Using machine learning to predict poor adherence to antiretroviral therapy among adolescents living with HIV in low resource settings.

Abstract

Objectives: Achieving optimal adherence to antiretroviral therapy (ART) and viral suppression is still insufficient for attaining the UNAIDS 95-95-95 target of 2030, especially among adolescents living with HIV (ALWHIV). This study sought to develop a model to predict poor adherence risk among ALWHIV and identify associated risk factors.

Design: We utilized machine learning (ML) to predict future ART adherence among ALWHIV leveraging its ability to analyze complex, multidimensional data.

Methods: We leveraged a dataset from a 6-year (2012-2018) longitudinal randomized control trial (RCT) with 635 ALWHIV in Uganda. We evaluated six ML models and retained one with the highest Area Under Receiver Operating Characteristic (AUROC), and Area Under Precision-Recall Curve (AUPRC). We further identified principal factors associated with ART adherence based on the best model.

Results: The random forest model outperformed others, with mean AUROC: 0·71 (BC 95% CI: [0·69-0·72]) and AUPRC: 0·55 (BC 95% CI: [0·53-0·58]). The principal risk factors of poor adherence were poor adherence history; poverty; biological relationship to caregiver; self-concept; savings confidence; duration on ART; frequency discussing sensitive topics with caregivers; household size; economic group assignment; and school enrollment.

Conclusions: Our findings support potential use of ML methods and socio-behavioral data for predicting poor ART adherence risk among ALWHIV. The predictive tool can help identify ALWHIV at the highest risk of treatment failure, and enable early targeted interventions. However, the tool is still preliminary and its accuracy could be improved by incorporating HIV phenotypic and clinical data.

Clinical trial number: ClinicalTrials.gov ID:NCT01790373.