http://orcid.org/0000-0002-0330-1824
Figures
Abstract
Background
Truenat is a novel molecular assay that rapidly detects tuberculosis (TB) and rifampicin-resistance. Due to the portability of its battery-powered testing platform, it may be valuable in peripheral healthcare settings in India.
Methods
Using a microsimulation model, we compared four TB diagnostic strategies for HIV-negative adults with presumptive TB: (1) sputum smear microscopy in designated microscopy centers (DMCs) (SSM); (2) Xpert MTB/RIF in DMCs (Xpert); (3) Truenat in DMCs (Truenat DMC); and (4) Truenat for point-of-care testing in primary healthcare facilities (Truenat POC). We projected life expectancy, costs, incremental cost-effectiveness ratios (ICERs), and 5-year budget impact of deploying Truenat POC in India’s public sector. We defined a strategy “cost-effective” if its ICER was <US$990/year-of-life saved (YLS). Model inputs included: TB prevalence, 15% (among those not previously treated for TB) and 27% (among those previously treated for TB); sensitivity for TB detection, 89% (Xpert) and 86% (Truenat); per test cost, $12.63 (Xpert) and $13.20 (Truenat); and linkage-to-care after diagnosis, 84% (DMC) and 95% (POC). We varied these parameters in sensitivity analyses.
Results
Compared to SSM, Truenat POC increased life expectancy by 0.39 years and was cost-effective (ICER $210/YLS). Compared to Xpert, Truenat POC increased life expectancy by 0.08 years due to improved linkage-to-care and was cost-effective (ICER $120/YLS). In sensitivity analysis, the cost-effectiveness of Truenat POC, relative to Xpert, depended on the diagnostic sensitivity of Truenat and linkage-to-care with Truenat. Deploying Truenat POC instead of Xpert increased 5-year expenditures by $270 million, due mostly to treatment costs. Limitations of our study include uncertainty in Truenat’s sensitivity for TB and not accounting for the “start-up” costs of implementing Truenat in the field.
Citation: Lee DJ, Kumarasamy N, Resch SC, Sivaramakrishnan GN, Mayer KH, Tripathy S, et al. (2019) Rapid, point-of-care diagnosis of tuberculosis with novel Truenat assay: Cost-effectiveness analysis for India’s public sector. PLoS ONE 14(7): e0218890. https://doi.org/10.1371/journal.pone.0218890
Editor: Selvakumar Subbian, Rutgers Biomedical and Health Sciences, UNITED STATES
Received: February 15, 2019; Accepted: June 11, 2019; Published: July 2, 2019
Copyright: © 2019 Lee 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.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. Data underlying the results of the study, “Use of Xpert MTB/RIF in Decentralized Public Health Settings and Its Effect on Pulmonary TB and DR-TB Case Finding in India” are publicly available at https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0126065 (Supporting Information File 2). All other relevant data are within the present manuscript and its Supporting Information files.
Funding: This work was supported by awards from the National Institute on Drug Abuse (K01 DA042687 [K.P.R.] and R01 DA015612 [A.D.P.]), the National Institute of Allergy and Infectious Diseases (R37 AI058736 [K.A.F.] and R37 AI093269 [K.P.R.]), and the National Institute of Mental Health (R01 MH105203 [A.D.P.]) of the National Institutes of Health (www.nih.gov). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health.
Competing interests: The authors have declared that no competing interests exist.
Introduction
With approximately 2.8 million cases annually, India has the world’s highest incidence of tuberculosis (TB) [1]. However, due to the widespread use of diagnostics with low sensitivity (e.g., sputum smear microscopy) and low linkage-to-care rates, over 25% of patients seeking care in India’s public sector are neither diagnosed nor started on treatment [2].
New rapid molecular diagnostics could dramatically increase TB detection and linkage-to-care, which are key components of both the World Health Organization’s (WHO) End TB Strategy and India’s National Strategic Plan for Tuberculosis Elimination 2017–2025 [1,3]. The Xpert MTB/RIF assay (Cepheid, Sunnyvale, CA, USA) is currently the only WHO-endorsed molecular test able to rapidly detect both TB and rifampicin (RIF)-resistance [1]. While there has been interest in deploying Xpert in peripheral laboratories [3–6], its decentralization may be limited by infrastructure requirements, including continuous power supply and air-conditioning [7,8].
Truenat (Molbio Diagnostics/Bigtec Labs, Goa/Bengaluru, India), is a new chip-based, micro real-time polymerase chain reaction (PCR) test that detects tubercle bacilli in sputum samples in approximately one hour [1,9,10]. Upon receiving a positive test result, an “add-on” chip can be used to detect RIF-resistance, adding another one hour of test time. The test is prepared and run on the battery-powered Truelab system, which includes the sample preparation device (i.e., machine for DNA extraction and purification from the sputum sample) and the PCR analyzer device, available in 1-, 2-, or 4-module configurations with the lattermost capable of testing four samples simultaneously. Due to the portability of this testing platform, Truenat may be valuable in peripheral healthcare settings, such as designated microscopy centers (DMCs) and primary healthcare facilities in India.
If used as a point-of-care (POC) test within primary healthcare facilities, Truenat could increase treatment initiation by reducing turnaround time for test results and decreasing the need for laboratory referrals [2,11]. However, uncertainties in parameter values, such as test characteristics and linkage-to-care, must be investigated. In resource-constrained settings, the potential benefits of Truenat must also be weighed against its costs. Using a mathematical model, therefore, we projected the clinical impact, costs, and cost-effectiveness of Truenat, as a replacement for smear microscopy or Xpert. We also evaluated the budget impact of deploying Truenat widely in India’s public sector.
