Potentially inappropriate medication prescribing in outpatient practices: Prevalence and patient characteristics based on electronic health records

Abstract

Background: Some older adults receive potentially inappropriate medications (PIMs), increasing their risk for adverse events. A literature search did not find any US multicenter studies that measured the prevalence of PIMs in outpatient practices based on data from electronic health records (EHRs), using both the Beers and Zhan criteria. Objectives: The aims of the present study were to compare the prevalence of PIMs using standard drug terminologies at 2 disparate institutions using EHRs and to identify characteristics of elderly patients who have a PIM on their active-medication lists. Methods: This cross-sectional study of outpatients' active-medication lists from April 1, 2006, was conducted using data from 2 outpatient primary care settings: Intermountain Healthcare, Salt Lake City, Utah (center 1), and the Cleveland Clinic, Cleveland, Ohio (center 2). Data were included from patients who were aged ≥65 years at the time of the last office visit and had ≥2 documented clinic visits within the previous 2 years. The primary end point was prevalence of PIMs, measured according to the 2002 Beers criteria or the 2001 Zhan criteria. Results: Data from 61,251 patients were included (36,663 women, 24,588 men; center 1: 37,247 patients; center 2: 24,004). A total of 8693 (23.3%) and 5528 (23.0%) patients at centers 1 and 2, respectively, were documented as receiving a PIM as per the Beers criteria; this difference was not statistically significant. Per the Zhan criteria ( P < 0.001), these values were 6036 (16.2%) and 4160 (17.3%). Eight of the most common PIMs were the same at both institutions, with propoxyphene and fluoxetine (once daily) being the most prescribed. Female sex, polypharmacy (≥6 medications), and multiple primary care visits were significantly associated with PIM prescribing. Conclusions: In this analysis of data from elderly patients at 2 outpatient centers, a small set of 8 medications accounted for the majority of PIMs at both centers, irrespective of geographic and demographic variations. Female sex, polypharmacy, and number of primary care visits were significantly associated with PIM prescribing. In this analysis of data from elderly patients at 2 outpatient centers, a small set of 8 medications accounted for the majority of PIMs at both centers, irrespective of geographic and demographic variations.