Back

Risk and determinants of tuberculosis recurrence: a 12-year population-based cohort study

Lepka de Lima, E.; Salles, I.; Cohn, S.; Lindoso, A. A.; Fukasava, S.; Golub, J.; Cox, S. R.

2025-11-20 infectious diseases
10.1101/2025.11.20.25340642 medRxiv
Show abstract

ABSTRACTO_ST_ABSBackgroundC_ST_ABSTuberculosis recurrence remains a challenge to elimination efforts, particularly in high-burden settings. A better understanding of incidence, timing, and risk factors is needed to inform policies and programs for post-treatment care. MethodsWe conducted a retrospective cohort study of tuberculosis survivors successfully treated for a first tuberculosis episode between 2013 and 2024 in Sao Paulo state, Brazil. Data were obtained from the state tuberculosis surveillance system (TBweb), and deaths were identified through linkage with the national mortality database. We applied competing risks survival analysis to estimate incidence and assess risk factors for recurrence. Covariates included demographic, clinical, behavioral, and social vulnerability indicators, all measured at treatment initiation for the first tuberculosis episode. Risk factor models were stratified by age (<15 vs. [&ge;]15 years). FindingsAmong 154,579 individuals who completed treatment for a first episode of tuberculosis, we identified 9,464 first recurrences over a median follow-up of 5.4 years (IQR 2.5-8.2). The overall recurrence incidence was 1,127.2 per 100,000 person-years. Rates were highest between 3-12 months after treatment completion, yet recurrence incidence remained over 1,000 per 100,000 person-years in the 2-5 years after treatment completion. In adults, hospitalization (subdistribution hazard ratio [SHR] 2.53, 95%CI 2.36-2.71), and incarceration during the initial episode (2.30, 2.13-2.48), together with pulmonary tuberculosis as initial diagnosis (2.10, 1.87-2.35) were the strongest independent risk factors for recurrence. In children, HIV (6.38, 2.64-15.40) was the strongest predictor. InterpretationThe high and persistent risk of tuberculosis recurrence reinforces the importance of post-treatment care with risk-based follow-up strategies.

Matching journals

The top 10 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.