Methods
Analytic overview
We expanded the Cost-Effectiveness of Preventing AIDS Complications-International (CEPAC-I) model, a validated and widely published, individual-based Monte Carlo state-transition model, developed and written in C++ [12–14], to account for TB natural history, diagnosis, and treatment. We simulated a cohort of adult, HIV-negative patients with presumptive pulmonary TB, defined as individuals aged ≥15 years with cough of ≥2 weeks duration [4,15], undergoing TB testing at DMCs and their attached primary healthcare facilities (e.g., primary health centers, hospital outpatient clinics, etc.) under India’s Revised National Tuberculosis Control Programme (RNTCP). Although HIV is an important risk factor for TB, we focused on the HIV-negative population, among whom 97% of new TB cases in India occur [1].
We compared the clinical and economic outcomes of four TB diagnostic strategies: (1) sputum smear microscopy in DMCs (SSM); (2) Xpert in DMCs (Xpert); (3) Truenat in DMCs (Truenat DMC); and (4) Truenat for point-of-care testing in primary healthcare facilities (Truenat POC). In each strategy, patients for whom clinical suspicion of TB remains high despite a negative test result may undergo additional testing. Patients previously treated for TB, and therefore at higher risk of drug-resistance, may receive additional tests to confirm multidrug-resistant TB (MDR-TB) (S1 Appendix; Figures A-C in S1 Appendix).
Model-generated outcomes included the correct detection and linkage-to-care of TB cases; life expectancy; lifetime TB-related healthcare costs; and incremental cost-effectiveness ratio (ICER), the difference between two strategies in costs (2017 US dollars) divided by the difference in life-years. We considered a strategy cost-effective if its ICER was less than US$990/year-of-life saved (YLS), an opportunity-based cost-effectiveness threshold that is 50% of India’s 2017 gross domestic product (GDP) per capita (S1 Appendix). We projected clinical and economic outcomes over patients’ lifetimes, while also considering shorter time horizons. We report outcomes discounted 3%/year for cost-effectiveness analysis and undiscounted outcomes for clinical and budget evaluations [16].
Model overview
A simulated cohort of patients enters the model upon seeking care in India’s public sector for symptoms suggestive of TB. They undergo a TB diagnostic protocol per national guidelines [4,15]. The model draws randomly from user-defined characteristics (e.g., age, sex, TB status), informed by a recent implementation study of Xpert in India for patients with presumptive TB [4]. As individuals transition monthly through “states” of TB progression and treatment, the model tracks clinical outcomes (e.g., cure, relapse, life-years accrued) and monthly TB-related healthcare costs (e.g., diagnostic tests, drugs, clinic visits). Throughout the simulation, all individuals are subject to age- and sex-stratified background mortality risks specific for India, while those with active, untreated TB have an excess mortality risk. Model details are in the S1 Appendix and at http://www.massgeneral.org/mpec/cepac/.
To initiate TB treatment, a simulated individual must: (1) complete the diagnostic pathway, including retrieving test results; (2) receive a diagnosis of TB and/or drug-resistance, as determined by test characteristics; and (3) link to a primary healthcare facility that will initiate and monitor TB treatment (“linkage-to-care”). In the diagnostic pathway for Truenat DMC or Truenat POC, an individual undergoes testing for RIF-resistance only if the individual first receives a positive test result for TB. Those who receive a TB diagnosis via Truenat POC have a higher probability of linking to care for TB treatment compared to those who are diagnosed in DMCs, who require a referral to a primary healthcare facility (S1 Appendix) [2].
Base case input parameters
Cohort characteristics and TB prevalence.
Cohort characteristics were derived from an implementation study of Xpert for individuals with presumptive TB in DMCs of 18 sub-district level TB program units, chosen for being geographically and demographically representative of the national population in India (Table 1) [4]. Mean age was 41 years, 36% were women, and 17% had been previously treated for TB. Among patients with presumptive TB, we assumed TB prevalence was 15% for patients with no prior TB treatment and 27% for previously treated patients (S1 Appendix) [4].
Diagnostic tests.
Sensitivity and specificity of smear microscopy for two sputum samples were 64% and 98% compared to culture (Table 1) [18]. Sensitivity and specificity of Xpert were 89% and 99% compared to culture, per the results of a meta-analysis [22]. One validation study in India reports Truenat’s sensitivity for TB detection as 96% compared to Xpert [9]. To compare Truenat to culture, we multiplied 96% (Truenat’s sensitivity against Xpert) by 89% (Xpert’s sensitivity against culture). Therefore, the sensitivity of Truenat was 86% compared to culture. We assumed 99% specificity for Truenat, similar to Xpert (Table 1) [22]. We varied Truenat sensitivity and specificity in sensitivity analyses.
Linkage-to-care and treatment.
Linkage-to-care was 84% for patients diagnosed in DMCs and 95% for those diagnosed by a POC test (Truenat POC only) [2]. The latter was an evidence-supported assumption (S1 Appendix) that we varied in sensitivity analyses. For patients who received a negative result after POC testing with Truenat but were subsequently diagnosed via culture, linkage-to-care remained 84%. Treatment-related parameters, including loss to follow-up (LTFU), were derived from Indian TB surveillance data (Table 1) [5]. The monthly probability of LTFU remained 1% regardless of the severity of TB-related symptoms. We varied this assumption in sensitivity analysis.
Costs.
