RESUMO
PURPOSE: Beers List drugs are potentially harmful in older adults and are grouped by level of risk. Over 9000 total knee arthroplasties (TKAs) are performed each year Veterans Affairs (VA) hospitals, primarily on older adults. Minimal data on the administration of Beers List drugs following arthroplasty currently exists in the literature. Our goal was to quantify the risks of these drugs following TKA. We hypothesized that increasing doses of Beers List drugs would be associated with increased risks for readmission, reoperation, emergency department (ED) visits, and mortality. METHODS: In this retrospective cohort study, data from TKAs performed in VA hospitals from 2010 to 2014 were examined, with complicated or bilateral procedures excluded. The data were obtained from the VA Corporate Data Warehouse. The outcomes examined were readmission, postoperative ED visits, reoperation on ipsilateral knee, and mortality. Beers List drugs were divided into 3 categories: medications to use with caution (Beers 0); medications to avoid in older adults (Beers 1); and medications to avoid in certain disease states (Beers 2). Beers 2 was not included in the final analysis due to an inability to verify appropriate diagnostic criteria without manual chart review. Logistic regression was performed looking at the total number of doses in the first 48 h after surgery compared to the above-mentioned outcomes. FINDINGS: Data from 12,639 TKAs were analyzed; the mean age of the patients was 65.06 years, and 77.8% of patients received Beers List drugs while admitted. The most frequently administered Beers List drugs were proton pump inhibitors, NSAIDs, insulin, α-blockers, benzodiazepines, antihistamines, muscle relaxants, and antipsychotics. There was a dose-dependent increase in readmission and ED visits in the Beers 1 group. The odds ratios were 1.03 for 30-day readmission and 1.02 at 90 days. The odds ratios for ED visits were 1.05 for 72-h ED visits and 1.04 for ED visits within 7 and 30 days. The odds ratios were set at 1-unit dose intervals. All results were found after control for VA facility, sex, age, American Society of Anesthesiologists class, Charlson score, case length, and body mass index. IMPLICATIONS: The group of medications to avoid (Beers 1) from the 2015 Beers List showed associations with increased frequency of readmission and postoperative ED visits. Reinforcement of the need to avoid those drugs during surgical care will hopefully reduce such complications. Limitations included not controlling for overall discharge drug count and reliance on the outpatient problem list for outpatient diagnoses. Additional subgroup analysis will be performed to see whether specific drugs pose a higher than risk others.
