A Canadian Cohort Study to Evaluate the Outcomes Associated with a Multicenter Initiative to Reduce Antipsychotic Use in Long-Term Care Homes.

Abstract Objectives To evaluate the impact of a multicenter intervention to reduce potentially inappropriate antipsychotic use in Canadian nursing homes at the individual and facility levels. Design Longitudinal, population-based cohort study to evaluate the Canadian Foundation for Healthcare Improvement's Spreading Healthcare Innovations Initiative to reduce potentially inappropriate antipsychotic use in 6 provinces/territories. Setting and Participants Adults in nursing homes in 6 provinces/territories in Canada between 2014 and 2016. The sample involved 4927 residents in 45 intervention homes and 122,570 residents in 1193 control homes in the first quarter of the study. Measures Assessment data based on the Resident Assessment Instrument 2.0 were used in both settings to track antipsychotic use and to obtain risk-adjusters for a quality indicator on potentially inappropriate use. Intervention Quality improvement teams in participating organizations were provided with education, training, and support to implement localized strategies intended to reduce antipsychotic medication use in residents without diagnosis of psychosis. Results At the resident level, we found that the odds of remaining on potentially inappropriate antipsychotics were 0.75 in intervention compared with control homes after adjusting for age, sex, aggressive behavior, and cognition. These findings were evident within the pooled Canadian data as well as within provinces. At the facility level, the intervention homes had greater improvements in risk-adjusted quality indicator performance than the control homes, and this was true for the worst, median, and best-performing homes at baseline. There was no major change in the quality indicator for worsening of behavior symptoms. Conclusions/Implications The Canadian Foundation for Healthcare Improvement intervention was associated with a reduction in potentially inappropriate antipsychotic use at both the individual and facility levels of analysis. This improvement in performance was independent of secular trends toward reduced antipsychotic use in participating provinces. This suggests that substantial improvements in medication use may be achieved through targeted, collaborative quality improvement initiatives in long-term care.