Using a microcosting approach, we derived unit costs of TB care from a health system perspective [19,21,23,24]. We multiplied unit costs by their expected quantities, either as indicated by published guidelines (e.g., number of clinic visits) or as reported in epidemiological studies (e.g., hospitalization rates) (S1 Appendix). The costs per test for sputum smear microscopy, Xpert, and Truenat were $0.86, $12.63, and $13.20 (Table 1), which included costs of overhead and building space, labor, reagents (e.g., cartridges for Xpert, chips for Truenat), and equipment (test instruments). The monthly costs of TB treatment were $28.13 (first-line), $32.25 (retreatment), and $104.23 (second-line), which included the costs of drugs, monitoring tests, clinic visits, and hospitalizations during treatment (S1 Appendix) [19,21,23,24].
We obtained price estimates for the Truenat platform, including the Truenat chip and the Truelab system, from the manufacturer (Sriram Natarajan, Director and CEO of Molbio, personal communication). The price of the Truelab system, which includes the sample preparation device and the 4-module PCR analyzer device (capable of testing four specimens simultaneously), is $14,150. We used this cost in the base case for comparison with the 4-module Xpert system (assumed to be the standard in India). This cost was annualized over the expected lifespan of the Truelab system, discounted 3%/year, and divided by the expected number of tests it would perform annually. The price of the Truenat chip for TB detection is $12.40, and the chip for RIF-resistance detection will be provided free of cost based on an average estimated TB-positive proportion of 20%. These initial price estimates for the public sector may change based on volume commitment by the government. We assumed that overhead, building space, and labor-related costs for Truenat strategies would be similar to those of Xpert [19].
Additional input parameters regarding TB prevalence, natural history, diagnostics, and treatment are in S1 Appendix. The majority of the data inputs in this study, including those related to population demographics, gaps in the TB cascade of care, and diagnostic test and treatment costs, specifically apply to the public sector in India.
Sensitivity and scenario analyses
To account for uncertainty, we varied key parameters (e.g., test sensitivity, linkage-to-care, costs) across a wide range of possible values, informed by literature whenever possible (Table 1). We evaluated the effect of empirical treatment on cost-effectiveness. Specifically, we considered the scenario in which 16% of those with a negative smear, Xpert, or Truenat result receive empirical treatment (S1 Appendix). We performed a scenario analysis in which patients who received a TB diagnosis under the Truenat POC strategy are twice as likely to be lost to follow-up during treatment compared to patients diagnosed under the SSM strategy (2% from 1%). This was intended to reflect that patients who are rapidly diagnosed and linked to care with POC testing are likely to have TB symptoms that are less severe and are, therefore, more likely to discontinue TB treatment.
We also evaluated scenarios in which the per-test costs of operating the Truenat platform are higher in primary healthcare facilities than they would be in DMCs due to economies of scale. This was intended to reflect that compared to DMCs, primary healthcare facilities are likely to experience lower test volumes, resulting in the less efficient use of certain material and human resources, such as overhead, building space, equipment (test instruments), and some elements of labor (S1 Appendix). Specifically, we considered the scenarios in which primary healthcare facilities have test volumes 5- and 10-fold lower than those of DMCs, increasing the per-test cost of Truenat POC from $13.20 (base case) to $15.32 and $17.96, respectively (S1 Appendix).
We conducted a two-way sensitivity analysis, simultaneously varying Truenat’s sensitivity for TB detection and linkage-to-care at a 5-year horizon, to define the combination of parameter values by which Truenat POC would be more economically efficient than Xpert. Because the public sector cost of Truenat may decrease based on volume commitment by the Indian government, we conducted this same analysis for a scenario in which the Truenat chip cost is negotiated to 60% of the current estimate (S1 Appendix).
Budget impact analysis
We projected costs associated with widespread deployment of diagnostic strategies in India’s public sector over a 2-year and 5-year period. We assumed 7.9 million adults with presumptive TB would be tested annually (S1 Appendix).
Results
Base case clinical outcomes
Truenat DMC, compared to SSM, increased life expectancy by 0.30 years (undiscounted) but, compared to Xpert, decreased life expectancy by 0.01 years (Table 2). Truenat POC was the most effective strategy, increasing life expectancy by 0.39 years compared to SSM and by 0.08 years compared to Xpert. Compared to SSM and Xpert, Truenat POC also increased the number of TB cases correctly detected and linked to care by 590 and 140, respectively, per 10,000 individuals with presumptive TB.
Base case lifetime costs and cost-effectiveness
Compared to SSM, Truenat DMC and Truenat POC strategies both increased discounted per-patient lifetime costs by ~$40 (Table 2). Compared to Xpert, Truenat DMC decreased discounted per-patient lifetime costs by $1 and Truenat POC increased costs by $5.
While Truenat DMC was cost-effective compared to SSM (ICER $240/YLS), it resulted in lower life expectancy and higher ICER than Xpert and was, therefore, “weakly dominated” (i.e., economically inefficient). Truenat POC was cost-effective compared to both SSM (ICER $210/YLS) and Xpert (ICER $120/YLS). When viewed over different time horizons, Truenat POC became cost-effective compared to Xpert and SSM after 4 and 6 years, and Xpert became cost-effective compared to SSM after 6 years (Figure E in S1 Appendix). The respective ICERs continued to decrease beyond these time horizons.
