GP-Led Deprescribing in Community-Living Older Australians: An Exploratory Controlled Trial

OBJECTIVE To assess feasibility, effectiveness, and safety of a multifaceted general practitioner (GP) led intervention to reduce potentially inappropriate polypharmacy in community-living older people. DESIGN Pragmatic, controlled, pre-post design. SETTING Five general practices in southeast Queensland, Australia. PARTICIPANTS Ten GPs from three practices, two pharmacists, and 78 patients comprised the intervention group, and 10 GPs from two practices and 67 patients comprised the usual care group. Patients were aged 65 years or older, receiving five or more regularly prescribed medicines, and capable of participating in telephone interviews in English. INTERVENTION A 5-hour interactive deprescribing training workshop for clinicians; an extended deprescribing consultation between GPs and enrolled patients, entailing a comprehensive review of their medicines using a standardized software template codesigned by GPs; and comprehensive medicine review by a pharmacist, at the GP's discretion. OUTCOME MEASURES Primary outcome was mean difference in number of regular medicines deprescribed (ie, ceased or dose reduced) per patient over an 18-week follow-up period. Medicine-specific and patient-reported outcomes, safety, and process measures were also evaluated. RESULTS At study completion, mean (SD) number of regular medicines deprescribed per patient was 0.99 (1.23) in the intervention group vs 0.43 (0.84) in the usual care group, equaling a mean difference of 0.55 (95% confidence interval = -0.90 to -0.21; P = .002). Crude totals showed 77 of 649 (11.9%) vs 29 of 571 (5.1%) regular medicines deprescribed in intervention and usual care groups, respectively (P < .001). Supplements, gastric acid suppressants, statins, oral hypoglycemics, and diuretics were medicine classes more frequently deprescribed. There were no statistically significant between-group differences in numbers of medicines commenced, self-reported unplanned hospitalizations, or worsened health-related quality of life. A subset of intervention patients reported greater certainty in the necessity and appropriateness of their medicines at study end. CONCLUSION The deprescribing intervention appears feasible, was modestly effective, and was not associated with any major safety events.