OBJECTIVES: To identify and synthesize evidence on impact of non-pharmacological interventions on medication adherence (MA) among patients with hypertension or dyslipidemia.

METHODS: A systematic review was conducted in EMBASE, PubMed-Medline and congress proceedings. Eligible studies were randomized clinical trials (RCTs) investigating ≥50 adult patients, published July 2011–July 2021, evaluating the effect of health education (HE), digital applications (DAPP), phone reminders, blood pressure monitors, singly or in combination, on MA. Interventions were categorized according to pre-defined definitions.

RESULTS: Sixty-four studies (54 RCTs, 10 cluster-RCTs) were eligible. Studies evaluated two (56 studies), three (6) or four (2) arms; 60 focused on hypertension and 4 on dyslipidemia. Statistical significance of change in MA in the intervention vs the comparator at the end of follow-up was assessed by 56/64 studies. Of these, 28 (50%) presented significant improvements in MA favoring the intervention versus comparator, six presented conflicting outcomes depending on the tool used to measure adherence, and 22 reported non-significant differences in MA between arms. HE, either as single intervention or in combination with other categories, was the most frequently investigated intervention category (60/64 studies). Among studies in which statistically significant improvements in MA were demonstrated, most included an intervention with an HE component (26/28), phone reminder component (10/28), or DAPP component (7/28). A broad range of methods were used to measure MA: Morisky medication adherence scale the most used tool (11/28 studies). Length of follow-up varied, most often between 1–6 months (14 studies), 7–12 months (11), and >12 months (3). Heterogeneity and incomplete reporting of patient populations, outcome measures, and interventions were key limitations in the evidence.

CONCLUSIONS: Non-pharmacological interventions (HE, digital applications, phone reminders, BP-monitors, or combinations) may improve adherence to medications in patients with hypertension or dyslipidemia. HE was the most frequently reported category that significantly improved MA.