ngwira LG, Khundi M, Harawa T, Corbett EL, et al. Screening for Tuberculosis Among Adults Newly Diagnosed With HIV in Sub-Saharan Africa: A Cost-Effectiveness Analysis. J Acquir Immune Defic Syndr. 2015; 70(1):83\u201390. Epub 2015/06/08. https://doi.org/10.1097/QAI. 0000000000000712 PMID: 26049281; PubMed Central PMCID: PMC4556591.\n29. Hsiang E, Little KM, Haguma P, Hanrahan CF, Katamba A, Cattamanchi A, et al. Higher cost of implementing Xpert((R)) MTB/RIF in Ugandan peripheral settings: implications for cost-effectiveness. Int J Tuberc Lung Dis. 2016; 20(9):1212\u20138. Epub 2016/08/12. https://doi.org/10.5588/ijtld.16.0200 PMID: 27510248; PubMed Central PMCID: PMC5018405.\n30. Walusimbi S, Kwesiga B, Rodrigues R, Haile M, de Costa A, Bogg L, et al. Cost-effectiveness analysis of microscopic observation drug susceptibility test versus Xpert MTB/Rif test for diagnosis of pulmonary tuberculosis in HIV patients in Uganda. BMC Health Serv Res. 2016; 16(1):563. Epub 2016/10/12. https://doi.org/10.1186/s12913-016-1804-9 PMID: 27724908; PubMed Central PMCID: PMC5057383.\n31. Pinto M, Entringer AP, Steffen R, A T. Cost analysis of nucleic acid amplification for diagnosing pulmonary tuberculosis, within the context of the Brazilian Unified Health Care System. Journal of Brasil Pneumol. 2015; 41(6):536\u20138. https://doi.org/10.1590/S1806-37562015000004524 PMID: 26785963\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0259056 October 25, 2021\n\n9 / 10\n\nPLOS ONE\n\nCost and cost-effectiveness of Gene Xpert and smear microscopy for the diagnosis of tuberculosis in Arsi zone\n32. Rupert S, Vassall A, Raizada N, Khaparde SD, Boehme C, Salhotra VS, et al. Bottom-up or top-down: unit cost estimation of tuberculosis diagnostic tests in India. Int J Tuberc Lung Dis. 2017; 21(4):375\u201380. Epub 2017/03/13. https://doi.org/10.5588/ijtld.16.0496 PMID: 28284251.\n33. Shah Maunank, Chihota Violet, Coetzee Gerrit, Churchyard Gavin, Dorman SE. Comparison of laboratory costs of rapid molecular tests and conventional diagnostics for the detection of tuberculosis and drug-resistant tuberculosis in South Africa. BMC Infectious Diseases 2013; 13(352). https://doi.org/10. 1186/1471-2334-13-352 PMID: 23895665\n34. Pantoja A, Fitzpatrick C, Vassall A, Weyer K, Floyd K. Xpert MTB/RIF for diagnosis of tuberculosis and drug-resistant tuberculosis: a cost and affordability analysis. Eur Respir J. 2013; 42(3):708\u201320. Epub 2012/12/22. https://doi.org/10.1183/09031936.00147912 PMID: 23258774.\n35. Agizew T, Boyd R, Ndwapi N, Auld A, Basotli J, Nyirenda S, et al. Peripheral clinic versus centralized laboratory-based Xpert MTB/RIF performance: Experience gained from a pragmatic, stepped-wedge trial in Botswana. PLoS One. 2017; 12(8):e0183237. Epub 2017/08/18. https://doi.org/10.1371/journal. pone.0183237 PMID: 28817643; PubMed Central PMCID: PMC5560557.\n36. Kebede A, Beyene D, Yenew B, Diriba G, Mehamd Z, Alemu A, et al. Monitoring quality indicators for the Xpert MTB/RIF molecular assay in Ethiopia. PLoS One. 2019; 14(11):e0225205. Epub 2019/11/13. https://doi.org/10.1371/journal.pone.0225205 PMID: 31714934; PubMed Central PMCID: PMC6850546.\n\nPLOS ONE | https://doi.org/10.1371/journal.pone.0259056 October 25, 2021\n\n10 / 10\n\n\n",
    "authors": [
      "Abdene Weya Kaso",
      "Alemayehu Hailu"
    ],
    "doi": "10.1371/journal.pone.0259056",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://dx.plos.org/10.1371/journal.pone.0259056"
  },
  {
    "key": "JFUV7RU9",
    "title": "Estimating the Cost of Point-of-Care Early Infant Diagnosis in a Program Setting: A Case Study Using Abbott m-PIMA and Cepheid GeneXpert IV in Zimbabwe",
    "abstract": "Background: Point-of-care early infant diagnosis (POC EID) increases access to HIV test results and shortens time to resultreturn and antiretroviral therapy initiation, as compared to central laboratory-based EID. However, to scale-up POC EID, governments need more information about programmatic costs.\nMethods: We evaluated POC EID costs from a health systems perspective. Our primary analysis assessed the Abbott m-PIMA and 2 versions of the Cepheid GeneXpert IV platforms\u2014with a solar battery or gel battery\u2014used in Zimbabwe, with instrument purchase. We also included the following 2 scenarios with zero upfront equipment purchase: (1) m-PIMA using a reagent rental model, with an all-inclusive price when the buyer commits to an average testing volume, and (2) GeneXpert IV, re\ufb02ecting contexts where GeneXpert is already in place for tuberculosis diagnosis or HIV viral load monitoring. We collected data from project expenditures, observations of health workers, and from government salary scales. We calculated cost per EID test based on number of EID tests performed on each machine per day.\nResults: The cost per successfully completed test was $44.55 for m-PIMA with platform purchase and $25.89 for m-PIMA reagent rental. Costs for GeneXpert IV with platform purchase were $25.70 using a solar battery, $25.29 using a gel battery, and $23.85 under a scenario assuming no equipment costs. In our primary analyses, materials costs comprised 73%\u201374% total costs, equipment 14%\u201320%, labor 5%\u20138%, training 1%, facility upgrades 1%, and monitoring 1%.\nConclusions: As countries consider scaling up POC EID, these data are important for budgeting and planning.",
    "full_text": "SUPPLEMENT ARTICLE\n\nDownloaded from http://journals.lww.com/jaids by BhDMf5ePHKbH4TTImqenVFpRRqarA4WBvXcKhjHvaWP9PSJBaJzk0cFimzvFKqug on 06/27/2020\n\nEstimating the Cost of Point-of-Care Early Infant Diagnosis in a Program Setting: A Case Study Using Abbott m-PIMA\nand Cepheid GeneXpert IV in Zimbabwe\nSushant Mukherjee, MA, MBA,a Jennifer Cohn, MD, MPH,b Andrea L. Ciaranello, MD, MPH,c,d Emma Sacks, PhD,a Oluwarantimi Adetunji, MS,a Addmore Chadambuka, MPH,e\nHaurovi Mafaune, MPH,e McMillan Makayi, MBA,e Nicole McCann, BA,c,d and Esther Turunga, MBAb\n\nBackground: Point-of-care early infant diagnosis (POC EID) increases access to HIV test results and shortens time to resultreturn and antiretroviral therapy initiation, as compared to central laboratory-based EID. However, to scale-up POC EID, governments need more information about programmatic costs.\nMethods: We evaluated POC EID costs from a health systems perspective. Our primary analysis assessed the Abbott m-PIMA and 2 versions of the Cepheid GeneXpert IV platforms\u2014with a solar battery or gel battery\u2014used in Zimbabwe, with instrument purchase. We also included the following 2 scenarios with zero upfront equipment purchase: (1) m-PIMA using a reagent rental model, with an all-inclusive price when the buyer commits to an average testing volume, and (2) GeneXpert IV, re\ufb02ecting contexts where GeneXpert is already in place for tuberculosis diagnosis or HIV viral load monitoring. We collected data from project expenditures, observations of health workers, and from government salary scales. We calculated cost per EID test based on number of EID tests performed on each machine per day.\nResults: The cost per successfully completed test was $44.55 for m-PIMA with platform purchase and $25.89 for m-PIMA reagent rental. Costs for GeneXpert IV with platform purchase were $25.70 using a solar battery, $25.29 using a gel battery, and $23.85 under a scenario assuming no equipment costs. In our primary analyses, materials costs comprised 73%\u201374% total costs, equipment 14%\u201320%, labor 5%\u20138%, training 1%, facility upgrades 1%, and monitoring 1%.\nConclusions: As countries consider scaling up POC EID, these data are important for budgeting and planning.\nReceived for publication March 10, 2020; accepted March 23, 2020. From the aElizabeth Glaser Pediatric AIDS Foundation, Washington DC;\nbElizabeth Glaser Pediatric AIDS Foundation, Geneva, Switzerland; cDivision of General Internal Medicine, Massachusetts General Hospital, Boston, MA; dDepartment of Medicine, Medical Practice Evaluation Center, Massachusetts General Hospital, Boston, MA; and eElizabeth Glaser Pediatric AIDS Foundation, Harare, Zimbabwe. Funded and supported by Unitaid, Geneva, Switzerland. The authors have no con\ufb02icts of interest to disclose. Correspondence to: Jennifer Cohn, MD, MPH, Elizabeth Glaser Pediatric AIDS Foundation, Geneva, Switzerland (e-mail: jcohn@pedaids.org). Copyright \u00a9 2020 Wolters Kluwer Health, Inc. All rights reserved.\n\nKey Words: early infant diagnosis, point-of-care, nucleic acid test, HIV, costing, resource utilization, Zimbabwe\n(J Acquir Immune De\ufb01c Syndr 2020;84:S63\u2013S69)\nINTRODUCTION Every year, nearly 1.4 million children are born to HIVpositive mothers, primarily in resource-limited settings.1 If infants perinatally infected with HIV are not diagnosed early and immediately initiated on treatment, as many as 50% will die by their second birthday, with peak mortality at 2 to 3 months of age.2 Yet, only 54% of children living with HIV received antiretroviral therapy (ART) in 2018, falling short of global targets.1 This shortfall can be attributed in part to the fact that early infant diagnosis (EID) uses high-throughput laboratory-based virologic assays, which require technology only available at national or provincial laboratories. Infrastructural challenges associated with transporting samples to these laboratories and returning results to health facilities means that caregivers wait several months to receive results,3 if received at all.4 Nearly half of infants tested never receive their results, and of those who test positive and receive results, only 50%\u201380% are eventually initiated on ART.5 The introduction of point-of-care (POC) infant HIV nucleic acid testing can help to solve this problem. If integrated into national EID networks, these POC assays increase the number of HIV-exposed infants diagnosed and dramatically reduce times for result-return and ART initiation, thereby decreasing infant mortality.6 POC platforms are simple, fast, do not require extensive training or infrastructure, and can be deployed in primary health care settings. A recent modeling study on POC EID in Zimbabwe shows that reduced result-return time associated with POC EID would lead to signi\ufb01cantly lower mortality. The study also found that POC EID was a cost-effective intervention for Zimbabwe with an incremental cost-effectiveness ratio of $630 per year of life saved, well below annual per capita gross domestic product.7 This analysis assumed costs for POC EID of $28 and central laboratory-based EID of $24. Although the \ufb01ndings were robust when costs were varied widely, a more detailed examination of the costs associated with this technology is necessary to facilitate budgeting and planning by policymakers. There are very few cost data\n\nJ Acquir Immune De\ufb01c Syndr   Volume 84, Supplement 1, July 1, 2020\n\nwww.jaids.com | S63\n\nCopyright \u00a9 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.\n\nMukherjee et al\n\nJ Acquir Immune De\ufb01c Syndr   Volume 84, Supplement 1, July 1, 2020\n\navailable in the published literature on POC EID. We sought to compile real-world resource utilization data to inform future cost-effectiveness analyses of POC EID in Zimbabwe and beyond.\nMETHODS A recent EID project, supported by Unitaid and implemented by the Elizabeth Glaser Pediatric AIDS Foundation (EGPAF), has expanded access to POC EID testing in 9 sub-Saharan African countries. The project used 2 types of platforms, Abbott m-PIMA and Cepheid GeneXpert IV. We evaluated comprehensive cost data for these 2 platforms for POC EID in Zimbabwe, where the project was implemented in collaboration with the Ministry of Health and Child Care.\nStudy Sites In Zimbabwe, the project uses a hub-and-spoke model\n(Fig. 1) for most sites, where lower-volume spoke sites send samples to a nearby higher-volume hub site where the POC platform is located, to optimize utilization of the platforms. We collected cost data from January\u2013June 2017 in Zimbabwe, at all 45 project sites where the project was being implemented.\nData Collection and Analytic Approach We collected data on 6 key components of POC EID\ncosts, based on published methods for costing the POC PIMA CD4 test on the PIMA Analyzer (Abbott).8 Components included materials and supplies; training; facility upgrades; site monitoring and quality assurance (QA); equipment, shipping, and freight; and labor.\nWe used a mixed methodology for collecting cost data for each component, based on the availability of cost data. For\n\ncertain components\u2014trainings and facility upgrades\u2014we used a top-down costing approach, whereby we compiled aggregated costs and then divided it by the relevant unit (eg, cost per site or cost per test). Primary sources of aggregate cost data were project \ufb01nancial records as captured in the accounting system software (Great Plains, QuickBooks) and project budgets.\nFor the other components\u2014materials, site monitoring, equipment, and labor\u2014we used a bottom-up costing approach, where we calculated costs for each input at a site level to estimate unit costs. The sources of bottom-up cost data included manufacturer contracts, data from other project studies (eg, direct observations and surveys of health workers), and other project documents, such as site visit schedules.\nAll cost data were compiled in Microsoft Excel by a central team in collaboration with Zimbabwe-based project managers and \ufb01nance team members.\nNo ethical approval was required for this costing study because we only collected secondary data from routine \ufb01nancial records and other secondary sources.\nUnit Costs and Cost Components To calculate unit costs, we calculated cost per test\nconducted, incorporating both EID throughput and overall utilization. We de\ufb01ned EID throughput as number of EID tests run per day on a machine, and EID utilization as the number of expected EID and other tests to be run on a machine per day (based on observed demand), divided by the machine\u2019s actual daily capacity. Utilization is expressed as a percentage, and it differs from throughput in that it includes the ability of the platforms to run tuberculosis (TB) and viral load (VL) assays in addition to EID. Because the GeneXpert can run TB and VL assays, whereas the m-PIMA was recently approved to run VL, it is appropriate to factor in total number of tests run on the machines to calculate equipment costs per test. We created the following 3 scenarios that combine observed and plausible future patterns of both throughput and utilization: base case, medium case, and high case. Importantly, variation in utilization only affects costs for equipment and not for other cost domains.\n\nFIGURE 1. Hub-and-spoke model.\nS64 | www.jaids.com\n\nMaterials and Supplies Required materials included gloves, single-use test\ncartridge (HIV 1/2 Detect for m-PIMA and HIV-1 Qual for GeneXpert), capillary blood collection tube, plastic transfer capillary, sample collection kit (gauze, lancet, swab, and plaster), and sealable plastic bag. In addition, cooler boxes with ice packs as well as envelopes used to transport samples and laboratory forms between hub and spoke have been included, based on proportion of samples collected at spoke sites. For waste management, we included additional costs for high temperature incineration of the GeneXpert cartridges (containing guanidinium thiocyanate) and transport of waste to an incinerator (no additional waste management costs were included for the m-PIMA HIV 1/2 Detect because it can be disposed of in a manner similar to other common medical products).\nCopyright \u00a9 2020 Wolters Kluwer Health, Inc. All rights reserved.\n\nCopyright \u00a9 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.\n\nJ Acquir Immune De\ufb01c Syndr   Volume 84, Supplement 1, July 1, 2020\n\nEarly Infant Diagnosis in a Program Setting\n\nTraining End-user training costs were already included as part of\nprocurement contracts. However, further training was required, relating to capturing data on EID forms, post hoc training of additional operators, and Training of Trainers (ToTs) for provincial laboratory managers to act as master trainers responsible for training new staff in the event of turnover. We divided total costs by 45 active sites, and then further divided this by 3 EID throughput scenarios, to estimate training cost per test. We assumed that over a 5year period, costs were incurred upfront and that further training needs would be met by master trainers during routine monitoring. To ensure trainings were resourced adequately, we assumed the ToT would be repeated annually.\nFacility Upgrades and Repairs Major infrastructural upgrades are not typically\nrequired to introduce POC EID, so costs involved minor upgrades, including the purchase of air conditioners where not already in place (especially relevant for GeneXpert IV, which requires a temperature of 30\u00b0C or lower to operate). These upgrades would only be required for testing sites and thus were only included for hub and standalone sites. Costs varied widely, based on the speci\ufb01c needs of each site. We compiled total costs for hub and standalone sites and applied them across all sites (because samples from spoke sites also rely on adequate POC EID infrastructure at hubs).\nSite Monitoring, Supervision, and QA Standalone testing sites and hubs were visited 2 and 6\nweeks after installation of POC EID and thereafter quarterly (spoke sites did not receive additional supervision, over and above routine supervision). We assumed Zimbabwe\u2019s Ministry of Health and Child Care would adopt a similar protocol once ownership was transferred. We further assumed that one year after installation, monitoring and QA would be integrated into routine supervision by District Health Management Teams, no longer requiring additional costs speci\ufb01cally for POC EID. We calculated costs for a site monitoring visit, based on average duration, salary and position of the individual conducting supervision (typically a district supervisor accompanied by a driver), average distance between district capitals and sites, as well as market prices for fuel. We calculated total site monitoring and QA costs over a 5-year window (where only the \ufb01rst year involved actual costs). Pro\ufb01ciency testing for QA (sending samples to a central laboratory for further testing) will likely be required in a POC context, but at the time of undertaking this analysis, it was unclear what the protocol for this would be, so those costs were excluded from this analysis; it is anticipated that they would comprise an extremely minor component of costs.\nEquipment, Shipping, and Freight We assumed that extended warranties would be pur-\nchased. For GeneXpert, we assumed 2 options. Option 1 uses an external gel battery that requires electricity, whereas option 2 uses a solar battery that would allow use in settings with limited access to electricity (the m-PIMA has a built-in battery that does not require a stable power supply). We included purchase costs for the platforms, as well as costs for\nCopyright \u00a9 2020 Wolters Kluwer Health, Inc. All rights reserved.