Research That Matters (January 17 - 20, 2008)


Blue Prefunction (Omni Shoreham)

Mental Illness and Substance Abuse in Explaining Jail Incarceration

Jeffrey Draine, PhD, University of Pennsylvania, Amy Blank, PhD, Rutgers University, Steve Metraux, PhD, University of the Sciences in Philadelphia, and Trevor Hadley, PhD, University of Pennsylvania.

PURPOSE: We examine the relative role of mental illness and substance abuse in explaining incarceration, and the trajectory through an urban jail. METHODS: A population of 44,397 jail intakes and releases in 2003 and 2004 were compared with the population of 206,987 Medicaid eligible individuals in 2003. Among these jail detainees, 14,064 were found to have MA eligibility. Eligibility records were then matched with claims data to determine history of treatment of serious mental illness and substance use disorders. RESULTS: Based on these claims, an estimated 5.1 to 5.5 percent of the detainees have a history of treatment for schizophrenia, 4.3 to 4.5 percent for affective disorders. Among substance use disorder, 8.7% to 9.4% show alcohol dependence, and about 11.5% show drug dependence. In multivariate analyses of risk for incarceration among the MA population, mental illnesses are generally not a risk factor when strong risk factors for substance use and co-occurring mental illness and substance use disorders are included in the model. These results also have implications for incarceration trajectories in the jail. CONCLUSIONS: These findings challenge the potential impact of services that target mental illness as the proximal risk factor for arrest in a poor, urban population. These results support a reframing of diversion and reentry to focus on substance use as the proximal cause for incarceration among people with mental illness. Such a reframing would challenge behavioral health service system mechanisms that inhibit access to effective treatment for co-occurring disorders, for which there are few evidence based models of treatment that account for a criminal justice context.