Mortality among patients with drug-resistant tuberculosis and HIV in Zambia, 2018–2022: an analysis of a programmatic cohort

Drug-resistant tuberculosis (DR-TB) and human immunodeficiency virus (HIV) co-infection remain a major public health challenge, particularly in sub-Saharan Africa. People with DR-TB and HIV co-infection experience poor treatment outcomes and higher mortality. However, evidence on the factors associated with mortality in this population in Zambia remains limited. This study aimed to determine the mortality proportion and identify factors associated with mortality among patients co-infected with DR-TB and HIV in Zambia. We conducted a retrospective analysis of a programmatic cohort using routinely collected data from patients co-infected with drug-resistant tuberculosis (DR-TB) and human immunodeficiency virus (HIV) who received care at health facilities across all ten (10) provinces of Zambia between 2018 and 2022 under the National Tuberculosis and Leprosy Programme (NTLP). Drug-resistant tuberculosis was defined as tuberculosis caused by Mycobacterium tuberculosis strains resistant to at least one anti-tuberculosis medicine. A total of 655 adult PLHIV with DR-TB and documented treatment outcomes were included in the analysis. Demographic and clinical parameters were collected. Modified Poisson regression analysis was performed to estimate crude (cRR) and adjusted (aRR) risk ratios with 95% confidence intervals (CIs) for factors associated with mortality. Owing to substantial missing data, the primary multivariable model was fitted as a complete-case analysis, and a sensitivity analysis excluding alcohol use was conducted in a larger sample. The median age was 37 years (IQR: 32–44), and 62.4% ( n = 408) were male. The overall mortality proportion was 11.0% ( n = 72; 95% CI: 8.7%–13.6%). In the complete-case multivariable analysis of 127 participants, brief treatment interruption (aRR 10.18, 95% CI: 2.18–47.46; p = 0.003), alcohol consumption (aRR 7.16, 95% CI: 1.29–39.70; p = 0.019), and female sex (aRR 4.63, 95% CI: 1.07–20.09; p = 0.040) were associated with higher mortality. However, these associations were inconsistent across analyses, and substantial missing data limit the reliability of the adjusted findings. Mortality among patients co-infected with DR-TB and HIV in Zambia remains substantial. Brief treatment interruption, alcohol use, and female sex were initially identified as associated with mortality in the complete-case analysis. However, these findings should be regarded as exploratory and require cautious interpretation due to analytical sample instability. The sensitivity analysis using a larger sample demonstrated that these initial associations were not consistent: female sex was no longer significant, while hospital admission and pallor emerged as associated factors. Further longitudinal studies are needed. Pending such evidence, strengthening adherence support, screening for alcohol use, and providing closer clinical monitoring of hospitalised patients and those presenting with pallor may help reduce mortality in this population. Not applicable (N/A).