Epidemiology and treatment outcomes of TB-HIV co-infection in a rural South African setting: an exploratory analysis of retreatment and contextual factors.

Authors

Luzuko Mkono, Ncomeka Sineke, Ntandazo Dlatu, Mojisola Clara Hosu, Teke Apalata, Lindiwe M Faye

Published in

Frontiers in public health. Volume 14. Pages 1792518. Epub Jun 26, 2026.

Abstract

Tuberculosis-HIV (TB-HIV) co-infection remains a major public health challenge in South Africa, particularly in rural settings where clinical, socioeconomic, and health-system factors may influence treatment outcomes. This study examined the epidemiological characteristics of TB-HIV co-infection and factors associated with tuberculosis treatment outcomes in a rural Eastern Cape cohort.
A retrospective cohort study was conducted using routinely collected programmatic data from 422 adult tuberculosis patients treated between 2018 and 2022. Baseline demographic and clinical characteristics were summarized descriptively. Associations between patient characteristics and treatment success were evaluated using univariable and multivariable logistic regression analyses. An exploratory Random Forest (RF) model was additionally applied to assess predictive performance and identify variables contributing to outcome classification. Model performance was evaluated using Precision-Recall curves and Average Precision (AP).
The prevalence of TB-HIV co-infection was 57.8%. Treatment success rates were similar between TB-HIV co-infected and TB-only patients, and HIV status was not independently associated with treatment success in adjusted analyses. Previous TB treatment was associated with lower treatment success in univariable analysis (OR = 0.48; 95% CI: 0.29-0.78), although this association did not remain statistically significant after adjustment. Mortality was 6.6% among TB-HIV co-infected patients and 10.7% among TB-only patients, with no statistically significant difference between groups. The RF model demonstrated higher predictive performance (AP = 0.907) than logistic regression (AP = 0.810). Feature importance analysis identified socioeconomic characteristics, including education, income source, and employment status, as important contributors to model performance.
TB-HIV co-infection was highly prevalent in this rural cohort; however, treatment success and mortality outcomes were comparable between TB-HIV co-infected and TB-only patients. While HIV status was not independently associated with treatment outcomes, the absence of detailed HIV-related clinical information limits the interpretation of the underlying mechanisms. The exploratory machine-learning analysis suggests that socioeconomic factors may contribute to outcome prediction, highlighting the importance of considering broader contextual influences alongside clinical characteristics when designing interventions to improve tuberculosis treatment outcomes.

PMID:
42433412
Bibliographic data and abstract were imported from PubMed on 11 Jul 2026.

Read full publication at:
Please sign in to see all details.

Stats

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.