Evidence-based decision support for managing the mentally ill inmate population

摘要

The proportion of incarcerated individuals in the United States who suffer from mental illness has increased significantly in recent years, and the treatment of these individuals within criminal justice systems has come under increased scrutiny. In Charlottesville and Albemarle County, as well as in jurisdictions across the country, there is a lack of coordination between the criminal justice system and mental health services, which leads to a gap in fully understanding the tendencies and trends within the inmate population suffering from mental illness. In an effort ameliorate this lack of coordination, Albemarle/Charlottesville seeks to become the second community in the country (behind Camden, NJ) to have data from community mental health service providers and criminal justice agencies analyzed together and the first to do so with a focus on a released jail cohort. The Albemarle/Charlottesville criminal justice system is a leader in evidence-based decision making practices, which aim to improve public safety and reduce jail overcrowding by using data and data-driven analysis in order to determine how to best manage the inmate population suffering from mental illness. This project contributes to the evidence-based practice by investigating treatment linkages and efficacy of treatments for individuals whom the Albemarle-Charlottesville Regional Jail has referred for further mental health evaluation and treatment. The project required merging data, covering an 18-month study period from July 2015 through December 2016, held by seven participating agencies, including the Albemarle-Charlottesville Regional Jail, Charlottesville Police Department, Emergency Communications Center, Region 10 Community Services, Offender Aid and Restoration, Virginia Department of Corrections, and University of Virginia Hospital. By using data from the Brief Jail Mental Health Screener administered to individuals during jail intake, this project tracks the inmates that have been referred for mental health evaluation or treatment in order to identify any patterns or similarities between these individuals as they are diverted from, move through, or exit the jail system. Identifying information was used to merge disparate datasets, with a core dataset containing 3,848 unique inmates, on a secure computer and server. The merged datasets, stripped of personally identifiable information, were used for analysis to answer research questions including characteristic differences between referred and non-referred cohorts, the efficacy of Crisis Intervention Teams, and the effectiveness of treatment on recidivism reduction. Of the 3,848 inmates in the jail data, 2,151 inmates took a brief mental health screener and 23% of screened individuals were referred. Of this referred cohort, there was roughly a 44% treatment linkage rate of receiving mental health, substance abuse, or other services from Region 10. This unique combination of datasets that merges information from multiple sources lays the foundation for future collaborative efforts among the participating agencies and provides a framework for exploring additional research questions beyond the scope of this effort.