Theresa I Shireman, Shekinah Fashaw-Walters, Tingting Zhang, Andrew R Zullo, Lauren B Gerlach, Antoinette B Coe, Lori Daiello, Derrick Lo, Julie Strominger, Julie P W Bynum
Abstract
Objectives:
Federal initiatives have been successful in reducing antipsychotic exposure in nursing home residents with dementia. We assessed if these initiatives were implemented equally across racial and ethnic minority groups.
Design:
Retrospective, cross-sectional trends study.
Setting and participants:
National long-stay nursing home residents with dementia from 2011 to 2017.
Methods:
We examined trends in psychotropic drug class exposures from the Minimum Data Set assessments for non-Hispanic Black (NHB), Hispanic, and non-Hispanic White (NHW) residents using interrupted time-series analyses with age-sex standardized quarterly outcomes and time points to denote the National Partnership (2012) and Five Star Rating changes (2015).
Results:
Initially, antipsychotic (33.0%) and sedative (6.8%) exposure was highest for Hispanic residents; antidepressant (59.8%) and anxiolytic (23.4%) exposure was highest for NHW residents; NHB residents had the lowest use of each. Antipsychotic use dropped at the time of the Partnership (β = -0.8807, P = .0023) and the slope declined further after the Partnership (β = -0.6611, P < .0001) for NHW. In comparison to NHW, the level and slope changes for NHB and Hispanics were not significantly different. The Five Star Rating change did not impact the level of antipsychotic use (β = 0.027, P = .9467), but the slope changed to indicate a slowed rate of decline (β = 0.1317, P = .4075) for NHW. As to the other psychotropic drug classes, there were few significant differences between trends seen in the racial and ethnic subgroups. The following exceptions were noted: antidepressant use decreased at a faster rate for NHB residents post-Partnership (β = -0.1485, P = .0371), and after the Five Star Rating change, NHB residents (β = -0.0428, P = .0312) and Hispanic residents (β = -0.0834, P < .0001) saw antidepressant use decrease faster than NHW. Sedative use in slope post-Partnership period (β = -0.086, P = .0275) and post-Five Star Rating (β = -0.0775, P < .0001) declined faster among Hispanic residents.
Conclusions and implications:
We found little evidence of clinically meaningful differences in changes to 4 classes of psychotropic medication use among racial and ethnic minority nursing home residents with dementia following 2 major federal initiatives.