Reducing Psychiatric Inpatient Readmissions Using an Organizational Change Model.

INTRODUCTION Thirty-day hospital readmission rates have become a quality indicator for many regulators and payers, but published accounts of reducing these rates across a patient population are lacking. OBJECTIVE This article describes and evaluates the Wisconsin Mental Health Readmissions Project, which aimed to reduce psychiatric inpatient 30-day readmission rates in Wisconsin. METHODS Nineteen county human services boards representing 23 of Wisconsin's 72 counties and 61% of the state's residential admissions participated in a statewide quality improvement collaborative from January 1, 2010 to December 31, 2013. Participants applied a standardized organizational change model, called NIATx, in the context of a multicounty quality improvement collaborative to reduce 30-day readmission rates. Readmission rates were tracked through national and state databases, using 2009 as a baseline, and analyzed using a chi-square analysis to test the proportion of means. The study team compared readmission rates of Wisconsin counties that participated in the statewide collaborative with those that did not. RESULTS Between 2009 and 2013, the 30-day readmission rates in Wisconsin declined significantly for counties that participated in the project when compared to those that did not (2009-2013) [Χ2(4) = 54.503, P < .001], based on a 2.5% decline for participants vs a 0.7% decline for nonparticipants. CONCLUSIONS Reductions to behavioral health inpatient readmission rates beyond individual case examples have been difficult to document. This analysis evaluates a method that Wisconsin behavioral health providers applied as part of a multicounty program addressing readmission rates. The findings highlight quality improvement program design elements and interventions to consider in reducing inpatient behavioral health readmissions, as well as the need for further research on this complex systems issue.