A Digital Adherence Tools randomised trial among children living with HIV: lessons learnt

Abstract Challenges that arise during the implementation of (clinical) trial activities in African settings can consume substantial resources for participants, investigators, and the funding agency. The literature on these aspects of the challenge is relatively scarce, although it could help prevent similar errors in future studies. We conducted a trial (registered as PACTR202301844164954) to test the effectiveness of digital adherence tools (DAT) on medicine adherence and viral suppression among children living with HIV (CLHIV). We encountered critical operational challenges, which we present as a case study to illustrate their consequences for primary outcomes and for predicting viral suppression. Then, we evaluate and extract lessons learnt to inform future DAT studies. Our case was a pragmatic two-arm trial investigating the effectiveness of DAT in Tanzania, in which we followed participants for 12-months. We measured self-reported adherence, pharmacy refill adherence, and viral suppression. We performed a chi-square test and a quasi-Poisson regression analysis to analyze the significance of differences between the two arms. We also determined the ability of our DAT data to predict viral suppression. We enrolled 154 children; 151 children had at least one follow-up visit during which refill information was collected from source documents. While 145 children had contact with the research team, during which self-reported adherence and viral load data were collected. Operational challenges related to the research team, infrastructure, data collection and management, and quality oversight led to inefficiency and inconsistency in implementing study procedures and to poor-quality adherence data. As a result, no conclusions could be drawn regarding the effect of DAT on adherence. However, there was a weak effect of on improving viral suppression among CLHIV. Adherence, as measured by DAT, was a strong predictor of viral suppression (AUC = 0.84 on the ROC curve). For the successful implementation of study procedures in pragmatic trials, it is essential to involve healthcare facility leadership in trial management from the planning stage, thereby facilitating the smooth integration of research activities into routine practice. Moreover, it is crucial to select measurable, objectively quantifiable endpoints sensitive enough to detect the intervention's effect, regardless of logistical constraints.