LAUSR

Supplemental Digital Content is available in the text. Objective: In 2012 Centers for Medicare and Medicaid Services (CMS) launched a multifaceted initiative aimed at reducing the unnecessary use of antipsychotic medications in nursing facilities due to evidence these medications are associated with little or uncertain benefit and substantial risk. Yet, little is known about whether efforts to reduce antipsychotic medication should be focused on residents with targeted characteristics, or on nursing facility regulation (eg, staffing levels). Our objective was to identify the relative contribution of resident and facility characteristics to potentially inappropriate antipsychotic use. Methods: We examined 1,156,875 long stay residents in 14,699 US nursing facilities in 2014 and predicted resident antipsychotic use controlling sequentially for resident and facility characteristics and calculated the incremental variation explained. Results: We found significant variability in unadjusted rates of potentially inappropriate antipsychotic use among nursing facilities (mean=18.0%; interquartile range: 11.3%–23.7%; SD: 11.1). Regression results indicated that 93% of the explained variation in antipsychotic use was attributed to resident characteristics and 7% was attributed to facility-level factors. At the facility level, worker hours per resident day was not significantly associated with antipsychotic use. Simulations indicated that applying the effect sizes achieved by the best performing facilities to the existing case mix across all nursing facilities would result in no more than a 1 percentage point change in population-level antipsychotic use. Conclusions: Efforts to reduce antipsychotic use may have greater impact by developing new clinical strategies to address specific diagnoses rather than regulations related to facility-level attributes.