One-way sensitivity and scenario analyses
In one-way sensitivity analyses comparing Truenat POC to SSM at a lifetime horizon, Truenat POC remained cost-effective compared to SSM across all parameter values analyzed (Fig 1; Figure F in S1 Appendix). Decreasing Truenat’s sensitivity for TB detection by an absolute 20% (from 86% to 66%) resulted in ~27% change in the base case ICER of Truenat POC versus SSM, at a lifetime horizon. Varying TB and MDR-TB prevalence had little influence on the ICER of Truenat POC. Varying the specificity of Truenat across a range of values (80% to 100%) also had little influence on the ICER of Truenat POC. While Truenat’s specificity for RIF-resistance detection was the most influential among the parameters considered, the ICER ($350/YLS) that resulted from decreasing the specificity by 10% remained well below the cost-effectiveness threshold of $990/YLS.
TB: tuberculosis. MDR-TB: multidrug-resistant tuberculosis. RIF: rifampicin. C&DST: culture and drug-susceptibility test. yr: year. POC: point-of-care. “previously treated”: previously treated for TB. GDP: gross domestic product. ICER: incremental cost-effectiveness ratio. USD: United States dollars. YLS: year-of-life saved. Horizontal bars represent ranges of ICERs when varying each model parameter across its plausible range. The vertical dashed line represents 50% of the GDP per capita of India in 2017 ($990), which we consider the cost-effectiveness threshold (see Methods). ICERs less than $990/YLS (left of dashed line) are considered cost-effective.
When accounting for empirical treatment in 16% of those with a negative smear, Xpert, or Truenat result, Truenat POC remained cost-effective compared to SSM (ICER $290/YLS) and compared to Xpert (ICER $200/YLS). Increasing the monthly probability of LTFU for patients who received a TB diagnosis under the Truenat POC strategy to twice the monthly probability of LTFU for patients diagnosed under the SSM strategy (2% from 1%) resulted in a minimal change (~3%) in the base case ICER of Truenat POC versus SSM. In the scenario that primary healthcare facilities have test volumes 5-fold lower than those of DMCs (increasing the per-test cost of Truenat POC to $15.32), Truenat POC remained cost-effective compared to SSM (ICER $220/YLS) and compared to Xpert (ICER $170/YLS). In the scenario that primary healthcare facilities have test volumes 10-fold lower than those of DMCs (increasing the per-test cost of Truenat POC to $17.96), Truenat POC remained cost-effective compared to SSM (ICER $240/YLS) and compared to Xpert ($240/YLS).
We compared Truenat POC to Xpert over a shorter 5-year horizon, varying Truenat’s sensitivity for TB detection (Fig 2). When Truenat’s sensitivity was ≥78%, Truenat POC increased life-years, increased costs, and was cost-effective compared to Xpert. The higher cost was driven mostly by the increased number of patients initiating treatment; however, these costs were offset by improvements in clinical outcomes, resulting in overall decreasing ICERs as sensitivity increased. When Truenat’s sensitivity was <75%, Truenat POC, compared to Xpert, resulted in fewer life-years and lower cost. In this scenario, the ICER of Xpert compared to Truenat POC was below the cost-effectiveness threshold, indicating that Xpert was cost-effective.
TB: tuberculosis. POC: point-of-care. ICER: incremental cost-effectiveness ratio. YLS: year-of-life saved. USD: United States dollars. This plot shows the differences in life expectancy and costs between Truenat POC and Xpert at a 5-year horizon when varying the sensitivity of Truenat for TB detection. The horizontal axis is the sensitivity of Truenat for TB detection. The blue line corresponds to the left vertical axis, which is the difference in life expectancy between Truenat POC and Xpert. The red line corresponds to the right vertical axis, which is the difference in per-person lifetime costs between Truenat POC and Xpert. The ICER (i.e., the difference in costs divided by the difference in life expectancy) is provided at regular intervals of test sensitivity values. For integer values of test sensitivity ≥78% (green panel), Truenat POC is cost-effective compared to Xpert (ICER <$990/YLS). For integer values <78% (red panel), Xpert is more efficient than Truenat POC. a“Xpert is more efficient than Truenat POC”: For Truenat sensitivity values <75%, Xpert was cost-effective compared to Truenat POC (ICER <$990/YLS). At Truenat sensitivity of 75–76%, Xpert was cost-saving (lower cost, higher clinical benefit [more life-years accrued]) compared to Truenat POC. At Truenat sensitivity of 77%, Xpert was decrementally cost-effective (lower cost and lower clinical benefit but with ICER >$990/year-of-life lost [YLL]—that is, at least $990 saved per year-of-life-lost) compared to Truenat POC.
Two-way sensitivity and scenario analyses
We simultaneously varied Truenat’s sensitivity for TB detection (68–100%) and linkage-to-care (84–100%), comparing Truenat POC to Xpert over a 5-year horizon (Fig 3A). We kept Xpert’s sensitivity and linkage-to-care at base case values. Truenat POC, at 86% sensitivity, was cost-effective when linkage-to-care was ≥88%. This linkage threshold for cost-effectiveness increased as Truenat sensitivity decreased. For sensitivity values ≤74%, Truenat POC was not cost-effective at any linkage value. Above 90% sensitivity, Truenat POC was cost-effective at linkage values as low as 84% (same linkage as Xpert).
TB: tuberculosis. POC: point-of-care. YLS: year-of-life saved. YLL: year-of- lost. These heat maps evaluate the incremental cost-effectiveness ratio of Truenat POC strategy relative to Xpert at a 5-year time horizon for different values of Truenat sensitivity for TB detection and linkage-to-care. Each panel displays different costs of Truenat, including the scenario (B), in which the price of the Truenat chip is negotiated to 60% of its current estimate for the public sector (S1 Appendix). Sensitivity of Truenat for TB detection increases from left to right on the horizontal axes. The probability of linking to care upon receiving a positive TB test result with Truenat increases up the vertical axes. a“Decrementally cost-effective”: Truenat POC results in lower cost and lower clinical benefit compared to Xpert, but with ICER >$990/year-of-life lost (YLL)—that is, at least $990 is saved per year-of-life lost. b“Xpert is more efficient than Truenat POC”: Xpert is either cost-effective (ICER <$990/YLS), cost-saving (lower cost, higher clinical benefit [more life-years accrued]), or decrementally cost-effective (ICER >$990/YLL), compared to Truenat POC.