\n\nmaintenance, related accessories, and freight. We calculated the per day cost of equipment under the appropriate discount rates and working life of the equipment.\nLabor In a separate evaluation, we conducted a time-use\nobservational study to evaluate the cost of labor by health workers.9 Data collectors conducted 15 observations of POC EID processes (9 using GeneXpert and 6 using m-PIMA) and administered surveys of health workers. We calculated average time spent by health workers to implement POC EID and converted this into dollars based on current salary information for each cadre of health worker. We also evaluated labor costs using upper and lower bounds of the 95% con\ufb01dence intervals of health worker time for total activity, multiplied by the proportion of time spent on each activity, which we de\ufb01ned as Minimal Time and Labor Intensive.\nError Rates Cost per test calculations also factored in error rates or\ninternal quality control failures for both platforms. These were taken directly from EGPAF project data and de\ufb01ned as number of internal quality control failures over total number of tests run.\nScenarios and Assumptions To calculate costs on a per test basis, we created 3\nthroughput and utilization cases (Table 1). Our base case assumed an EID throughput of 1.5 tests per site per day, based on observed POC EID throughput in Zimbabwe from October 2017 to September 2018. We evaluated the following 2 additional cases with higher throughput: medium throughput (2.5 tests/day) and high throughput (3.5 tests/day). These re\ufb02ect increases in EID demand that may arise from testing at additional ages (eg, birth or at age 9 months).10\nWe de\ufb01ned the following 3 levels of EID utilization (proportion of maximum test capacity per day used for EID): 60% (base case), 75% (medium), and 90% (high). The maximum capacity of GeneXpert IV is 20 tests per day (run time of 90 minutes and the ability to run 4 tests concurrently) and 8 tests per day for m-PIMA (52 minutes run time and no option to run tests concurrently).11 However, once we factor in hands-on health worker time of 27 minutes per EID test across both platforms9 and an 8-hour work day, a more realistic optimal capacity is 16 tests for GeneXpert IV and 6 tests per day for the m-PIMA. A study of GeneXpert in Zimbabwe estimated that 9% of tests run on the machine were EID, compared with VL (41%) and TB (50%), suggesting that GeneXpert and m-PIMA have the potential to reach near-optimal utilization once they are running non-EID assays in addition to EID.12\n\nTABLE 1. Throughput and Utilization Cases\n\nBase Case Medium Case\n\nEID throughput\n\n1.5 tests/d\n\nEquipment utilization (%)\n\n60%\n\n(includes EID, VL, and TB)\n\n2.5 tests/d 75%\n\nHigh Case 3.5 tests/d\n90%\n\nwww.jaids.com | S65\n\nCopyright \u00a9 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.\n\nMukherjee et al\n\nJ Acquir Immune De\ufb01c Syndr   Volume 84, Supplement 1, July 1, 2020\n\nTABLE 2. Total Costs of POC EID (Base Case)\nItem Cartridges Capillary tubes (blood collection) Transfer capillary tubes (blood transfer) Sample collection kit (gloves, lancet, swab, gauze, plaster, etc) Plastic ziploc bags, envelopes, and cooler boxes Waste management (incineration and waste transport) Freight (insurance and customs clearance) Storage and distribution Total\n\nm-PIMA\n$25.00 $0.16 $0.18 $0.48 $0.15 N/A $1.74 $2.86 $30.56\n\nMaterials and Supplies Per Assay\nm-PIMA Reagent Rental\n$20.00 $0.16 $0.18 $0.48 $0.15 N/A $0.52 N/A $21.49\n\nGeneXpert IV\n$14.90 $0.16 $0.18 $0.48 $0.15 $0.40 $1.06 $1.79 $19.12\n\nVenue, participants, and travel ToTs Cost per site Training cost per test (5 yrs)\n\nTraining\n$16,437 $5910 $497 $0.27\n\nFacility cost per site (5 yrs) Facility cost per test (5 yrs)\n\nFacility Upgrades and Repairs $404 $0.22\n\nLabor Per diem (meals and incidentals) per site Vehicle fuel per site Total per visit (per site) Monitoring and supervision cost per test (5 yrs)\n\nSite Monitoring and Supervision\n$39.20 $30\n$35.50 $104.70 $0.34\n\nPlatform Type\nPlatform price Extended warranty cost Gel battery Solar battery Software connectivity per machine Freight (insurance and customs clearance) per machine Storage and distribution per machine Total per machine Present value (discounted at 6%) Equipment cost per day Equipment cost per test (including EID, VL, and TB\nassays)\n\nm-PIMA\n$25,000 $9000 N/A N/A N/A $1675 $2750 $38,425 $37,221 $30.26 $8.41\n\nEquipment, Shipping, and Freight\n\nm-PIMA Reagent Rental GeneXpert IV Solar\n\nN/A\n\n$17,500\n\nN/A\n\n$6840\n\nN/A\n\nN/A\n\nN/A\n\n$9714\n\nN/A\n\n$5,50015\n\nN/A\n\n$1823\n\nN/A\n\n$1925\n\n$43,302\n\nN/A\n\n$42,212\n\nN/A\n\n$34.32\n\nN/A\n\n$3.57\n\nGeneXpert IV Gel\n$17,500 $6840 $5222 N/A $5500 $1522 $1925 $38,509 $37,418 $30.42 $3.17\n\nPlatform Type Staff cost per test\n\nm-PIMA $1.95\n\nLabor\n\nGeneXpert IV $2.18\n\nPlatform type\nTotal cost per test Error or failure rate Total cost per valid test\n\nm-PIMA\n$41.75 6.7%\n$44.55\n\nm-PIMA Reagent Rental\n$24.27 6.7%\n$25.89\n\nTotal Costs\u2013Base Case\n\nGeneXpert IV Solar GeneXpert IV Gel\n\n$25.70 7.8%\n$27.70\n\n$25.29 7.8%\n$27.27\n\nGeneXpert IV (No Equipment)\n$22.13 7.8%\n$23.85\n\nS66 | www.jaids.com\n\nCopyright \u00a9 2020 Wolters Kluwer Health, Inc. All rights reserved.\n\nCopyright \u00a9 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.\n\nJ Acquir Immune De\ufb01c Syndr   Volume 84, Supplement 1, July 1, 2020\n\nEarly Infant Diagnosis in a Program Setting\n\nThe 60% utilization in the base case is derived from actual current utilization data of the GeneXpert in Zimbabwe (where most tests being run are for TB); we have applied this base case utilization to m-PIMA as well, even though current utilization is lower because of anticipated increase in uptake of EID and the recent approval of POC VL. The 75% and 90% utilization cases assume that demand will increase once all platforms are running EID and non-EID assays.\nThe base, medium, and high throughput and utilization cases are summarized in Table 1.\nWe also evaluated a scenario with no equipment costs for GeneXpert IV, re\ufb02ecting contexts in which GeneXpert IV is already in place for TB diagnosis or HIV VL monitoring.\nFinally, we considered a scenario for a reagent rental offered by m-PIMA where a testing cartridge is purchased at $20.00, with the costs of platform, maintenance, data, and connectivity all included, if an average of 1300 tests per platform per year over 3 years can be attained (these would include VL in addition to EID).\nPerspective and Time Horizon This cost analysis assumed a health systems perspec-\ntive, incorporating all costs associated with POC EID that accrued to health care providers and facilities. Other societal costs, such as wait times or transport costs for caregivers, although critical, are not included. According to the manufacturers, the useful life of the m-PIMA and GeneXpert is 5 years with extended warranty, so we chose to analyze all costs over a 5-year window. Zimbabwe formally adopted the US dollar in 2009, so all costs are re\ufb02ected in 2017 USD.\nEquipment costs were discounted based on current lending rates and in\ufb02ation in Zimbabwe, to calculate present value of the stream of equipment costs over the course of time. A discount rate is applied to costs to capture time value of money, which is an important consideration when making a public or private investment. The prime interest rate for borrowers in Zimbabwe in 2017\u20132018 was estimated at 10%,13 and in\ufb02ation was recorded at approximately 4%.14\n\nGiven this, we used a real discount rate (nominal borrowing rate minus in\ufb02ation) of 6% for the base case, and examined rates of 0% and 10% in sensitivity analyses. We used a standard methodology applying the appropriate discount rate to calculate the present value of equipment costs.8 For example, for m-PIMA, given an annual warranty of $2250 and an upfront cost of $28,825 (cost of platform plus freight, storage, and distribution) the present value of costs over 5 years would be as follows: $28,825 + $2250 + $2250/(1.06) + $2250/(1.06)^2 + $2250/(1.06)^3 + $2250/(1.06)^4 = $36,621.\nRESULTS Total costs for POC EID using base case (based on observed project data), assumptions for EID throughput, utilization of equipment, and staff time are summarized in Table 2. Total costs per valid test, after factoring in error rates from project data, vary from $27.27 for GeneXpert with a gel battery to $44.55 for m-PIMA in our base case. Error rates can vary signi\ufb01cantly and may be higher with new or undertrained users. For the scenario where we assume no equipment costs for GeneXpert, cost per valid test is $23.85, and for the m-PIMA reagent rental model, cost per valid test is $25.89. The difference across platforms is driven by supplies and equipment. The proportional component costs for each platform is similar, with supplies and equipment comprising the vast majority. Figure 2 illustrates the breakdown for both platforms, assuming equipment is purchased and not rented. Labor costs depended on the cadre of health worker conducting the test and duration of the EID process. Salary cost per test is shown in Table 3. Total hands-on labor time for POC EID was 27 minutes. We did not incorporate test processing time for m-PIMA and GeneXpert because we observed that health workers worked on other tasks while the machine processed the test. Because the sample size was too small for the difference between m-PIMA (23 minutes) and GeneXpert (30 minutes) to reach\n\nFIGURE 2. Cost breakdown by component assuming equipment is purchased.\nCopyright \u00a9 2020 Wolters Kluwer Health, Inc. All rights reserved.\n\nwww.jaids.com | S67\n\nCopyright \u00a9 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.\n\nMukherjee et al\n\nJ Acquir Immune De\ufb01c Syndr   Volume 84, Supplement 1, July 1, 2020\n\nTABLE 3. Average Labor Cost per Test\nActivity\nStep 1: Counseling Step 2: Blood draw or specimen collection\nand labeling Step 3: Specimen transport (within facility) Step 4: Hands-on time for processing test Step 5: Recording or logging results Step 6: Result communication or counseling\nfor caregiver Total time (mins) Total staff time (h) per patient or EID test\n(minutes divided by 60)\n\nMinutes Per Patient or EID Test 5 7\n1 4 3 7\n27 0.45\n\nStaff Salaries\n\nStaff\n\nSenior\n\nStaff Cadre/Position Units Nurse Nurse/Midwife\n\nSalary (including bene\ufb01ts)\nWorking d/yr Annual leave National holidays Actual working d/yr Actual working d/mo Salary per working d Working h/d Salary per working h\n\nUSD/mo\nd/yr d/yr d/yr d/yr d/mo USD/d h/d USD/h\n\n$584\n260 42 14 204 17 $34.35 8 $4.29\n\n$669\n260 42 14 204 17 $39.35 8 $4.92\n\nLab Technician\n$864\n260 42 14 204 17 $50.82 8 $6.35\n\nGeneXpert IV Labor Cost Per Test\n\nStaff time per POC EID test Salary cost per test Total salary cost per test\n(including laboratory technician)\n\nH USD USD\n\n0.34 $1.48 $2.18\n\n0.34 $1.69 $2.39\n\n0.11 $0.70\n\nm-PIMA Labor Cost Per Test\n\nStaff time per POC EID test\n\nH\n\n0.45\n\n0.45\n\nN/A\n\nTotal salary cost per EID test\n\nUSD\n\n$1.95\n\n$2.24\n\nN/A\n\nstatistical signi\ufb01cance, we assigned average labor time of 27 minutes to both platforms. It is important to note that the GeneXpert is considered a laboratory-based platform requiring a laboratory technician, so we assumed that a laboratory technician completes steps 4 and 5, whereas a nurse may complete other steps. For m-PIMA, laboratory technicians are not required and a nurse may undertake all steps.\nFinally, Table 4 summarizes total cost per valid test under the medium and high cases, as compared to the base case.\n\nDISCUSSION\nAs countries seek to introduce and scale-up POC EID into decentralization, they will need to understand the resource implications so that they can plan appropriately. We hope that the costs compiled here, although speci\ufb01c to Zimbabwe, will shed light on cost drivers, and present a methodology that could be adapted for use elsewhere.\nOur analysis indicates that most costs relate to supplies, namely the single-use testing cartridges. These costs may be negotiated downward with volume discounts and bulk purchasing, and therefore, this is an area where stakeholders should focus most attention as they seek to scale-up POC EID. This can be achieved through increased demand, from the adoption of birth testing and repeat con\ufb01rmatory testing, and through the ability to run multiple assays on the same platform. By offering POC VL to pregnant and breastfeeding women on the same platform, volume can more than double. If POC VL is extended to wider populations, such as adolescents, those with advanced disease, or a second VL for those with an initial elevated VL, volume could increase by an order of magnitude, thereby creating possibilities for price reductions, when appropriately balanced against the need to expand capacity by adding machines.\nThe second largest contributor to costs is equipment, where increased testing volume could signi\ufb01cantly bring down per test costs. Integrated testing on m-PIMA (VL in addition to EID) is now a reality, and multiple assay testing (TB, VL, and EID) on GeneXpert is already happening in Zimbabwe.12 The impact of multiple assay testing is particularly signi\ufb01cant on GeneXpert, given the higher maximum capacity of that platform.11 When POC EID is integrated into an existing base of GeneXpert, equipment costs could be removed altogether (recognizing that these are still real costs to the health system), until current capacity is exceeded. These data emphasize that in low-resource settings, multianalyte testing (using one machine for multiple purposes) may offer considerable value for money. Reagent rental agreements, such as the one negotiated for m-PIMA, further reduce the burden of capital investment while simultaneously incentivizing suppliers to respond quickly to maintenance needs to keep machines running.\nFurther reductions in capital costs could be realized from an increased lifespan of the platforms. Our current assumption that platforms last 5 years may be conservative, as early experience suggests that 7 years may be more accurate. Again, a reagent rental agreement can play a critical role here by shifting the responsibility of placing a functioning platform to the supplier.\nGiven that materials and equipment comprise most costs, they will be the most likely source of future cost\n\nTABLE 4. Total Cost per Valid Test Under Medium and High Cases\n\nm-PIMA\n\nm-PIMA Reagent Rental\n\nGeneXpert IV Solar\n\nBase case Medium case High case\n\n$44.55 $42.40 $41.05\n\n$25.89 $25.54 $25.39\n\n$27.70 $26.40 $25.66\n\nGeneXpert IV Gel\n$27.27 $26.05 $25.37\n\nGeneXpert IV (No Equipment)\n$23.85 $23.32 $23.10\n\nS68 | www.jaids.com\n\nCopyright \u00a9 2020 Wolters Kluwer Health, Inc. All rights reserved.\n\nCopyright \u00a9 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.\n\nJ Acquir Immune De\ufb01c Syndr   Volume 84, Supplement 1, July 1, 2020\n\nEarly Infant Diagnosis in a Program Setting\n\nreductions. Other cost components\u2014training, site monitoring, facility upgrades, and labor\u2014comprised a much smaller fraction of total implementation costs. Ef\ufb01ciencies such as integrating monitoring for POC into existing site monitoring visits and use of existing staff, who perform EID activities, to support POC EID can help contain these costs.\nIt is worth noting that although we have attempted to be comprehensive, certain costs have been excluded. Sample transport costs have been excluded because it is hoped that transport will be integrated into existing networks for TB sputum samples using motorcycle riders. This is realistic, given that some POC EID specimens are already being transported using these networks. However, it is possible that integrating EID into routine TB sample transport might require addition of new pick-up locations from spoke sites not already covered by TB networks, which would result in additional costs. An internal analysis of sample transport costs during the reporting period under the hub-and-spoke model, where 85% of samples were transported by the project, and 15% using existing transport networks, yielded an average cost per sample transported of $3.33. In other words, even if existing sample transport networks for TB could not absorb POC EID, it is likely that POC EID would remain cost effective relative to central laboratory-based EID.\nAlso, the cost of adding more machines, where volume exceeds the capacity of a single platform, has not been factored in explicitly into this analysis. Although the ability of platforms to perform multiple types of tests does create potential for an upward trend in volume, for most health facilities, there is still signi\ufb01cant existing capacity, and therefore, this scenario was not included. However, if an additional machine would be required in the future, the data collected here would still shed light on the costs of such a scenario.\nFinally, an all-in cost analysis of central EID is beyond the scope of this article, whose primary purpose is to inform policymakers on the resource implications of adopting POC EID. However, costs of central laboratory-based EID are robustly estimated at $24.18 per assay16 for the 2 most commonly used central laboratory-based EID tests. This estimate is conservative because it excludes labor costs, retraining, and site monitoring, all costs that are included in our costing of POC EID. And yet, a recent modeling study of POC EID in Zimbabwe using this lower estimate for central EID, still found that POC EID cost per test would have to rise above $60 for POC to no longer be cost effective.7 In comparison, even under the highest cost scenario we considered, POC EID cost per test was estimated at $44.55,\n\nwith other scenarios well below $30. Given this benchmark, the cost data compiled here would support recent analysis indicating that POC EID is of excellent value relative to central laboratory-based EID.\nREFERENCES\n1. UNAIDS. Global HIV & AIDS Statistics\u20142019 Fact Sheet (Based on 2018 Data). Available at: https://www.unaids.org/en/resources/factsheet. Accessed October 31, 2019.\n2. Newell ML, Coovadia H, Cortina-Borja M, et al. Mortality of infected and uninfected infants born to HIV-infected mothers in Africa: a pooled analysis. Lancet. 2004;364:1236\u20131243.\n3. Hsiao NY, Stinson K, Myer L. Linkage of HIV-infected infants from diagnosis to antiretroviral therapy services across the Western Cape, South Africa. PLoS One. 2013;8:e55308.\n4. Ciaranello AL, Park JE, Ramirez-Avila L, et al. Early infant HIV-1 diagnosis programs in resource-limited settings: opportunities for improved outcomes and more cost-effective interventions. BMC Med. 2011;9:59.\n5. Dube Q, Dow A, Chawanangwa C, et al. Implementing early infant diagnosis of HIV infection at the primary care level: experiences and challenges in Malawi. Bull World Health Organ. 2012;90:699\u2013704.\n6. Bianchi F, Cohn J, Sacks E, et al. Evaluation of a routine point-of-care intervention for early infant diagnosis of HIV: an observational study in eight African countries. Lancet HIV. 2019;6:e373\u2013e381.\n7. Frank S, Cohn J, Dunning L, et al. The clinical impact and costeffectiveness of incorporating point-of-care (POC) assays into early infant HIV diagnosis (EID) programs at 6 weeks of age in Zimbabwe. Lancet HIV. 2019;5:2019.\n8. Larson B, Schnippel K, Ndibongo B, et al. How to estimate the cost of point-of-care CD4 testing in program settings: an example using the Alere Pima Analyzer in South Africa. PLoS One. 2012;7:e35444.\n9. Adetunji O, Mukherjee S, Sacks E, et al. Front line human resource time use for early infant HIV diagnosis: a comparative time-motion study at centralized and point-of-care health facilities in Zimbabwe. J Acquired Immune De\ufb01c Syndr. 2019.\n10. HIV Diagnosis and ARV Use in HIV-Exposed Infants: A Programmatic Update. Geneva, Switzerland: World Health Organization Technical Report; 2018.\n11. Guidance Note on Product Selection, Site Upgrades, and Sample Transportation. EGPAF and Unitaid. Available at: http://childrenandaids.org/sites/ default/\ufb01les/poc-toolkit/Guidance%20Note%20on%20Product%20SelectionFacility%20Upgrades%20and%20Sample%20Transportation.pdf. Accessed June 4, 2019.\n12. Ndlovu Z, Fajardo E, Mbofana E, et al. Multidisease testing for HIV and TB using the GeneXpert platform: a feasibility study in rural Zimbabwe. PLoS One. 2018;13:e0193577.\n13. Available at: https://tradingeconomics.com/zimbabwe/interest-rate. Accessed December 4, 2018.\n14. Available at: https://tradingeconomics.com/zimbabwe/in\ufb02ation-cpi. Accessed December 4, 2018.\n15. Cepheid. Pricing Sheet: 2018 Pricing for GxAlert by SystemOne. 2018. 16. Global Fund to \ufb01ght AIDS, Tuberculosis, and Malaria. HIV Viral Load and\nEarly Infant Diagnosis Selection and Procurement Tool. Available at: https:// www.theglobalfund.org/media/5765/psm_viralloadearlyinfantdiagnosis_ content_en.pdf. Accessed March 24, 2019.\n\nCopyright \u00a9 2020 Wolters Kluwer Health, Inc. All rights reserved.\n\nwww.jaids.com | S69\n\nCopyright \u00a9 2020 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.\n\n\n",
