Treatment outcomes and associated factors among patients with drug-resistant tuberculosis in Ethiopia: A retrospective cohort study.

Drug-resistant tuberculosis (DR-TB) remains a major public health challenge globally. Despite diagnostic advances and expanded access to all-oral regimens in Ethiopia, treatment outcomes continue to vary across the country. Evidence on factors affecting treatment outcomes under programmatic setups is needed to inform targeted interventions. This study aimed to assess the treatment outcomes and associated factors of patients with DR-TB at Adama Hospital Medical College from January 2016 to December 2024. We conducted a retrospective cohort study among patients treated for DR-TB. Treatment outcomes were classified according to the World Health Organization definitions as successful (cured/completed) or unsuccessful (treatment failure/death/loss to follow-up). Sociodemographic, behavioral, and clinical characteristics were extracted from routine records using an online form administered via Kobo Toolbox. The data set was later exported in CSV format and analyzed using R statistical software (version 4.3.2). Descriptive statistics were used to summarize the characteristics of the study population. Bivariate logistic regression was initially performed. Variables with a P < 0.25 in the bivariate analysis were included into a multivariable Firth penalized logistic regression model addressing sparse data and separation to identify independent predictors of treatment outcomes. Statistical significance was determined at P < 0.05, with a 95% confidence interval (CI). Sensitivity analysis used a model controlling for collapsed regimen categories. Among 249 eligible patients with MDR/RR-TB, 197 (79.1%) achieved successful treatment outcomes, whereas 27 (10.8%) died and 22 (8.9%) were lost to follow-up and three (1.2%) were treatment failures. Of the 249 patients included in the cohort, 238 were included in the multivariable analysis due to complete data availability. In the complete-case analysis using multivariable Firth model (n = 238), age ≥45 years was independently associated with an unsuccessful outcome (adjusted odds ratio [AOR] 3.43, 95% CI 1.23-9.89). Patients who were underweight at enrollment (AOR 2.47, 95% CI 1.19-5.36) had increased odds of unsuccessful treatment outcome. Previous defaulters (loss to follow-up) had substantially higher odds of unsuccessful outcome than new patients (AOR 7.13, 95% CI 1.02-84.26), although CIs were wide due to sparse data. Human immunodeficiency virus (HIV) co-infection was associated with increased odds of unsuccessful treatment outcome in the primary analysis (AOR 3.20, 95% CI 1.04-10.21). In the sensitivity analysis, the association remained in the same direction but with borderline significance (AOR 3.01, 95% CI 0.97-9.62). Despite high overall treatment success, older adults, undernourished, and previous defaulters remain at substantially increased risk of unsuccessful DR-TB treatment outcomes. HIV co-infection was also associated with unsuccessful treatment outcomes, with a directionally consistent association in sensitivity analysis. Targeted adherence support, strengthened TB/HIV integration, and improved programmatic monitoring are essential to further reduce mortality and loss to follow-up.