In the scenario that Truenat’s chip cost is reduced to 60% of its current estimate, the range of parameter values for which Truenat POC was cost-effective (or cost-saving) compared to Xpert broadened (Fig 3B). At 95% linkage (as assumed for POC test), Truenat POC was cost-effective when sensitivity was ≥88%, cost-saving when sensitivity was 77–87%, and decrementally cost-effective (“much less costly and almost as good”) when sensitivity was ≤76%. When linkage was 84% (typical of DMC), Truenat POC was cost-saving when sensitivity was >88% and decrementally cost-effective when sensitivity was 74–88%.
Budget impact analysis
Compared to country-wide use of SSM in India’s public sector, scaling up Xpert increased cumulative TB-related healthcare expenditures by $580 million (81% increase) over 2 years and by $1.58 billion (80% increase) over 5 years (Fig 4). Most of the difference in costs over 5 years was due to increased spending on MDR-TB treatment (56% of the increase), followed by diagnostic tests (37% of the increase). Deploying Truenat POC instead of Xpert increased cumulative healthcare expenditures by $100 million (7% increase) over 2 years and by $270 million (8% increase) over 5 years. Most of the difference in costs over 5 years was due to increased spending on MDR-TB treatment (63% of the increase), followed by drug-susceptible TB treatment (22% of the increase).
TB: tuberculosis. ds-TB: drug-susceptible tuberculosis. MDR-TB: multidrug-resistant tuberculosis. SSM: sputum smear microscopy. POC: point-of-care. mill: million. bill: billion. Budget impact analysis of full public sector implementation of sputum smear microscopy (SSM), Xpert, and Truenat POC strategies over 2- and 5-year time horizons. Cumulative TB-related costs (2017 USD, billions) are on the vertical axis. This analysis assumes that 7.9 million adults in India are tested each year for symptoms suggestive of TB (S1 Appendix) [4]. All calculations were made using exact numbers before rounding to the nearest $10 million (for costs) and 1% (for percentages). aEach treatment regimen is associated with a frequency of clinic visits and rate of hospitalization during the course of TB treatment, as reported by published guidelines and/or epidemiological data (S1 Appendix). These clinical costs are incorporated into the budget impact projection for each category.
Discussion
A major WHO priority for TB diagnostics is to implement a rapid, sputum-based molecular test to replace smear microscopy at the peripheral level (i.e., microscopy centers and attached primary healthcare facilities) [7,8]. Our model-based analysis shows that in India, Truenat, when replacing smear microscopy and used at point-of-care, increases the number of TB cases correctly detected and linked to care by 590 per 10,000 individuals with presumptive TB. It also increases life expectancy by nearly 0.4 years and is cost-effective. While Truenat DMC was economically inefficient among the four strategies, it was cost-effective when compared directly to SSM. The cost-effectiveness of Truenat POC, compared to SSM, was consistent across a wide range of clinical and cost parameter values.
The WHO’s target product profile (TPP) of the “smear replacement test” includes a set of minimal and optimal requirements [7,8]. Truenat fits many minimal TPP standards, including battery-powered operation and <2 hours to result. However, it currently falls short of optimal TPP standards for test characteristics (i.e., sensitivity for TB detection ideally better than Xpert) and minimal standards for price (i.e., <$6 per reagent and <$1,400 per instrument) [7,8]. Our analysis shows that despite these limitations, Truenat POC increases life expectancy and is cost-effective compared to SSM or Xpert.
Our analysis, however, reveals an important relationship between TB detection sensitivity and linkage-to-care. As Truenat’s TB detection sensitivity decreases, the linkage-to-care level necessary for Truenat POC to be cost-effective compared to Xpert increases. This interplay is consistent with other modeling work showing that a theoretical, peripheral-level TB test with inferior performance characteristics but improved treatment initiation rates, compared to district-level Xpert testing, would reduce TB transmission and incidence [25]. Our results show that such a test can also be cost-effective. These findings together imply that increasing case detection (via improved test performance characteristics) and increasing the number of patients on life-prolonging treatment (via improved linkage-to-care) may yield synergistic results from a cost-effectiveness standpoint.
Truenat POC will be even more economically efficient if the price of the Truenat chip is reduced to 60% of its current estimate (with Truenat, therefore, becoming less expensive than Xpert). This reduced price would still be higher than the <$6 TPP stipulation. The 4-module Truelab system is also currently ten times more expensive than the TPP price stipulation for a test instrument. Even so, our scenario analysis shows Truenat POC is cost-effective, cost-saving, or decrementally cost-effective (“much less costly and almost as good”), relative to Xpert, across the wide range of parameter values considered.
Results from our scenario analysis are notable in that improvements in Truenat’s TB detection sensitivity and linkage-to-care yield favorable but higher ICERs. For example, Truenat POC, at 86% sensitivity and 90% linkage-to-care, is cost-saving (more life-years, lower cost) compared to Xpert, but becomes cost-effective (more life-years, higher cost) if linkage-to-care improves to 100%. This occurs because TB treatment costs, not diagnostic test costs, are the main driver of lifetime costs. Factors leading to higher treatment initiation increase the cumulative costs per patient. Our analysis, nonetheless, shows that this increase in cumulative costs is justified, from a cost-effectiveness perspective, by the clinical benefit of more patients receiving life-prolonging treatment (i.e., more life-years are saved at an acceptable additional cost). Such findings have also been observed in HIV screening programs, where a major driver of cost is the treatment pathway triggered when a previously undetected case is identified and the patient is linked to life-prolonging care [26,27].
The RNTCP can consider these findings as it pursues its ambitious National Strategic Plan (NSP) to eliminate TB in India by 2025 [3]. The NSP’s projected budget of $2.49 billion for the 2017–2020 period assumes approximately half of patients with TB receive molecular testing. Additional diagnostics are being planned for another 4.5 million patients during this period [3]. Our analysis shows that scaling up molecular diagnostics will increase the required budget but the majority of the cost will be from MDR-TB treatment. A recent economic analysis for India similarly found that full replacement of smear microscopy with Xpert would substantially increase budget requirements but would result in lower cost per MDR-TB case initiated on treatment [21]. As the RNTCP plans its NSP budget for 2020–2025, it should consider MDR-TB treatment costs as much as, if not more than, the prices of diagnostic tests.
A “real-world” economic evaluation by Vassall et al. showed that Xpert did not improve the cost-effectiveness of drug-susceptible TB diagnosis and treatment over a 6-month horizon, contrary to previous projections [28]. This finding underscores the need for cost-effectiveness analyses to account for uncertainties in implementation constraints [29]. In modeling our diagnostic and treatment pathway, we adjusted for many demand-side constraints, including care-seeking, test uptake, and adherence. We also adjusted for some supply-side constraints, including infrastructure limitations. While these parameters vary with geographical region (e.g., rural, urban, etc.), we drew on aggregated sources, including national surveillance data and studies using demographically and geographically representative population samples from India [2,4,6], to adjust for the “average” peripheral healthcare setting in India. We found that neither Xpert nor Truenat POC was cost-effective compared to SSM until 6 years after initial testing, well beyond the 6-month time horizon considered by Vassall et al [28].
Our findings apply to adult, HIV-negative individuals with presumptive pulmonary TB. In populations with high prevalence of undertreated HIV [28], the potential clinical benefits of Xpert and Truenat POC could be offset by high HIV-related mortality. HIV is also associated with lower sensitivity of TB diagnostics and substantial costs of screening and treatment [1,22,26,27]. Thus, studies of Truenat POC testing in this vulnerable population would be valuable.
We focused our analysis closely on the comparison of Truenat POC and Xpert because the decentralization of rapid, molecular TB diagnostics to the peripheral healthcare setting is currently of policy relevance to India and other high-burden countries [4,6,21,30,31]. We, however, did not consider a POC strategy for Xpert. Studies conducted primarily within well-resourced clinics in South Africa have demonstrated the feasibility of Xpert POC testing [32–36]. We are aware of only one study in India evaluating Xpert POC testing, within an outpatient clinic of a well-resourced tertiary hospital [31]. It remains unclear whether this strategy can be replicated universally across diverse settings, including rural and tribal/hilly areas of India [30,31,37]. While infrastructure concerns may have, to an extent, deterred interest in a POC strategy for Xpert in many high-burden countries [30,38], they have also spurred substantial research and development in a new generation of molecular diagnostics, such as Truenat, specifically intended for the primary healthcare setting [7,8]. Our study is focused on this latter area of policy interest.
Limitations of our study include those related to parameter inputs and model structure. We did not account for reduced transmission from faster diagnosis and treatment initiation, which could improve the cost-effectiveness of POC testing at a broader scale. Because preference weights for health states were not incorporated into the model, our study did not account for the additional benefits of reduced TB-related morbidity that could result from POC testing. We also did not consider some supply-side constraints that could disrupt successful POC testing and treatment, such as irregularities in test reagent and drug supply chains, potential lack of human resources, and variations in provider adherence to the diagnostic and treatment pathway [29]. Due to the lack of microcosting data regarding the unit costs of POC testing within the primary healthcare setting, our base case analysis was limited in its ability to adjust for the cost disadvantages of conducting tests at a lower operational scale, though we evaluated this in scenario analysis.
Importantly, more data are needed regarding Truenat test characteristics. In our analysis, the estimated sensitivity of Truenat for TB detection was based on a single study, and RIF-resistance detection performance was based on a clinical validation study conducted by the manufacturer. We, therefore, varied these parameters widely in sensitivity analyses. Test characteristics may improve with newer versions of Truenat under evaluation. Furthermore, as Truenat chips for other infectious diseases (e.g., HIV, malaria, HCV), which utilize the same Truelab system, become validated for India, the wide implementation of the Truenat platform could eventually help strengthen and improve the clinical and economic efficiency of the general health system.
Our study was limited to a health system perspective and does not include economic costs or savings to patients. Our healthcare expenditure projections should also be interpreted only for diagnostic test costs, drug costs, and treatment-associated clinic, monitoring, and hospitalization costs. We did not include “start-up” costs of establishing Xpert or Truenat in the field, such as costs of training staff to utilize the machines and costs of maintaining steady supply chains and reporting systems. Implementation studies are needed and ongoing to demonstrate Truenat’s efficacy under real-world conditions. As results become available, dedicated studies will be needed to estimate the implementation costs of Truenat on a larger scale, as there have been for Xpert [31,39–41].
