Characteristics of Participants in Jail Diversion and Prison Reentry Programs: Implications for Forensic ACT

The study setting was a large, nonprofit mental health agency—Greater Cincinnati Behavioral Health Services, Inc. (GCB), a nationally accredited agency in Cincinnati, Ohio, that serves approximately 3,500 adults with severe mental illness annually. GCB’s jail diversion FACT team began in March 2003 and serves nonviolent misdemeanants. GCB’s prison reentry FACT team began in September 2002 and serves consumers who return to Hamilton County (Cincinnati) from one of the state’s prison facilities. Both FACT teams adhere to national ACT fidelity standards and incorporate all of the core elements of ACT (14). Moreover, the teams adhere to Ohio standards for ACT eligibility as a common starting place for identifying FACT-eligible consumers. Eligible consumers have had a serious mental illness, typically schizophrenia or bipolar disorder, for at least a year and experience considerable disability from their disorders, typically evidenced by two or more hospitalizations in a calendar year.

The jail diversion FACT team accepts referrals from the local criminal justice center. Most of these referrals are for misdemeanants, although persons who have committed felonies and violent crimes are assessed on a case-by-case basis. All prison reentry referrals originate from the state prison system. Acceptance of referrals of consumers with a history of violence is assessed on a case-by-case basis.

An observational study design and administrative data were used to compare 212 jail diversion FACT consumers and 192 prison reentry FACT consumers who had received FACT services between 2002 and 2008. Data were collected as part of a larger study of transitions from ACT to less intensive services (15).

The study was approved by the Institutional Review Board at the University of North Carolina at Chapel Hill.

The data contained demographic and clinical information, as well as FACT staff impressions of current clinical risk factors and those in the recent past. Demographic variables included gender, race (white, black, or other), and age at the time of the study. Behavioral health and general medical diagnoses came from GCB psychiatrists or psychiatric nurses during routine intake and follow-up assessment processes. Dichotomous variables were created to indicate schizophrenia or mood disorders, substance use disorders, general medical problems, receipt of Supplemental Security Income (SSI) or Social Security Disability Insurance (SSDI), housing (assisted living, homeless, or independent living), and current or recent-past history of the following: treatment noncompliance, suicide attempt, sexual abuse, interpersonal violence, homelessness, and health complications in extreme temperatures (underdressing in cold temperatures or overdressing in hot temperatures).

Collecting information about these risk factors was part of the agency’s overall data collection efforts. Agency staff members were required to document any clinical risk factors that needed to be addressed in treatment (for example, general medical problems) or that might be a barrier to recovery (for example, noncompliance or a history of trauma). Risk factors could be acute or in the recent past; clinicians were free to define recent past. Typically, recent past entailed a risk factor that occurred within the last several months. All FACT staff were required to document clinical risk factors.

The dependent variable was a dichotomous variable called group status, coded 0 for jail diversion and 1 for prison reentry. Chi square tests were used to examine the relationship between group status (jail diversion or prison reentry) and all dichotomous variables, and t tests were used to examine group differences in age. These analyses were repeated to examine the relationship between gender and all variables within teams. Analyses were conducted with SAS 9.2, and two-tailed tests with alpha set at .05 were conducted for all tests.

Table 1 presents data on the characteristics of the jail diversion FACT consumers and the prison reentry FACT consumers. Compared with jail diversion consumers, prison reentry consumers were more often male (46% and 57%, respectively; χ²=5.37, df=1, p≤.05), were more likely to have a diagnosis of schizophrenia (43% and 82%; χ²=62.74, df=1, p≤.001), and were older (36.9 and 45.9 years; t=7.12, df=368, p≤.001). Also, prison reentry consumers were more likely than jail diversion consumers to be in assisted living arrangements (36% and 8%, respectively; χ²=46.26, df=1, p≤.001) and to be SSI or SSDI recipients (77% and 59%; χ²=14.38, df=1, p≤.001).

Compared with prison reentry consumers, jail diversion consumers were more likely to be female (43% and 54%, respectively; χ²=5.37, df=1, p≤.05), to have diagnoses of mood disorders (13% and 53%, respectively; χ²=72.69, df=1, p≤.001) and substance use disorders (30% and 55%; χ²=25.61, df=1, p≤.001), and to have histories of sexual abuse (3% and 10%; χ²=8.92, df=1, p≤.05) and interpersonal violence (4% and 12%; χ²=7.81, df=1, p≤.01).

