Aims The aim of this systematic review and meta‐analysis was to synthesise the evidence relating to medication non‐adherence and its association with health outcomes in people aged ≥50 years. Methods Seven databases were searched up to February 2019 for observational studies that measured medication (non‐)adherence as a predictor of the following health outcomes in adults aged ≥50 years: healthcare utilisation (hospitalisation, emergency department visits, outpatient visits and general practitioner visits), mortality, adverse clinical events and quality of life. Screening and quality assessment using validated criteria were completed by 2 reviewers independently. Random effects models were used to generate pooled estimates of association using adjusted study results. The full methodological approach was published on PROSPERO (ID: CRD42017077264). Results Sixty‐six studies were identified for qualitative synthesis, with 11 of these studies eligible for meta‐analyses. A meta‐analysis including 3 studies measuring medication non‐adherence in adults aged ≥55 years showed a significant association with all‐cause hospitalisation (adjusted odds ratio 1.17, 95% confidence interval [CI] 1.12, 1.21). A meta‐analysis including 2 studies showed that medication non‐adherence was not significantly associated with an emergency department visit (adjusted odds ratio 1.05, 95% CI 0.90, 1.22). Good adherence was associated with a 21% reduction in long‐term mortality risk in comparison to medication non‐adherence (adjusted hazard ratio 0.79, 95% CI 0.63, 0.98). Conclusion Medication non‐adherence may be significantly associated with all‐cause hospitalisation and mortality in older people. Medication adherence should be monitored and addressed in this cohort to minimise hospitalisation, improve clinical outcomes and reduce healthcare costs.

No abstract

No abstract