Abstract DP300: Digital Health Interventions to Improve Medication Adherence After Stroke: A Systematic Review

Introduction: Medication adherence is a cornerstone of secondary stroke prevention, yet non-adherence remains common and contributes to recurrent events. Digital health technologies offer promising self-driven options, but their effectiveness in stroke survivors is not well established. We conducted a systematic review and pooled analysis to evaluate the existing interventions. Methods: We systematically reviewed MEDLINE, Embase, and Cochrane (inception–Feb 2025) using terms related to stroke, TIA, secondary prevention, digital health, and medication adherence. Eligible studies reporting quantitative adherence outcomes in stroke or TIA patients were included. Data on demographics, stroke subtype, intervention characteristics, study design, and adherence outcomes were extracted. Weighted demographics were calculated where possible. Risk of bias was assessed using RoB-2 for randomized clinical trials (RCTs) and ROBINS-I for non-randomized studies. Results: Six studies with total 943 subjects reported on adherence (intervention n=516, mean age 55.4 years, 25% female, 98% stroke; control n=427, mean age 55.6 years, 28% female, 99% stroke). Interventions included smartphone app–based reminders (n=5) and SMS/text messaging (n=1). Designs comprised 3 RCTs, 2 single-arm trials, and 1 observational study. Adherence was assessed with the Morisky Medication Adherence Scale, other custom questionnaires, and app-based entries. Despite heterogeneity, all studies demonstrated improved adherence in the intervention group compared with control or baseline. Reported effects ranged from 10%–123% relative improvement in adherence, with higher Morisky scores, questionnaire-based, or app entry compliance (Figure). Risk of bias was low–moderate in RCTs; among non-randomized studies, one was moderate, one serious, and one at critical risk of bias. Intervention content, adherence definitions, and follow-up duration varied, limiting meta-analysis for most outcomes. We therefore present descriptive pooled summaries of relative improvements. Conclusions: Digital health interventions to improve medication adherence after stroke are feasible, acceptable, and generally effective. However, heterogeneity in design and outcome measurement limits comparability. Larger RCTs with standardized adherence endpoints are needed to establish efficacy and guide the integration of digital adherence tools into secondary stroke prevention.