Comparison of six methods to estimate adherence in an ART-naïve cohort in a resource-poor setting: which best predicts virological and resistance outcomes?

摘要: Incomplete adherence to antiretroviral therapy (ART) results in virologic failure and resistance. It remains unclear which measure best predicts these outcomes. We compared six patient-reported objective measures one ART-naive cohort South Africa. recruited 230 participants from a community ART clinic prospectively collected demographic data, CD4 count HIV-RNA at weeks 0, 16 48. quantified using 3-day self-report (SR), clinic-based pill (CPC), average by pharmacy refill (PR-average), calculation of medication-free days (PR-gaps), efavirenz therapeutic drug monitoring (TDM) an electronic device (EAMD). Associations between genotypic outcomes were modelled logistic regression, with the area under curve (AUC) receiver operator characteristic (ROC) analyses derived assess performance predicting At week 48 median (IQR) was: SR 100% (100–100), CPC (95–107), PR-average 103% (95–105), PR-gaps (95–100) EAMD 86% (59–94), concentrations (>1 mg/L) 92%. EAMD, PR-average, predicted virological outcome AUC ROC 0.73 (95% CI 0.61–0.83), 0.61–0.85), 0.72 0.59–0.84) 0.64 0.52–0.76) respectively. highly predictive detection resistance mutations 48, 0.92 0.87–0.97), 0.86 (0.67–1.0) 0.83 0.65–1.0) TDM poorly both PR similarly. Pharmacy data is pragmatic resource-limited settings where unavailable. Trial registration The trial was retrospectively registered Pan African Clinical Trials Registry, number PACTR201311000641402, on 13 Sep 2013 ( www.pactr.org ). first participant enrolled 12th July 2012. last patient visit (week 48) 15 April 2014