Open Forum Infectious Diseases MAJOR ARTICLE Cost-effectiveness of WHO-Recommended Algorithms for TB Case Finding at Ethiopian HIV Clinics 1 2 3 4 5 2,5,6 Max W. Adelman, Deborah A. McFarland, Mulugeta Tsegaye, Abraham Aseffa, Russell R. Kempker, and Henry M. Blumberg 1 2 Department of Medicine, Massachusetts General Hospital, Boston, Massachusetts; Hubert Department of Global Health, Rollins School of Public Health, Emory University, Atlanta, Georgia; 3 4 5 Training Division, ALERT Hospital, Addis Ababa, Ethiopia; Armauer Hansen Research Institute, Addis Ababa, Ethiopia; Division of Infectious Diseases, Department of Medicine, Emory University School of Medicine, Atlanta, Georgia; Department of Epidemiology, Rollins School of Public Health, Emory University, Atlanta, Georgia Background: e W Th orld Health Organization (WHO) recommends active tuberculosis (TB) case finding and a rapid molecular diagnostic test (Xpert MTB/RIF) to detect TB among people living with HIV (PLHIV) in high-burden settings. Information on the cost-effectiveness of these recommended strategies is crucial for their implementation. Methods: We conducted a model-based cost-effectiveness analysis comparing 2 algorithms for TB screening and diagnosis at Ethiopian HIV clinics: (1) WHO-recommended symptom screen combined with Xpert for PLHIV with a positive symptom screen and (2) current recommended practice algorithm (CRPA; based on symptom screening, smear microscopy, and clinical TB diagno- sis). Our primary outcome was US$ per disability-adjusted life-year (DALY) averted. Secondary outcomes were additional true-pos- itive diagnoses, and false-negative and false-positive diagnoses averted. Results: Compared with CRPA, combining a WHO-recommended symptom screen with Xpert was highly cost-effective (incre- mental cost of $5 per DALY averted). Among a cohort of 15 000 PLHIV with a TB prevalence of 6% (900 TB cases), this algorithm detected 8 more true-positive cases than CRPA, and averted 2045 false-positive and 8 false-negative diagnoses compared with CRPA. e WH Th O-recommended algorithm was marginally costlier ($240 000) than CRPA ($239 000). In sensitivity analysis, the symptom screen/Xpert algorithm was dominated at low Xpert sensitivity (66%). Conclusions: In this model-based analysis, combining a WHO-recommended symptom screen with Xpert for TB diagnosis among PLHIV was highly cost-effective ($5 per DALY averted) and more sensitive than CRPA in a high-burden, resource-limited setting. Keywords. cost-effectiveness; developing countries; Ethiopia; modeling; TB/HIV co-infection. Tuberculosis (TB) is the leading cause of death among people estimated 191 000 new cases of TB in Ethiopia in 2015, and living with HIV (PLHIV) globally . In 2015, there were an about 10% of new TB cases are co-infected with HIV . In estimated 1.2 million new TB cases and 400 000 TB-related Addis Ababa, the prevalence of active TB disease among deaths among PLHIV worldwide . Although the global PLHIV is estimated to be as high as 17% [6–10]. Ethiopia is one burden of TB is enormous, TB control efforts are substan- of the world’s lowest-income countries (2013 GDP per capita of tially underfunded. The World Health Organization (WHO) US$505) and had a government health care budget of $73 per estimates that of the $4.8 billion per year required to combat person in 2014 [11, 12]. TB disease, there is a $1.6 billion (33%) funding gap . Any Because of the impact of HIV co-infection on TB incidence, attempt to inform TB health policy must consider the economic morbidity, and mortality, the WHO recommends “intensified impact of policy changes given limited resources and health-re- case finding” of TB among all PLHIV in high–TB burden areas lated budgets in many TB high-burden countries [4, 5]. . A 2011 meta-analysis determined that absence of current Ethiopia is one of 30 “high–TB burden” countries that cough, fever, night sweats, and weight loss in PLHIV identi- account for nearly 90% of global TB cases . There were an fied persons with a very low probability of active TB (negative predictive value of 98% at 5% TB prevalence) . WHO rec- ommends this “symptom-based screen” for PLHIV, followed Received 12 February 2017; editorial decision 11 December 2017; accepted 18 December 2017. by a diagnostic work-up for TB in those with a positive symp- Correspondence: M. W. Adelman, MD, MSc, Department of Medicine, Massachusetts tom screen (ie, at least 1 of the 4 symptoms) [1, 14]. However, General Hospital, 55 Fruit Street, Boston, MA 02114 (email@example.com). the diagnostic test currently most commonly available in Open Forum Infectious Diseases © The Author(s) 2017. Published by Oxford University Press on behalf of Infectious Diseases resource-limited areas, acid-fast bacillus (AFB) smear, has sev- Society of America. This is an Open Access article distributed under the terms of the Creative eral limitations, including very poor sensitivity (≤30%) for TB Commons Attribution-NonCommercial-NoDerivs licence (http://creativecommons.org/licenses/ by-nc-nd/4.0/), which permits non-commercial reproduction and distribution of the work, in any among PLHIV . medium, provided the original work is not altered or transformed in any way, and that the work e X Th pert MTB/RIF assay (“Xpert,” Cepheid, Sunnyvale, is properly cited. For commercial re-use, please contact firstname.lastname@example.org DOI: 10.1093/ofid/ofx269 CA) is a rapid molecular diagnostic test that can detect TB Case Finding in HIV Patients • OFID • 1 Downloaded from https://academic.oup.com/ofid/article-abstract/5/1/ofx269/4773940 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Modeling Strategies M. tuberculosis (MTB) and rifampin resistance in less than 2 We modeled 2 strategies for intensified TB case finding among hours . A multicenter study conducted among HIV+ and PLHIV (Figure 1) and compared hypothetical cohorts