Factors associated with self-reported tuberculosis medication adherence in Jinja Regional Referral Hospital

Abstract

Tuberculosis (TB) remains a major public health challenge, with non-adherence to TB medication contributing to treatment failure, continued disease transmission, drug resistance, and mortality. This study examined factors associated with self-reported TB medication adherence among patients attending Jinja Regional Referral Hospital. A quantitative cross-sectional study was conducted among 118 eligible TB patients drawn from 172 patients registered for treatment at Jinja Regional Referral Hospital. Data were collected using a structured interviewer-administered questionnaire. Descriptive statistics were used to summarize the study variables, while binary logistic regression was used to identify independent predictors of self-reported TB medication adherence. Adjusted Odds Ratios (AORs) with 95% confidence intervals were estimated, with statistical significance assessed at the 5% level. Overall, 70 (59.3%) respondents were adherent to TB medication, while 48 (40.7%) were non-adherent. The binary logistic regression model was statistically significant overall (LR χ²(18) = 99.52, p < 0.001), with a Pseudo R² of 0.6806. After adjustment for other factors, challenges in accessing the health facility remained the only statistically significant independent predictor of self-reported adherence (AOR = 30.24; 95% CI: 2.83-322.84; p = 0.005). Respondents who did not experience challenges in accessing the health facility had approximately 30 times higher odds of being adherent to TB medication compared to those who experienced challenges. The study concludes that challenges in accessing health facilities are an important barrier to self-reported TB medication adherence among patients attending Jinja Regional Referral Hospital. The study recommends strengthening decentralized TB treatment services at lower-level health facilities to improve access to treatment. Further studies involving larger and more diverse populations are recommended to obtain more precise estimates and improve comparability across settings.