A systematic review and meta-analysis pooled data from 52 studies involving adults with various chronic diseases to assess the effectiveness of digital self-management technologies—including mobile health apps, SMS reminders, and web-based platforms—on medication adherence. The comparator was standard care or other control groups, though specifics were not reported. The pooled random-effects analysis showed a statistically significant, small-to-moderate benefit, with an effect size of Cohen's d = 0.268 (95% CI 0.123-0.414, p = 0.0003). Effects were largest for medium-duration interventions (d = 0.50) and varied markedly by country, ranging from d = 2.29 in Iran to d = -0.94 in Taiwan. A trim-and-fill adjustment for potential publication bias increased the pooled effect to d = 0.366.
Safety and tolerability data were not reported across the included studies. The analysis has important limitations, including high statistical heterogeneity (I² = 89%), which suggests substantial variation in true effects across studies, and possible publication bias as indicated by Begg's test. The evidence base included RCTs, quasi-experimental, and controlled before-after studies, establishing an association but not definitive causality.
For practice, this meta-analysis indicates that digital self-management tools can be associated with improved medication adherence in chronic disease populations, but the effect is modest and highly context-dependent. The significant variation by geography and intervention characteristics underscores that these technologies are not uniformly effective. Clinicians should consider local evidence and patient preferences when recommending specific digital tools, recognizing that the overall evidence strength is tempered by heterogeneity and potential bias.
View Original Abstract ↓
BACKGROUND: Chronic diseases require sustained medication adherence, yet nonadherence remains common, leading to poor outcomes and increased healthcare costs. Digital self-management technologies such as mobile health (mHealth) apps, SMS reminders, and web-based platforms offer scalable ways to support adherence, but evidence on their overall effectiveness across diverse contexts is fragmented.
AIM: To systematically review and meta-analyze the effectiveness of self-management technologies in improving medication adherence among adults with chronic diseases and to examine potential moderators of intervention impact.
METHODS: Following PRISMA guidelines, we searched PubMed, Scopus, Web of Science, CINAHL, and JMIR for peer-reviewed studies (January 2010-June 2025) evaluating digital self-management interventions with adherence outcomes and comparator groups. Eligible designs included RCTs, quasi-experimental, and controlled before-after studies in adults with chronic disease. Random-effects meta-analysis estimated pooled effect sizes (Cohen's d). Heterogeneity (I), subgroup analyses, and publication bias (Egger's, Begg's, trim-and-fill) were assessed.
RESULTS: Fifty-two studies were included, spanning 2015-2025. Early interventions (2015-2019) focused on feasibility, using SMS and basic web tools; later years (2021-2025) showed technological maturity, dominated by mHealth apps integrating monitoring, reminders, and education. The pooled random-effects effect size was d = 0.268 (95% CI 0.123-0.414, p = 0.0003), indicating a small-to-moderate benefit. Heterogeneity was high (I = 89%). Medium-duration (10.8-24 weeks) interventions had the largest effect (d = 0.50), and effects varied markedly by country (e.g., Iran d = 2.29; Taiwan d = -0.94). Begg's test suggested possible publication bias; trim-and-fill adjustment increased the pooled effect to d = 0.366.
LINKING EVIDENCE TO ACTION: Digital self-management technologies yield a statistically significant, small-to-moderate improvement in medication adherence across chronic diseases, with potential underestimation due to selective reporting. Effectiveness is moderated by temporal trends, geography, intervention duration, and study design, underscoring the need for context-specific adaptation and methodological rigor. Future research should prioritize large, well-controlled trials, pre-registration, and exploration of cultural and systemic determinants to optimize intervention impact.