Associations between specialized dementia care, COVID-19 and central nervous system medication use in assisted living: a population-based repeated cross-sectional study
Colleen J Maxwell,Hana Dampf,Jillian P Squires,David B Hogan,Cecilia A Cotton,Erik Youngson MMath,Zoe Hsu,Matthias Hoben
DOI: https://doi.org/10.1186/s12877-024-05274-w
2024-08-14
Abstract:Background: Assisted living (AL) is an increasingly common residential setting for persons with dementia; yet concerns exist about sub-optimal care of this population in AL given its lower levels of staffing and services. Our objectives were to (i) examine associations between AL setting (dementia care vs. other), COVID-19 pandemic waves, and prevalent antipsychotic, antidepressant, anti-dementia, benzodiazepine, and anticonvulsant drug use among residents with dementia/cognitive impairment, and (ii) explore associations between resident and home characteristics and prevalent medication use. Methods: We conducted a population-based, repeated cross-sectional study using linked clinical and health administrative databases for all publicly funded AL homes in Alberta, Canada, examined between January 2018 - December 2021. The quarterly proportion of residents dispensed a study medication was examined for each setting and period (pandemic vs. comparable historical [2018/2019 combined]) focusing on four pandemic waves (March-May 2020, September 2020-February 2021, March-May 2021, September-December 2021). Log-binomial GEE models estimated prevalence ratios (PR) for period (pandemic vs. historical periods), setting (dementia care vs. other) and period-setting interactions, adjusting for resident (age, sex) and home (COVID-19 cases, health region, ownership) characteristics. Results: On March 1, 2020, there were 2,779 dementia care and 3,013 other AL residents (mean age 83, 69% female) with dementia/cognitive impairment. Antipsychotic use increased during waves 2-4 in both settings, but this was more pronounced in dementia care than other AL during waves 3 and 4 (e.g., adjusted [adj]PR 1.20, 95% CI 1.14-1.27 vs. adjPR 1.09, 95% CI 1.02-1.17, interaction p = 0.023, wave 3). Both settings showed a statistically significant but modest increase in antidepressant use and decrease in benzodiazepine use. For dementia care AL residents only, there was a statistically significant increase in gabapentinoid use during several waves (e.g., adjPR 1.32, 95% CI 1.10-1.59, wave 3). Other than a modest decrease in prevalent anti-dementia drug use for both settings in wave 2, no other significant pandemic effects were observed. Conclusions: The persistence of the pandemic-associated increase in antipsychotic and antidepressant use in AL residents coupled with a greater increase in antipsychotic and gabapentinoid use for dementia care settings raises concerns about the attendant risks for residents with cognitive impairment.