Evaluating motivational interviewing and peer adherence support as strategies to improve oral PrEP

Consistent adherence is essential for the effectiveness of daily oral pre-exposure prophylaxis (PrEP) in preventing HIV. Motivational interviewing (MI) and peer adherence support (PAS) are adherence support strategies, providing individualized support and may enhance both adherence to oral PrEP and continued use over time. This study aimed to evaluate the impact of implementation-adapted MI and PAS approaches on oral PrEP adherence and identify predictors of referral for adherence support. This MI/PAS intervention was embedded in the CAPRISA 084 PrEP demonstration study (2016–2018) and operated as a sequential, stepped adherence support approach rather than participant-selected alternatives: poor adherence triggered referral for MI, while participants with sustained/persistent adherence challenges (< 90% adherence on pill count or low drug levels) were escalated to PAS, a multi-component programme including SMS reminders, telephone support, and peer-support meetings. MI was delivered pragmatically and tailored to real-world clinical contexts rather than as a standardized intervention with no formal fidelity monitoring conducted. Tenofovir diphosphate levels from dried blood spots were measured in a subset of participants. We used Cox proportional hazards models to identify baseline predictors of MI/PAS referral; a paired t-test to assess changes in drug levels following referral, and a multivariable mixed-effects logistic model to evaluate associations between MI/PAS referral and subsequent visit attendance. Of 895 participants, 764 initiated PrEP, of whom 573 (75.0%) were referred for MI/PAS at a median of 88 (IQR:35–272) days after PrEP initiation. Multivariable analysis showed a higher likelihood of MI/PAS referral among urban residents (aHR 1.80, 95% CI 1.50–2.16), while being in a relationship > 5 years was protective (aHR 0.71, 95% CI 0.55–0.92). Participants referred for MI/PAS had nearly twice the odds of missing their subsequent visit compared with those not referred (OR 1.96; 95% CI 1.56–2.47). In a subset of 45 (5.9%) participants with drug-level data, tenofovir diphosphate concentrations increased significantly after MI/PAS (mean difference 154.7 fmol/punch; 95% CI 69.4–240; t(44) = 3.66; p < 0.001). MI/ PAS were implemented as targeted adherence support strategies within a PrEP implementation programme. Referral was influenced by relationship stability and site of care, and although recipients had poorer visit attendance, drug-level data showed meaningful improvements in adherence, supporting integration of MI/PAS-based support into PrEP delivery. Embedding MI/PAS alongside strengthened retention strategies, such as appointment reminders, community-based refills, and active follow-up of missed visits, may improve adherence, particularly for mobile urban populations.