Drug-Resistant Tuberculosis Outcomes in South Kivu Province, Democratic Republic of Congo, During a Humanitarian Crisis (2018-2024).

BACKGROUND

Drug-resistant tuberculosis (DR-TB) management in fragile, conflict-affected settings remains challenging. South Kivu Province has faced protracted insecurity and recent outbreaks (COVID-19, mpox). We assessed treatment outcomes and predictors among DR-TB patients managed under programmatic conditions.

METHODS

We conducted a retrospective provincial cohort study of all patients registered in electronic Tuberculosis registry with DR-TB in South Kivu from January 2018 to June 2024. Primary outcome was treatment success; unsuccessful outcomes. We summarized characteristics, performed Kaplan-Meier functions for time to outcome, and fitted multivariable logistic regression and Cox proportional hazards models to identify predictors of unsuccessful outcomes.

RESULTS

Among 277 DR-TB patients (median age 37 years; 3-75), overall treatment success was 77.6%, mortality 13.4%, and loss to follow-up (LTFU) 8.6%. HIV status was negative in 79.1% and positive in 9.4%. Bedaquiline-containing regimens were used in 100%. In adjusted analyses, rural residence was protective against unsuccessful outcome (aOR 0.42; 95% CI 0.21-0.82; P = .011), whereas retreatment other than relapse was associated with higher odds of unsuccessful outcome (aOR 2.04; 95% CI 1.09-3.81; P = .026). Findings were consistent in time-to-event models (rural aHR 0.50; 95% CI 0.29-0.85; P = .011; retreatment aHR 1.71; 95% CI 1.03-2.83; P = .038).

CONCLUSIONS

In a humanitarian crisis context, programmatic all-oral DR-TB care achieved success exceeding the WHO ≥ 75% target. Persistently high mortality and LTFU-concentrated among urban residents and retreatment (non-relapse) cases-highlight the need for targeted urban patient-tracing, strengthened referral/continuity mechanisms, and uninterrupted drug supply. Results support scale-up of short, all-oral regimens with context-specific adherence support in fragile settings.