Assessing polypharmacy in the older population: Comparison of a self-reported and prescription based method

Purpose To explore differences in the prevalence and determinants of polypharmacy in the older general population in Belgium between self-reported and prescription based estimates and assess the relative merits of each data source. Methods Data were used from participants aged >= 65 years of the Belgian national health survey 2013 (n = 1950). Detailed information was asked on the use of medicines in the past 24 h and linked with prescription data from the Belgian compulsory health insurance (BCHI). Agreement between polypharmacy (use or prescription >= 5 medicines) and excessive polypharmacy (>= 10 medicines) between both sources was assessed with kappa statistics. Multinomial logistic regression was used to study determinants of moderate (5-9 medicines) and excessive polypharmacy (>= 10 medicines) and over- and underestimation of prescription based compared to self-reported polypharmacy. Results Self-reported and prescription based polypharmacy prevalence estimates were respectively 27% and 32%. Overall agreement was moderate, but better in men (kappa 0.60) than in women (0.45). Determinants of moderate polypharmacy did not vary substantially by source of outcome indicator, but restrictions in activities of daily living (ADL), living in an institution and a history of a hospital admission was associated with self-reported based excessive polypharmacy only. Conclusions Surveys and prescription data measure polypharmacy from a different perspective, but overall conclusions in terms of prevalence and determinants of polypharmacy do not differ substantially by data source. Linking survey data with prescription data can combine the strengths of both data sources resulting in a better tool to explore polypharmacy at population level.