حجم مشروط تخصصی بهداشت روانی: آیا آنها موثر باشد؟

With the large and growing number of persons with mental illnesses under probation supervision, a form of specialized probation called specialized mental health caseload (SMHC) has been implemented. This study explores the effectiveness of a prototypic SMHC implemented statewide. A quasi-experimental design was used to compare criminal justice, mental health, and community engagement outcomes among three caseloads: a newly established SMHC supervising no more than 30 clients per officer (N = 1367); an established SMHC supervising roughly 50 clients per officer (N = 495); and a traditional caseload of clients receiving mental health treatment and supervised by officers with average caseloads of over 130 clients (N = 5453). Using a mixed methods approach, we found that the SMHC was implemented with high adherence to fidelity, and comparisons based on different caseload samples generally support the effectiveness of the specialized mental health caseload, particularly on criminal justice outcomes. Future studies using random assignment are needed to examine the connection among mental health symptoms, compliance with treatment and probation supervision, and recidivism.

Over the past 20 years, specialized interventions for justice-involved persons with mental illnesses have proliferated (Epperson et al., 2014). The majority of these interventions were implemented at the front-end of the criminal justice system — police, courts, and jail, with the intention of diverting people with mental illnesses to the mental health system (Sirotich, 2009). The design and development pattern of specialized interventions reflect the now-contested presumption, most commonly referred to as the criminalization hypothesis (Abramson, 1972), that individuals with mental illnesses were involved in the criminal justice system primarily because of symptoms related to un- or under-treated mental illness (Epperson et al., 2014 and Skeem et al., 2011). For this reason, intervention development and dissemination waned in other areas of the criminal justice system, particularly probation supervision. While presumption bias historically hampered development, the facts clearly support the need for specialized interventions at the back-end of the criminal justice system. Three facts are relevant here. First, the bulk of criminal justice supervision occurs in the area of community supervision and, more specifically, under the responsibility of probation. In 2011, of the 7.1 million people under correctional supervision in the US, 4.9 million (66.8%) were supervised in the community and the majority (4.1 million or 82.9%) by the probation service (Glaze, 2011). Second, mental disorder is over-represented among justice-involved persons. Approximately 1 in 7 people involved in the criminal justice system in the US is estimated to have serious mental illnesses (SMI) — major depression, bipolar, schizophrenia (Fazel and Danesh, 2002 and Steadman et al., 2009), nearly three times the rate found in the general population (NIMH, 2010). Taken together, on an average day, approximately 574,000 people with SMI in the US are under probation supervision. In addition to SMI, these individuals are very likely to have a co-occurring substance abuse problem (Lurigio et al., 2003, Roskes and Feldman, 1999, Sirdifield, 2012, Teplin, 1994 and Teplin et al., 1996), further complicating their supervision. And, third, while under probation supervision, people with SMI, compared to those without SMI, are more likely to fail — that is, have their community supervision term revoked because of a violation of the special conditions of probation (i.e., technical violation) or a new offense (Cloyes et al., 2010, Dauphinot, 1996, Eno Louden and Skeem, 2011 and Porporino and Motiuk, 1995). Failure on probation pulls people with SMI deeper into the criminal justice system, further disrupting their already precarious living situations (Draine et al., 2002 and Fisher et al., 2006) and fragile community ties (Draine et al., 2005 and Skeem et al., 2009). In short, the stated aims of probation, namely rehabilitation and providing an alternative to incarceration, are less likely to be achieved among persons with SMI. As early as 1994, attention was drawn to a perfect storm within community corrections. Veysey (1994) rang the alarm, describing a parole system that under-identified and underserved clients with SMI in large part because most parole officers were ineffective in responding to their unique set of needs. This was followed two years later by a call for a specialized response to people with SMI within community corrections. In a monograph entitled Community Corrections in America: New Directions and Sounder Investments for Persons with Mental Illness and Codisorders ( Lurigio, 1996), the collective of authors recommended: (1) specialized units with smaller and exclusive caseloads for offenders with SMI ( Latessa, 1996); (2) specialized training in mental illness for officers ( Clear, 1996); (3) ongoing specialized training ( Latessa, 1996) and cross-training ( Veysey, 1996); (4) boundary spanning — collaboration among systems ( Corbett, 1996, Davidson, 1996 and Veysey, 1996); (5) services integration ( Veysey, 1996); and (6) normalization of punitiveness — supervision that was no more punitive for clients with SMI who are supervised more intensively. It was further advanced that the “purpose of correctional services for offenders with mental illness should always be to maximize their potential for living and functioning effectively in the community” ( Clear, 1996, p. 16). A philosophical shift from exclusively punishment and social control to conjointly include (re)habilitation and community engagement was seen as key to effectively supervise clients with SMI. Community corrections remained largely unchanged, with responses to clients with SMI repeatedly deemed “woefully inadequate,” although there was evidence of embryonic development of specialized efforts in Cook County Illinois, California, Baltimore, Maryland (Lurigio, 2001 and Lurigio and Swartz, 2000) and Ohio (Latessa, 1996). In 2002, reform was recommended again, this time by the Council of State Governments (2002), which proposed a more systematic and focused response to clients under community supervision who have SMI; a response that targeted the special needs and issues of clients with SMI that were hindering their compliance with the special conditions of supervision. The recommended venue was specialized mental health caseloads (also referred to as specialized probation) — comprised of smaller, exclusive caseloads supervised by officers with special training in mental health and related areas of needs. Smaller more specialized caseloads managed by expert officers were expected to be more effective in securing needed resources (treatment, as well as social, housing, and public benefits) through advocacy and collaboration with collateral community agencies; working with mental health providers and clients towards goals of treatment compliance; and engaging clients through styles and patterns of interpersonal interactions that would increase compliance with the general and special conditions of supervision