of 15 000 HIV– patients in 5 high–TB incidence countries found Xpert to PLHIV (equivalent to the ALERT HIV Clinic cohort) progress- be 98% sensitive for smear microscopy–positive, culture-pos- ing through each diagnostic strategy. We assumed that each itive specimens and 73% for smear microscopy–negative TB of the 15 000 PLHIV in each cohort would progress through compared with AFB culture . Specifically, among PLHIV in the model only once, thus not accounting for repeat visits. a high–TB prevalence area in South Africa, Xpert had a sensi- Models were constructed with TreeAge (TreeAge Software, Inc., tivity of 100% for smear-positive TB and 62% for smear-neg- Williamstown, MA, USA), and additional calculations were ative TB, compared with smear microscopy sensitivity of 28% performed using SAS v9.4 (SAS Institute, Cary, NC, USA). The . In 2013, the WHO recommended Xpert as the initial diag- 2 diagnostic algorithms compared were: nostic test for TB in adult PLHIV . Although Xpert improves TB diagnosis compared with 1. WHO-recommended TB symptom screen plus WHO- the current standard of care (ie, smear microscopy) in most recommended Xpert diagnostic algorithm (“SSX”). PLHIV resource-limited countries [16–18], its cost remains an issue. In were screened for TB using the WHO-recommended symp- developing countries, the Xpert platform costs US$17 000 (for tom screening algorithm (cough, fever, night sweats, or a 4-channel instrument) and each test cartridge costs approxi- weight loss) [1, 13]. If the symptom screen was negative (ie, mately US$10, prices substantially discounted from those in absence of all 4 symptoms), the patient was considered to high-income countries . Even with this discount, cost not have active TB disease and was prescribed isoniazid pre- may restrict its availability . We conducted a model-based ventive therapy (IPT) for presumed latent TB infection, per cohort study to determine the cost-effectiveness of combining WHO guidelines . However, a proportion of these PLHIV a WHO-recommended symptom-based screen with a molec- were assumed to have active TB (ie, false-negative diagnoses) ular diagnostic test, Xpert, at an HIV clinic in Ethiopia, a based on symptom screen test characteristics. resource-limited, high–TB burden country. METHODS If a patient had a positive symptom screen (ie, at least 1 of Parent Study and Clinical Inputs the 4 symptoms), sputum samples would be collected and Clinical inputs were derived in large part from an operational tested initially for MTB with Xpert, per WHO guidelines research project we conducted on implementation of the WHO- for diagnostic tests for PLHIV, and would not be tested with recommended TB symptom screen for PLHIV (cough, fever, smear microscopy . Sputum samples obtained from night sweats, weight loss) followed by diagnostic testing with patients with a positive symptom screen were either pos- Xpert among sputum specimens collected from PLHIV with a itive or negative for MTB (and rifampin resistance) based positive symptom screen [1, 9, 13]. Methods of the parent study on Xpert results. Of those PLHIV with active TB, some had are described in detail elsewhere . That study took place drug-susceptible TB (DS-TB) and others had multidrug-re- from July to October 2013 at the ALERT Hospital HIV Clinic, sistant TB (MDR-TB). PLHIV with presumptive MDR-TB which cares for approximately 15 000 PLHIV in Addis Ababa, based on a positive Xpert test for rifampin resistance had Ethiopia. The Armauer Hansen Research Institute (AHRI)/ their sputum specimens tested with AFB culture and drug ALERT and Emory University Institutional Review Boards susceptibility testing (DST) and received treatment for approved the parent study. Results from the parent study  MDR-TB. Patients with negative Xpert results were pre- were used to estimate base case clinical characteristics (Table 1). scribed IPT. Where these were not available, inputs were determined from relevant literature on PLHIV in Ethiopia or in sub-Saharan This algorithm also considers the cost of an Xpert MTB/RIF Africa when Ethiopia-specific inputs were not available. Xpert platform (a 1-year payment based on machine cost of $17 MTB/RIF sensitivity and specificity were determined from a 000 amortized over 10 years at a 3% interest rate) and cost meta-analysis on test characteristics in PLHIV . For sensi- of laboratory maintenance (Table 1). The number of Xpert tivity analysis, we used lower and upper estimates from relevant instruments needed was based on clinic volume (patients literature; input references are listed in Table 1. In our parent per day) and proportion of patients with a positive WHO- study, 89% of PLHIV at the clinic were on antiretroviral therapy recommended symptom screen (ie, patients who would (ART), and the mean CD4 count was 420 cells/μL . We did require TB diagnostic testing with Xpert). We assumed the not assume differential test characteristics of any of the diag- laboratory would utilize 4-channel Xpert machines operat- nostic tests based on CD4 count or whether the patient was on ing at maximum capacity of 4 simultaneous tests and run- ART. Further details on cost and disability-adjusted life-year ning for 8 hours per day. For calculating incremental cost of inputs are described in the Supplementary Appendix. the SSX algorithm, the total cost of the Xpert platform and 2 • OFID • Adelman et al. Downloaded from https://academic.oup.com/ofid/article-abstract/5/1/ofx269/4773940 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Table 1. Model Parameters for Cost-effectiveness Analysis, Base Case, and Ranges for Sensitivity Analyses Parameter Base Case Range, References Reference (Base Case) Cost inputs, laboratory, US$ a a Smear microscopy 1.20 0.60 –2.40 ALERT Labor cost per AFB smear 0.63 0.31–0.94 (AHRI) AHRI AFB culture 4.80 2.40 –8.75 (EPHI) EPHI a a DST 1.8 0.90 –12 AHRI b a a Xpert MTB/RIF, machine 1480 740 –2960  Xpert MTB/RIF, cartridge 9.98 9.98–72.87   a a Xpert MTB/RIF, yearly maintenance 1088.86 544.43 –2177.72  a a Chest x-ray 3.50 1.75 –7 ALERT Cost inputs, medication, US$ Drug-susceptible TB 33 25.17 –66 EFMOH a a MDR-TB 4856 2428 –9712 EFMOH a a IPT 5 2.50 –10 EFMOH Clinical characteristics TB prevalence 6% 4% –17%  Parent study  Clinic volume, patients/d 135 50%–250 Parent study  Proportion with positive WHO symptom screen 53% 25%–75% Parent study  Proportion of TB cases that are MDR 2.8% 0%–2.8%   Symptom screen sensitivity 72% 52% –93%   Symptom screen specificity 50% 33%–56%   Smear microscopy sensitivity 30% 19% –33%  Parent study  Smear microscopy specificity 100% 99.7% –100%  Parent study  Xpert MTB/RIF sensitivity 72% 66% –94%   Xpert MTB/RIF specificity 98% 95% –99%   Xpert RIF resistance sensitivity 95% 95% –100%   Xpert RIF resistance specificity 98% 98% –100%   Clinical diagnosis sensitivity, AFB negative TB 61% 55%–67%   Clinical diagnosis specificity, AFB negative TB 69% 66%–72%   Abbreviations: AFB, acid-fast bacillus; AHRI, Armauer Hansen Research Institute, Addis Ababa, Ethiopia; ALERT, ALERT Hospital, Addis Ababa, Ethiopia; ART, antiretroviral therapy; DST, drug susceptibility testing; EPHI, Ethiopian Public Health Institute, Addis Ababa, Ethiopia; EFMOH, Ethiopian Federal Ministry of Health, Addis Ababa, Ethiopia; MDR, multidrug-resistant; IPT, isoniazid preventive therapy; RIF, rifampin; TB, tuberculosis; US$, 2014 US dollars; WHO, World Health Organization. Where values could not be found in literature searches, we assumed lower bounds of ½× base case costs and upper bounds of 2× base case costs. We determined 1-year cost of Xpert MTB/RIF machine based on amortizing a US$17 000 payment for the machine over 10 years (expected useful life of machine) at a 3% interest rate. In the cited reference, TB screening using a symptom screen of one of cough, fever, or night sweats had 93% sensitivity for active TB among PLHIV. Adding a fourth symptom, as in our study (ie, weight loss), either would not change or would increase sensitivity. We therefore included this reported sensitivity as the upper limit for sensitivity analysis. In the cited reference, Xpert sensitivity was 61% among PLHIV with smear-negative TB and 97% among PLHIV with smear-positive TB. To calculate the listed sensitivity, we weighted the Xpert sensitivity by proportion of patients who had smear-negative and -positive disease (ie, pooled Xpert sensitivity = (p * Xpert sensitivity ) + (p * Xpert smear negative smear negative smear positive sensitivity ) = (0.7 * 0.61) + (0.3 * 0.97) = 0.72). smear positive maintenance were divided by the number of patients under- Of those with negative smear microscopy, 64% were tested going analysis (modeled as n = 15 000). with AFB culture and 92.8% with chest radiography . Further details on this assessment, including case definitions, 2. Current recommended practice algorithm (“CRPA”). As in have been previously described . All patients not diag- the SSX strategy, all PLHIV were initially screened for active nosed with active TB were prescribed IPT. TB with the WHO-recommended symptom screen. PLHIV with a negative symptom screen were assumed to not have Further details of the modeling structure are provided in the active TB and were prescribed IPT. Those who had a posi- Appendix. tive symptom screen provided sputum samples, which were tested for MTB with 3 separate smear microscopy tests, and Outcomes not with AFB cultures. PLHIV with negative smear micros- Our primary outcome was the incremental cost-effective- copy results were clinically diagnosed to be TB positive or ness ratio (ICER) of the SSX algorithm compared with CRPA negative. We modeled clinical assessment based on an oper- (US$ per disability-adjusted life-year [DALY] averted). Inputs ational research study carried out at the ALERT HIV clinic for DALY calculations are provided in Table 2. SSX was con- . In this study, PLHIV undergoing diagnostic workup for sidered cost-effective if the ICER was less than 3 times the pulmonary TB were initially tested with smear microscopy. Ethiopian gross domestic product (GDP) per capita and highly TB Case Finding in HIV Patients • OFID • 3 Downloaded from https://academic.oup.com/ofid/article-abstract/5/1/ofx269/4773940 by Ed 'DeepDyve' Gillespie user on 16 March 2018 RIF + MDR-TB (RIF_sens) Xpert + (MDR_prev) RIF – (Xpert_sens) (1 – RIF_sens) RIF + Symptom screen + (1 – RIF_spec) (screen_sens) DS-TB MDR-TB (1 – MDR_prev) RIF – (MDR_prev) Xpert – (RIF_spec) TB + (1 – Xpert_sens) DS-TB (TB_prev) f (1 - MDR_prev) MDR-TB (MDR_prev) Symptom screen – SSX (1 – screen_sens) DS-TB (1 - MDR_prev) Xpert + (1 – Xpert_spec) Symptom screen + (1 - screen_spec) TB – Xpert – (1 - TB_prev) (Xpert_spec) Symptom screen – MDR-TB (screen_spec) (MDR_prev) PLHIV Smear + (smear_sens) DS-TB N = 15 000 (1 – MDR_prev) MDR-TB (MDR_prev) Symptom screen + Clinical dx + (screen_sens) (clin_sens) Smear – DS-TB TB + (1 – smear_sens) (1 – MDR_prev) (TB_prev) MDR-TB CRPA MDR-TB (MDR_prev) Symptom screen – (MDR_prev) Clinical dx – (1 - screen_sens) (1 – clin_sens) DS-TB DS-TB (1 - MDR_prev) (1 – MDR_prev) Smear + (1 - smear_spec) Symptom screen + Clinical dx + (1 - screen_spec) (1 – clin_spec) Smear – TB – (smear_spec) Clinical dx – (1 - TB_prev) (clin_spec) Symptom screen – (screen_spec) Figure 1. Decision analysis model for tuberculosis screening and diagnosis among patients at Ethiopian HIV clinics. Decision analytic model with 2 different strategies for TB screening and diagnosis among PLHIV: (1) Symptom screen/Xpert (“SSX”) combines a World Health Organization–recommended symptom screen (cough, fever, night sweats, weight loss) with Xpert as the initial diagnostic test for PLHIV with a positive symptom screen (ie, having at least 1 symptom) [1, 12]. (2) Current practice screens patients with