Prevalence and Average Annual Percent Change of Potentially Inappropriate Medication Use in Older Adults before and during the COVID-19 National Emergency

OBJECTIVES: Potentially inappropriate medication (PIM) use in older adults increases the risk of injury, hospitalization, and adverse outcomes. The COVID-19 pandemic contributed to increases in antipsychotics and other PIMs in nursing homes, though less is known about the pandemic's impact on PIM use in the community-dwelling older adult population. In this study we investigated select antidepressant, anticonvulsant, antipsychotic, and benzodiazepine use in older adults before and during the COVID-19 national emergency.

METHODS: This is a retrospective, yearly, repeated cross-sectional study using Medicare and CVS Health linked data between January 2017 and December 2022. Community-dwelling individuals 65 years and older with Medicare Part D coverage and community pharmacy records were included. We estimated prevalence and average annual percent change (AAPC) of PIM use during the pre-emergency period (2017-2019), emergency period (2020-2022), and by discrete years.

RESULTS: We identified 19,275,777 eligible individuals across all years. During our study period, the prevalence of PIMs generally increased for all PIM categories except benzodiazepine which generally decreased.

In 2017 the prevalence of PIMs was 9.19%(95% CI 9.17-9.20), 13.71%(95% CI 13.69-13.73) 1.69%(95% CI 1.69-1.70) , and 7.75%(95% CI 7.74-7.77); yet in 2022 the prevalence of PIMs was 10.74%(95% CI 10.73-10.76), 14.88%(95% CI 14.87-14.90), 2.39%(95% CI 2.38-2.40), and 6.02%(95% CI 6.01-6.03) for anticonvulsant, antidepressant, antipsychotic, and benzodiazepine use respectively.

Pre-emergency AAPC was 3.92%(95% CI 3.80-4.05), 2.70%(95% CI 2.60-2.80), 11.88%(95% CI 11.58-12.18), and -5.65%(95% CI -5.78—5.52); yet during the emergency AAPC was 2.96%(95% CI 2.85-3.07), 1.26%(95% CI 1.17-1.34), 0.46%(95% CI 0.24-0.68), and -4.76%(95% CI -4.88—4.63) for anticonvulsant, antidepressant, antipsychotic, and benzodiazepine use respectively.

CONCLUSIONS: Sustained decreases of benzodiazepine use during our study period is a positive finding. However, increases in anticonvulsant, antidepressant, and antipsychotic use is concerning. Annual change in prevalence was less pronounced in all PIM categories during the emergency, flattening both positive and negative prevalence trends.