Epidemiology and association with outcomes of polypharmacy in patients undergoing surgery : retrospective, population-based cohort study

Abstract

Background: The aim of this study was to determine the prevalence of preoperative polypharmacy and the incidence of postoperative polypharmacy/hyper-polypharmacy in surgical patients and their association with adverse outcomes. Methods: This was a retrospective, population-based cohort study among patients older than or equal to 18 years undergoing surgery at a university hospital between 2005 and 2018. Patients were categorized based on the number of medications: non-polypharmacy (fewer than 5); polypharmacy (5–9); and hyper-polypharmacy (greater than or equal to 10). The 30-day mortality, prolonged hospitalization (greater than or equal to 10 days), and incidence of readmission were compared between medication-use categories. Results: Among 55 997 patients, the prevalence of preoperative polypharmacy was 32.3 per cent (95 per cent c.i. 33.5 to 34.3) and the prevalence of hyper-polypharmacy was 25.5 per cent (95 per cent c.i. 25.2 to 25.9). Thirty-day mortality was higher for patients exposed to preoperative hyper-polypharmacy (2.3 per cent) and preoperative polypharmacy (0.8 per cent) compared with those exposed to non-polypharmacy (0.6 per cent) (P < 0.001). The hazards ratio (HR) of long-term mortality was higher for patients exposed to hyper-polypharmacy (HR 1.32 (95 per cent c.i. 1.25 to 1.40)) and polypharmacy (HR 1.07 (95 per cent c.i. 1.01 to 1.14)) after adjustment for patient and procedural variables. The incidence of longer hospitalization (greater than or equal to 10 days) was higher for hyper-polypharmacy (11.3 per cent) and polypharmacy (6.3 per cent) compared with non-polypharmacy (4.1 per cent) (P < 0.001). The 30-day incidence of readmission was higher for patients exposed to hyper-polypharmacy (10.2 per cent) compared with polypharmacy (6.1 per cent) and non-polypharmacy (4.8 per cent) (P < 0.001). Among patients not exposed to polypharmacy, the incidence of new postoperative polypharmacy/hyper-polypharmacy was 33.4 per cent (95 per cent c.i. 32.8 to 34.1), and, for patients exposed to preoperative polypharmacy, the incidence of postoperative hyper-polypharmacy was 16.3 per cent (95 per cent c.i. 16.0 to 16.7). Conclusion: Preoperative polypharmacy and new postoperative polypharmacy/hyper-polypharmacy are common and associated with adverse outcomes. This highlights the need for increased emphasis on optimizing medication usage throughout the perioperative interval.