the symptom screen, and then combines smear microscopy with clinical diagnosis for those with negative smear microscopy results. Squares represent deci- sion nodes, circles represent chance nodes, and triangles represent terminal nodes. The number listed under each arm is the probability of progressing to that arm (from prior node), calculated under base case conditions. Abbreviations: CP, current practice algorithm; DALY, disability-adjusted life-year; DS, drug-susceptible; Dx, diagnosis; HIV, human immunodeficiency virus; MDR, multidrug-resistant; PLHIV, people living with HIV; SSX, symptom screen/Xpert algorithm; RIF, rifampin resistance; TB, tuberculosis. Sensitivity Analyses cost-effective if less than the Ethiopian GDP per capita ($505) We conducted 1-way sensitivity analyses by varying model [11, 22]. Secondary outcomes were true-positive, false-positive, inputs over ranges determined from the literature (Table 1). and false-negative TB diagnoses under each algorithm. We cal- culated ICERs of US$ per additional true-positive case detected RESULTS with SSX, and false-positive and false-negative TB diagnoses averted with SSX. Cost-effectiveness In a simulated cohort of 15 000 PLHIV attending an HIV Clinic in Ethiopia, an algorithm combining the WHO- Table 2. Inputs for Disability-Adjusted Life-Year Calculations recommended symptom screen and Xpert (SSX) was highly cost-effective. The ICER comparing the SSX algorithm with Condition Mortality (Range) Disability Weight (Range) the current recommended practice algorithm (CRPA) was HIV, TB negative 0.05 (0–0.3)  0.053 (0.034–0.079)  $5 per DALY averted (less than current Ethiopian GDP per HIV, untreated TB 1 (0.5–1) [30, 40] 0.399 (0.267–0.547)  HIV, treated drug- 0.105 (0.04–0.3) [40, 41] 0.1 (0.085–0.115)  capita of $505), highly cost-effective per WHO thresholds for susceptible TB cost-effectiveness . The SSX algorithm averted 243 DALYs HIV, treated MDR-TB 0.2 (0.04–0.37) [40–42] 0.2  compared with CRPA (Table 3). The SSX algorithm was esti- Inputs were used to calculate disability-adjusted life-years, as previously described . mated to be marginally costlier than CRPA. With base case Abbreviations: HIV, human immunodeficiency virus; MDR, multidrug-resistant; TB, tuberculosis. inputs (Tables 1 and 2), SSX costs $240 300, compared with 4 • OFID • Adelman et al. Downloaded from https://academic.oup.com/ofid/article-abstract/5/1/ofx269/4773940 by Ed 'DeepDyve' Gillespie user on 16 March 2018 $239 000 under the CRPA algorithm, an incremental cost of Under several scenarios, SSX was less costly than CRPA. $1300 per 15 000 PLHIV. With a high DS-TB treatment cost of $66, the total cost of the SSX had a higher case detection rate than CRPA (Table 3). In SSX algorithm was $260 000, vs $327 000 for CRPA (cost sav- the SSX cohort, there were 466 true-positive TB cases compared ings of $67 000 per 15 000 PLHIV). When care is more expen- with 458 in the CRPA cohort (8 additional true-positive diag- sive (eg when DS-TB treatment cost is expensive), SSX becomes noses, ICER = $157 per true-positive diagnosis). Additionally, more valuable due to its reduction in false-positive cases, which SSX averted both false-negative and false-positive TB diagnoses. prevents patients from receiving more expensive treatment. er Th e were 434 false-negative cases with SSX compared with Additionally, with a low MDR prevalence of 0% , the total 442 with CRPA (8 false-negative cases averted, ICER = $157 per cost of the SSX algorithm was $180 000, vs $239 000 for CRPA false-negative case averted). er Th e were 141 false-positive cases (cost savings of $59 000 per 15 000 PLHIV). Similarly, with a low with SSX compared with 2186 with CRPA (2045 false-positive MDR-TB treatment cost of $2428, the cost of SSX was $210 000, cases averted, ICER = $1 per false-positive case averted). compared with $239 000 for CRPA (cost savings of $29 000 per 15 000 PLHIV). Sensitivity Analyses SSX was not more effective than CRPA in all sensitivity anal- The SSX algorithm was highly cost-effective under a range of yses. At a low Xpert sensitivity of 66% , there were mar- parameter estimates (Figure 2). It was least cost-effective with ginally more DALYs in the SSX cohort (31 145) than in the a high Xpert cartridge cost of $73 (ICER = $1995), the current CRPA cohort (30 523). With Xpert sensitivity of 66%, SSX was price in developed countries (although we used $73 as the high both costlier and less effective than CRPA. Additionally, with end of the range for sensitivity analysis, the current price in high clinical diagnosis sensitivity of 67% , there were more developing countries is $9.98) . DALYs in the SSX cohort (30 279) than CP (29 931), similar to Under several conditions, the SSX algorithm was less the low MDR-TB rate of 0% (30 234 DALYs in SSX vs 30 227 cost-effective than it was in base case conditions. With a high in CRPA). MDR-TB treatment cost of $9712 (cost data provided by the Ethiopian Federal Ministry of Health), the ICER was $253 per DISCUSSION DALY averted. Similarly, SSX was less cost-effective than base case with a high TB prevalence of 17% (ICER = $236 per DALY In this cost-effectiveness analysis of a WHO-recommended TB averted) , low sputum smear cost of $0.60 (ICER = $62), case finding algorithm of symptom screening in combination and high sputum microscopy sensitivity of 33% (ICER = $15) with a molecular diagnostic test, Xpert MTB/RIF, at an Ethiopian . The full results of the primary sensitivity analysis are shown HIV clinic (SSX algorithm), we found SSX to be highly cost-ef- in Figure 2. fective ($5 per DALY averted) compared with the current recom- SSX was most costly, at a high Xpert cost of $73 ($724 000 mended practice (CRPA). This ICER is less the than Ethiopian per 15 000 PLHIV). SSX was least costly, at a low MDR-TB GDP per capita of $505, the WHO’s threshold for a highly prevalence