Truenat is, nonetheless, the first TB test with capacity comparable to the 4-module Xpert but with operational features suited for the peripheral level. Our model-based analysis shows that, when used at point-of-care for TB diagnosis, Truenat improves linkage-to-care, increases life expectancy, and is cost-effective compared with smear microscopy or Xpert. Appropriate diagnostics are needed at every level of the healthcare system [8]. Truenat deployed at the peripheral level may be complementary to other diagnostic technologies, such as Xpert and Xpert Ultra [42], which are appropriate for the district and sub-district levels, and Xpert Omni and other compact, 1-module testing platforms, which may be valuable for community-based active case-finding (clinicaltrials.gov, NCT 03168945) [43]. In this way, Truenat should contribute to TB control and, thus, should be more widely utilized in India.
Acknowledgments
The authors thank Liyang Yu (Massachusetts General Hospital) for statistical assistance, Ramnath Subbaraman (Tufts University School of Medicine) for insights on the TB care pathway in India, and Claudia Denkinger (Foundation for Innovative New Diagnostics) for guidance on policy-relevant considerations for TB diagnostics.
References
- 1.
World Health Organization. Global tuberculosis report 2017 [Internet]. Geneva, Switzerland: World Health Organization; 2017. Available from: http://www.who.int/tb/publications/global_report/en/
- 2. Subbaraman R, Nathavitharana RR, Satyanarayana S, Pai M, Thomas BE, Chadha VK, et al. The tuberculosis cascade of care in India’s public sector: a systematic review and meta-analysis. PLoS Med. 2016;13: e1002149. pmid:27780217
- 3.
Central TB Division. Revised National Tuberculosis Control Programme: national strategic plan for tuberculosis elimination 2017–2025 [Internet]. New Delhi, India: Directorate General of Health Services, Ministry of Health and Family Welfare; 2017. Available from: https://tbcindia.gov.in/index1.php?lang=1&level=1&sublinkid=4768&lid=3266
- 4. Sachdeva KS, Raizada N, Sreenivas A, van’t Hoog AH, van den Hof S, Dewan PK, 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. pmid:25996389
- 5.
Central TB Division. TB India 2017: Revised National Tuberculosis Control Programme: annual status report [Internet]. New Delhi, India: Directorate General of Health Services, Ministry of Health and Family Welfare; 2017. Available from: https://tbcindia.gov.in/index1.php?lang=1&level=2&sublinkid=4728&lid=3275
- 6. Raizada N, Sachdeva KS, Sreenivas A, Vadera B, Gupta RS, Parmar M, et al. Feasibility of decentralised deployment of Xpert MTB/RIF test at lower level of health system in India. PLoS One. 2014;9: e89301. pmid:24586675
- 7.
World Health Organization. High-priority target product profiles for new tuberculosis diagnostics: report of a consensus meeting [Internet]. Geneva, Switzerland: World Health Organization; 2014 Apr. Available from: http://www.who.int/tb/publications/tpp_report/en/
- 8. Denkinger CM, Kik SV, Cirillo DM, Casenghi M, Shinnick T, Weyer K, et al. Defining the needs for next generation assays for tuberculosis. J Infect Dis. 2015;211: S29–S38. pmid:25765104
- 9. Nikam C, Kazi M, Nair C, Jaggannath M, M M, R V, et al. Evaluation of the Indian TrueNAT micro RT-PCR device with GeneXpert for case detection of pulmonary tuberculosis. Int J Mycobacteriology. 2014;3: 205–210. pmid:26786489
- 10. Nikam C, Jagannath M, Narayanan MM, Ramanabhiraman V, Kazi M, Shetty A, et al. Rapid diagnosis of Mycobacterium tuberculosis with Truenat MTB: a near-care approach. PLoS One. 2013;8: e51121. pmid:23349670
- 11. Drain PK, Hyle EP, Noubary F, Freedberg KA, Wilson D, Bishai WR, et al. Diagnostic point-of-care tests in resource-limited settings. Lancet Infect Dis. 2014;14: 239–249. pmid:24332389
- 12. Walensky RP, Ross EL, Kumarasamy N, Wood R, Noubary F, Paltiel AD, et al. Cost-effectiveness of HIV treatment as prevention in serodiscordant couples. N Engl J Med. 2013;369: 1715–1725. pmid:24171517
- 13. Zheng A, Kumarasamy N, Huang M, Paltiel AD, Mayer KH, Rewari BB, et al. The cost-effectiveness and budgetary impact of a dolutegravir-based regimen as first-line treatment of HIV infection in India. J Int AIDS Soc. 2018;21: e25085. pmid:29603882
- 14. 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 model-based analysis. AIDS. 2012;26: 987–995. pmid:22333751
- 15.
Central TB Division. Revised National Tuberculosis Control Programme: training module for medical practitioners [Internet]. New Delhi, India: Directorate General of Health Services, Ministry of Health and Family Welfare; 2010. Available from: https://tbcindia.gov.in/showfile.php?lid=2908
- 16. Weinstein MC, Siegel JE, Gold MR, Kamlet MS, Russell LB. Recommendations of the panel on cost-effectiveness in health and medicine. JAMA. 1996;276: 1253–1258. pmid:8849754
- 17. Goyal V, Kadam V, Narang P, Singh V. Prevalence of drug-resistant pulmonary tuberculosis in India: systematic review and meta-analysis. BMC Public Health. 2017;17: 817. pmid:29041901
- 18. Davis JL, Cattamanchi A, Cuevas LE, Hopewell PC, Steingart KR. Diagnostic accuracy of same-day microscopy versus standard microscopy for pulmonary tuberculosis: a systematic review and meta-analysis. Lancet Infect Dis. 2013;13: 147–154. pmid:23099183
- 19. 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: 375–380. pmid:28284251
- 20. 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: e1001120. pmid:22087078
- 21. Khaparde S, Raizada N, Nair SA, Denkinger C, Sachdeva KS, Paramasivan CN, et al. Scaling-up the Xpert MTB/RIF assay for the detection of tuberculosis and rifampicin resistance in India: an economic analysis. PLoS One. 2017;12: e0184270. pmid:28880875
- 22. Steingart KR, Schiller I, Horne DJ, Pai M, Boehme CC, Dendukuri N. Xpert® MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev. 2014;1: CD009593. pmid:24448973
- 23.
