Abstract Body

Novel approaches to identify incomplete adherence are necessary to realise the full benefits of HIV treatment. We compared real-time electronic methods with patient-reported and objective adherence measures in an ART-naïve cohort in South Africa.

We recruited ART-naïve participants from a community ART clinic. We collected demographic and disease data, CD4 count and HIV-RNA at weeks 0, 16 and 48. HIV-RNA >500 copies/ml triggered a genotype. We quantified adherence using self-report (SR), tablet return (TR), average adherence by pharmacy refill (PR-average), calculation of medication-free days (PR-gaps), therapeutic drug monitoring (TDM) and an electronic adherence monitoring device (EAMD). We modelled associations between adherence measures and virologic and genotypic outcomes using logistic regression, and constructed receiver operator curves (ROC) to assess performance of adherence measures in predicting outcomes.

The 230 participants had median (IQR) adherence: SR 100% (100-100), TR 100% (95-107), PR-average 103% (95-105), PR-gaps 100% (95-100) and EAMD 86% (59-94) at week 48. Efavirenz concentrations were therapeutic (>1mg/ml) in 92%. Retention in care was 81% (186/230), with 83% (155/186) achieving HIV-RNA <40 copies/ml. EAMD, PR-average and PR-gaps best predicted virological outcome at week 48 with area under the ROC (AUC ROC) of 0.73 (095%CI 0.61-0.83), 0.73 (95%CI 0.61-0.85) and 0.72 (95%CI 0.59-0.84) respectively. EAMD, PR-gaps and PR-average were all highly predictive of resistance at week 48, with AUC ROC of 0.92 (95%CI 0.87-0.97), 0.86 (0.67-1.0) and 0.83 (95%CI 0.65-1.0) respective. SR, TR and EFV concentrations were poorly predictive of virological or resistance outcomes.

Adherence data from EAMD and pharmacy refill measures predicted resistance and virological failure similarly. Pharmacy refill data is a reasonable option for monitoring adherence in resource-limited settings where electronic monitoring is unavailable.