of 0% ($180 000 per 15 000 PLHIV)  due to fewer cost-effective intervention . Additionally, the SSX algorithm patients requiring costly MDR-TB treatment than the base case detected more TB cases than CRPA (466 vs 458, difference = 8, MDR-TB rate of 2.8% of TB cases. MDR-TB treatment was ICER = $157 per true-positive case) and averted both false-posi- more than 100 times costlier than DS-TB treatment. tive (141 vs 2186, difference = 2045, ICER = $1 per false-positive Table 3. Expected Outcomes and Cost-effectiveness of 2 Strategies for Active Tuberculosis Case Finding at Ethiopian HIV Clinics Algorithm (n = 15 000 per Algorithm) Current Practice, ICER b b c Outcome Symptom Screen/ Xpert, No. (Range ) No. (range ) (Maximum ) TB cases, actual 900 (600–2550) 900 (600–2550) -- Symptom screen positive 7950 (3750–11 250) 7950 (3750–11 250) -- Cost, 1000 US$ 240 (180–724) 239 (208–327) -- DALYs, thousands 30.3 (12.9–117) 30.5 (13.1–117) 5.2 (1995) TP TB diagnoses 466 (310–1320) 458 (305–1297) 157 (60 700) FN TB diagnoses 434 (289–1230) 442 (295–1250) 157 (60 700) FP TB diagnoses 141 (71–353) 2200 (1920–2930) 1 (240) Abbreviations: DALY, disability-adjusted life-year; FN, false-negative; FP, false-positive; ICER, incremental cost-effectiveness ratio; TB, tuberculosis; TP, true positive; US$, 2014 US dollars. ICERs were calculated for each row according to the following formula (eg, for DALY): ICER = [Cost – Cost ]/[DALY – DALY ]. ICER units are Symptom screen/Xpert Current practice Current practice Symptom screen/Xpert US$ per 1 additional outcome, eg, US$ per DALY averted. Ranges are minimum and maximum values determined from sensitivity analyses. We did not report minimum ICERs; under several cases, Xpert was either dominated by or less costly than SSX (please see “Results: Sensitivity Analyses” and Figure 2). TB Case Finding in HIV Patients • OFID • 5 Downloaded from https://academic.oup.com/ofid/article-abstract/5/1/ofx269/4773940 by Ed 'DeepDyve' Gillespie user on 16 March 2018 ICER (2014 US$/DALY averted) 0 100 200 300 400 Base case ICER = $5 Xpert MTB/RIF cartridge cost (10–73 US$) MDR-TB treatment cost (2428 –9712 US$) TB prevalence (4 –17%) DS-TB treatment cost (25–66 US$) Symptom screen sensitivity (52 –93%) Smear microscopy cost (0.60–2.40 US$) Chest x-ray cost (1.75–7.00 US$) AFB culture cost (2.40–8.75 US$) b b IPT cost (2.50 –10 US$) Clinic volume (50 –250 patients per day) Symptom screen speciﬁcity (33 –56%) Labor cost per AFB smear (0.31–0.94 US$) Xpert MTB/RIF machine cost (740–2960 US$) Clinical diagnosis speciﬁcity (66 –72%) Xpert MTB/RIF speciﬁcity (95–99 %) Xpert MTB/RIF maintenance cost (544–2173 US$) Positive WHO symptom screen (25 –75%) Smear microscopy sensitivity (19–33%) Xpert RIF resistance sensitivity (95–100%) Untreated MDR-TB mortality (0.5–1.0) Clinical diagnosis sensitivity (55–67 %) Treated MDR-TB motality (0.04–0.37) TB-negative disability weight (0.034–0.079) Xpert MTB/RIF sensitivity (66 –94%) Treated DS-TB disability weight (0.085–0.115) Smear microscopy speciﬁcity (99.7–100%) DST cost (0.90–12.00 US$) Xpert RIF resistance speciﬁcity (98–100%) Untreated DS-TB mortality (0.5–1.0) Treated DS-TB mortality (0.04–0.30) Untreated MDR-TB disability weight (0.267–0.547) Untreated DS-TB disability weight (0.267–0.547) TB cases that are MDR (0 –2.8%) TB-negative mortality (0–0.3) Low estimate High estimate Figure 2. Ranges of incremental cost-effectiveness ratio (US$ per disability-adjusted life-year averted) of a World Health Organization–recommended tuberculosis diag- a b nostic algorithm vs current recommended practice at Ethiopian HIV clinics. Graph truncated for space reasons; in this case, ICER = $1995. Under these conditions, the symptom screen/Xpert algorithm was cost-saving; ICERs are not reported in these cases. Under these conditions, the symptom screen/Xpert algorithm was less effective than current practice; the ICER was not reported in this case. At low Xpert sensitivity of 66% , the symptom screen/Xpert algorithm was dominated (more costly and less effective) at current practice; the ICER was not reported in this case. Abbreviations: AFB, acid-fast bacillus; DALY, disability-adjusted life-year; DS, drug-susceptible; DST, drug susceptibility testing; HIV, human immunodeficiency virus; ICER, incremental cost-effectiveness ratio; IPT, isoniazid preventive therapy; MDR, multidrug-resistant; TB, tuberculosis; WHO, World Health Organization. diagnosis averted) and false-negative diagnoses (434 vs 442, dif- high rate of empiric treatment. In our model-based study, there ference = 8, ICER = $157 per false-negative diagnosis averted) were 2200 (1920–2930) false-positive TB cases in the CRPA compared with CRPA. Based on our results, the value of SSX cohort. A large proportion of patients (53%) with a positive comes from increased diagnostic accuracy (given enhanced case WHO-recommended symptom screen were therefore recom- finding with the use of the combined WHO algorithms), espe- mended to undergo further testing . Given the low specific- cially given the poor test characteristics of those tests used in ity of the symptom screen (50%) , many of these patients CRPA (ie, smear microscopy and clinical diagnosis), which leads with a positive symptom screen did not have active TB. When to more false-positive TB cases. Specifically, the value of the SSX followed by testing with a more specific diagnostic test such as algorithm is more related to its ability to avert false-positive diag- Xpert (98% specificity)  in the SSX algorithm, this is able to noses than to detect additional true-positive cases. limit the number of false-positive cases (141 false-positives in PLHIV in high-burden settings in low- and middle-income SSX algorithm). However, when symptom screening is part of countries with a positive WHO-recommended TB symptom an algorithm followed by diagnostic tests with lower