Predicting treatment adherence in patients with drug-resistant tuberculosis: insights from socioeconomic, demographic, and clinical factors of patients in the rural Eastern Cape
Retrospective analysis identifies socioeconomic and clinical factors influencing treatment adherence, suggesting machine learning models may improve patient outcomes.
Key Points
The Random Forest model achieved an accuracy of 53.3% in predicting treatment adherence, highlighting machine learning's potential role.
Patients with higher incomes, education levels, and fewer comorbidities exhibited better adherence to drug-resistant tuberculosis treatment regimens.
Analysis showed age, income, education, social history, patient category, and comorbidities were significant factors affecting adherence.
These findings underscore the importance of understanding socioeconomic and clinical factors for improving treatment outcomes in drug-resistant tuberculosis patients.