and, consequently, stabilize community living and promote public safety. With these recommendations in mind, and given the dearth of information about specialized probation practices, a national survey of specialized probation agencies was conducted (Skeem, Emke-Francis, & Eno Louden, 2006). Of the 137 probation agencies found nationwide with a dedicated response to clients with mental illnesses, only 73 were “pure” forms of supervision, i.e., did not mix clients (e.g., sex offenders or other special needs groups), or comprised the efforts of a single officer. The defining characteristics of “prototypic” specialized probation were narrowed to five: (1) caseloads inclusive only of clients with mental illnesses; (2) smaller than average caseloads (average caseload of 43 to 45, compared to the normal caseload of over 100 for traditional officers); (3) specialized and continuous training of officers; (4) collaboration with and integration of internal and external resources; and (5) use of problem-solving methods to engage clients in compliance with special conditions (Skeem et al., 2006). These characteristics reflect the recommendations within the Community Corrections in America monograph and the report by the Council of State Governments, although neither of these sets of recommendations was based on research, but rather, on an understanding of the special needs of people with SMI who were failing at community supervision and needing help “to maximize their potential for living and functioning effectively in the community” ( Clear, 1996, p. 16). What remains unclear is whether these specialized caseloads work to achieve the twin-goals of enhanced compliance and improved community living, as research on specialized probation for persons with SMI is quite limited. Key characteristics and processes of specialized probation units have been studied only in recent years (Eno Louden et al., 2008, Eno Louden et al., 2012, Skeem et al., 2003 and Skeem et al., 2007). Evaluation of specialized probation targeting people with SMI is in its infancy, and existing evidence to date bears numerous empirical limitations. Latessa (1996), using administrative data and a non-experimental comparison design, compared probation outcomes among clients from five specialized units: mental illness (supervised either by a specialized officer or unit); sex offender unit; drug offender unit; high risk offender unit; and traditional unit (i.e., those supervised by standard probation), with an average time under supervision of 13.5 months (mentally disordered group) to 18 months (sex offender group). Over the period of observation, clients with SMI had more officer contacts per month (4.1) than those clients under traditional supervision (1.3) but less than those in the drug offender group (5.1); were less likely to be arrested (26%) compared to all groups (ranging from 43 to 32%) except the traditional group (25%); were roughly equally likely to be charged with a technical violation (44%) compared to 41 to 45% for the other groups except for the sex offender group (23%); were more likely to be deemed successful on probation (63%), compared to the drug (41%) and high risk (46%) groups, but less successful than the sex offender group (78%). One of the first evaluations of specialized probation was conducted by Roskes and Feldman (1999), using a pre-post-study of 16 probationers with SMI. They found reduced rates of violation for the 13 probationers who remained on a specialized program. However, given non-random sample selection and the fact that 3 of the 16 participants were removed from the program due to non-compliance, this study's finding is quite preliminary and not generalizable. A second, larger study by Burke and Keaton (2004) involved random assignment of probationers with SMI to specialty probation and case management (n = 225) versus traditional probation (n = 224). Although the study demonstrated greater access to mental health services and lower likelihood of re-arrest for those on specialized probation, only 58% of the specialty probation sample completed the program, and “non-completers” were excluded from outcomes analyses, whereas non-completers of traditional probation were not excluded. Additionally, the follow-up period was limited (six months), and the specialized probation with case management was quite costly, more approximating assertive community treatment, and not likely to be replicated by probation jurisdictions (Skeem & Eno Louden, 2006). An unpublished study cited by Skeem and Eno Louden (2006) involved 800 probationers with mental illnesses randomized to either specialized probation, traditional probation, or no probation with or without additional treatment. Initial results indicate that those on specialty probation received more mental health services, but were equally as likely to return to jail as those on traditional probation or case management services. However, because this study is as yet unpublished, important aspects of study design, such as measurement instruments, follow-up period, and use of control groups, are unknown. A retrospective, non-randomized study based on three years of arrest data of 241 probationers with mental illnesses participating in a specialized probation unit located in Chicago demonstrated reduced arrest rates comparing the period prior to and after specialized probation (Jesse, Bishop, Thomas, & Dudish-Poulsen, 2008). While all these studies suggest that specialized probation may hold promise in producing positive mental health and criminal justice outcomes for probationers with SMI (Skeem & Eno Louden, 2006), these studies have significant limitations: a less than ideal comparison group based on a pre-post-non-randomized design or the exclusion of non-completers from the experimental group analyses. None of these evaluation studies addressed issues of selection bias or program fidelity, which limits the interpretation and generalizability of study findings. The present study explores the effectiveness of a prototypic specialty mental health caseload (SMHC) implemented statewide. A quasi-experimental design was used to compare supervision, mental health, and community engagement outcomes among three caseloads: a newly established SMHC supervising no more than 30 clients per officer; an established SMHC supervising roughly 50 clients per officer; and traditional caseloads of clients receiving mental health treatment (diagnosis unknown) and supervised by officers with average caseloads of over 130 clients. Data for these comparisons are drawn from administrative data sets. Additional self-report data were obtained from a research sample of clients enrolled in the newly established SMHC and are used to verify findings from inter-group comparisons. For this study, within the Oxford Centre for Evidence-based Medicine's evidence hierarchy, we are using a Level 2b design: quasi-experimental design — non-equivalent comparison group (Oxford Centre, 2012), the most common design used to study specialized interventions targeting justice-involved persons with mental illnesses (Sirotich, 2009). Issues of program fidelity and selection bias are explored as well. Our study, using both administrative and self-report data, tests three hypotheses: