The Impact of Preventive Treatment for Multidrug- and Rifampin-Resistant Tuberculosis Exceeds Trial-Based Estimates

Abstract

Abstract Background Several clinical trials of tuberculosis preventive treatment (TPT) for household contacts of patients with multidrug- or rifampin-resistant tuberculosis (MDR/RR-TB) are nearing completion. The potential benefits of delivering TPT to MDR/RR-TB contacts extend beyond the outcomes that clinical trials can measure. Methods We developed an agent-based, household-structured TB and MDR/RR-TB transmission model, calibrated to an illustrative setting in India. We simulated contact investigation in households of patients with MDR/RR-TB, comparing an MDR/RR-TPT regimen (assuming 6-month duration, 70% efficacy) and associated active case finding against alternatives of contact investigation without TPT or no household intervention. We simulated the TB and MDR/RR-TB incidence averted relative to placebo over 2 years, as measurable by a typical trial, as well as the incidence averted over a longer time horizon, in the broader population, and relative to no contact investigation. Results Observing TPT and placebo recipients for 2 years as in a typical trial, MDR/RR-TPT was measured to prevent 72% (interquartile range, 45%–100%) of incident MDR/RR-TB among recipients; the median number needed to treat (NNT) to prevent 1 MDR/RR-TB case was 73, compared to placebo. This NNT decreased to 54 with 13–18 years of observation, to 27 when downstream transmission effects were also considered, and to 12 when the effects of active TB screening were included by comparing to a no-household-contact-intervention scenario. Conclusions If forthcoming trial results demonstrate efficacy, the long-term population impact of TPT for MDR/RR-TB—including the large effect of increased active TB detection among MDR/RR-TB contacts—could be much greater than suggested by trial outcomes alone.