Also shown in Table 1, among jail diversion consumers, females were more likely than males to be white (54% and 36%, respectively; χ²=6.74, df=1, p≤.01), to be living independently (63% and 45%; χ²=6.32, df=1, p≤.01), and to have a history of sexual abuse (17% and 2%; χ²=12.33, df=1, p≤.001) and interpersonal violence (19% and 3%; χ²=13.01, df=1, p≤.001). Among jail diversion consumers, males were more likely than females to have schizophrenia (55% and 34%; χ²= 9.20, df=1, p≤.01), substance use disorders (65% and 47%; χ²=6.89, df=1, p≤.01), and a history of treatment noncompliance (52% and 30%; χ²=10.63, df=1, p≤.001) and homelessness (48% and 27%; χ²=10.46, df=1, p≤.001).

This study contributes to the literature about FACT and provides important information for the frontline providers of jail diversion and prison reentry services. It is important to note that all consumers in the study received FACT services. That is, the same service was provided to potentially different populations—consumers who were diverted from jail and those who were released from prison. Thus this study was well positioned to assess the extent to which population differences imply the need to adjust the engagement and intervention strategies used by jail diversion FACT teams and by prison reentry FACT teams.

Justice-involved consumers are not a homogeneous group, and this study underscores the need for FACT teams to tailor services to meet population- and gender-specific clinical needs. Prison reentry consumers in this study appeared to be more disabled and to be more likely to be living in assisted living settings than jail diversion consumers. That is, the former group had higher rates of schizophrenia and receipt of SSI and SSDI. Thus prison reentry FACT teams might need to focus more on increasing independence, building skills for activities of daily living, providing education about mental illness and the importance of adhering to medication, and identifying early signs of decompensation (crisis management). It could be that a more casual rather than assertive or coercive engagement approach is warranted, or teams may run the risk of recreating the power inequalities and rigid environments experienced by these consumers during imprisonment, which could inhibit engagement and stunt the development of therapeutic rapport. Also, connection to benefits, housing, and primary care to address chronic general medical problems should be a fundamental part of a reentry strategy for persons with severe mental illness.

Jail diversion consumers appear to be a younger, less chronic population that likely has developed well-honed street survival skills, which suggests that jail diversion FACT staff should focus on moving consumers from street survival to more effective community living. The success of outreach and engagement strategies could depend on finding the right hook (for example, shelter, financial assistance, transportation, or food vouchers). Moreover, intervention strategies should focus on problem solving and symptom recognition; stagewise harm reduction approaches to substance abuse; accessing general medical care, especially to address women’s health issues; using motivational interviewing to engage consumers; and regulating emotions. Staff should also recognize that the population’s lack of chronicity may mean that benefits are harder to obtain, which introduces a perverse incentive to prematurely end assertive engagement efforts, particularly for teams that are dependent on billing Medicaid for services.

In addition, among jail diversion consumers, the findings appeared to indicate that there are two somewhat distinct caseloads: males with schizophrenia and substance use disorders who are homeless and noncompliant and female consumers with mood disorders and histories of sexual abuse and interpersonal violence. For female consumers, jail diversion FACT teams should focus on managing mood disorders and addressing trauma and interpersonal violence, whereas strategies to facilitate compliance, reduce substance use, and improve housing stability are particularly important for male consumers.

This is the first comparative study of relatively large samples of jail diversion and prison reentry FACT consumers. The generalizability of findings from a small, local study is unclear, and the limitations of using agency data are well known. Data errors may have affected results in unknown ways; however, it is unlikely that data issues affected results differentially given uniform data collection protocols among FACT staff from the same agency. Currently, these are the best available data to answer important questions about jail diversion and prison reentry FACT consumers. Nevertheless, this study should be replicated across other settings.

Jail diversion and prison reentry consumers present different demographic and clinical profiles that warrant tailored services from FACT teams.

This study was supported by research funding from the Health Foundation of Greater Cincinnati and from the Ohio Department of Mental Health (grant 08-1241). The authors are grateful to Shirley Richards for assistance with data programming.

The authors report no competing interests.