    "authors": [
      "Sushant Mukherjee",
      "Jennifer Cohn",
      "Andrea L. Ciaranello",
      "Emma Sacks",
      "Oluwarantimi Adetunji",
      "Addmore Chadambuka",
      "Haurovi Mafaune",
      "McMillan Makayi",
      "Nicole McCann",
      "Esther Turunga"
    ],
    "doi": "10.1097/QAI.0000000000002371",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://journals.lww.com/10.1097/QAI.0000000000002371"
  },
  {
    "key": "EMP3TBGD",
    "title": "Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF \u00ae",
    "abstract": "",
    "full_text": "G Model\nEIMC-1562; No. of Pages 8\n\nARTICLE IN PRESS\nEnferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\nwww.elsevier.es/eimc\nOriginal\nEstudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae\u0b1d\n\u00d3scar Herr\u00e1ez a, Mar\u00eda \u00c1ngeles Asencio-Egea b,\u2217, Mar\u00eda Huertas-Vaquero b, Rafael Carranza-Gonz\u00e1lez b, Jes\u00fas Castellanos-Monedero c, Mar\u00eda Franco-Huerta c, Jos\u00e9 Ram\u00f3n Barber\u00e1-Farr\u00e9 c y Jos\u00e9 Mar\u00eda Ten\u00edas-Burillo d\na Laboratorio de An\u00e1lisis Cl\u00ednicos, Hospital General La Mancha Centro, Alc\u00e1zar de San Juan, Ciudad Real, Espa\u02dcna b Laboratorio de Microbiolog\u00eda, Hospital General La Mancha Centro, Alc\u00e1zar de San Juan, Ciudad Real, Espa\u02dcna c Servicio de Medicina Interna, Hospital General La Mancha Centro, Alc\u00e1zar de San Juan, Ciudad Real, Espa\u02dcna d Unidad de Apoyo a la Investigaci\u00f3n, Hospital General La Mancha Centro, Alc\u00e1zar de San Juan, Ciudad Real, Espa\u02dcna\n\ninformaci\u00f3n del art\u00edculo\nHistoria del art\u00edculo: Recibido el 10 de febrero de 2016 Aceptado el 19 de junio de 2016 On-line el xxx\nPalabras clave: Tuberculosis Diagn\u00f3stico Reacci\u00f3n en cadena de la polimerasa Sensibilidad y especi\ufb01cidad An\u00e1lisis de coste-efectividad Evaluaci\u00f3n econ\u00f3mica\n\nr e s u m e n\nIntroducci\u00f3n/Objetivo: Evaluar mediante un an\u00e1lisis de coste-efectividad la aplicaci\u00f3n de una t\u00e9cnica de biolog\u00eda molecular al diagn\u00f3stico de tuberculosis frente a la alternativa diagn\u00f3stica cl\u00e1sica. M\u00e9todos: Se realiz\u00f3 un an\u00e1lisis de coste-efectividad para evaluar la aplicaci\u00f3n te\u00f3rica de un procedimiento de biolog\u00eda molecular que incluye 2 alternativas de una t\u00e9cnica para la detecci\u00f3n precoz de Mycobacterium tuberculosis Complex y resistencia a rifampicina (alternativa 1: una determinaci\u00f3n a pacientes seleccionados; alternativa 2: 2 determinaciones a todos los pacientes). Ambas alternativas se compararon con el procedimiento habitual de diagn\u00f3stico microbiol\u00f3gico de tuberculosis realizado a 1972 pacientes durante 2008-2012 (microscopia y cultivo). La medida de la efectividad se hizo en QALY y la incertidumbre se trat\u00f3 mediante an\u00e1lisis de sensibilidad univariable, multivariable y probabil\u00edstico. Resultados: Para el m\u00e9todo habitual se obtuvo un valor de 8.588 D /QALY. En la alternativa 1 el gasto fue de 8.487 D /QALY, mientras que en la alternativa 2 el cociente coste-efectivo ascendi\u00f3 a 2.960 D /QALY. La alternativa 2 fue la de mayor e\ufb01ciencia diagn\u00f3stica, alcanzando una reducci\u00f3n del 75% del n\u00famero de d\u00edas que un paciente con tuberculosis permanece sin tratamiento adecuado, as\u00ed como una reducci\u00f3n del 70% del n\u00famero de d\u00edas que un paciente sin tuberculosis permanece ingresado. Conclusi\u00f3n: La aplicaci\u00f3n de una t\u00e9cnica microbiol\u00f3gica molecular en el diagn\u00f3stico de tuberculosis es sumamente coste-efectiva frente al m\u00e9todo habitual. Su introducci\u00f3n en el procedimiento diagn\u00f3stico de rutina supondr\u00eda una mejora en la calidad asistencial de los pacientes al evitar ingresos y tratamientos innecesarios, re\ufb02ej\u00e1ndose en un ahorro econ\u00f3mico al hospital.\n\u00a9 2016 Elsevier Espan\u02dc a, S.L.U. y Sociedad Espan\u02dc ola de Enfermedades Infecciosas y Microbiolog\u0131\u00b4a Cl\u0131\u00b4nica. Todos los derechos reservados.\n\nCost-effectiveness study of the microbiological diagnosis of tuberculosis using geneXpert MTB/RIF\u00ae\n\na b s t r a c t\n\nKeywords: Tuberculosis Diagnosis Real-time polymerase chain reaction Sensitivity and speci\ufb01city Cost-effectiveness analysis Economic evaluation\n\nIntroduction/Objective: To perform a cost-effectiveness analysis of a molecular biology technique for the diagnosis of tuberculosis compared to the classical diagnostic alternative. Methods: A cost-effectiveness analysis was performed to evaluate the theoretical implementation of a molecular biology method including two alternative techniques for early detection of Mycobacterium tuberculosis Complex, and resistance to rifampicin (alternative 1: one determination in selected patients; alternative 2: two determinations in all the patients). Both alternatives were compared with the usual procedure for microbiological diagnosis of tuberculosis (staining and microbiological culture), and was\n\naccomplished on 1,972 patients in the period in 2008-2012. The effectiveness was measured in QALYs,\n\nand the uncertainty was assessed by univariate, multivariate and probabilistic analysis of sensitivity.\n\n\u0b1d Este proyecto ha sido premiado en el V Premio AEFA a la Calidad y a la Innovaci\u00f3n. \u2217 Autor para correspondencia.\nCorreo electr\u00f3nico: marian asencio@yahoo.es (M.\u00c1. Asencio-Egea).\n\nhttp://dx.doi.org/10.1016/j.eimc.2016.06.009 0213-005X/\u00a9 2016 Elsevier Espan\u02dc a, S.L.U. y Sociedad Espan\u02dc ola de Enfermedades Infecciosas y Microbiolog\u0131\u00b4a Cl\u0131\u00b4nica. Todos los derechos reservados.\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n2\n\nARTICLE IN PRESS\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\nResults: A value of D 8,588/QALYs was obtained by the usual method. Total expenditure with the alternative 1 was D 8,487/QALYs, whereas with alternative 2, the cost-effectiveness ratio amounted to D 2,960/QALYs. Greater diagnostic ef\ufb01ciency was observed by applying the alternative 2, reaching a 75% reduction in the number of days that a patient with tuberculosis remains without an adequate treatment, and a 70% reduction in the number of days that a patient without tuberculosis remains in hospital. Conclusion: The implementation of a molecular microbiological technique in the diagnosis of tuberculosis is extremely cost-effective compared to the usual method. Its introduction into the routine diagnostic procedure could lead to an improvement in quality care for patients, given that it would avoid both unnecessary hospitalisations and treatments, and re\ufb02ected in economic savings to the hospital.\n\u00a9 2016 Elsevier Espan\u02dc a, S.L.U. and Sociedad Espan\u02dc ola de Enfermedades Infecciosas y Microbiolog\u0131\u00b4a Cl\u0131\u00b4nica. All rights reserved.\n\nIntroducci\u00f3n\nEspan\u02dc a es un pa\u00eds de baja incidencia de tuberculosis (TB), aunque se diagnostican aproximadamente 5.000 casos al an\u02dc o1. Por ello, es necesario desarrollar actividades que aseguren un diagn\u00f3stico preciso y precoz, as\u00ed como el seguimiento y el cumplimiento de un tratamiento adecuado2,3. Sin embargo, el diagn\u00f3stico microbiol\u00f3gico de TB es complejo. La baciloscopia (BK) es una t\u00e9cnica r\u00e1pida, sencilla y econ\u00f3mica, pero presenta una sensibilidad baja y, por tanto, un n\u00famero elevado de falsos negativos, con el consecuente retraso diagn\u00f3stico, as\u00ed como algunos falsos positivos2. Un retraso en el diagn\u00f3stico incrementa el riesgo de transmitir la infecci\u00f3n y prolongar la enfermedad de los pacientes. Por el contrario, un resultado falso positivo puede causar la prescripci\u00f3n de un tratamiento innecesario, toxicidad farmacol\u00f3gica y selecci\u00f3n de cepas resistentes, as\u00ed como el retraso del diagn\u00f3stico correcto de enfermedades cr\u00f3nicas, condicionando un incremento de la morbilidad y de los costes4.\nLa aplicaci\u00f3n de t\u00e9cnicas de ampli\ufb01caci\u00f3n de \u00e1cidos nucleicos (TAAN) a tiempo real directamente sobre muestras cl\u00ednicas, como Xpert MTB/RIF\u00ae (Xpert), permite la detecci\u00f3n precoz del complejo Mycobacterium tuberculosis (MTBC) y de la resistencia a rifampicina (RIF) con sensibilidad y especi\ufb01cidad elevadas en 2 h5,6. Por esta raz\u00f3n, tanto la Organizaci\u00f3n Mundial de la Salud7 como el European Centre for Disease Prevention and Control8 recomiendan su uso en los casos pulmonares bacil\u00edferos. Pese al relativamente elevado coste econ\u00f3mico de estas t\u00e9cnicas frente a las convencionales, su aplicaci\u00f3n sistem\u00e1tica en pacientes con elevada sospecha de TB permitir\u00eda un ahorro econ\u00f3mico, sobre todo por reducir estancias hospitalarias9.\nDado que los costes del Xpert se presentan como la principal barrera para su implantaci\u00f3n en los procesos diagn\u00f3sticos de TB, el objetivo de este estudio ha sido evaluar mediante un an\u00e1lisis de coste-efectividad 2 alternativas diagn\u00f3sticas de TB que incluyen la tecnolog\u00eda Xpert MTB/RIF\u00ae frente al m\u00e9todo convencional utilizado en nuestro centro.\nM\u00e9todos\nSe hizo un estudio retrospectivo de los 1.972 pacientes que acudieron al Hospital General La Mancha Centro (HGMC) con sospecha de TB entre el 1 de enero de 2008 y el 31 de diciembre de 2012. Se ha considerado como criterio de inclusi\u00f3n la solicitud a pacientes ingresados de 3 esputos en un periodo inferior a 8 d\u00edas, de los cuales se dispone del resultado de BK y cultivo para micobacterias, obteni\u00e9ndose en caso positivo la identi\ufb01caci\u00f3n de especie y la susceptibilidad a los distintos antituberculosos.\nSe ha realizado un an\u00e1lisis de coste-efectividad para evaluar 3 estrategias de diagn\u00f3stico microbiol\u00f3gico de TB. La primera de ellas se basa en el m\u00e9todo habitual de diagn\u00f3stico de TB en nuestro hospital, mientras que las otras 2 estrategias (v\u00edas alternativas 1 y 2) se eval\u00faan mediante la aplicaci\u00f3n te\u00f3rica de la tecnolog\u00eda Xpert MTB/RIF\u00ae (Sunnyvale, CA, EE. UU.) sobre los mismos pacientes. Para\n\ncomparar los costes y la efectividad de las 3 estrategias se construy\u00f3 un \u00e1rbol de decisi\u00f3n usando la aplicaci\u00f3n TreeAge Pro 2011\u00ae (TreeAge Software Inc. Williamstown, MA, EE. UU.).\nEl m\u00e9todo habitual de diagn\u00f3stico de TB en el HGMC consiste en solicitar una radiograf\u00eda de t\u00f3rax, una prueba de Mantoux y obtener de cada paciente 3 esputos consecutivos y realizar BK (auramina-rodamina) y cultivo en el medio l\u00edquido MGIT (BD, S.A.), considerando como patr\u00f3n de oro diagn\u00f3stico el aislamiento de MTBC3. Mientras se obtienen los resultados de la BK, el paciente permanece hospitalizado en aislamiento respiratorio. Si el resultado de alguna BK es positivo, el paciente puede recibir el alta precoz (antes de 7 d\u00edas) con tratamiento emp\u00edrico y seguimiento en consulta en espera de los resultados del cultivo y antibiograma. Por el contrario, en caso de que las 3 BK sean negativas se distinguen 2 situaciones:\n1. En pacientes con sospecha cl\u00ednica elevada de TB (SCETB) (contexto cl\u00ednico-epidemiol\u00f3gico compatible y radiolog\u00eda sugerente de TB) se completa el estudio con la realizaci\u00f3n de una tomograf\u00eda axial computarizada (TAC) y la obtenci\u00f3n de un lavado broncoalveolar y/o aspirado bronquial mediante \ufb01brobroncoscopia, manteniendo al paciente sin tratamiento antituberculoso; en caso de que alguna BK sea positiva, el paciente es tratado emp\u00edricamente, habiendo permanecido en aislamiento una media de 2 semanas, mientras que los pacientes que presentan SCETB y resultados de BK negativos comienzan igualmente el tratamiento emp\u00edrico, pero permanecen ingresados sin aislamiento una media de 3 semanas hasta recibir el resto de pruebas diagn\u00f3sticas (demora de pruebas de imagen, anatom\u00eda patol\u00f3gica, cultivos).\n2. En el caso de que las 3 BK sean negativas sin SCETB, el paciente es ingresado sin condiciones de aislamiento una media de 11 d\u00edas en los que se observa su evoluci\u00f3n, siendo dado de alta sin tratamiento antituberculoso, con diagn\u00f3stico o no de otra enfermedad respiratoria.\nLos tiempos de ingreso en cada situaci\u00f3n han sido facilitados por el Servicio de Medicina Interna y son una estimaci\u00f3n basada en la pr\u00e1ctica cl\u00ednica habitual tras 20 an\u02dc os de experiencia en el manejo de pacientes con TB.\nLa primera alternativa propuesta consiste en incorporar el diagn\u00f3stico precoz de TB mediante Xpert de una muestra de esputo en pacientes con BK positiva o SCETB. El modelo sigue el mismo procedimiento que el m\u00e9todo habitual, pero sin considerar la obtenci\u00f3n de muestras por m\u00e9todos invasivos. En el caso de que alguna BK fuera positiva se realizar\u00eda una determinaci\u00f3n de Xpert para con\ufb01rmar que realmente es MTBC. En caso de que el resultado del Xpert sea negativo y exista una SCETB, se considerar\u00eda la posibilidad de tratar al paciente emp\u00edricamente, permaneciendo ingresado sin aislamiento una media de 3 semanas hasta la obtenci\u00f3n de los resultados microbiol\u00f3gicos. En el caso de que las 3 BK sean negativas y el paciente presente SCETB se realizar\u00eda el an\u00e1lisis mediante Xpert del\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n\nARTICLE IN PRESS\n\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\nTabla 1 Variables utilizadas en la construcci\u00f3n del \u00e1rbol de decisi\u00f3n\nVariable\nCoste por d\u00eda de estancia (5 primeros d\u00edas) Coste por d\u00eda de estancia (a partir del 6.\u25e6 d\u00eda) Coste por equipos de protecci\u00f3n respiratoria por d\u00eda Coste por consulta sucesiva Coste por BK Coste por cultivo de micobacterias Coste por \ufb01brobroncoscopia Coste por tomograf\u00eda Coste identi\ufb01caci\u00f3n/antibiograma Coste laboratorio por consultas sucesivas Coste por determinaci\u00f3n Xpert Coste tratamiento TB resistente a rifampicina por d\u00eda Coste tratamiento TB multirresistente por d\u00eda Coste tratamiento TB habitual (sensible) por d\u00eda Coste tratamiento TB sensible a rifampicina y resistente a\nisoniazida por d\u00eda Sensibilidad del Xpert para la detecci\u00f3n de M. tuberculosis\nen pacientes con BK negativa Sensibilidad del Xpert para la detecci\u00f3n de M. tuberculosis\nen pacientes con BK positiva Sensibilidad del Xpert en la segunda determinaci\u00f3n para la\ndetecci\u00f3n de M. tuberculosis en pacientes con BK negativa Sensibilidad del Xpert en la segunda determinaci\u00f3n para la\ndetecci\u00f3n de M. tuberculosis en pacientes con BK positiva Especi\ufb01cidad del Xpert para la detecci\u00f3n de M. tuberculosis Especi\ufb01cidad del Xpert en la segunda determinaci\u00f3n para\nla detecci\u00f3n de M. tuberculosis Especi\ufb01cidad del Xpert para la detecci\u00f3n de resistencia a\nrifampicina Sensibilidad del Xpert para la detecci\u00f3n de resistencia a\nrifampicina Probabilidad de toxicidad tratamiento multirresistente Probabilidad de toxicidad tratamiento no multirresistente Nivel de utilidad en poblaci\u00f3n general Utilidad de hospitalizaci\u00f3n Utilidad de hospitalizaci\u00f3n en condiciones de aislamiento Utilidad de estancia domiciliaria Utilidad en pacientes con TB activa tratada correctamente Utilidad en pacientes con TB activa no tratada\ncorrectamente Utilidad de toxicidad por tratamiento prescrito en\npacientes sin TB Utilidad de toxicidad por tratamiento prescrito en\npacientes con TB\n\nValor medio\n524,56 D 472,10 D 2,1 D 71,91 D 0,48 D 12,34 D 134,02 D 113 D 0 D 20 D 65 D 0,4 D 5,5 D 0,3 D 0,4 D\n0,775\n0,987\n0,775\n0,987\n0,982 0,982\n0,97\n0,941\n0,3 0,045 0,86 U general-0,4 U general-0,5 U general U general-0,1 U general-0,39\nU general-0,16\nU general-0,25\n\nValor inferior\n0,77 0,98 0,77 0,98 0,978 0,978 0,96 0,916\n\nValor superior\n0,778 0,992 0,778 0,992 0,992 0,992 0,977 0,96\n\nDistribuci\u00f3n\nTriangular Triangular Triangular Triangular Triangular Triangular Triangular Triangular LogNormal\n\n3\nFuente DOCM18 DOCM18 DOCM18 DOCM18 HGMC HGMC DOCM18 DOCM18 HGMC HGMC HGMC Caminero19 Caminero19 Caminero19 Caminero19\nChang et al.20\nChang et al.20\nAsumido\nAsumido\nChang20 Asumido\nChang et al.20\nChang et al.20\nHughes et al.11 Hughes et al.11 Kind et al.21 Holland et al.22 Estimaci\u00f3n Estimaci\u00f3n Tan et al.23 Khan et al. 24\nHolland et al.22\nHolland et al.22\n\n\u00faltimo esputo, mientras que en pacientes con 3 BK negativas y sin SCETB se aplicar\u00eda el m\u00e9todo convencional. Si el resultado del Xpert es positivo, independientemente del resultado de la BK, el paciente recibir\u00eda tratamiento emp\u00edrico y el alta precoz, tal como suced\u00eda con el m\u00e9todo habitual, pero con la posibilidad de adecuar el tratamiento antituberculoso al disponer del dato de resistencia a RIF.