specificity, screen (at least one of cough, fever, night sweats, or weight loss) such as clinical diagnosis (69%) , patients are more likely are recommended to undergo TB diagnostic testing with Xpert to be diagnosed empirically with TB. Given the high specific- . However, TB diagnosis in resource-limited settings gen- ity of smear microscopy (100%) , this indicates that many erally relies on smear microscopy, and the poor sensitivity of of the false-positive cases are related to clinical diagnosis. That smear microscopy among PLHIV (≤30%) leads to low rates of clinical diagnosis specificity is not a major driver of the ICER microbiologically confirmed disease . As a result, there is a (Figure 1) may be related to the narrow range of specificity in 6 • OFID • Adelman et al. Downloaded from https://academic.oup.com/ofid/article-abstract/5/1/ofx269/4773940 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Model parameter (range) sensitivity analysis; further data are needed on clinical diagno- cost-effective compared with smear microscopy (ICER = $58), sis in Ethiopia and other resource-limited settings. which is similar to the results of our current study, indicating that er Th e are few data on rate of clinical diagnosis and result- a diagnostic algorithm using Xpert may be cost-effective across ant empiric treatment of TB in PLHIV, largely due to inability a range of model parameters (as this study used different inputs to diagnose TB (ie, no good gold standard) in such patients. than ours) . In a model-based study of PLHIV being screened However, a study in Uganda found that among PLHIV who for TB prior to ART initiation in South Africa, a diagnostic algo- were suspected of having TB, 33% of smear-negative patients rithm using 2 Xpert samples was cost-effective (ICER = $6700 were initiated on empiric TB treatment . In one randomized per year of life saved) . Our study differs from the South controlled trial, a large proportion of PLHIV (46%) with CD4 African study because we assumed screening and Xpert testing counts <150 cells/µL had a “high suspicion” of TB; there was no for all patients regardless of ART status, as recommended by the mortality benefit when these patients were assigned to a nurse- WHO . However, a 2017 analysis of Xpert roll-out in South driven protocol that emphasized early TB treatment, indicating Africa found that using Xpert as the initial diagnostic test was not that empiric diagnosis may not be accurate in these patients more expensive than using smear microscopy, but showed that . Further studies are needed to explore rates of empiric Xpert was unlikely to be cost-effective . Further real-world treatment and its potential consequences. implementation studies will be important to determine for which Given the high rates of empiric TB treatment, an accurate populations Xpert is likely to be cost-effective. diagnostic test such as Xpert would be valuable for HIV clinics Our cost-effectiveness analysis is subject to several limi- in resource-limited areas where TB/HIV co-infection is preva- tations. We did not quantify costs associated with scale-up of lent. However, for clinics such as the ALERT HIV clinic, with Xpert; we took the position of a diagnostic system (including a high volume (135 patients per day) and high rate of symp- Xpert machines) that was already implemented. Additionally, tom screen positivity (53%) , the WHO recommendations we did not consider Xpert’s potential impact on TB transmis- require a large number of PLHIV to undergo routine TB test- sion, further downstream effects of delays in TB diagnosis and ing with Xpert. This calls into question the feasibility of routine treatment, efficacy of IPT and TB treatment, and issues regard- Xpert rollout in resource-limited settings. Substantial additional ing delivering Xpert results and linkages to care. These consid- resource expenditure from either local governments or outside erations will be important for policy makers considering Xpert organizations (eg, PEPFAR, Global Fund) will be required to implementation. We did not consider patients who are unable implement enhanced TB case finding with symptom screening to provide a sputum sample or otherwise lost to follow-up; these and Xpert. Although nongovernmental organizations currently proportions may differ between groups. Similarly, there was no fund a large proportion of TB control and care in developing empiric treatment following a negative Xpert test. However, as countries , cheaper, more efficient point-of-care tests are the CRPA algorithm requires more sputum samples than SSX needed to reduce this funding burden. and did include clinical diagnosis, consideration of these factors Local factors, such as country-level cost inputs and clinic vol- would likely make SSX more cost-effective. We additionally did ume, are important for determining cost-effectiveness of various not vary diagnostic test characteristics by CD4 count or ART TB diagnostic strategies and therefore the extent to which they status; in practice, these clinical characteristics may ae ff ct these will be adopted locally . This is one of the first cost-effectiveness tests’ diagnostic yield. We attempted to reflect uncertainty in analyses of symptom screening and Xpert at Ethiopian HIV clin- clinical inputs in sensitivity analyses. For some sensitivity anal- ics, and therefore these data may be valuable for scaling up Xpert yses, small changes in input parameters significantly impacted in Ethiopia. Although we conducted sensitivity analyses by vary- the results, which would have important consequences for ing model inputs with data taken from the literature on Ethiopia implementation of any diagnostic strategy. and sub-Saharan Africa where available (Table 1), characteristics In conclusion, a TB diagnostic algorithm that combines a of the local population may vary beyond the ranges we have used WHO-recommended