World Health Organization. CHOosing Interventions that are Cost Effective (WHO-CHOICE): country-specific unit costs [Internet]. Geneva, Switzerland: World Health Organization; 2008. Available from: http://www.who.int/choice/country/country_specific/en/
- 24.
Global Drug Facility. Global Drug Facility: product catalogue, 2016 [Internet]. Vernier, Switzerland: Global Drug Facility, Stop TB Partnership; 2016. Available from: http://www.stoptb.org/assets/documents/gdf/drugsupply/GDF%20product%20catalog_25%20Jul%202016_final.pdf
- 25. Sun AY, Pai M, Salje H, Satyanarayana S, Deo S, Dowdy DW. Modeling the impact of alternative strategies for rapid molecular diagnosis of tuberculosis in Southeast Asia. Am J Epidemiol. 2013;178: 1740–1749. pmid:24100953
- 26. Walensky RP, Weinstein MC, Kimmel AD, Seage GR, Losina E, Sax PE, et al. Routine human immunodeficiency virus testing: an economic evaluation of current guidelines. Am J Med. 2005;118: 292–300. pmid:15745728
- 27. Baggaley RF, Irvine MA, Leber W, Cambiano V, Figueroa J, McMullen H, et al. Cost-effectiveness of screening for HIV in primary care: a health economics modelling analysis. Lancet HIV. 2017;4: e465–e474. pmid:28768604
- 28. 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: e710–e719. pmid:28619229
- 29. Vassall A, Mangham‐Jefferies L, Gomez GB, Pitt C, Foster N. Incorporating demand and supply constraints into economic evaluations in low-income and middle-income countries. Health Econ. 2016;25: 95–115. pmid:26786617
- 30. Denkinger CM, Nicolau I, Ramsay A, Chedore P, Pai M. Are peripheral microscopy centres ready for next generation molecular tuberculosis diagnostics? Eur Respir J. 2013;42: 544–547. pmid:23904551
- 31. 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: 516–525. pmid:27418550
- 32. Theron G, Zijenah L, Chanda D, Clowes P, Rachow A, Lesosky 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: 424–435. pmid:24176144
- 33. Hanrahan CF, Clouse K, Bassett J, Mutunga L, Selibas K, Stevens W, et al. The patient impact of point-of-care vs. laboratory placement of Xpert® MTB/RIF. Int J Tuberc Lung Dis. 2015;19: 811–816. pmid:26056107
- 34. Hanrahan CF, Selibas K, Deery CB, Dansey H, Clouse K, Bassett J, 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. pmid:23762367
- 35. Clouse K, Page-Shipp L, Dansey H, Moatlhodi B, Scott L, Bassett J, et al. Implementation of Xpert MTB/RIF for routine point-of-care diagnosis of tuberculosis at the primary care level. S Afr Med J. 2012;102: 805–807. pmid:23034211
- 36. Lessells RJ, Cooke GS, McGrath N, Nicol MP, Newell ML, Godfrey-Faussett P. Impact of point-of-care Xpert MTB/RIF on tuberculosis treatment initiation: a cluster-randomized trial. Am J Respir Crit Care Med. 2017;196: 901–910. pmid:28727491
- 37. Ardizzoni E, Fajardo E, Saranchuk P, Casenghi M, Page AL, Varaine F, et al. Implementing the Xpert® MTB/RIF diagnostic test for tuberculosis and rifampicin resistance: outcomes and lessons learned in 18 countries. PLoS One. 2015;10: e0144656. pmid:26670929
- 38. García-Basteiro AL, DiNardo A, Saavedra B, Silva DR, Palmero D, Gegia M, et al. Point of care diagnostics for tuberculosis. Pulmonology. 2018;24: 73–85. pmid:29426581
- 39. Hsiang E, Little KM, Haguma P, Hanrahan CF, Katamba A, Cattamanchi A, et al. Higher cost of implementing Xpert® MTB/RIF in Ugandan peripheral settings: implications for cost-effectiveness. Int J Tuberc Lung Dis. 2016;20: 1212–1218. pmid:27510248
- 40. Schnippel K, Meyer‐Rath G, Long L, MacLeod W, Sanne I, Stevens WS, 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: 1142–1151. pmid:22686606
- 41. Abdurrahman ST, Emenyonu N, Obasanya OJ, Lawson L, Dacombe R, Muhammad M, et al. The hidden costs of installing Xpert machines in a tuberculosis high-burden country: experiences from Nigeria. Pan Afr Med J. 2014;18: 1–5. pmid:25489371
- 42. Kendall EA, Schumacher SG, Denkinger CM, Dowdy DW. Estimated clinical impact of the Xpert MTB/RIF Ultra cartridge for diagnosis of pulmonary tuberculosis: a modeling study. PLoS Med. 2017;14: e1002472. pmid:29240766
- 43. Drain PK, Garrett NJ. The arrival of a true point-of-care molecular assay-ready for global implementation? Lancet Glob Health. 2015;3: e663–664. pmid:26475005