Minimum requirements and optimal testing strategies of a diagnostic test for leprosy as a tool towards zero transmission: A modeling study

BACKGROUND: The availability of a diagnostic test to detect subclinical leprosy cases is crucial to interrupt the transmission of M. leprae. In this study we assessed the minimum sensitivity level of such a (hypothetical) diagnostic test and the optimal testing strategy in order to effectively reduce the new case detection rate (NCDR) of leprosy. METHODS AND FINDINGS: We used the individual-based model SIMCOLEP, and based it on previous quantification using COLEP data, a cohort study of leprosy cases in Bangladesh. The baseline consisted of treatment with Multidrug therapy of clinically diagnosed leprosy cases, passive case detection and household contact tracing. We examined the use of a leprosy diagnostic test for subclinical leprosy in four strategies: testing in 1) household contacts, 2) household contacts with a 3-year follow-up, 3) a population survey with coverage 50%, and 4) a population survey (100%). For each strategy, we varied the test sensitivity between 50% and 100%. All analyses were conducted for a high, medium, and low (i.e. 25, 5 and 1 per 100,000) endemic setting over a period of 50 years. In all strategies, the use of a diagnostic test further reduces the NCDR of leprosy compared to the no test strategy. A substantial reduction could already be achieved at a test sensitivity as low as 50%. In a high endemic setting, a NCDR of 10 per 100,000 could be reached within 8–10 years in household contact testing, and 2–6 years in a population testing. Testing in a population survey could also yield the highest number of prevented new cases, but requires a large number needed to test and treat. In contrast, household contact testing has a smaller impact on the NCDR but requires a substantially lower number needed to test and treat. CONCLUSIONS: A diagnostic test for subclinical leprosy with a sensitivity of at least 50% could substantially reduce M. leprae transmission. To effectively reduce NCDR in the short run, a population survey is preferred over household contact tracing. However, this is only favorable in high endemic settings.