A systematic review of the prevalence, determinants, and impact of potentially inappropriate prescribing in middle-aged adults

Purpose: Potentially inappropriate prescribing (PIP) is common in older adults and is associated with adverse clinical outcomes, such as adverse drug events, and increased economic costs. However, PIP in middle-aged adults (MAA) has not been well-characterised. Objective: This review reports the prevalence, determinants, and impact of PIP in MAA. Methods: A systematic review and meta-analysis were conducted following PRISMA reporting guidelines (PROSPERO CRD42020206617–7 October 2020). Searches were performed of OVID MEDLINE, Embase, Cumulative Index to Nursing and Allied Health Literature, Cochrane Library, Web of Science, ProQuest Dissertation and Thesis Database, ClinicalTrials.gov, World Health Organization International Clinical Trials Registry Platform, and OpenGrey. Studies that included adults aged 45–64 years and applied explicit PIP criteria were eligible through to September 2020. Data collected from studies were pre-specified and recorded using a data extraction sheet, details of which are available in our PROSPERO protocol. Meta-analysis of PIP prevalence was conducted using Stata version 16. All other outcomes were examined by narrative synthesis. Risk of bias and overall certainty of findings were assessed with the QUIPS (Quality in Prognosis Study) and GRADE (grading of recommendations assessment, development, and evaluation) tools, respectively. Results: Of 8183 citations screened, 22 studies were included in the review. PIP was defined using 15 different criteria, with only one specifically designed for MAA (PROMPT criteria). Four studies that had disaggregated prevalence data for MAA (n = 753,193) were included in a meta-analysis (PIP prevalence 38%; 95% confidence interval [CI] 25–52; p < 0.01; I² > 99%], although heterogeneity was high. Female sex and polypharmacy were associated with PIP. Only one study reported healthcare utilisation and quality of life and found no association with PIP. One cohort study reported that adverse drug reactions had a significant association with PIP (adjusted odds ratio 3.65; 95% CI 1.13–11.78; p < 0.05). Conclusion: PIP is common in MAA. There was moderate certainty for our findings for the prevalence of PIP and its association with female sex and polypharmacy. Certainty for other findings was either low or very low. No data were available to assess some of our pre-specified outcomes. Further studies are needed to examine whether PIP is associated with negative outcomes in MAA.