\nLa segunda alternativa propuesta incluye la posibilidad de realizar 2 determinaciones con Xpert a todos los pacientes del estudio, con independencia de la sospecha cl\u00ednica y del resultado de la BK. Si el resultado del Xpert es positivo se procede igual que en la alternativa anterior. En el caso de un resultado Xpert negativo se realizar\u00eda una segunda determinaci\u00f3n con otra muestra. Los pacientes con BK positiva y 2 determinaciones Xpert negativas ser\u00edan tratados emp\u00edricamente hasta la obtenci\u00f3n del resultado del cultivo y del antibiograma. Los pacientes con las 3 BK negativas y las 2 determinaciones de Xpert negativas ser\u00edan dados de alta sin tratamiento hasta la obtenci\u00f3n del resultado del cultivo.\nEl modelo incluye varios tipos de variables. Por una parte, se de\ufb01nen variables de car\u00e1cter bibliogr\u00e1\ufb01co aplicadas a todos los pacientes como costes, utilidades y probabilidades (tabla 1). Se de\ufb01nen tambi\u00e9n variables num\u00e9ricas estimadas por aplicaci\u00f3n de los procedimientos cl\u00ednicos habituales a cada tipo de paciente y de\ufb01nidas para cada rama del \u00e1rbol de decisi\u00f3n (tabla 2). Por \u00faltimo, se han calculado variables que permiten estimar el coste y la utilidad para cada rama del \u00e1rbol y que utilizan para su de\ufb01nici\u00f3n\n\nvariables bibliogr\u00e1\ufb01cas y num\u00e9ricas (tabla 2). No se ha incluido la mortalidad como variable a medir debido a que en la poblaci\u00f3n estudiada no se dio ning\u00fan caso de fallecimiento asociado a la patolog\u00eda tuberculosa.\nLa medida de la efectividad se hizo en Quality Adjusted-Life-Years o an\u02dc os de vida ajustados por calidad (QALY). El c\u00e1lculo de la utilidad \ufb01nal (equivalente a la efectividad) de cada rama del \u00e1rbol se realiza mediante la ponderaci\u00f3n anual de las utilidades diarias para cada paciente desde la sospecha diagn\u00f3stica hasta la \ufb01nalizaci\u00f3n del tratamiento, o bien hasta que el diagn\u00f3stico de TB es descartado. Para hacer este c\u00e1lculo se estima la utilidad del paciente por d\u00eda hasta completar un an\u02dc o. En el caso de pacientes con una duraci\u00f3n del proceso superior al an\u02dc o se considera que la utilidad anual presenta el mismo valor que la utilidad total. Se identi\ufb01can 2 utilidades \ufb01nales: la utilidad en los pacientes con TB y en aquellos sin TB.\nLa incertidumbre se trat\u00f3 mediante an\u00e1lisis de sensibilidad univariable, multivariable y de tipo tornado10.\nResultados\nDe los 1.972 pacientes estudiados se con\ufb01rm\u00f3 la enfermedad mediante cultivo en 69 de ellos (3,5%), aunque hubo 177 cultivos positivos; por tanto, se aislaron 108 micobacterias at\u00edpicas. El 71% (49/69) de los casos de TB se pudieron detectar precozmente mediante BK. Hubo 4 cepas de MTBC resistentes a los\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n4\n\nARTICLE IN PRESS\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\n1.781 no SCETB\n\n1.673 cultivos negativos\n108 cultivos positivos\n\n93 MNT 15 MTBC\n\n1.908 BK negativa\n\n2 confirmados por FBC\n\n1 MNT 1 MTBC\n\n1.972 pacientes 64 BK positiva\n\n127 SCETB\n58 cultivos positivos 6 cultivos negativos\n\n125 no confirmados por\nFBC\n9 MNT\n49 MTBC\n\n9 cultivos positivos\n116 cultivos negativos\n\n5 MNT 4 MTBC\n\nFigura 1. Distribuci\u00f3n de pacientes en el m\u00e9todo habitual. BK: baciloscopia; FBC: \ufb01brobroncoscopia; MNT: micobacteria no tuberculosa; MTBC: Mycobacterium tuberculosis Complex; SCETB: sospecha cl\u00ednica elevada de tuberculosis.\n\nantituberculosos de primera l\u00ednea (8%), 2 de ellas resistentes a isoniazida, una resistente a RIF y otra multirresistente. El grupo m\u00e1s numeroso fue el de pacientes con sospecha no con\ufb01rmada de TB, sin SCETB y con resultado de BK negativo, que permanecieron ingresados una media de 11 d\u00edas (3 en aislamiento). En la \ufb01gura 1 se muestra una descripci\u00f3n detallada de los resultados obtenidos mediante el m\u00e9todo habitual.\nTabla 2 Variables medidas para cada paciente del \u00e1rbol de decisi\u00f3n\nVariable N\u00famero de d\u00edas con estancia en condiciones de aislamiento N\u00famero de BK realizadas para el diagn\u00f3stico N\u00famero de consultas sucesivas N\u00famero de cultivos realizados en diagn\u00f3stico N\u00famero medio de d\u00edas desde la llegada al laboratorio de la primera muestra\nhasta el resultado del cultivo N\u00famero total de d\u00edas de estancia hospitalaria N\u00famero de d\u00edas en estancias inferiores a 6 d\u00edas N\u00famero de estancias a partir de 6 d\u00edas N\u00famero de \ufb01brobroncoscopias y TAC realizadas en el proceso diagn\u00f3stico N\u00famero de d\u00edas desde la emisi\u00f3n del cultivo positivo por el laboratorio de\nmicrobiolog\u00eda hasta la obtenci\u00f3n de resultados de identi\ufb01caci\u00f3n y antibiograma N\u00famero de identi\ufb01caciones y antibiogramas realizados en el paciente N\u00famero de d\u00edas con tratamiento para MTBC sensible a todos los f\u00e1rmacos de primera l\u00ednea salvo rifampicina N\u00famero de d\u00edas con tratamiento para MTBC multirresistente N\u00famero de d\u00edas con tratamiento habitual sensible a todos los f\u00e1rmacos de primera l\u00ednea N\u00famero de d\u00edas con tratamiento para MTBC sensible a todos los f\u00e1rmacos de primera l\u00ednea salvo isoniazida N\u00famero de determinaciones realizadas por Xpert D\u00edas con el paciente a la espera de tratamiento y/o diagn\u00f3stico D\u00edas de tratamiento para MTBC multirresistentes en los que el tratamiento es incorrecto D\u00edas de tratamiento para MTBC no multirresistente en los que el tratamiento es incorrecto D\u00edas de tratamiento para MTBC multirresistentes con tratamiento correcto D\u00edas de tratamiento para MTBC no multirresistente con tratamiento correcto\n\nE\ufb01ciencia diagn\u00f3stica\nEl \u00e1rbol de decisi\u00f3n con las 3 alternativas se muestra en la \ufb01gura 2. Una vez calculadas las probabilidades para cada rama del \u00e1rbol de decisi\u00f3n se estimaron los consumos de recursos por los 1.972 pacientes al aplicar las distintas alternativas (tabla 3). De los 3 m\u00e9todos evaluados, las 2 alternativas al m\u00e9todo habitual presentaron mayor e\ufb01ciencia diagn\u00f3stica, con una disminuci\u00f3n del porcentaje de pacientes con TB que han permanecido sin tratamiento hasta la obtenci\u00f3n del resultado del cultivo (porcentaje de falsos negativos). Adem\u00e1s, en la alternativa 2 se observa una disminuci\u00f3n del n\u00famero de pacientes sin TB que reciben tratamiento inadecuado hasta la obtenci\u00f3n de los resultados del cultivo y/o antibiograma (\ufb01g. 3).\nEvaluaci\u00f3n coste-efectividad en pacientes con sospecha de tuberculosis\nPara el m\u00e9todo habitual se ha obtenido un cociente de 8.588 D /QALY. La estimaci\u00f3n del coste en pacientes con TB fue de 6.329 D /QALY y de 8.659 D /QALY en pacientes sin TB. En la alternativa 1 se obtiene un coste total de 8.487 D /QALY, siendo el gasto por paciente con TB de 7.222 D /QALY, y por paciente sin TB, de 8.532 D /QALY. En la alternativa 2 se observa un cociente costeefectivo de 2.969 D /QALY. Los pacientes con TB presentaron un gasto de 5.510 D /QALY, mientras que el gasto para aquellos en los que se descart\u00f3 el diagn\u00f3stico de TB fue de 2.878 D /QALY. Por tanto, la alternativa 2 presenta un ratio coste-efectivo dominante.\nEstudio de sensibilidad\nUn an\u00e1lisis de sensibilidad tipo tornado indic\u00f3 que el par\u00e1metro con mayor in\ufb02uencia en el gasto por paciente fue el coste por estancia durante los 5 primeros d\u00edas. La disminuci\u00f3n de 300 D /d\u00eda\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n\nARTICLE IN PRESS\n\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\n5\n\nsupone una reducci\u00f3n del coste por paciente de 1.614 D /d\u00eda al aplicar la alternativa 2, frente a los 6.194 D obtenidos al aplicar el m\u00e9todo habitual. La segunda variable con mayor in\ufb02uencia en el modelo fue el coste de las medidas de protecci\u00f3n. Se han tomado como valores posibles el intervalo entre 2,1 y 200 D , en donde el m\u00ednimo valor se corresponde con el coste de las medidas de\n\nbarrera (mascarillas) y el m\u00e1ximo a permanecer en una habitaci\u00f3n en aislamiento. La variaci\u00f3n en este par\u00e1metro tuvo una in\ufb02uencia mayor en la alternativa 2, si bien esta contin\u00faa siendo la alternativa dominante. La variaci\u00f3n en el coste de la determinaci\u00f3n de Xpert in\ufb02uy\u00f3 de manera m\u00e1s notable, aunque moderada, en la alternativa 2. Tambi\u00e9n se ha realizado el estudio de la sensibilidad para los\n\nM\u00e9todo habitual\n\nNo SCETB\n\nCultivo de micobacterias negativo\n\nCultivo +\n\nSensible rifampicina Micobacterias no tubeculosas\n\nBaciloscopia \u2013 SCETB\n\nFibrobronocoscopia +\n\nCultivo de micobacterias negativo\n\nCultivo +\n\nSensible rifampicina Micobacterias no tubeculosas\n\nCultivo de micobacterias negativo\n\nFibrobronocoscopia \u2013\n\nSen Rif - Res INH\n\nCultivo +\n\nSensible rifampicina\n\nMicobacterias no tubeculosas\n\nSensible rifampicina Res Rif\n\nCultivo +\n\nRes Rif - multirresistente\n\nBaciloscopia +\n\nSen Rif - Res INH Micobacterias no tubeculosas\n\nCultivo de micobacterias negativo\n\nSensible rifampicina\n\nXpert + Xpert \u2013\n\nXpertRRif XpertSRif\n\nBaciloscopia +\n\nPacientes con TB\n\nDiagn\u00f3stico con GenXpert\n\nBaciloscopia \u2013\n\nRes Rif Res Rif - multirresistente Sen Rif - Res INH SCETB\nNo SCETB\n\nClone 1: Determinacion simple Xpert\n\nClone 1: Determinacion simple Xpert\n\nClone 1: Determinacion simple Xpert\n\nSensible rifampicina\n\nClone 1: Determinacion simple Xpert\n\nSen Rif - Res INH\n\nClone 1: Determinacion simple Xpert\n\nSensible rifampicina\n\nBaciloscopia +\nPacientes sin TB\nBaciloscopia \u2013\n\nClone 1: Determinacion simple Xpert\n\nSCETB\n\nClone 1: Determinacion simple Xpert\n\nNo SCETB\n\nCultivo de micobacterias negativo Micobacterias no tuberculosas\n\nFigura 2. \u00c1rbol de decisi\u00f3n con las 3 alternativas diagn\u00f3sticas: a) m\u00e9todo habitual; b) alternativa 1; c) alternativa 2.\n\nTotCos_total \\ U_NO_TB TotCos_total \\ U_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_TB\nTotCos_total \\ U_TB\nTotCos_total \\ U_NO_TB TotCos_total \\ U_NO_TB\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n6\n\nARTICLE IN PRESS\n\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\nXpert +\n\nXpertRRif XpertSRif\n\nBaciloscopia + Pacientes con TB\n\nDiagn\u00f3stico con dobel genXpert\n\nBaciloscopia \u2013\n\nBaciloscopia +\nPacientes sin TB\nBaciloscopia \u2013\n\nSensible rifampicina\n\n2 Xpert \u2013\n\n2Xpert + 2Xpert \u2013\n\nRes Rif\n\nClone 2: Determinacion doble Xpert\n\nRes Rif - multirresistente\n\nClone 2: Determinacion doble Xpert\n\nSen Rif - Res INH\n\nClone 2: Determinacion doble Xpert\n\nSensible rifampicina\n\nClone 2: Determinacion doble Xpert\n\nSen Rif - Res INH\n\nClone 2: Determinacion doble Xpert\n\nCultivo de micobacterias negativo\n\nClone 2: Determinacion doble Xpert\n\nMicobacterias no tuberculosas Clone 2: Determinacion doble Xpert\n\nCultivo de micobacterias negativo\n\nClone 2: Determinacion doble Xpert\n\nMicobacterias no tuberculosas Clone 2: Determinacion doble Xpert\n\nFigura 2. (Continuaci\u00f3n )\n\nXpertRRif XpertSRif\n\nTabla 3 Estimaciones de recursos consumidos por las 3 alternativas evaluadas\nN\u00famero total de d\u00edas de pacientes con TB sin tratamiento N\u00famero total de d\u00edas con prescripci\u00f3n a los pacientes de\ntratamiento habitual (para MTBC no multirresistente) incorrecto N\u00famero total de d\u00edas en los que el paciente est\u00e1 en aislamiento N\u00famero de consultas sucesivas D\u00edas totales de estancia hospitalaria N\u00famero de determinaciones de Xpert\n\nM\u00e9todo habitual\n522 5.974\n7.533 462\n29.124 0\n\nAlternativa 1\n403 3.177\n7.039 460\n28.807 206\n\nTotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_TB\nAlternativa 2 126\n2.233 8.200 2.135 8.588 3.847\n\nvalores de la utilidad (QALY) de la poblaci\u00f3n general, analizando el intervalo comprendido entre 0,80 y 1 QALY.\nCon la aplicaci\u00f3n del m\u00e9todo de Monte Carlo se obtuvo una curva de aceptabilidad que indica que la alternativa 2 es la m\u00e1s rentable, con independencia del umbral elegido, con una probabilidad del 100%.\n\n25\n\nPorcentaje de pacientes\n\n20\n\n15\n\n10\n\n5\n\n0 M\u00e9todo\n\nAlternativa 1\n\nAlternativa 2\n\n% FN\n\n21,7\n\n2,6\n\n1,5\n\n% FP\n\n7,2\n\n7,2\n\n3,5\n\nFigura 3. Porcentaje de pacientes con tratamiento emp\u00edrico incorrecto.\n\nDiscusi\u00f3n\nLos resultados de este trabajo sugieren que, en nuestro medio y con las suposiciones realizadas en el modelo, el diagn\u00f3stico de TB con la tecnolog\u00eda Xpert es m\u00e1s coste-efectiva que el procedimiento convencional. De las 2 alternativas al m\u00e9todo actual ha resultado dominante la que incluye la posibilidad de realizar 2 determinaciones de Xpert, de manera que la aplicaci\u00f3n de esta tecnolog\u00eda como m\u00e9todo de cribado evidenciar\u00eda una mejora en la calidad de vida de los pacientes con TB por permitir iniciar un tratamiento adecuado sin demoras, as\u00ed como una disminuci\u00f3n del gasto hospitalario derivado de la reducci\u00f3n del n\u00famero de estancias de los pacientes sin TB.\nEl estudio de las 3 ramas del \u00e1rbol de decisi\u00f3n muestra que hay diferencias entre los QALY medidos en los 3 procedimientos diagn\u00f3sticos, con un mayor n\u00famero de QALY ganados por aplicaci\u00f3n de la alternativa 2 (reducci\u00f3n del 70% de las estancias hospitalarias y del 75% de d\u00edas sin tratamiento adecuado). Sin embargo, la diferencia m\u00e1s notable reside en la disminuci\u00f3n del coste por paciente, un 65% menor en la alternativa 2 respecto al m\u00e9todo habitual. Esta disminuci\u00f3n de costes permitir\u00eda un ahorro te\u00f3rico anual de 1,8 millones de euros debido, sobre todo, a una reducci\u00f3n del n\u00famero de estancias hospitalarias asociadas a la sospecha de un proceso tuberculoso.\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n\nARTICLE IN PRESS\n\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\n7\n\nDebido a que la sensibilidad de Xpert en pacientes con BK negativa es relativamente baja6,10-12, no se recomienda su aplicaci\u00f3n a no ser que exista una SCETB, puesto que de lo contrario se reducir\u00eda el valor predictivo positivo de la t\u00e9cnica13. Algunos estudios muestran un aumento de la sensibilidad de este test al realizar una segunda determinaci\u00f3n en pacientes con resultado negativo14. De esta manera se justi\ufb01car\u00eda la utilizaci\u00f3n de Xpert como t\u00e9cnica de cribado, con el inconveniente de aumentar el gasto de laboratorio y el n\u00famero de falsos positivos con respecto a la alternativa 1, aunque todav\u00eda resultar\u00eda inferior a los falsos positivos obtenidos con el m\u00e9todo actual (75 vs. 200). Adem\u00e1s, seg\u00fan algunos autores el dan\u02dc o de no tratar a un paciente con TB es mayor que tratar con antituberculosos a un paciente sin TB15. Por otra parte, tras 2 resultados negativos de Xpert se podr\u00eda descartar el diagn\u00f3stico de TB, e incluso podr\u00eda dejar de realizarse el cultivo microbiol\u00f3gico, ya que el valor predictivo negativo de la prueba es cercano al 100%16, si bien esta posibilidad no ha sido evaluada en nuestro estudio. Por esta raz\u00f3n podr\u00eda parecer contradictorio haber considerado en la alternativa 2 prescribir tratamiento emp\u00edrico a los pacientes con BK+ y 2 resultados del Xpert negativos; sin embargo, se ha considerado esta posibilidad por tratarse del peor escenario que podr\u00edamos encontrarnos, que adem\u00e1s incluye la posible resistencia de algunos cl\u00ednicos a dejar de tratar a un paciente con BK+, as\u00ed como la posibilidad de una infecci\u00f3n por una micobacteria no tuberculosa (en este caso habr\u00eda que individualizar la terapia). La realizaci\u00f3n de una tercera determinaci\u00f3n de Xpert, tras 2 resultados negativos, podr\u00eda ser desaconsejable, ya que el n\u00famero de falsos positivos se estima superior a 100.\nOtros estudios de coste-efectividad de Xpert realizados en pa\u00edses en v\u00edas de desarrollo con elevada carga de TB muestran que Xpert es coste-efectivo y su introducci\u00f3n podr\u00eda suponer un cambio importante en la morbimortalidad de la infecci\u00f3n mediante la mayor detecci\u00f3n de casos y, por tanto, de tratamientos dirigidos5,6. La aplicaci\u00f3n de esta tecnolog\u00eda en zonas de baja prevalencia como Espan\u02dc a reducir\u00eda su rentabilidad, si bien su potencial en situaciones de mayor prevalencia ser\u00eda considerable (poblaci\u00f3n inmigrante, sin techo, usuarios de drogas por v\u00eda parenteral y pacientes VIH y, en general, cuando existe elevada sospecha de TB)10. Nuestro estudio muestra que la aplicaci\u00f3n te\u00f3rica de esta t\u00e9cnica a cualquier paciente con sospecha de TB es m\u00e1s coste-efectiva que su aplicaci\u00f3n solo a pacientes seleccionados, ya que el hecho de obtener un resultado \ufb01able y r\u00e1pido que descarte la infecci\u00f3n da lugar a altas hospitalarias, independientemente de la verdadera enfermedad del paciente. As\u00ed, la diferencia fundamental entre la alternativa 2 y el m\u00e9todo habitual radica en la forma de considerar a un paciente sin SCETB, de manera que, seg\u00fan el m\u00e9todo habitual, este tipo de pacientes permanecen ingresados, mientras que los pacientes con 2 Xpert negativos son dados de alta precozmente.\nUna limitaci\u00f3n del estudio es la consideraci\u00f3n del coste de la estancia hospitalaria sin condiciones de aislamiento respiratorio. Sin embargo, en el estudio de sensibilidad se ha tenido en cuenta la posibilidad de un aumento del coste hospitalario debido al establecimiento de condiciones de aislamiento, comport\u00e1ndose este factor como el segundo m\u00e1s in\ufb02uyente en la estimaci\u00f3n de costes. No obstante, la cuanti\ufb01caci\u00f3n del ahorro obtenido con la alternativa 2 ha de tomarse con cautela, ya que no se ha considerado la posibilidad de hospitalizaci\u00f3n del paciente para realizar el diagn\u00f3stico y tratamiento de la enfermedad no tuberculosa. Adem\u00e1s, se observa que la aplicaci\u00f3n de la alternativa 2 supone un aumento importante del n\u00famero de consultas sucesivas, para asegurar la revisi\u00f3n de los pacientes que, a pesar de la sospecha de TB, fueron dados de alta sin diagn\u00f3stico y, por ello, sin tratamiento. Por otra parte, podr\u00eda interpretarse como una mejora en la atenci\u00f3n sanitaria al paciente por una mayor vigilancia a un coste total menor, contribuyendo adem\u00e1s a la descongesti\u00f3n de los centros hospitalarios si el seguimiento se realizara en los centros de atenci\u00f3n primaria2.