BACKGROUND: Polypharmacy, the simultaneous use of multiple drugs, is associated with adverse drug reactions, medication errors, and increased risk of hospitalization. When the number of concurrently used drugs totals five or more (major polypharmacy), a significant risk may be present. AIM: To analyse the interpractice variation in the prevalence of major polypharmacy among listed patients, and to identify possible predictors of major polypharmacy related to the practice. METHOD: Prescription data were retrieved from the Odense Pharmacoepidemiological Database, and individuals subject to major polypharmacy were identified. The age- and sex-standardized prevalence rate of major polypharmacy was calculated for each practice in the County of Funen in Denmark (n = 173), using the distribution of age and sex of the background population as a reference. The practice characteristics were retrieved from the Regional Health Insurance System. Possible predictors of major polypharmacy related to the general practitioners (GPs) were analysed using backward stepwise linear multiple regression. RESULTS: A six-fold variation between the practices in the prevalence of major polypharmacy was found (16 to 96 per 1000 listed patients; median = 42). Predictors related to the practice structure, workload, clinical work profile, and prescribing profile could explain 56% of the variation. CONCLUSION: A substantial part of the variation in major polypharmacy between practices can be explained by predictors related to practice.
BACKGROUND: Polypharmacy, the simultaneous use of multiple drugs, is associated with adverse drug reactions, medication errors, and increased risk of hospitalization. When the number of concurrently used drugs totals five or more (major polypharmacy), a significant risk may be present. AIM: To analyse the interpractice variation in the prevalence of major polypharmacy among listed patients, and to identify possible predictors of major polypharmacy related to the practice. METHOD: Prescription data were retrieved from the Odense Pharmacoepidemiological Database, and individuals subject to major polypharmacy were identified. The age- and sex-standardized prevalence rate of major polypharmacy was calculated for each practice in the County of Funen in Denmark (n = 173), using the distribution of age and sex of the background population as a reference. The practice characteristics were retrieved from the Regional Health Insurance System. Possible predictors of major polypharmacy related to the general practitioners (GPs) were analysed using backward stepwise linear multiple regression. RESULTS: A six-fold variation between the practices in the prevalence of major polypharmacy was found (16 to 96 per 1000 listed patients; median = 42). Predictors related to the practice structure, workload, clinical work profile, and prescribing profile could explain 56% of the variation. CONCLUSION: A substantial part of the variation in major polypharmacy between practices can be explained by predictors related to practice.