Potentially Inappropriate Medication Prescriptions for Older Adults with Painful Conditions and Association with Return Emergency Department Visits

Objectives: To describe the frequency and risk of return visit to the emergency department (ED) by older adults after prescription of any of four potentially inappropriate medication (PIM) classes included in the 2015 Beers Criteria commonly used for the relief of acute pain in the ED. Design: Retrospective cohort study. Setting: Large urban academic ED from January 1, 2013, to December 31, 2015. Participants: Patients age 65 and older discharged from the ED with an initial pain score of 1 or higher (11 822 visits). Measurements: Prescriptions for PIM classes were collected from the medical record: nonsteroidal anti-inflammatory drugs (NSAIDs), benzodiazepines, skeletal muscle relaxants, and opioids. The proportion of patients with ED returns within 9 days were compared by medication class and pain severity (mild, moderate, or severe). Multivariable logistic regression was performed for each pain category to determine adjusted odds ratios (aORs) of ED return. Results: Of 11 822 included patients, PIMs were prescribed in 3392 (28.7%): 2550 (21.6%) opioids, 826 (7.0%) NSAIDs, 277 (2.3%) benzodiazepines, and 68 (0.6%) nonbenzodiazepine skeletal muscle relaxants. Total 9-day ED returns were 1125 (9.5%): mild 7.0%, moderate 8.3%, and severe pain 11.7%. Opioids were not associated with more frequent ED returns for mild or moderate pain, and they were associated with less frequent ED returns for severe pain (9.2% vs 12.7%; p <.001; aOR 0.69; 95% confidence interval [CI] = 0.54-0.87). Benzodiazepines were associated with more frequent ED returns for patients with moderate pain (15.5% vs 8.2%; p <.01; aOR = 2.01; 95%CI = 1.10-3.70). Conclusions: These results are consistent with recommendations to limit benzodiazepine prescriptions for older adults and that among older adults with severe pain, opioid prescribing is associated with less frequent ED visits within 9 days of discharge. However, this study was not designed to evaluate safety, adverse events, or other important patient-centered outcomes. J Am Geriatr Soc 67:719–725, 2019.