Community-engaged clinical governance and machine learning for optimizing tuberculosis management in rural Eastern Cape.

Tuberculosis (TB) remains a major global health challenge, particularly in high-burden, resource-limited settings. Community-Engaged Clinical Governance (CE-CG) has emerged as a promising framework for strengthening accountability, adherence, and continuity of care by integrating clinical governance and community participation. This study examined the alignment between CE-CG implementation and TB treatment outcomes in the rural Eastern Cape, South Africa, using patient data from 2018 to 2020. Descriptive statistics, correlation analysis, and explanatory machine-learning models (logistic regression, random forest, and decision tree) were applied to address distinct research objectives, along with scenario-based projections. CE-CG was retrospectively operationalized as a binary programmatic indicator reflecting periods of structured governance implementation, including community health worker tracing, digital adherence monitoring, integrated TB-HIV care, and governance dashboard oversight. Machine-learning models were intentionally used as explanatory tools rather than predictive models to assess the internal coherence of the CE-CG framework. The observed perfect classification performance reflects deterministic alignment between governance implementation and treatment outcomes within this cohort rather than generalizable predictive accuracy. Treatment success improved substantially over the study period, increasing from 41.6% in 2018 to 68.3% in 2020. Scenario-based projections indicate that under a slow intervention trajectory (3.5% annual growth), treatment success would reach only 76.6% by 2030. In contrast, a sustained governance strategy (5.34% annual growth) could achieve the World Health Organization (WHO) target of 95%. Correlation analysis revealed a perfect positive association between CE-CG and treatment success, which was interpreted as an artifact of retrospective coding rather than a causal effect. Loss to follow-up and multidrug-resistant TB demonstrated weaker associations with outcomes, while extensively drug-resistant TB remained negatively associated. Overall, the findings support CE-CG as a policy-relevant, programmatic framework for strengthening adherence, retention, and accountability in high-burden rural TB settings. Embedding CE-CG within TB programmes offers a sustainable pathway toward achieving the WHO treatment success targets and accelerating progress toward TB elimination.