symptom screen with Xpert for PLHIV in sensitivity analyses. Other specific local factors may ae ff ct the with a positive symptom screen was highly cost-effective in an cost-effectiveness of a diagnostic algorithm that includes Xpert; Ethiopian HIV clinic compared with current recommended local clinicians and policy makers will need to consider these fac- practice (ICER = $5 per DALY averted). Adoption of the symp- tors when considering optimal diagnostic strategies. tom screen and Xpert in Ethiopian HIV clinics remains lim- Several studies have examined use of Xpert in other sub-Saha- ited, and our data suggest that clinicians and policy makers in ran countries. The only other cost-effectiveness analysis based in Ethiopia could consider a similar diagnostic algorithm for TB Ethiopia similarly found diagnostic algorithms that concluded diagnosis among PLHIV. Xpert to be highly cost-effective; however, this analysis did not Acknowledgments specify symptom screening of presumptive TB cases and included Financial support. This work was supported in part by the both HIV+ and HIV- patients . In a model-based study of National Institutes of Health (NIH) National Center for Advancing PLHIV in Uganda, a TB diagnostic algorithm using Xpert was Translational Sciences (grant numbers UL1TR000454, UL1TR002378 TB Case Finding in HIV Patients • OFID • 7 Downloaded from https://academic.oup.com/ofid/article-abstract/5/1/ofx269/4773940 by Ed 'DeepDyve' Gillespie user on 16 March 2018 and TL1TR000456); Fogarty International Center at the NIH (grant num- 19. FIND Diagnostics. Negotiated Product Pricing. Available at https://www.finddx. org/find-negotiated-product-pricing/. Accessed 8 January 2018. ber D43TW009127); the NIH National Institute of Allergy and Infectious 20. Lawn SD, Mwaba P, Bates M, et al. Advances in tuberculosis diagnostics: the Xpert Diseases (grant number K23AI103044); the Emory Global Health Institute; MTB/RIF assay and future prospects for a point-of-care test. Lancet Infect Dis the Infectious Diseases Society of America; and the American Medical 2013; 13:349–61. Association Foundation. e Th funders had no role in study design, data col- 21. Steingart KR, Schiller I, Horne DJ, et al. Xpert® MTB/RIF assay for pulmonary lection and analysis, decision to publish, or preparation of the manuscript. tuberculosis and rifampicin resistance in adults. Cochrane database Syst Rev Potential conifl cts of interest. All authors: no reported conflicts of 2014; CD009593. interest. All authors have submitted the ICMJE Form for Disclosure of 22. World Health Organization. Cost-effectiveness thresholds. Cost effectiveness and Potential Conflicts of Interest. Conflicts that the editors consider relevant to strategic planning (WHO-CHOICE). 2015. Available at: http://www.who.int/ the content of the manuscript have been disclosed. choice/costs/CER_thresholds/en/. Accessed 1 February 2017. 23. World Health Organization. WHO endorses new rapid tuberculosis test. 2010. Available at: http://www.who.int/mediacentre/news/releases/2010/tb_ Supplementary Data test_20101208/en/. Accessed 24 August 2014. Supplementary materials are available at Open Forum Infectious Diseases online. 24. Walusimbi S, Bwanga F, De Costa A, et al. Meta-analysis to compare the accuracy Consisting of data provided by the authors to benefit the reader, the posted of GeneXpert, MODS and the WHO 2007 algorithm for diagnosis of smear-neg- ative pulmonary tuberculosis. BMC Infect Dis 2013; 13:507. materials are not copyedited and are the sole responsibility of the authors, so 25. Ahmad Khan F, Verkuijl S, Parrish A, et al. Performance of symptom-based tuber- questions or comments should be addressed to the corresponding author. culosis screening among people living with HIV: not as great as hoped. AIDS 2014; 28:1463–72. References 26. Nakiyingi L, Bwanika JM, Kirenga B, et al. Clinical predictors and accuracy of 1. World Health Organization. Guidelines for intensified tuberculosis case-finding empiric tuberculosis treatment among sputum smear-negative HIV-infected and isoniazid preventive therapy for people living with HIV in resource-con- adult TB suspects in Uganda. Neyrolles O, ed. PLoS One 2013; 8:e74023. strained settings. 2011. Available at: http://whqlibdoc.who.int/publica- 27. Grant A, Charalambous S, Tlali M, et al. Empirical TB treatment in advanced tions/2011/9789241500708_eng.pdf. Accessed 20 August 2014. HIV disease: results of the TB fast track trial. Paper presented at: CROI 2016. 2. World Health Organization. Global tuberculosis report 2016. 2016. Available at: Boston, MA; February 22−25, 2016. http://www.croiconference.org/sessions/ http://apps.who.int/iris/bitstream/10665/250441/1/9789241565394-eng.pdf ?ua=1. empirical-tb-treatment-advanced-hiv-dis eas e-results-tb-fast-track-trial. Accessed 2 February 2017. Accessed 8 January 2018. 3. World Health Organization. Tuberculosis financing and funding gaps. Available at: 28. Floyd K, Fitzpatrick C, Pantoja A, Raviglione M. Domestic and donor financing http://www.who.int/tb/WHO_GF_TB_financing_factsheet.pdf. Published 2013. for tuberculosis care and control in low-income and middle-income countries: an Accessed 20 August 2014. analysis of trends, 2002–11, and requirements to meet 2015 targets. Lancet Glob 4. Dowdy DW, Cattamanchi A, Steingart KR, et al. Is scale-up worth it? Challenges Heal 2013; 1:e105–15. in economic analysis of diagnostic tests for tuberculosis. PLoS Med 2011; 29. Tesfaye A, Fiseha D, Assefa D, et al. Modeling the patient and health system 8:e1001063. impacts of alternative xpert® MTB/RIF algorithms for the diagnosis of pulmonary 5. Mann G, Squire SB, Bissell K, et al. Beyond accuracy: creating a comprehensive tuberculosis in Addis Ababa, Ethiopia. BMC Infect Dis 2017; 17:318. evidence base for TB diagnostic tools. Int J Tuberc Lung Dis 2010; 14:1518–24. 