Cross-sectional study quantifies ART adherence in people living with HIV, suggesting targeted interventions are needed.
Key Points
This research aims to quantify adherence to antiretroviral therapy (ART) among people living with HIV at Pantang Hospital.
Cross-sectional design assessing ART adherence among 151 participants at Pantang Hospital, Ghana.
Data collected through interviewer-administered questionnaires on sociodemographic characteristics and adherence behaviors.
Six machine learning models were applied to identify predictors of non-adherence and compare model performance.
Poor adherence to ART was found, with 66.9% forgetting to take medication and 64.2% exhibiting carelessness.
Elastic Net regression was the top-performing model for predicting adherence.
Key predictors of non-adherence included not taking medication over the past weekend (utility estimate 0.3471, 95% CI 0.2292–0.4650) and missing medication 6–10 times in the past week (utility estimate 0.2693, 95% CI 0.1538–0.3848).