Predicting treatment adherence in patients with drug-resistant tuberculosis: insights from socioeconomic, demographic, and clinical factors of patients in the rural Eastern Cape

Background Drug-resistant tuberculosis (DR-TB) poses a serious challenge to global health. Patients must follow complex medication regimens over long periods, and any failure to comply with these treatment plans can result in treatment failure, higher mortality rates, and an increased risk of developing additional drug resistance. Setting The study was conducted in the rural Eastern Cape. Aim This study aims to identify the key factors influencing treatment adherence among patients with DR-TB. Furthermore, it rigorously evaluates the predictive accuracy of machine learning models in assessing treatment adherence, with a strong focus on socioeconomic, demographic, and clinical factors. Methods A retrospective analysis was conducted on patients with DR-TB. Data were collected from medical records. Four different models were developed and tested to evaluate their effectiveness in predicting treatment adherence: Random Forest, Logistic regression, Support Vector Machine (SVM), and Gradient Boosting. Results The Random Forest model achieved an accuracy of 53.3% in predicting treatment adherence. An analysis of feature importance indicated that age, income, education, social history, patient category, and comorbidities were the most significant factors influencing adherence. Patients with higher incomes, higher levels of education, and fewer comorbidities were more likely to follow their treatment plans. Conclusion Adhering to treatment for DR-TB involves a range of socioeconomic and clinical factors. Income, education level, and pre-existing health conditions significantly influence how well patients follow their prescribed treatment regimens. Understanding these influences is crucial for enhancing treatment outcomes and facilitating patients' journey toward improved health. Contribution These findings suggest that machine-learning models, especially Random Forest algorithms, can effectively support clinical decision-making by identifying patients at risk of non-adherence to their treatment.