30. Shah M, Dowdy D, Joloba M, et al. Cost-effectiveness of novel algorithms for 6. Zaeh S, Kempker R, Stenehjem E, et al. Improving tuberculosis screening and iso- rapid diagnosis of tuberculosis in HIV-infected individuals in Uganda. AIDS niazid preventive therapy in an HIV clinic in Addis Ababa, Ethiopia. Int J Tuberc 2013; 27:2883–92. Lung Dis 2013; 17:1396–401. 31. Andrews JR, Lawn SD, Rusu C, et al. The cost-effectiveness of routine tuberculo- 7. Abebe G, Deribew A, Apers L, et al. Evaluation of the 2007 WHO guideline to sis screening with Xpert MTB/RIF prior to initiation of antiretroviral therapy: a diagnose smear negative tuberculosis in an urban hospital in Ethiopia. BMC model-based analysis. AIDS 2012; 26:987–95. Infect Dis. 2013; 13:427. 32. Vassall A, Siapka M, Foster N, et al. Cost-effectiveness of Xpert MTB/RIF for 8. Shah S, Demissie M, Lambert L, et al. Intensified tuberculosis case finding among tuberculosis diagnosis in South Africa: a real-world cost analysis and economic HIV-Infected persons from a voluntary counseling and testing center in Addis evaluation. Lancet Glob Health 2017; 5:e710–9. Ababa, Ethiopia. J Acquir Immune Defic Syndr 2009; 50:537–45. 33. Datiko DG, Lindtjørn B. Cost and cost-effectiveness of smear-positive tubercu- 9. Adelman MW, Tsegaye M, Kempker RR, et al. Intensified tuberculosis case find- losis treatment by Health Extension Workers in Southern Ethiopia: a community ing among HIV-infected persons using a WHO symptom screen and Xpert(®) randomized trial. PLoS One 2010; 5:e9158. MTB/RIF. Int J Tuberc Lung Dis 2015; 19:1197–203. 34. Cain KP, McCarthy KD, Heilig CM, et al. An algorithm for tuberculosis screening 10. Balcha TT, Sturegård E, Winqvist N, et al. Intensified tuberculosis case-finding and diagnosis in people with HIV. N Engl J Med 2010; 362:707–16. in HIV-positive adults managed at Ethiopian health centers: diagnostic yield of 35. Alland D, Rowneki M, Smith L, et al. Xpert MTB/RIF Ultra: a new near-patient Xpert MTB/RIF compared with smear microscopy and liquid culture. PLoS One TB test with sensitivity equal to culture. Paper presented at: CROI 2015. Seattle, 2014; 9:e85478. WA; February 23−26, 2015. Available at: http://www.croiwebcasts.org/console/ 11. The World Bank. GDP per capita (current US$). 2015. Available at: http://data. player/25681?mediaType=slideVideo&. Accessed 1 June 2015. worldbank.org/indicator/NY.GDP.PCAP.CD. Accessed 9 June 2016. 36. Find Diagnostics. Report for WHO: a multicentre non-inferiority diagnositc 12. World Health Organization. Ethiopia. Available at: http://www.who.int/coun- accuracy study of the Ultra Assay compared to the Xpert MTB/RIF Assay. 2017. tries/eth/en/. Accessed 20 August 2014. Available at: https://www.finddx.org/wp-content/uploads/2017/03/Ultra-WHO- 13. Getahun H, Kittikraisak W, Heilig CM, et al. Development of a standardized report_24MAR2017_FINAL.pdf. Accessed 8 January 2018. screening rule for tuberculosis in people living with HIV in resource-constrained 37. Shinnick TM, Starks AM, Alexander HL, Castro KG. Evaluation of the Cepheid settings: individual participant data meta-analysis of observational studies. PLoS Xpert MTB/RIF assay. Expert Rev Mol Diagn 2015; 15:9–22. Med 2011; 8:e1000391. 38. Tadesse M, Aragaw D, Dimah B, et al. Xpert MTB/RIF for rapid detection of 14. World Health Organization. Xpert MTB/RIF assay for the diagnosis of pulmo- rifampicin-resistant Mycobacterium tuberculosis from pulmonary tuberculosis nary and extrapulmonary TB in adults and children. 2013. Available at: http:// patients in Southwest Ethiopia. Int J Mycobacteriol 2016; 5 (Suppl 1):S48–9. apps.who.int/iris/bitstream/10665/112472/1/9789241506335_eng.pdf ?ua=1. 39. Salomon JA, Vos T, Hogan DR, et al. Common values in assessing health out- Accessed 20 August 2014. comes from disease and injury: disability weights measurement study for the 15. Lawn SD, Brooks SV, Kranzer K, et al. Screening for HIV-associated tuberculosis Global Burden of Disease Study 2010. Lancet 2012; 380:2129–43. and rifampicin resistance before antiretroviral therapy using the Xpert MTB/RIF 40. Vassall A, van Kampen S, Sohn H, et al. Rapid diagnosis of tuberculosis with the assay: a prospective study. PLoS Med 2011; 8:e1001067. Xpert MTB/RIF assay in high burden countries: a cost-effectiveness analysis. 16. Lawn SD, Nicol MP. Xpert® MTB/RIF assay: development, evaluation and imple- PLoS Med 2011; 8:e1001120. mentation of a new rapid molecular diagnostic for tuberculosis and rifampicin 41. World Health Organization. Global tuberculosis report 2012. 2012. Available resistance. Future Microbiol 2011; 6:1067–82. at: http://apps.who.int/iris/bitstream/10665/75938/1/9789241564502_eng.pdf. 17. Boehme CC, Nabeta P, Hillemann D, et al. Rapid molecular detection of tubercu- Accessed 24 August 2014. losis and rifampin resistance. N Engl J Med 2010; 363:1005–15. 42. Seung KJ, Omatayo DB, Keshavjee S, et al. Early outcomes of MDR-TB treatment 18. Theron G, Zijenah L, Chanda D, et al; TB-NEAT team. Feasibility, accuracy, and in a high HIV-prevalence setting in Southern Africa. PLoS One 2009; 4:e7186. clinical effect of point-of-care Xpert MTB/RIF testing for tuberculosis in pri- 43. World Health Organization. Metrics: disability-adjusted life year (DALY). Health mary-care settings in Africa: a multicentre, randomised, controlled trial. Lancet statistics and information systems. Available at: http://www.who.int/healthinfo/ 2014; 383:424–35. global_burden_disease/metrics_daly/en/. Accessed 1 September 2014. 8 • OFID • Adelman et al. Downloaded from https://academic.oup.com/ofid/article-abstract/5/1/ofx269/4773940 by Ed 'DeepDyve' Gillespie user on 16 March 2018
Open Forum Infectious Diseases – Oxford University Press
Published: Jan 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
All the latest content is available, no embargo periods.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud