Supporting the Art: Medication Adherence Patterns in Persons Prescribed Ingestible Sensor-enabled Oral Pre-Exposure Prophylaxis to Prevent Human Immunodeficiency Virus (HIV) Infection

Timely, accurate adherence data may support oral pre-exposure prophylaxis (PrEP) success and inform prophylaxis choice. This study evaluated ingestible-sensor-enabled (IS) Truvada (R) use in persons starting oral PrEP. Individual behavior patterns were captured in real-time, distinguishing adherent persons from those at risk of prophylactic failure. Background Timely, accurate adherence data may support oral pre-exposure prophylaxis (PrEP) success and inform prophylaxis choice. We evaluated a Food and Drug Administration (FDA)-approved digital health feedback system (DHFS) with ingestible-sensor-enabled (IS) tenofovir-disoproxil-fumarate plus emtricitabine (Truvada (R)) in persons starting oral PrEP. Methods Human immunodeficiency virus (HIV)-negative adults were prescribed IS-Truvada (R) with DHFS for 12 weeks to observe medication taking behavior. Baseline demographics, urine toxicology, and self-report questionnaires were obtained. Positive detection accuracy and adverse events were computed as percentages, with Kaplan Meier Estimate for persistence-of-use. In participants persisting >= 28 days, adherence patterns (taking and timing) were analyzed, and mixed-effects logistic regression modeled characteristics associated with treatment adherence. Results Seventy-one participants were enrolled, mean age 37.6 years (range 18-69), 90.1% male, 77.5% White, 33.8% Hispanic, 95.8% housed, and 74.6% employed. Sixty-three participants (88.7%) persisted >= 28 days, generating 4987 observation days, average 79.2 (29-105). Total confirmed doses were 86.2% (95% confidence interval [CI] 82.5, 89.4), decreasing over time, odds ratio (OR) 0.899 (95% CI .876, .923) per week, P < .001; 79.4% (95% CI 66.7%, 87.3%) of participants had >= 80% adherence. Pattern analysis showed days without confirmed doses clustered (P = .003); regular dose timing was higher among participants with >= 80% confirmed doses (0.828, 95% CI .796 to .859) than among those with <80% (0.542, 95% CI95 .405 to .679) P < .001. In multi-predictor models, better adherence was associated with older age, OR 1.060 (95% CI 1.033, 1.091) per year, P < .001; negative vs positive methamphetamine screen, OR 5.051 (95% CI 2.252, 11.494), P < .001. Conclusions DHFS with IS-Truvada (R) distinguished adherent persons from those potentially at risk of prophylactic failure. Ongoing methamphetamine substance use may impact oral PrEP success.