\n\nConsideramos que en estudios futuros ser\u00eda conveniente evaluar el bene\ufb01cio real que la pr\u00e1ctica cl\u00ednica con la aplicaci\u00f3n de la tecnolog\u00eda Xpert tendr\u00eda en los costes y la calidad de vida de los pacientes con sospecha de TB. Asimismo, existen en el mercado otras TAAN disponibles para ser evaluadas como alternativas17.\nEn conclusi\u00f3n, este estudio sugiere que la aplicaci\u00f3n de la tecnolog\u00eda Xpert en el diagn\u00f3stico de TB es sumamente coste-efectiva comparada con el m\u00e9todo convencional. El impacto de introducir la tecnolog\u00eda Xpert abarca el \u00e1mbito econ\u00f3mico y sanitario, de manera que su aplicaci\u00f3n supondr\u00eda una mejora en la calidad asistencial de los pacientes por evitar estancias y tratamientos innecesarios, permitiendo adem\u00e1s iniciar un tratamiento precoz dirigido, romper la cadena de transmisi\u00f3n de la infecci\u00f3n y conseguir un ahorro econ\u00f3mico considerable para al hospital.\nCon\ufb02icto de intereses\nLos autores declaran no tener con\ufb02icto de intereses.\nBibliograf\u00eda\n1. Rodr\u00edguez E, D\u00edas O, Hern\u00e1ndez G, Tello O. Situaci\u00f3n de la tuberculosis en Espan\u02dc a. Casos de tuberculosis declarados a la Red Nacional de Vigilancia Epidemiol\u00f3gica en 2010. Bolet\u00edn Epidemiol\u00f3gico Semanal. 2012;20:26\u201341.\n2. Gu\u00eda de Pr\u00e1ctica Cl\u00ednica sobre el Diagn\u00f3stico, el Tratamiento y la Prevenci\u00f3n de la Tuberculosis. Plan de Calidad para el Sistema Nacional de Salud del Ministerio de Sanidad, Pol\u00edtica Social e Igualdad. En: Ag\u00e8ncia d\u2019Informaci\u00f3 AiQeSAdC, ed. Gu\u00edas de Pr\u00e1ctica Cl\u00ednica en el SNS 2009.\n3. Gonz\u00e1lez-Mart\u00edn J, Garc\u00eda-Garc\u00eda JM, Anibarroc L, Vidald R, Estebane J, Blanquerf R, et al. Documento de consenso sobre diagn\u00f3stico, tratamiento y prevenci\u00f3n de la tuberculosis. Enferm Infecc Microbiol Clin. 2010;28:297.e1\u201320.\n4. Davies PD, Pai M. The diagnosis and misdiagnosis of tuberculosis. Int J Tuberc Lung Dis. 2008;12:1226\u201334.\n5. Steingart KR, Sohn H, Schiller I, Kloda LA, Boehme CC, Pai M, et al. Xpert MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Sys Rev. 2013;1(CD009593).\n6. Vassall A, van Kampen S, Sohn H, Michael JS, John KR, den Boon S, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: A cost-effectiveness analysis. PLoS medicine. 2011;8:e1001120.\n7. WHO. Automated Real-Time Nucleic Acid Ampli\ufb01 Cation Technology for Rapid and Simultaneous Detection of Tuberculosis and Rifampicin Resistance: Xpert MTB/RIF System: Policy Statement. Geneva: World Health Organization; 2011.\n8. European Centre for Disease Prevention and Control. ERLN-TB expert opinion on the use of the rapid molecular assays for the diagnosis of tuberculosis and detection of drug-resistance. Stockholm: ECDC; 2013.\n9. Menzies NA, Cohen T, Lin HH, Murray M, Salomon JA. Population health impact and cost-effectiveness of tuberculosis diagnosis with Xpert MTB/RIF: a dynamic simulation and economic evaluation. PLoS medicine. 2012;9: e1001347.\n10. Drummond MF, Sculpher MJ, Claxton K, Stoddart G, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes. Oxford: Oxford University Press; 2015.\n11. Hughes R, Wonderling D, Li B, Higgins B. The cost effectiveness of nucleic acid ampli\ufb01cation techniques for the diagnosis of tuberculosis. Respir Med. 2012;106:300\u20137.\n12. Roos BR, van Cleeff MR, Githui WA, Kivihya-Ndugga L, Odhiambo JA, Kibuga DK, et al. Cost-effectiveness of the polymerase chain reaction versus smear examination for the diagnosis of tuberculosis in Kenya: A theoretical model. Int J Tuberc Lung Dis. 1998;2:235\u201341.\n13. Catanzaro A, Perry S, Clarridge JE, Dunbar S, Goodnight-White S, LoBue PA, et al. The role of clinical suspicion in evaluating a new diagnostic test for active tuberculosis: Results of a multicenter prospective trial. JAMA. 2000;283: 639\u201345.\n14. Lawn SD, Brooks SV, Kranzer K, Nicol MP, Whitelaw A, Vogt M, et al. Screening for HIV-associated tuberculosis and rifampicin resistance before antiretroviral therapy using the Xpert MTB/RIF assay: A prospective study. PLoS medicine. 2011;8:e1001067.\n15. Moreira J, Bisig B, Muwawenimana P, Basinga P, Bisof\ufb01 Z, Haegeman F, et al. Weighing harm in therapeutic decisions of smear-negative pulmonary tuberculosis. Med Decis Making. 2009;29:380\u201390.\n16. Dorman SE, Chihota VN, Lewis JJ, Shah M, Clark D, Grant AD, et al. Performance characteristics of the Cepheid Xpert MTB/RIF test in a tuberculosis prevalence survey. PloS one. 2012;7:e43307.\n17. Armand S, Vanhuls P, Delcroix G, Courcol R, Lemaitre N. Comparison of the Xpert MTB/RIF test with an IS6110-TaqMan real-time PCR assay for direct detection of Mycobacterium tuberculosis in respiratory and nonrespiratory specimens. J Clin Microbiol. 2011;49:1772\u20136.\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n\nARTICLE IN PRESS\n\n8\n\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\n18. DOCM. Resoluci\u00f3n de 03/09/2012, de la Direcci\u00f3n Gerencia, sobre precios a aplicar por sus centros sanitarios a terceros obligados al pago o a los usuarios sin derecho a asistencia sanitaria 2012:28987-95.\n19. Caminero JA. Treatment of tuberculosis according to the different pattern of resistance. Med Clin (Barc.). 2010;134:173\u201381.\n20. Chang K, Lu W, Wang J, Zhang K, Jia S, Li F, et al. Rapid and effective diagnosis of tuberculosis and rifampicin resistance with Xpert MTB/RIF assay: A metaanalysis. J Infect. 2012;64:580\u20138.\n21. Kind P, Macran S, Hardman G. UK population norms for EQ-D5. 1999. Discussion paper 172.\n\n22. Holland DP, Sanders GD, Hamilton CD, Stout JE. Costs and cost-effectiveness of four treatment regimens for latent tuberculosis infection. Am J Respir Crit Care Med. 2009;179:1055\u201360.\n23. Tan MC, Marra CA, Sadatsafavi M, Marra F, Moran-Mendoza O, Moadebi S, et al. Cost-effectiveness of LTBI treatment for TB contacts in British Columbia. Value Health. 2008;11:842\u201352.\n24. Khan K, Muennig P, Behta M, Zivin JG. Global drug-resistance patterns and the management of latent tuberculosis infection in immigrants to the United States. N Engl J Med. 2002;347:1850\u20139.\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\n\n",
    "authors": [
      "\u00d3scar Herr\u00e1ez",
      "Mar\u00eda \u00c1ngeles Asencio-Egea",
      "Mar\u00eda Huertas-Vaquero",
      "Rafael Carranza-Gonz\u00e1lez",
      "Jes\u00fas Castellanos-Monedero",
      "Mar\u00eda Franco-Huerta",
      "Jos\u00e9 Ram\u00f3n Barber\u00e1-Farr\u00e9",
      "Jos\u00e9 Mar\u00eda Ten\u00edas-Burillo"
    ],
    "doi": "10.1016/j.eimc.2016.06.009",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://linkinghub.elsevier.com/retrieve/pii/S0213005X16301550"
  },
  {
    "key": "Q9TUM49N",
    "title": "Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF \u00ae",
    "abstract": "",
    "full_text": "G Model\nEIMC-1562; No. of Pages 8\n\nARTICLE IN PRESS\nEnferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\nwww.elsevier.es/eimc\nOriginal\nEstudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae\u0b1d\n\u00d3scar Herr\u00e1ez a, Mar\u00eda \u00c1ngeles Asencio-Egea b,\u2217, Mar\u00eda Huertas-Vaquero b, Rafael Carranza-Gonz\u00e1lez b, Jes\u00fas Castellanos-Monedero c, Mar\u00eda Franco-Huerta c, Jos\u00e9 Ram\u00f3n Barber\u00e1-Farr\u00e9 c y Jos\u00e9 Mar\u00eda Ten\u00edas-Burillo d\na Laboratorio de An\u00e1lisis Cl\u00ednicos, Hospital General La Mancha Centro, Alc\u00e1zar de San Juan, Ciudad Real, Espa\u02dcna b Laboratorio de Microbiolog\u00eda, Hospital General La Mancha Centro, Alc\u00e1zar de San Juan, Ciudad Real, Espa\u02dcna c Servicio de Medicina Interna, Hospital General La Mancha Centro, Alc\u00e1zar de San Juan, Ciudad Real, Espa\u02dcna d Unidad de Apoyo a la Investigaci\u00f3n, Hospital General La Mancha Centro, Alc\u00e1zar de San Juan, Ciudad Real, Espa\u02dcna\n\ninformaci\u00f3n del art\u00edculo\nHistoria del art\u00edculo: Recibido el 10 de febrero de 2016 Aceptado el 19 de junio de 2016 On-line el xxx\nPalabras clave: Tuberculosis Diagn\u00f3stico Reacci\u00f3n en cadena de la polimerasa Sensibilidad y especi\ufb01cidad An\u00e1lisis de coste-efectividad Evaluaci\u00f3n econ\u00f3mica\n\nr e s u m e n\nIntroducci\u00f3n/Objetivo: Evaluar mediante un an\u00e1lisis de coste-efectividad la aplicaci\u00f3n de una t\u00e9cnica de biolog\u00eda molecular al diagn\u00f3stico de tuberculosis frente a la alternativa diagn\u00f3stica cl\u00e1sica. M\u00e9todos: Se realiz\u00f3 un an\u00e1lisis de coste-efectividad para evaluar la aplicaci\u00f3n te\u00f3rica de un procedimiento de biolog\u00eda molecular que incluye 2 alternativas de una t\u00e9cnica para la detecci\u00f3n precoz de Mycobacterium tuberculosis Complex y resistencia a rifampicina (alternativa 1: una determinaci\u00f3n a pacientes seleccionados; alternativa 2: 2 determinaciones a todos los pacientes). Ambas alternativas se compararon con el procedimiento habitual de diagn\u00f3stico microbiol\u00f3gico de tuberculosis realizado a 1972 pacientes durante 2008-2012 (microscopia y cultivo). La medida de la efectividad se hizo en QALY y la incertidumbre se trat\u00f3 mediante an\u00e1lisis de sensibilidad univariable, multivariable y probabil\u00edstico. Resultados: Para el m\u00e9todo habitual se obtuvo un valor de 8.588 D /QALY. En la alternativa 1 el gasto fue de 8.487 D /QALY, mientras que en la alternativa 2 el cociente coste-efectivo ascendi\u00f3 a 2.960 D /QALY. La alternativa 2 fue la de mayor e\ufb01ciencia diagn\u00f3stica, alcanzando una reducci\u00f3n del 75% del n\u00famero de d\u00edas que un paciente con tuberculosis permanece sin tratamiento adecuado, as\u00ed como una reducci\u00f3n del 70% del n\u00famero de d\u00edas que un paciente sin tuberculosis permanece ingresado. Conclusi\u00f3n: La aplicaci\u00f3n de una t\u00e9cnica microbiol\u00f3gica molecular en el diagn\u00f3stico de tuberculosis es sumamente coste-efectiva frente al m\u00e9todo habitual. Su introducci\u00f3n en el procedimiento diagn\u00f3stico de rutina supondr\u00eda una mejora en la calidad asistencial de los pacientes al evitar ingresos y tratamientos innecesarios, re\ufb02ej\u00e1ndose en un ahorro econ\u00f3mico al hospital.\n\u00a9 2016 Elsevier Espan\u02dc a, S.L.U. y Sociedad Espan\u02dc ola de Enfermedades Infecciosas y Microbiolog\u0131\u00b4a Cl\u0131\u00b4nica. Todos los derechos reservados.\n\nCost-effectiveness study of the microbiological diagnosis of tuberculosis using geneXpert MTB/RIF\u00ae\n\na b s t r a c t\n\nKeywords: Tuberculosis Diagnosis Real-time polymerase chain reaction Sensitivity and speci\ufb01city Cost-effectiveness analysis Economic evaluation\n\nIntroduction/Objective: To perform a cost-effectiveness analysis of a molecular biology technique for the diagnosis of tuberculosis compared to the classical diagnostic alternative. Methods: A cost-effectiveness analysis was performed to evaluate the theoretical implementation of a molecular biology method including two alternative techniques for early detection of Mycobacterium tuberculosis Complex, and resistance to rifampicin (alternative 1: one determination in selected patients; alternative 2: two determinations in all the patients). Both alternatives were compared with the usual procedure for microbiological diagnosis of tuberculosis (staining and microbiological culture), and was\n\naccomplished on 1,972 patients in the period in 2008-2012. The effectiveness was measured in QALYs,\n\nand the uncertainty was assessed by univariate, multivariate and probabilistic analysis of sensitivity.\n\n\u0b1d Este proyecto ha sido premiado en el V Premio AEFA a la Calidad y a la Innovaci\u00f3n. \u2217 Autor para correspondencia.\nCorreo electr\u00f3nico: marian asencio@yahoo.es (M.\u00c1. Asencio-Egea).\n\nhttp://dx.doi.org/10.1016/j.eimc.2016.06.009 0213-005X/\u00a9 2016 Elsevier Espan\u02dc a, S.L.U. y Sociedad Espan\u02dc ola de Enfermedades Infecciosas y Microbiolog\u0131\u00b4a Cl\u0131\u00b4nica. Todos los derechos reservados.\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n2\n\nARTICLE IN PRESS\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\nResults: A value of D 8,588/QALYs was obtained by the usual method. Total expenditure with the alternative 1 was D 8,487/QALYs, whereas with alternative 2, the cost-effectiveness ratio amounted to D 2,960/QALYs. Greater diagnostic ef\ufb01ciency was observed by applying the alternative 2, reaching a 75% reduction in the number of days that a patient with tuberculosis remains without an adequate treatment, and a 70% reduction in the number of days that a patient without tuberculosis remains in hospital. Conclusion: The implementation of a molecular microbiological technique in the diagnosis of tuberculosis is extremely cost-effective compared to the usual method. Its introduction into the routine diagnostic procedure could lead to an improvement in quality care for patients, given that it would avoid both unnecessary hospitalisations and treatments, and re\ufb02ected in economic savings to the hospital.\n\u00a9 2016 Elsevier Espan\u02dc a, S.L.U. and Sociedad Espan\u02dc ola de Enfermedades Infecciosas y Microbiolog\u0131\u00b4a Cl\u0131\u00b4nica. All rights reserved.\n\nIntroducci\u00f3n\nEspan\u02dc a es un pa\u00eds de baja incidencia de tuberculosis (TB), aunque se diagnostican aproximadamente 5.000 casos al an\u02dc o1. Por ello, es necesario desarrollar actividades que aseguren un diagn\u00f3stico preciso y precoz, as\u00ed como el seguimiento y el cumplimiento de un tratamiento adecuado2,3. Sin embargo, el diagn\u00f3stico microbiol\u00f3gico de TB es complejo. La baciloscopia (BK) es una t\u00e9cnica r\u00e1pida, sencilla y econ\u00f3mica, pero presenta una sensibilidad baja y, por tanto, un n\u00famero elevado de falsos negativos, con el consecuente retraso diagn\u00f3stico, as\u00ed como algunos falsos positivos2. Un retraso en el diagn\u00f3stico incrementa el riesgo de transmitir la infecci\u00f3n y prolongar la enfermedad de los pacientes. Por el contrario, un resultado falso positivo puede causar la prescripci\u00f3n de un tratamiento innecesario, toxicidad farmacol\u00f3gica y selecci\u00f3n de cepas resistentes, as\u00ed como el retraso del diagn\u00f3stico correcto de enfermedades cr\u00f3nicas, condicionando un incremento de la morbilidad y de los costes4.\nLa aplicaci\u00f3n de t\u00e9cnicas de ampli\ufb01caci\u00f3n de \u00e1cidos nucleicos (TAAN) a tiempo real directamente sobre muestras cl\u00ednicas, como Xpert MTB/RIF\u00ae (Xpert), permite la detecci\u00f3n precoz del complejo Mycobacterium tuberculosis (MTBC) y de la resistencia a rifampicina (RIF) con sensibilidad y especi\ufb01cidad elevadas en 2 h5,6. Por esta raz\u00f3n, tanto la Organizaci\u00f3n Mundial de la Salud7 como el European Centre for Disease Prevention and Control8 recomiendan su uso en los casos pulmonares bacil\u00edferos. Pese al relativamente elevado coste econ\u00f3mico de estas t\u00e9cnicas frente a las convencionales, su aplicaci\u00f3n sistem\u00e1tica en pacientes con elevada sospecha de TB permitir\u00eda un ahorro econ\u00f3mico, sobre todo por reducir estancias hospitalarias9.\nDado que los costes del Xpert se presentan como la principal barrera para su implantaci\u00f3n en los procesos diagn\u00f3sticos de TB, el objetivo de este estudio ha sido evaluar mediante un an\u00e1lisis de coste-efectividad 2 alternativas diagn\u00f3sticas de TB que incluyen la tecnolog\u00eda Xpert MTB/RIF\u00ae frente al m\u00e9todo convencional utilizado en nuestro centro.\nM\u00e9todos\nSe hizo un estudio retrospectivo de los 1.972 pacientes que acudieron al Hospital General La Mancha Centro (HGMC) con sospecha de TB entre el 1 de enero de 2008 y el 31 de diciembre de 2012. Se ha considerado como criterio de inclusi\u00f3n la solicitud a pacientes ingresados de 3 esputos en un periodo inferior a 8 d\u00edas, de los cuales se dispone del resultado de BK y cultivo para micobacterias, obteni\u00e9ndose en caso positivo la identi\ufb01caci\u00f3n de especie y la susceptibilidad a los distintos antituberculosos.\nSe ha realizado un an\u00e1lisis de coste-efectividad para evaluar 3 estrategias de diagn\u00f3stico microbiol\u00f3gico de TB. La primera de ellas se basa en el m\u00e9todo habitual de diagn\u00f3stico de TB en nuestro hospital, mientras que las otras 2 estrategias (v\u00edas alternativas 1 y 2) se eval\u00faan mediante la aplicaci\u00f3n te\u00f3rica de la tecnolog\u00eda Xpert MTB/RIF\u00ae (Sunnyvale, CA, EE. UU.) sobre los mismos pacientes. Para\n\ncomparar los costes y la efectividad de las 3 estrategias se construy\u00f3 un \u00e1rbol de decisi\u00f3n usando la aplicaci\u00f3n TreeAge Pro 2011\u00ae (TreeAge Software Inc. Williamstown, MA, EE. UU.).\nEl m\u00e9todo habitual de diagn\u00f3stico de TB en el HGMC consiste en solicitar una radiograf\u00eda de t\u00f3rax, una prueba de Mantoux y obtener de cada paciente 3 esputos consecutivos y realizar BK (auramina-rodamina) y cultivo en el medio l\u00edquido MGIT (BD, S.A.), considerando como patr\u00f3n de oro diagn\u00f3stico el aislamiento de MTBC3. Mientras se obtienen los resultados de la BK, el paciente permanece hospitalizado en aislamiento respiratorio. Si el resultado de alguna BK es positivo, el paciente puede recibir el alta precoz (antes de 7 d\u00edas) con tratamiento emp\u00edrico y seguimiento en consulta en espera de los resultados del cultivo y antibiograma. Por el contrario, en caso de que las 3 BK sean negativas se distinguen 2 situaciones:\n1. En pacientes con sospecha cl\u00ednica elevada de TB (SCETB) (contexto cl\u00ednico-epidemiol\u00f3gico compatible y radiolog\u00eda sugerente de TB) se completa el estudio con la realizaci\u00f3n de una tomograf\u00eda axial computarizada (TAC) y la obtenci\u00f3n de un lavado broncoalveolar y/o aspirado bronquial mediante \ufb01brobroncoscopia, manteniendo al paciente sin tratamiento antituberculoso; en caso de que alguna BK sea positiva, el paciente es tratado emp\u00edricamente, habiendo permanecido en aislamiento una media de 2 semanas, mientras que los pacientes que presentan SCETB y resultados de BK negativos comienzan igualmente el tratamiento emp\u00edrico, pero permanecen ingresados sin aislamiento una media de 3 semanas hasta recibir el resto de pruebas diagn\u00f3sticas (demora de pruebas de imagen, anatom\u00eda patol\u00f3gica, cultivos).\n2. En el caso de que las 3 BK sean negativas sin SCETB, el paciente es ingresado sin condiciones de aislamiento una media de 11 d\u00edas en los que se observa su evoluci\u00f3n, siendo dado de alta sin tratamiento antituberculoso, con diagn\u00f3stico o no de otra enfermedad respiratoria.\nLos tiempos de ingreso en cada situaci\u00f3n han sido facilitados por el Servicio de Medicina Interna y son una estimaci\u00f3n basada en la pr\u00e1ctica cl\u00ednica habitual tras 20 an\u02dc os de experiencia en el manejo de pacientes con TB.\nLa primera alternativa propuesta consiste en incorporar el diagn\u00f3stico precoz de TB mediante Xpert de una muestra de esputo en pacientes con BK positiva o SCETB. El modelo sigue el mismo procedimiento que el m\u00e9todo habitual, pero sin considerar la obtenci\u00f3n de muestras por m\u00e9todos invasivos. En el caso de que alguna BK fuera positiva se realizar\u00eda una determinaci\u00f3n de Xpert para con\ufb01rmar que realmente es MTBC. En caso de que el resultado del Xpert sea negativo y exista una SCETB, se considerar\u00eda la posibilidad de tratar al paciente emp\u00edricamente, permaneciendo ingresado sin aislamiento una media de 3 semanas hasta la obtenci\u00f3n de los resultados microbiol\u00f3gicos. En el caso de que las 3 BK sean negativas y el paciente presente SCETB se realizar\u00eda el an\u00e1lisis mediante Xpert del\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n\nARTICLE IN PRESS\n\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\nTabla 1 Variables utilizadas en la construcci\u00f3n del \u00e1rbol de decisi\u00f3n\nVariable\nCoste por d\u00eda de estancia (5 primeros d\u00edas) Coste por d\u00eda de estancia (a partir del 6.\u25e6 d\u00eda) Coste por equipos de protecci\u00f3n respiratoria por d\u00eda Coste por consulta sucesiva Coste por BK Coste por cultivo de micobacterias Coste por \ufb01brobroncoscopia Coste por tomograf\u00eda Coste identi\ufb01caci\u00f3n/antibiograma Coste laboratorio por consultas sucesivas Coste por determinaci\u00f3n Xpert Coste tratamiento TB resistente a rifampicina por d\u00eda Coste tratamiento TB multirresistente por d\u00eda Coste tratamiento TB habitual (sensible) por d\u00eda Coste tratamiento TB sensible a rifampicina y resistente a\nisoniazida por d\u00eda Sensibilidad del Xpert para la detecci\u00f3n de M. tuberculosis\nen pacientes con BK negativa Sensibilidad del Xpert para la detecci\u00f3n de M. tuberculosis\nen pacientes con BK positiva Sensibilidad del Xpert en la segunda determinaci\u00f3n para la\ndetecci\u00f3n de M. tuberculosis en pacientes con BK negativa Sensibilidad del Xpert en la segunda determinaci\u00f3n para la\ndetecci\u00f3n de M. tuberculosis en pacientes con BK positiva Especi\ufb01cidad del Xpert para la detecci\u00f3n de M. tuberculosis Especi\ufb01cidad del Xpert en la segunda determinaci\u00f3n para\nla detecci\u00f3n de M. tuberculosis Especi\ufb01cidad del Xpert para la detecci\u00f3n de resistencia a\nrifampicina Sensibilidad del Xpert para la detecci\u00f3n de resistencia a\nrifampicina Probabilidad de toxicidad tratamiento multirresistente Probabilidad de toxicidad tratamiento no multirresistente Nivel de utilidad en poblaci\u00f3n general Utilidad de hospitalizaci\u00f3n Utilidad de hospitalizaci\u00f3n en condiciones de aislamiento Utilidad de estancia domiciliaria Utilidad en pacientes con TB activa tratada correctamente Utilidad en pacientes con TB activa no tratada\ncorrectamente Utilidad de toxicidad por tratamiento prescrito en\npacientes sin TB Utilidad de toxicidad por tratamiento prescrito en\npacientes con TB\n\nValor medio\n524,56 D 472,10 D 2,1 D 71,91 D 0,48 D 12,34 D 134,02 D 113 D 0 D 20 D 65 D 0,4 D 5,5 D 0,3 D 0,4 D\n0,775\n0,987\n0,775\n0,987\n0,982 0,982\n0,97\n0,941\n0,3 0,045 0,86 U general-0,4 U general-0,5 U general U general-0,1 U general-0,39\nU general-0,16\nU general-0,25\n\nValor inferior\n0,77 0,98 0,77 0,98 0,978 0,978 0,96 0,916\n\nValor superior\n0,778 0,992 0,778 0,992 0,992 0,992 0,977 0,96\n\nDistribuci\u00f3n\nTriangular Triangular Triangular Triangular Triangular Triangular Triangular Triangular LogNormal\n\n3\nFuente DOCM18 DOCM18 DOCM18 DOCM18 HGMC HGMC DOCM18 DOCM18 HGMC HGMC HGMC Caminero19 Caminero19 Caminero19 Caminero19\nChang et al.20\nChang et al.20\nAsumido\nAsumido\nChang20 Asumido\nChang et al.20\nChang et al.20\nHughes et al.11 Hughes et al.11 Kind et al.21 Holland et al.22 Estimaci\u00f3n Estimaci\u00f3n Tan et al.23 Khan et al. 24\nHolland et al.22\nHolland et al.22\n\n\u00faltimo esputo, mientras que en pacientes con 3 BK negativas y sin SCETB se aplicar\u00eda el m\u00e9todo convencional. Si el resultado del Xpert es positivo, independientemente del resultado de la BK, el paciente recibir\u00eda tratamiento emp\u00edrico y el alta precoz, tal como suced\u00eda con el m\u00e9todo habitual, pero con la posibilidad de adecuar el tratamiento antituberculoso al disponer del dato de resistencia a RIF.\nLa segunda alternativa propuesta incluye la posibilidad de realizar 2 determinaciones con Xpert a todos los pacientes del estudio, con independencia de la sospecha cl\u00ednica y del resultado de la BK. Si el resultado del Xpert es positivo se procede igual que en la alternativa anterior. En el caso de un resultado Xpert negativo se realizar\u00eda una segunda determinaci\u00f3n con otra muestra. Los pacientes con BK positiva y 2 determinaciones Xpert negativas ser\u00edan tratados emp\u00edricamente hasta la obtenci\u00f3n del resultado del cultivo y del antibiograma. Los pacientes con las 3 BK negativas y las 2 determinaciones de Xpert negativas ser\u00edan dados de alta sin tratamiento hasta la obtenci\u00f3n del resultado del cultivo.\nEl modelo incluye varios tipos de variables. Por una parte, se de\ufb01nen variables de car\u00e1cter bibliogr\u00e1\ufb01co aplicadas a todos los pacientes como costes, utilidades y probabilidades (tabla 1). Se de\ufb01nen tambi\u00e9n variables num\u00e9ricas estimadas por aplicaci\u00f3n de los procedimientos cl\u00ednicos habituales a cada tipo de paciente y de\ufb01nidas para cada rama del \u00e1rbol de decisi\u00f3n (tabla 2). Por \u00faltimo, se han calculado variables que permiten estimar el coste y la utilidad para cada rama del \u00e1rbol y que utilizan para su de\ufb01nici\u00f3n\n\nvariables bibliogr\u00e1\ufb01cas y num\u00e9ricas (tabla 2). No se ha incluido la mortalidad como variable a medir debido a que en la poblaci\u00f3n estudiada no se dio ning\u00fan caso de fallecimiento asociado a la patolog\u00eda tuberculosa.\nLa medida de la efectividad se hizo en Quality Adjusted-Life-Years o an\u02dc os de vida ajustados por calidad (QALY). El c\u00e1lculo de la utilidad \ufb01nal (equivalente a la efectividad) de cada rama del \u00e1rbol se realiza mediante la ponderaci\u00f3n anual de las utilidades diarias para cada paciente desde la sospecha diagn\u00f3stica hasta la \ufb01nalizaci\u00f3n del tratamiento, o bien hasta que el diagn\u00f3stico de TB es descartado. Para hacer este c\u00e1lculo se estima la utilidad del paciente por d\u00eda hasta completar un an\u02dc o. En el caso de pacientes con una duraci\u00f3n del proceso superior al an\u02dc o se considera que la utilidad anual presenta el mismo valor que la utilidad total. Se identi\ufb01can 2 utilidades \ufb01nales: la utilidad en los pacientes con TB y en aquellos sin TB.\nLa incertidumbre se trat\u00f3 mediante an\u00e1lisis de sensibilidad univariable, multivariable y de tipo tornado10.\nResultados\nDe los 1.972 pacientes estudiados se con\ufb01rm\u00f3 la enfermedad mediante cultivo en 69 de ellos (3,5%), aunque hubo 177 cultivos positivos; por tanto, se aislaron 108 micobacterias at\u00edpicas. El 71% (49/69) de los casos de TB se pudieron detectar precozmente mediante BK. Hubo 4 cepas de MTBC resistentes a los\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n4\n\nARTICLE IN PRESS\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\n1.781 no SCETB\n\n1.673 cultivos negativos\n108 cultivos positivos\n\n93 MNT 15 MTBC\n\n1.908 BK negativa\n\n2 confirmados por FBC\n\n1 MNT 1 MTBC\n\n1.972 pacientes 64 BK positiva\n\n127 SCETB\n58 cultivos positivos 6 cultivos negativos\n\n125 no confirmados por\nFBC\n9 MNT\n49 MTBC\n\n9 cultivos positivos\n116 cultivos negativos\n\n5 MNT 4 MTBC\n\nFigura 1. Distribuci\u00f3n de pacientes en el m\u00e9todo habitual. BK: baciloscopia; FBC: \ufb01brobroncoscopia; MNT: micobacteria no tuberculosa; MTBC: Mycobacterium tuberculosis Complex; SCETB: sospecha cl\u00ednica elevada de tuberculosis.\n\nantituberculosos de primera l\u00ednea (8%), 2 de ellas resistentes a isoniazida, una resistente a RIF y otra multirresistente. El grupo m\u00e1s numeroso fue el de pacientes con sospecha no con\ufb01rmada de TB, sin SCETB y con resultado de BK negativo, que permanecieron ingresados una media de 11 d\u00edas (3 en aislamiento). En la \ufb01gura 1 se muestra una descripci\u00f3n detallada de los resultados obtenidos mediante el m\u00e9todo habitual.\nTabla 2 Variables medidas para cada paciente del \u00e1rbol de decisi\u00f3n\nVariable N\u00famero de d\u00edas con estancia en condiciones de aislamiento N\u00famero de BK realizadas para el diagn\u00f3stico N\u00famero de consultas sucesivas N\u00famero de cultivos realizados en diagn\u00f3stico N\u00famero medio de d\u00edas desde la llegada al laboratorio de la primera muestra\nhasta el resultado del cultivo N\u00famero total de d\u00edas de estancia hospitalaria N\u00famero de d\u00edas en estancias inferiores a 6 d\u00edas N\u00famero de estancias a partir de 6 d\u00edas N\u00famero de \ufb01brobroncoscopias y TAC realizadas en el proceso diagn\u00f3stico N\u00famero de d\u00edas desde la emisi\u00f3n del cultivo positivo por el laboratorio de\nmicrobiolog\u00eda hasta la obtenci\u00f3n de resultados de identi\ufb01caci\u00f3n y antibiograma N\u00famero de identi\ufb01caciones y antibiogramas realizados en el paciente N\u00famero de d\u00edas con tratamiento para MTBC sensible a todos los f\u00e1rmacos de primera l\u00ednea salvo rifampicina N\u00famero de d\u00edas con tratamiento para MTBC multirresistente N\u00famero de d\u00edas con tratamiento habitual sensible a todos los f\u00e1rmacos de primera l\u00ednea N\u00famero de d\u00edas con tratamiento para MTBC sensible a todos los f\u00e1rmacos de primera l\u00ednea salvo isoniazida N\u00famero de determinaciones realizadas por Xpert D\u00edas con el paciente a la espera de tratamiento y/o diagn\u00f3stico D\u00edas de tratamiento para MTBC multirresistentes en los que el tratamiento es incorrecto D\u00edas de tratamiento para MTBC no multirresistente en los que el tratamiento es incorrecto D\u00edas de tratamiento para MTBC multirresistentes con tratamiento correcto D\u00edas de tratamiento para MTBC no multirresistente con tratamiento correcto\n\nE\ufb01ciencia diagn\u00f3stica\nEl \u00e1rbol de decisi\u00f3n con las 3 alternativas se muestra en la \ufb01gura 2. Una vez calculadas las probabilidades para cada rama del \u00e1rbol de decisi\u00f3n se estimaron los consumos de recursos por los 1.972 pacientes al aplicar las distintas alternativas (tabla 3). De los 3 m\u00e9todos evaluados, las 2 alternativas al m\u00e9todo habitual presentaron mayor e\ufb01ciencia diagn\u00f3stica, con una disminuci\u00f3n del porcentaje de pacientes con TB que han permanecido sin tratamiento hasta la obtenci\u00f3n del resultado del cultivo (porcentaje de falsos negativos). Adem\u00e1s, en la alternativa 2 se observa una disminuci\u00f3n del n\u00famero de pacientes sin TB que reciben tratamiento inadecuado hasta la obtenci\u00f3n de los resultados del cultivo y/o antibiograma (\ufb01g. 3).\nEvaluaci\u00f3n coste-efectividad en pacientes con sospecha de tuberculosis\nPara el m\u00e9todo habitual se ha obtenido un cociente de 8.588 D /QALY. La estimaci\u00f3n del coste en pacientes con TB fue de 6.329 D /QALY y de 8.659 D /QALY en pacientes sin TB. En la alternativa 1 se obtiene un coste total de 8.487 D /QALY, siendo el gasto por paciente con TB de 7.222 D /QALY, y por paciente sin TB, de 8.532 D /QALY. En la alternativa 2 se observa un cociente costeefectivo de 2.969 D /QALY. Los pacientes con TB presentaron un gasto de 5.510 D /QALY, mientras que el gasto para aquellos en los que se descart\u00f3 el diagn\u00f3stico de TB fue de 2.878 D /QALY. Por tanto, la alternativa 2 presenta un ratio coste-efectivo dominante.\nEstudio de sensibilidad\nUn an\u00e1lisis de sensibilidad tipo tornado indic\u00f3 que el par\u00e1metro con mayor in\ufb02uencia en el gasto por paciente fue el coste por estancia durante los 5 primeros d\u00edas. La disminuci\u00f3n de 300 D /d\u00eda\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n\nARTICLE IN PRESS\n\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\n5\n\nsupone una reducci\u00f3n del coste por paciente de 1.614 D /d\u00eda al aplicar la alternativa 2, frente a los 6.194 D obtenidos al aplicar el m\u00e9todo habitual. La segunda variable con mayor in\ufb02uencia en el modelo fue el coste de las medidas de protecci\u00f3n. Se han tomado como valores posibles el intervalo entre 2,1 y 200 D , en donde el m\u00ednimo valor se corresponde con el coste de las medidas de\n\nbarrera (mascarillas) y el m\u00e1ximo a permanecer en una habitaci\u00f3n en aislamiento. La variaci\u00f3n en este par\u00e1metro tuvo una in\ufb02uencia mayor en la alternativa 2, si bien esta contin\u00faa siendo la alternativa dominante. La variaci\u00f3n en el coste de la determinaci\u00f3n de Xpert in\ufb02uy\u00f3 de manera m\u00e1s notable, aunque moderada, en la alternativa 2. Tambi\u00e9n se ha realizado el estudio de la sensibilidad para los\n\nM\u00e9todo habitual\n\nNo SCETB\n\nCultivo de micobacterias negativo\n\nCultivo +\n\nSensible rifampicina Micobacterias no tubeculosas\n\nBaciloscopia \u2013 SCETB\n\nFibrobronocoscopia +\n\nCultivo de micobacterias negativo\n\nCultivo +\n\nSensible rifampicina Micobacterias no tubeculosas\n\nCultivo de micobacterias negativo\n\nFibrobronocoscopia \u2013\n\nSen Rif - Res INH\n\nCultivo +\n\nSensible rifampicina\n\nMicobacterias no tubeculosas\n\nSensible rifampicina Res Rif\n\nCultivo +\n\nRes Rif - multirresistente\n\nBaciloscopia +\n\nSen Rif - Res INH Micobacterias no tubeculosas\n\nCultivo de micobacterias negativo\n\nSensible rifampicina\n\nXpert + Xpert \u2013\n\nXpertRRif XpertSRif\n\nBaciloscopia +\n\nPacientes con TB\n\nDiagn\u00f3stico con GenXpert\n\nBaciloscopia \u2013\n\nRes Rif Res Rif - multirresistente Sen Rif - Res INH SCETB\nNo SCETB\n\nClone 1: Determinacion simple Xpert\n\nClone 1: Determinacion simple Xpert\n\nClone 1: Determinacion simple Xpert\n\nSensible rifampicina\n\nClone 1: Determinacion simple Xpert\n\nSen Rif - Res INH\n\nClone 1: Determinacion simple Xpert\n\nSensible rifampicina\n\nBaciloscopia +\nPacientes sin TB\nBaciloscopia \u2013\n\nClone 1: Determinacion simple Xpert\n\nSCETB\n\nClone 1: Determinacion simple Xpert\n\nNo SCETB\n\nCultivo de micobacterias negativo Micobacterias no tuberculosas\n\nFigura 2. \u00c1rbol de decisi\u00f3n con las 3 alternativas diagn\u00f3sticas: a) m\u00e9todo habitual; b) alternativa 1; c) alternativa 2.\n\nTotCos_total \\ U_NO_TB TotCos_total \\ U_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_NO_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_TB\nTotCos_total \\ U_TB\nTotCos_total \\ U_NO_TB TotCos_total \\ U_NO_TB\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n6\n\nARTICLE IN PRESS\n\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\nXpert +\n\nXpertRRif XpertSRif\n\nBaciloscopia + Pacientes con TB\n\nDiagn\u00f3stico con dobel genXpert\n\nBaciloscopia \u2013\n\nBaciloscopia +\nPacientes sin TB\nBaciloscopia \u2013\n\nSensible rifampicina\n\n2 Xpert \u2013\n\n2Xpert + 2Xpert \u2013\n\nRes Rif\n\nClone 2: Determinacion doble Xpert\n\nRes Rif - multirresistente\n\nClone 2: Determinacion doble Xpert\n\nSen Rif - Res INH\n\nClone 2: Determinacion doble Xpert\n\nSensible rifampicina\n\nClone 2: Determinacion doble Xpert\n\nSen Rif - Res INH\n\nClone 2: Determinacion doble Xpert\n\nCultivo de micobacterias negativo\n\nClone 2: Determinacion doble Xpert\n\nMicobacterias no tuberculosas Clone 2: Determinacion doble Xpert\n\nCultivo de micobacterias negativo\n\nClone 2: Determinacion doble Xpert\n\nMicobacterias no tuberculosas Clone 2: Determinacion doble Xpert\n\nFigura 2. (Continuaci\u00f3n )\n\nXpertRRif XpertSRif\n\nTabla 3 Estimaciones de recursos consumidos por las 3 alternativas evaluadas\nN\u00famero total de d\u00edas de pacientes con TB sin tratamiento N\u00famero total de d\u00edas con prescripci\u00f3n a los pacientes de\ntratamiento habitual (para MTBC no multirresistente) incorrecto N\u00famero total de d\u00edas en los que el paciente est\u00e1 en aislamiento N\u00famero de consultas sucesivas D\u00edas totales de estancia hospitalaria N\u00famero de determinaciones de Xpert\n\nM\u00e9todo habitual\n522 5.974\n7.533 462\n29.124 0\n\nAlternativa 1\n403 3.177\n7.039 460\n28.807 206\n\nTotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_TB TotCos_total \\ U_TB\nAlternativa 2 126\n2.233 8.200 2.135 8.588 3.847\n\nvalores de la utilidad (QALY) de la poblaci\u00f3n general, analizando el intervalo comprendido entre 0,80 y 1 QALY.\nCon la aplicaci\u00f3n del m\u00e9todo de Monte Carlo se obtuvo una curva de aceptabilidad que indica que la alternativa 2 es la m\u00e1s rentable, con independencia del umbral elegido, con una probabilidad del 100%.\n\n25\n\nPorcentaje de pacientes\n\n20\n\n15\n\n10\n\n5\n\n0 M\u00e9todo\n\nAlternativa 1\n\nAlternativa 2\n\n% FN\n\n21,7\n\n2,6\n\n1,5\n\n% FP\n\n7,2\n\n7,2\n\n3,5\n\nFigura 3. Porcentaje de pacientes con tratamiento emp\u00edrico incorrecto.\n\nDiscusi\u00f3n\nLos resultados de este trabajo sugieren que, en nuestro medio y con las suposiciones realizadas en el modelo, el diagn\u00f3stico de TB con la tecnolog\u00eda Xpert es m\u00e1s coste-efectiva que el procedimiento convencional. De las 2 alternativas al m\u00e9todo actual ha resultado dominante la que incluye la posibilidad de realizar 2 determinaciones de Xpert, de manera que la aplicaci\u00f3n de esta tecnolog\u00eda como m\u00e9todo de cribado evidenciar\u00eda una mejora en la calidad de vida de los pacientes con TB por permitir iniciar un tratamiento adecuado sin demoras, as\u00ed como una disminuci\u00f3n del gasto hospitalario derivado de la reducci\u00f3n del n\u00famero de estancias de los pacientes sin TB.\nEl estudio de las 3 ramas del \u00e1rbol de decisi\u00f3n muestra que hay diferencias entre los QALY medidos en los 3 procedimientos diagn\u00f3sticos, con un mayor n\u00famero de QALY ganados por aplicaci\u00f3n de la alternativa 2 (reducci\u00f3n del 70% de las estancias hospitalarias y del 75% de d\u00edas sin tratamiento adecuado). Sin embargo, la diferencia m\u00e1s notable reside en la disminuci\u00f3n del coste por paciente, un 65% menor en la alternativa 2 respecto al m\u00e9todo habitual. Esta disminuci\u00f3n de costes permitir\u00eda un ahorro te\u00f3rico anual de 1,8 millones de euros debido, sobre todo, a una reducci\u00f3n del n\u00famero de estancias hospitalarias asociadas a la sospecha de un proceso tuberculoso.\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n\nARTICLE IN PRESS\n\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\n7\n\nDebido a que la sensibilidad de Xpert en pacientes con BK negativa es relativamente baja6,10-12, no se recomienda su aplicaci\u00f3n a no ser que exista una SCETB, puesto que de lo contrario se reducir\u00eda el valor predictivo positivo de la t\u00e9cnica13. Algunos estudios muestran un aumento de la sensibilidad de este test al realizar una segunda determinaci\u00f3n en pacientes con resultado negativo14. De esta manera se justi\ufb01car\u00eda la utilizaci\u00f3n de Xpert como t\u00e9cnica de cribado, con el inconveniente de aumentar el gasto de laboratorio y el n\u00famero de falsos positivos con respecto a la alternativa 1, aunque todav\u00eda resultar\u00eda inferior a los falsos positivos obtenidos con el m\u00e9todo actual (75 vs. 200). Adem\u00e1s, seg\u00fan algunos autores el dan\u02dc o de no tratar a un paciente con TB es mayor que tratar con antituberculosos a un paciente sin TB15. Por otra parte, tras 2 resultados negativos de Xpert se podr\u00eda descartar el diagn\u00f3stico de TB, e incluso podr\u00eda dejar de realizarse el cultivo microbiol\u00f3gico, ya que el valor predictivo negativo de la prueba es cercano al 100%16, si bien esta posibilidad no ha sido evaluada en nuestro estudio. Por esta raz\u00f3n podr\u00eda parecer contradictorio haber considerado en la alternativa 2 prescribir tratamiento emp\u00edrico a los pacientes con BK+ y 2 resultados del Xpert negativos; sin embargo, se ha considerado esta posibilidad por tratarse del peor escenario que podr\u00edamos encontrarnos, que adem\u00e1s incluye la posible resistencia de algunos cl\u00ednicos a dejar de tratar a un paciente con BK+, as\u00ed como la posibilidad de una infecci\u00f3n por una micobacteria no tuberculosa (en este caso habr\u00eda que individualizar la terapia). La realizaci\u00f3n de una tercera determinaci\u00f3n de Xpert, tras 2 resultados negativos, podr\u00eda ser desaconsejable, ya que el n\u00famero de falsos positivos se estima superior a 100.\nOtros estudios de coste-efectividad de Xpert realizados en pa\u00edses en v\u00edas de desarrollo con elevada carga de TB muestran que Xpert es coste-efectivo y su introducci\u00f3n podr\u00eda suponer un cambio importante en la morbimortalidad de la infecci\u00f3n mediante la mayor detecci\u00f3n de casos y, por tanto, de tratamientos dirigidos5,6. La aplicaci\u00f3n de esta tecnolog\u00eda en zonas de baja prevalencia como Espan\u02dc a reducir\u00eda su rentabilidad, si bien su potencial en situaciones de mayor prevalencia ser\u00eda considerable (poblaci\u00f3n inmigrante, sin techo, usuarios de drogas por v\u00eda parenteral y pacientes VIH y, en general, cuando existe elevada sospecha de TB)10. Nuestro estudio muestra que la aplicaci\u00f3n te\u00f3rica de esta t\u00e9cnica a cualquier paciente con sospecha de TB es m\u00e1s coste-efectiva que su aplicaci\u00f3n solo a pacientes seleccionados, ya que el hecho de obtener un resultado \ufb01able y r\u00e1pido que descarte la infecci\u00f3n da lugar a altas hospitalarias, independientemente de la verdadera enfermedad del paciente. As\u00ed, la diferencia fundamental entre la alternativa 2 y el m\u00e9todo habitual radica en la forma de considerar a un paciente sin SCETB, de manera que, seg\u00fan el m\u00e9todo habitual, este tipo de pacientes permanecen ingresados, mientras que los pacientes con 2 Xpert negativos son dados de alta precozmente.\nUna limitaci\u00f3n del estudio es la consideraci\u00f3n del coste de la estancia hospitalaria sin condiciones de aislamiento respiratorio. Sin embargo, en el estudio de sensibilidad se ha tenido en cuenta la posibilidad de un aumento del coste hospitalario debido al establecimiento de condiciones de aislamiento, comport\u00e1ndose este factor como el segundo m\u00e1s in\ufb02uyente en la estimaci\u00f3n de costes. No obstante, la cuanti\ufb01caci\u00f3n del ahorro obtenido con la alternativa 2 ha de tomarse con cautela, ya que no se ha considerado la posibilidad de hospitalizaci\u00f3n del paciente para realizar el diagn\u00f3stico y tratamiento de la enfermedad no tuberculosa. Adem\u00e1s, se observa que la aplicaci\u00f3n de la alternativa 2 supone un aumento importante del n\u00famero de consultas sucesivas, para asegurar la revisi\u00f3n de los pacientes que, a pesar de la sospecha de TB, fueron dados de alta sin diagn\u00f3stico y, por ello, sin tratamiento. Por otra parte, podr\u00eda interpretarse como una mejora en la atenci\u00f3n sanitaria al paciente por una mayor vigilancia a un coste total menor, contribuyendo adem\u00e1s a la descongesti\u00f3n de los centros hospitalarios si el seguimiento se realizara en los centros de atenci\u00f3n primaria2.\n\nConsideramos que en estudios futuros ser\u00eda conveniente evaluar el bene\ufb01cio real que la pr\u00e1ctica cl\u00ednica con la aplicaci\u00f3n de la tecnolog\u00eda Xpert tendr\u00eda en los costes y la calidad de vida de los pacientes con sospecha de TB. Asimismo, existen en el mercado otras TAAN disponibles para ser evaluadas como alternativas17.\nEn conclusi\u00f3n, este estudio sugiere que la aplicaci\u00f3n de la tecnolog\u00eda Xpert en el diagn\u00f3stico de TB es sumamente coste-efectiva comparada con el m\u00e9todo convencional. El impacto de introducir la tecnolog\u00eda Xpert abarca el \u00e1mbito econ\u00f3mico y sanitario, de manera que su aplicaci\u00f3n supondr\u00eda una mejora en la calidad asistencial de los pacientes por evitar estancias y tratamientos innecesarios, permitiendo adem\u00e1s iniciar un tratamiento precoz dirigido, romper la cadena de transmisi\u00f3n de la infecci\u00f3n y conseguir un ahorro econ\u00f3mico considerable para al hospital.\nCon\ufb02icto de intereses\nLos autores declaran no tener con\ufb02icto de intereses.\nBibliograf\u00eda\n1. Rodr\u00edguez E, D\u00edas O, Hern\u00e1ndez G, Tello O. Situaci\u00f3n de la tuberculosis en Espan\u02dc a. Casos de tuberculosis declarados a la Red Nacional de Vigilancia Epidemiol\u00f3gica en 2010. Bolet\u00edn Epidemiol\u00f3gico Semanal. 2012;20:26\u201341.\n2. Gu\u00eda de Pr\u00e1ctica Cl\u00ednica sobre el Diagn\u00f3stico, el Tratamiento y la Prevenci\u00f3n de la Tuberculosis. Plan de Calidad para el Sistema Nacional de Salud del Ministerio de Sanidad, Pol\u00edtica Social e Igualdad. En: Ag\u00e8ncia d\u2019Informaci\u00f3 AiQeSAdC, ed. Gu\u00edas de Pr\u00e1ctica Cl\u00ednica en el SNS 2009.\n3. Gonz\u00e1lez-Mart\u00edn J, Garc\u00eda-Garc\u00eda JM, Anibarroc L, Vidald R, Estebane J, Blanquerf R, et al. Documento de consenso sobre diagn\u00f3stico, tratamiento y prevenci\u00f3n de la tuberculosis. Enferm Infecc Microbiol Clin. 2010;28:297.e1\u201320.\n4. Davies PD, Pai M. The diagnosis and misdiagnosis of tuberculosis. Int J Tuberc Lung Dis. 2008;12:1226\u201334.\n5. Steingart KR, Sohn H, Schiller I, Kloda LA, Boehme CC, Pai M, et al. Xpert MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Sys Rev. 2013;1(CD009593).\n6. Vassall A, van Kampen S, Sohn H, Michael JS, John KR, den Boon S, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: A cost-effectiveness analysis. PLoS medicine. 2011;8:e1001120.\n7. WHO. Automated Real-Time Nucleic Acid Ampli\ufb01 Cation Technology for Rapid and Simultaneous Detection of Tuberculosis and Rifampicin Resistance: Xpert MTB/RIF System: Policy Statement. Geneva: World Health Organization; 2011.\n8. European Centre for Disease Prevention and Control. ERLN-TB expert opinion on the use of the rapid molecular assays for the diagnosis of tuberculosis and detection of drug-resistance. Stockholm: ECDC; 2013.\n9. Menzies NA, Cohen T, Lin HH, Murray M, Salomon JA. Population health impact and cost-effectiveness of tuberculosis diagnosis with Xpert MTB/RIF: a dynamic simulation and economic evaluation. PLoS medicine. 2012;9: e1001347.\n10. Drummond MF, Sculpher MJ, Claxton K, Stoddart G, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes. Oxford: Oxford University Press; 2015.\n11. Hughes R, Wonderling D, Li B, Higgins B. The cost effectiveness of nucleic acid ampli\ufb01cation techniques for the diagnosis of tuberculosis. Respir Med. 2012;106:300\u20137.\n12. Roos BR, van Cleeff MR, Githui WA, Kivihya-Ndugga L, Odhiambo JA, Kibuga DK, et al. Cost-effectiveness of the polymerase chain reaction versus smear examination for the diagnosis of tuberculosis in Kenya: A theoretical model. Int J Tuberc Lung Dis. 1998;2:235\u201341.\n13. Catanzaro A, Perry S, Clarridge JE, Dunbar S, Goodnight-White S, LoBue PA, et al. The role of clinical suspicion in evaluating a new diagnostic test for active tuberculosis: Results of a multicenter prospective trial. JAMA. 2000;283: 639\u201345.\n14. Lawn SD, Brooks SV, Kranzer K, Nicol MP, Whitelaw A, Vogt M, et al. Screening for HIV-associated tuberculosis and rifampicin resistance before antiretroviral therapy using the Xpert MTB/RIF assay: A prospective study. PLoS medicine. 2011;8:e1001067.\n15. Moreira J, Bisig B, Muwawenimana P, Basinga P, Bisof\ufb01 Z, Haegeman F, et al. Weighing harm in therapeutic decisions of smear-negative pulmonary tuberculosis. Med Decis Making. 2009;29:380\u201390.\n16. Dorman SE, Chihota VN, Lewis JJ, Shah M, Clark D, Grant AD, et al. Performance characteristics of the Cepheid Xpert MTB/RIF test in a tuberculosis prevalence survey. PloS one. 2012;7:e43307.\n17. Armand S, Vanhuls P, Delcroix G, Courcol R, Lemaitre N. Comparison of the Xpert MTB/RIF test with an IS6110-TaqMan real-time PCR assay for direct detection of Mycobacterium tuberculosis in respiratory and nonrespiratory specimens. J Clin Microbiol. 2011;49:1772\u20136.\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\nG Model\nEIMC-1562; No. of Pages 8\n\nARTICLE IN PRESS\n\n8\n\n\u00d3. Herr\u00e1ez et al. / Enferm Infecc Microbiol Clin. 2016;xxx(xx):xxx\u2013xxx\n\n18. DOCM. Resoluci\u00f3n de 03/09/2012, de la Direcci\u00f3n Gerencia, sobre precios a aplicar por sus centros sanitarios a terceros obligados al pago o a los usuarios sin derecho a asistencia sanitaria 2012:28987-95.\n19. Caminero JA. Treatment of tuberculosis according to the different pattern of resistance. Med Clin (Barc.). 2010;134:173\u201381.\n20. Chang K, Lu W, Wang J, Zhang K, Jia S, Li F, et al. Rapid and effective diagnosis of tuberculosis and rifampicin resistance with Xpert MTB/RIF assay: A metaanalysis. J Infect. 2012;64:580\u20138.\n21. Kind P, Macran S, Hardman G. UK population norms for EQ-D5. 1999. Discussion paper 172.\n\n22. Holland DP, Sanders GD, Hamilton CD, Stout JE. Costs and cost-effectiveness of four treatment regimens for latent tuberculosis infection. Am J Respir Crit Care Med. 2009;179:1055\u201360.\n23. Tan MC, Marra CA, Sadatsafavi M, Marra F, Moran-Mendoza O, Moadebi S, et al. Cost-effectiveness of LTBI treatment for TB contacts in British Columbia. Value Health. 2008;11:842\u201352.\n24. Khan K, Muennig P, Behta M, Zivin JG. Global drug-resistance patterns and the management of latent tuberculosis infection in immigrants to the United States. N Engl J Med. 2002;347:1850\u20139.\n\nC\u00f3mo citar este art\u00edculo: Herr\u00e1ez \u00d3, et al. Estudio de coste-efectividad del diagn\u00f3stico microbiol\u00f3gico de tuberculosis mediante geneXpert MTB/RIF\u00ae. Enferm Infecc Microbiol Clin. 2016. http://dx.doi.org/10.1016/j.eimc.2016.06.009\n\n\n",
    "authors": [
      "\u00d3scar Herr\u00e1ez",
      "Mar\u00eda \u00c1ngeles Asencio-Egea",
      "Mar\u00eda Huertas-Vaquero",
      "Rafael Carranza-Gonz\u00e1lez",
      "Jes\u00fas Castellanos-Monedero",
      "Mar\u00eda Franco-Huerta",
      "Jos\u00e9 Ram\u00f3n Barber\u00e1-Farr\u00e9",
      "Jos\u00e9 Mar\u00eda Ten\u00edas-Burillo"
    ],
    "doi": "10.1016/j.eimc.2016.06.009",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://linkinghub.elsevier.com/retrieve/pii/S0213005X16301550"
  },
  {
    "key": "4EB8DJ5Z",
    "title": "Cost-effectiveness analysis of the Xpert MTB/RIF assay for rapid diagnosis of suspected tuberculosis in an intermediate burden area",
    "abstract": "Objectives: We examined, from a Hong Kong healthcare providers\u2019 perspective, the cost-effectiveness of rapid diagnosis with Xpert in patients hospitalized for suspected active pulmonary tuberculosis (PTB).",
    "full_text": "Journal of Infection (2015) 70, 409e414\n\nwww.elsevierhealth.com/journals/jinf\nCost-effectiveness analysis of the Xpert MTB/RIF assay for rapid diagnosis of suspected tuberculosis in an intermediate burden area\nJoyce H.S. You a,*, Grace Lui b, Kai Man Kam c, Nelson L.S. Lee b\na School of Pharmacy, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong b Division of Infectious Diseases, Department of Medicine & Therapeutics, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong c Stanley Ho Centre for Emerging Infectious Diseases, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong\nAccepted 30 December 2014 Available online 6 January 2015\n\nKEYWORDS Rapid testing; Tuberculosis; Decision-analytic model; Cost-effectiveness; Hong Kong\n\nSummary Objectives: We examined, from a Hong Kong healthcare providers\u2019 perspective, the cost-effectiveness of rapid diagnosis with Xpert in patients hospitalized for suspected active pulmonary tuberculosis (PTB). Methods: A decision tree was designed to simulate outcomes of three diagnostic assessment strategies in adult patients hospitalized for suspected active PTB: conventional approach, sputum smear plus Xpert for acid-fast bacilli (AFB) smear-negative, and a single sputum Xpert test. Model inputs were derived from the literature. Outcome measures were direct medical cost, one-year mortality rate, quality-adjusted life-years (QALYs) and incremental cost per QALY (ICER). Results: In the base-case analysis, Xpert was more effective with higher QALYs gained and a lower mortality rate when compared with smear plus Xpert by an ICER of USD99. A conventional diagnostic approach was the least preferred option with the highest cost, lowest QALYs gained and highest mortality rate. Sensitivity analysis showed that Xpert would be the most cost-effective option if the sensitivity of sputum AFB smear microscopy was 74%. The probabilities of Xpert, smear plus Xpert and a conventional approach to be cost-effective were 94.5%, 5.5% and 0%, respectively, in 10,000 Monte Carlo simulations.\n\n* Corresponding author. School of Pharmacy, Faculty of Medicine, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong. Tel.: \u00fe852 3943 6830; fax: \u00fe852 2603 5295.\nE-mail address: joyceyou@cuhk.edu.hk (J.H.S. You).\nhttp://dx.doi.org/10.1016/j.jinf.2014.12.015 0163-4453/\u00aa 2015 The British Infection Association. Published by Elsevier Ltd. All rights reserved.\n\n410\n\nJ.H.S. You et al.\n\nConclusions: The Xpert sputum test appears to be a highly cost-effective diagnostic strategy for patients with suspected active PTB in an intermediate burden area like Hong Kong. \u00aa 2015 The British Infection Association. Published by Elsevier Ltd. All rights reserved.\n\nIntroduction\nHong Kong is a developed city with intermediate tuberculosis (TB) burden of 90 per 100,000 population and low HIV prevalence of <0.1%.1,2 The diagnosis of TB in hospitalized patients was reported to be dif\ufb01cult due to atypical manifestation and low sensitivity of sputum acid-fast bacilli (AFB) smear microscopy examination.3 False negative results of sputum smear examination would delay the initiation of early anti-TB therapy. A recent outcome study in Hong Kong found high mortality in adults hospitalized for active TB, and failure to receive early anti-TB treatment during initial assessment was identi\ufb01ed to be an independent predictor associated with increased risk of mortality.4\nXpert MTB/RIF (GeneXpert, Cepheid Inc, Sunnyvale, CA, USA) is a rapid molecular TB diagnostic system for Mycobacterium tuberculosis and resistance to rifampin, allowing diagnosis at the point of treatment. The sensitivity and speci\ufb01city of Xpert to detect M. tuberculosis and rifampin resistance were examined in over 1,700 patients with suspected drug-sensitive or multidrug-resistant (MDR) pulmonary tuberculosis (PTB).5 Xpert was found to provide susceptibility testing of PTB to rifampin resistance in less than 2 h.\nThe decision of whether to administer the rapid test to all patients with suspected PTB or only to patients with negative AFB sputum smears requires consideration of costeffectiveness of each strategy. To facilitate positioning Xpert in the diagnostic algorithm of PTB in hospital setting of Hong Kong, we have conducted this analysis to evaluate the potential economic and clinical outcomes of rapid diagnosis with Xpert in patients hospitalized for suspected active PTB from the perspective of Hong Kong healthcare providers.\nMaterials and methods\nModel design\nA decision tree (Fig. 1) was designed to simulate the outcomes of three clinical assessment strategies in a hypothetical cohort of adult patients hospitalized for suspected\n\nactive PTB based on compatible signs and symptoms or radiographic \ufb01ndings, including: (1) conventional approach (control group, details were described below), (2) sputum microscopy examination plus Xpert (AFB smear plus Xpert group), and (3) single sputum test by Xpert (Xpert group). Three tiers of outcomes were simulated for each study arm: (1) Direct medical cost, (2) one-year mortality rate, and (3) quality-adjusted life-years (QALYs) gained in patients who survived active PTB.\nIn the present model, the patients might or might not be infected with active PTB. The PTB patients might be infected with MDR-TB. Mycobacterial culture and (if culture was positive) drug susceptibility testing would be conducted on sputum specimen for patients in all three arms.\nIn the control group, two sputum microscopy examinations would be initially performed. Smear-negative patients would be followed by clinical diagnosis including (but not limited to) diagnostic antibiotic trial using \ufb01rst-line empirical treatment (broad spectrum beta-lactams and macrolides) for pneumonia. Smear-positive patients and those with positive clinical diagnosis would receive early anti-TB treatment with \ufb01rst-line anti-TB agents (isoniazid, rifampicin, pyrazinamide, ethambutol, or aminoglycoside) according to TB treatment guidelines in Hong Kong.1\nIn the smear plus Xpert group, two initial sputum microscopy examinations would be conducted. AFB smear-negative patients would be tested by Xpert on a single sputum specimen. Smear-positive and Xpert-positive patients would receive early anti-TB treatment. Those who were test-negative would not receive early treatment because of the high negative-predictive value of Xpert.\nIn the Xpert group, all patients would be tested by Xpert on a single sputum specimen. The initiation of anti-TB treatment would be based upon the Xpert results as described above.\nIn all three arms, patients who did not receive early anti-TB treatment would receive late treatment when positive sputum culture of M. tuberculosis was reported. Patients with MDR-TB would receive early second-line anti-TB agents if their specimen was tested to be rifampin-resistant by Xpert or they would receive late second-line anti-TB agents as indicated by the results of drug susceptibility testing. Those without active PTB but\n\nFigure 1 Simpli\ufb01ed decision tree.\n\nXpert MTB/RIF assay for tuberculosis CEA\nreceived early \ufb01rst-line anti-TB treatment as indicated by false positive initial testing or clinical diagnosis would receive anti-TB treatment until PTB was ruled out by culture results. The patients with active PTB, received early or late treatment, might survive or die.\nClinical inputs\nThe clinical inputs of the model were shown in Table 1. A literature search on MEDLINE over the period of 2004e2014 was performed. The selection criteria of clinical studies on diagnosis and treatment of PTB were: (1) reports were written in English; (2) etiology of respiratory illnesses was identi\ufb01ed to be M. tuberculosis, and (3) mortality rate was reported. All articles retrieved by this process were screened for relevance to our model. A manuscript was included if it had data pertaining to the model inputs.\nThe prevalence of active PTB among patients hospitalized with compatible symptoms or radiological abnormalities was estimated to be 35.2%, with median age of 47 years, from a prospective clinical study on various diagnostic approaches for PTB in a Hong Kong acute hospital.6 A\n\n411\nglobal epidemiology study on anti-tuberculosis drug resistance surveillance in 1999e2002 indicated that the prevalence of MDR-TB among TB cases in Hong Kong was 0.8%.7 The sensitivity and speci\ufb01city of microscopic examination of sputum for AFB, clinical diagnosis and Xpert were retrieved from systematic review,8 clinical diagnostic guideline validation study9 and prospective study,5 respectively.\nAn outcome analysis of adults hospitalized for active tuberculosis in a Hong Kong acute hospital reported the one-year mortality rate to be 24.1%. Failure to receive early anti-TB treatment during initial assessment was associated with higher hazard ratio (1.8) of death at one year. Early anti-TB treatment was de\ufb01ned as empirical therapy started during initial hospital admission. Late treatment was started when mycobacterial infection was con\ufb01rmed.4 In the present model, patients with susceptible-PTB would receive early treatment if the treatment was initiated for positive sputum smear, clinical diagnosis or Xpert. Patients with MDR-PTB would receive early treatment if they received second-line anti-TB treatment as indicated by positive Xpert test as rifampin-resistant. It\n\nTable 1 Model inputs.\nClinical inputs Prevalence of active PTB among hospitalized suspects (%) Percentage of multi-drug resistance in TB cases (%) One-year mortality rate (%) Hazard ratio of one-year mortality with late anti-TB treatment Sensitivity of diagnostic test Sputum AFB smear Xpert in culture positive Xpert for smear negative and culture positive Clinical diagnosis for smear negative Xpert for multi-drug resistance Xpert for rifampin-sensitive bacteria Speci\ufb01city of diagnostic test Sputum AFB smear Xpert in culture negative Clinical diagnosis for AFB smear negative\nUtility inputs Utility score 18e64 years 65e85 years Age of patients with active PTB (years) Life expectancy in Hong Kong\nCost inputs (USD)a Cost per test Xpert Sputum smear Mycobacterial culture Drug susceptibility test Diagnostic antibiotic trial Anti-TB treatment per month First-line treatment Second-line treatment\na 1 USD Z 7.8 HK.\n\nBase-case value\n32.5% 0.8% 24.1% 1.83\n50% 92.2% 72.5% 59% 97.5% 98.1%\n97% 99.2% 57%\n0.92 0.84 47 83\n128 7.5 45 86 340\n27 769\n\nRange of sensitivity analysis\n0%e100% 0.5%e1.1% 19.3%e28.9% 1.13e2.96\n40%e80% 90.0%e93.9% 65.4%e78.7% 46%e66% 94.3%e98.9% 96.5%e98.9%\n91%e100% 98.1%e99.6% 25%e75%\ne e 29e73 e\n100e150 6e9 36e54 69e103 272e408\n20e30 615e923\n\nReferences\n6 7 4 4\n8 5 5 9 5 5\n8 5 9\n10 10 6 11 Hong Kong gazette\nEstimation\n\n412\nwas considered as late treatment for patients with MDRPTB who were treated initially with \ufb01rst-line anti-TB treatment and later changed to second-line treatment when the results of drug susceptibility testing became available.\nUtility inputs\nThe expected QALYs gained by patients who survived active PTB within one-year of admission were calculated using agespeci\ufb01c utility value, age of patient and life years gained. The utilities of adults aged 18e64 years and 65e85 years were retrieved from health-related quality of life scores reported in literature.10 The life-years gained were estimated using patient\u2019s age and life expectancy of Hong Kong population.11\nCost inputs\nThe economic analysis was conducted from the perspective of Hong Kong healthcare providers, consisted of direct medical costs of the three diagnostic approaches for suspected active PTB. The medical resources for diagnosis of TB included sputum AFB smear microscopy examinations, sputum culture for M. tuberculosis, drug susceptibility testing if culture was positive for M. tuberculosis, Xpert test, diagnostic antibiotic trial using \ufb01rst-line empirical treatment (broad spectrum beta-lactams and macrolides) for pneumonia. The costs of \ufb01rst-line and second-line anti-TB treatment were also included. There was no published data on the association between early anti-TB treatment and duration of hospitalization (or use of intensive care unit), the cost of hospitalization was therefore not included in the economic analysis.\nHospital Authority is the largest, non-pro\ufb01t-making public health organization in Hong Kong. Its service is almost completely subsidized by the government of Hong Kong. Patients who are non-residents of Hong Kong are charged by Hospital Authority based upon the charges of healthcare services of public hospitals posted in legislation in the Hong Kong Gazette. Assuming the charges listed represent only the cost components (including labor costs) with no addition of pro\ufb01ts, the costs of diagnostic tests and drug acquisition were therefore approximated using the Hospital Authority charges. Xpert is currently not available in the Hospital Authority and the cost per test was estimated from consumable and manpower costs. All costs were discounted to year 2014 costs with 3% discount rate.\nCost-effectiveness analysis and sensitivity analysis\nA diagnostic approach was dominated when it was more costly with lower QALYs gained than another option. After\n\nJ.H.S. You et al.\nexclusion of the dominated strategy, the incremental cost per QALY gained (ICER) of each remaining study arm, comparing to the next less costly arm, was calculated using the following equation: Dcost/DQALYs. Using the threshold of USD50,000 as the willingness-to-pay per QALY,12 the most effective strategy with ICER USD50,000 or less was considered as cost-effective.\nSensitivity analysis was performed by TreeAge Pro 2009 (TreeAge Software, Inc., Williamstown, MA, USA) and Microsoft Excel 2010 (Microsoft Corporation, Redmond, WA, USA) to examine the robustness of the model results. All the parameters were examined over the upper and lower limits of the variables, if available. Otherwise, a range of variation by \u00c620% of the base-case value was used.\nOne-way sensitivity analysis on all variables was performed to screen for potential in\ufb02uential factors. To evaluate the impact of the uncertainty in all of the variables simultaneously, a probabilistic sensitivity analysis was performed using Monte Carlo simulation. The cost and QALYs of each study arm were recalculated 10,000 times by simultaneously varying the values of each model input through the ranges of sensitivity analysis to determine the percentage of time in which each study arm would be the most cost-effective option.\nResults\nBase-case analysis\nThe results of base-case analysis were shown in Table 2. Both the Xpert and smear plus Xpert groups were more effective and less costly than the control group. The control group was the most costly (USD417 per patient) option with lowest (7.992) QALYs gained and highest mortality rate (10 per 100 suspected cases), and it was dominated by both Xpert and smear plus Xpert groups. Excluding the dominated option (conventional approach), the Xpert arm was more effective with higher QALYs and lower mortality rate when compared to the smear plus Xpert group by an ICER per QALY gained of USD99. Using USD50,000 as the threshold of willingness-to-pay per additional QALY saved, Xpert was highly cost-effective in the base-case scenario.\nSensitivity analysis\nThe base-case results were robust to one-way sensitivity analysis of all except one factor. The Xpert group would remain to be the most cost-effective option if the sensitivity of sputum smear microscopy examination was 74%. The smear plus Xpert group would become the cost-\n\nTable 2 Results of base-case analysis on cost, one-year mortality rate and QALYs gained.\n\nStrategy\n\nCost (USD)\n\nMortality ratea\n\nQALYs gainedb\n\nAFB smear plus Xpert\n\n246\n\n9.5\n\nXpert\n\n260\n\n9.0\n\nControl\n\n417\n\n10\n\na One-year mortality rate per 100 patients presented with suspected PTB. b Quality-adjusted life-years (QALYs) gained in patients who survived active PTB.\n\n8.148 8.289 7.992\n\nICER (USD)\ne 99 Dominated\n\nXpert MTB/RIF assay for tuberculosis CEA\neffective alternative if the sensitivity of sputum smear was >74%.\nThe prevalence of active PTB among hospitalized suspects was identi\ufb01ed to be the most in\ufb02uential factor on the ICER of the Xpert group (Fig. 2). ICER of Xpert remained below the threshold of willingness-to-pay (USD50,000) throughout the variation of active PTB prevalence, and it became the least costly group with highest QALYs gained when the prevalence among hospitalized suspects was <17%, indicated by the zero value of ICER in Fig. 2.\nIn the 10,000 Monte Carlo simulations generated by probabilistic sensitivity analysis, the control group remained to be the least effective option with lowest QALYs gained (7.346 \u00c6 2.135 QALYs) and highest cost (USD407 \u00c6 25 per patient), comparing to Xpert (7.621 \u00c6 2.202 QALYs and USD262 \u00c6 15 per patient) (p < 0.001) and smear plus Xpert (7.512 \u00c6 2.175 QALYs and USD243 \u00c6 12 per patient) (p < 0.001). Comparing smear plus Xpert and Xpert alone, the Xpert arm was signi\ufb01cantly (p < 0.001) more effective than smear plus Xpert by 0.109 QALYs (95% CI Z 0.107e0.111) with higher cost of USD19.2 (95% CI Z 19.1e19.3). The probabilities of each strategy to be cost-effective were examined in acceptability curves over a wide range of willingness-to-pay per QALY loss reduced, from USD0-50,000 (Fig. 3). Using USD50,000 as the threshold of willingness-to-pay, the probabilities of Xpert, smear plus Xpert and conventional approach to be cost-effective were 94.5%, 5.5% and 0%, respectively.\nDiscussion\nThe present study examined the cost-effectiveness of Xpert for all and sputum microscopy examination plus Xpert (for smear-negative) versus the conventional diagnostic approach in hospitalized adults with suspected active PTB. Our results suggested that using a single sputum test of Xpert at initial assessment was the most cost-effective\nFigure 2 One-way sensitivity analysis of prevalence of active PTB among hospitalized suspects on ICER per QALY gained by Xpert.\n\n413\nFigure 3 Acceptability curves of three diagnostic approaches to be cost-effective versus willingness-to-pay per QALY.\noption. The \ufb01nding of this option to be the preferred choice was highly robust to the variation of all except one model input. If the sensitivity of sputum microscopy examination could be enhanced from 50% (base-case value) to >74%, using sputum smear examination for screening and plus Xpert for smear-negative patients would become the most cost-effective option.\nWe found that in times of lower active PTB prevalence (<17%) among hospitalized patients with suspected TB, the diagnostic approach using a single sputum test by Xpert became the least costly while remaining the most effective option with highest QALYs gained, owing to the high sensitivity and speci\ufb01city of the test and therefore no subsequent testing was performed. Apparently, the major cost driver was the subsequent test for smear negative patients (due to the low sensitivity of 50%) in the two diagnostic arms initiated with sputum smear examination. Higher prevalence of non-PTB cases resulted in higher percentage of smear negative cases and therefore increased subsequent testing and the total diagnostic cost.\nThe robustness of Xpert being the preferred costeffective diagnostic test was indicated by the Monte Carlo 10,000 simulations that its probability to be cost-effective in 94.5% of suspected cases when the threshold of willingness-to-pay was USD50,000 per QALY. As indicated by the acceptability curve (Fig. 3), the probability of Xpert to be cost-effective was over 90% when the threshold of willingness-to-pay was USD1,800 or above per QALY. These \ufb01ndings suggested that Xpert is highly cost-effective even when the willingness-to-pay threshold was low.\nThe cost-effectiveness of replacing sputum smear examination with Xpert was compared with sputum smear plus Xpert and conventional diagnostic approach in high and low TB burden countries. Vassall et al. reported that replacement of smear with Xpert was cost-effective in India, South Africa and Uganda, comparing to sputum smear plus Xpert for smear-negative cases, with ICERs below the threshold of willingness-to-pay.13 Choi et al. found Xpert to\n\n414\nbe less costly and gain higher QALYs than the conventional approach of sputum microscopy and culture in the United States.14 Our \ufb01ndings showed that the cost-effectiveness of Xpert in Hong Kong, a developed city with intermediate TB burden, was similar to the cost-effectiveness of Xpert in high burden developing countries as well as low burden developed country.\nOur decision analysis compared the potential changes in economic and clinical outcomes of three diagnostic approaches in patients hospitalized with suspected active PTB at a wide range of PTB prevalence with varying combinations of MDR-TB and rifampin susceptible mycobacteria, as well as board ranges of sensitivity and speci\ufb01city of various diagnostic tests. The decision model provides a framework to examine the in\ufb02uential factors and the corresponding threshold values (if any) for each strategy to translate into a cost-effective option. The present \ufb01ndings, in combination with real-time epidemiologic data through continuous surveillance and sensitivity of TB diagnostic tools, may provide better insights into the cost-effectiveness of different diagnostic strategies for healthcare providers to consider in their clinical practice.\nOne of the limitations of this study was the source of key clinical model input (the association of late anti-TB treatment to mortality) that it was obtained from a retrospective observational study in Hong Kong. The model input was subjected to the same limitations applied to retrospective observational study. All model inputs were therefore examined over a wide range in the sensitivity analyses to identify in\ufb02uential factors that would alter the base-case \ufb01ndings. The model simpli\ufb01ed the diagnostic algorithm for smear-negative cases in conventional approach to smear examination, diagnostic antibiotic trial and clinical judgment. The costs of possible additional radiologic examinations and bronchoscopy were not included due to the variability in their usage. Despite the conservative estimation on utilization of diagnostic resources, the conventional approach was showed to be the least cost-effective option in the present model. Our analysis took on the perspective of healthcare providers and the cost of premature death (loss of productivity) was not included. The potential bene\ufb01ts of reduced risk of mortality as a result of rapid diagnostic test and early anti-TB treatment might be underestimated.\nIn conclusion, a single sputum test by Xpert during initial assessment of hospitalized patients with suspected active PTB appears to be a highly cost-effective diagnostic strategy in Hong Kong with intermediate TB burden.\nFunding\nThis project has received no funding support.\n\nPotential con\ufb02icts of interest\n\nJ.H.S. You et al.\n\nAll authors have no con\ufb02icts of interest.\nReferences\n\n1. Tuberculosis and chest service of the Department of Health annual report. Hong Kong: Department of Health; 2009.\n2. Surveillance Team, Special Preventive Programme, Centre for Health Protection. HIV surveillance report e 2010 Update. Hong Kong: Department of Health.\n3. Greenaway C, Menzies D, Fanning A, Grewal R, Yuan L, FitzGerald JM, et al. Delay in diagnosis among hospitalized patients with active tuberculosisdpredictors and outcomes. Am J Respir Crit Care Med 2002;165:927e33.\n4. Lui G, Wong RYK, Li F, Lee MKP, Lai RWM, Li TCM, et al. High mortality in adults hospitalized for active tuberculosis in a low HIV prevalence setting. PLoS One 2014. http: //dx.doi.org/10.1371/journal.pone.0092077.\n5. Boehme CC, Nabeta P, Hillemann D, Nicol MP, Shenai S, Krapp F, et al. Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med 2010;363:1005e15.\n6. Lui G, Lee N, Cheung SW, Lam JS, Wong BC, Choi KW, et al. Interferon gamma release assay for differentiating tuberculosis among pneumonia cases in acute healthcare setting. J Infect 2011;62:440e7.\n7. Aziz MA, Wright A, Laszlo A, De Muynck A, Portaels F, Van Deun A, et al. Epidemiology of antituberculosis drug resistance (the Global Project on Anti-tuberculosis Drug Resistance Surveillance): an updated analysis. Lancet 2006 December 16; 368:2142e54.\n8. Steingart KR, Ng V, Henry M, Hopewell PC, Ramsay A, Cunningham J, et al. Sputum processing methods to improve the sensitivity of smear microscopy for tuberculosis: a systematic review. Lancet Infect Dis 2006;6:664e74.\n9. Siddiqi K, Walley J, Khan MA, Shah K, Safdar N. Clinical guidelines to diagnose smear-negative pulmonary tuberculosis in Pakistan, a country with low-HIV prevalence. Trop Med Int Health 2006;11:23e31.\n10. Gold MR, Franks P, McCoy KI, Fryback DG. Toward consistency in cost-utility analyses: using national measures to create condition-speci\ufb01c values. Med care 1998;36:778e92.\n11. Census and Statistics Department, The Government of Hong Kong SAR. Accessed on 31 March 2014: www.censtatd.gov.hk/.\n12. Weinstein MC, Siegel JE, Gold MR, Kamlet MS, Russel LB. Recommendations of the panel on cost-effectiveness in health and medicine. JAMA 1996;276:1253e8.\n13. Vassall A, van Kampen S, Sohn H, Michael JS, John KR, den Boon S, et al. Rapid diagnosis of tuberculosis with the Xpert MTB/RIF assay in high burden countries: a costeffectiveness analysis. PLoS Med 2011;8:e1001120. http: //dx.doi.org/10.1371/journal.pmed.1001120.\n14. Choi HW, Miele K, Dowdy D, Shah M. Cost-effectiveness of Xpert\u00d2 MTB/RIF for diagnosing pulmonary tuberculosis in the United States. Int J Tuberc Lung Dis 2013;17:1328e35. http: //dx.doi.org/10.5588/ijtld.13.0095.\n\n\n",
    "authors": [
      "Joyce H.S. You",
      "Grace Lui",
      "Kai Man Kam",
      "Nelson L.S. Lee"
    ],
    "doi": "10.1016/j.jinf.2014.12.015",
    "year": null,
    "item_type": "journalArticle",
    "url": "https://linkinghub.elsevier.com/retrieve/pii/S016344531500002X"
  }
]