The Medication Possession Ratio as a Predictor of Longitudinal HIV-1 Viral Suppression

Background: Antiretroviral adherence is essential to achieve viral suppression and limit HIV-related morbidity and mortality; however, antiretroviral adherence thresholds to achieve viral suppression in clinical practice have not been fully characterized using administrative claims data. Objective: The purpose of this study was to assess the relationship between medication adherence and viral suppression among adult persons with HIV/AIDS (PWH) receiving antiretroviral therapy (ART) for >= 6 months. Methods: This historical cohort, real-world investigation assessed maintenance of viral load suppression and viral load area-under-the-curve (vAUC) in PWH >= 18 years of age based on ART adherence. A marginal effects model was used to determine the predicted probabilities of final plasma HIV-1 RNA <50 copies/mL or vAUC <1,000 copy-days/mL according to the medication possession ratio (MPR), estimated using a Jackknife model variance estimator and a delta-method for marginal effects standard error. Tests for statistical significance used a Sidak method to correct for multiple comparisons. Results: The mean MPR for ART was 86.7% (95% CI: 85.0%-88.4%) for the 372 PWH included in the study. The marginal effects analysis indicated that an MPR >= 82% was associated with a predicted probability of viral suppression <50 copies/mL (P < 0.05). Significant predicted probabilities for vAUC <1,000 copy-days/mL were observed with an MPR >= 90% (P < 0.05). Conclusion and Relevance: Medication possession ratio as a proxy for drug exposure was significantly and consistently associated with viral suppression using a longitudinal measure of HIV viremia. These findings can aid clinicians in the clinical management of PWH and inform future studies of adherence-viral suppression relationships with contemporary antiretroviral regimens.