Adapting Evidence-Based Practices for Persons With Mental Illness Involved With the Criminal Justice System

The prevalence of high rates of co-occurring mental and substance-related disorders within the general population has been well documented ( 5, 6 ). Even the highest estimates of co-occurring disorders by condition in the general population are low compared with best data about the prevalence of co-occurring disorders in jails and prisons. Abram and Teplin ( 11 ) found that about 8% of the jail inmates have a current serious mental illness and that 72% of both male and female jail detainees with serious mental illness had a co-occurring substance use disorder. In one study, almost one-quarter of veterans with co-occurring disorders released from inpatient facilities were incarcerated within 12 months of discharge ( 12 ), and rates of incarceration are generally higher for persons with co-occurring disorders, compared with those with mental illness alone ( 13 ).

In 2000 nonwhites constituted 25% of the general U.S. population but 62% of the prison population and 57% of the jail population ( 14 ). In a study of nonwhite persons with mental illness involved in the criminal justice system, inadequate attention to cultural issues was reported at every point of contact with the justice system ( 15 ).

Finding and keeping a job after incarceration is critical to address basic needs and to structure daily life for persons with mental illnesses. Research has shown that persons with serious mental illness want to work and can work, but as a group they are typically unemployed. Only 10%–20% of persons with serious mental illness are currently working ( 16 ). Persons released from incarceration compete with other low-income and low-education workers for low-wage jobs, yet they face policies and discriminatory hiring practices that make it more difficult to succeed. Some research supports vocational training and employment as antidotes to reincarceration ( 17 ), but little specific research has addressed this dimension among persons with serious mental illness.

Persons with mental illness who are homeless are at high risk of incarceration, and when incarcerated, they stay in custody for longer periods ( 18 ). They are more likely to come in contact with law enforcement officers and be charged with a crime ( 19 ). In jails 30.3% of inmates with mental illnesses were homeless in the year before arrest, compared with 17.3% of other inmates ( 20 ). Not having a home upon release from jail or prison also increases the risk of rearrest.

Critical to contemplating effective interventions for persons involved in the criminal justice system is an awareness of the potential for them to have an extensive history of trauma before incarceration. Women are a fast-growing segment of arrestees and currently account for 12.3% of the U.S. jail population ( 21 ). A majority of these women are persons of color, in their 20s and 30s, and unmarried; seven of ten of these women have dependent children in the community; and a majority of these women have a history of irregular education and employment ( 22 ). They have experienced high rates of childhood and adult physical and sexual abuse ( 3, 22, 23 ). In 2002 a history of physical or sexual abuse was reported by 55% of women in jail ( 23 ), and 27% reported rape ( 24 ). Women with a history of abuse have an increased risk of alcohol and substance abuse ( 22 ).

Two aspects that clearly distinguish persons with mental illness involved in the criminal justice system from those without such a history are high rates of criminal behavior and experiences within jail and prison settings. Although persons with mental illness are not necessarily more violent than those without such an illness ( 25 ) and although many arrests of persons with mental illnesses are for misdemeanors associated with "crimes of survival" (that is, panhandling or urinating in public), the association between comorbid antisocial personality disorders and high rates of conviction for criminal offenses creates a cohort of individuals with unique needs. "Criminogenic thinking" is the criminal justice term used to convey cognitive differences between criminals and law-abiding citizens; having a mental illness does not immunize one from these patterns of thought. Second, people who spend extended periods of time in custody are exposed to a different culture and experience stressful periods that include threats to their well-being and periods of isolation. In response to these circumstances, persons with mental illnesses may develop adaptive behaviors that interfere with subsequent community adjustment ( 26 ).

Coercion is a consideration in the application of all evidence-based practices to persons involved in the criminal justice system. First, measurable public safety outcomes include abstinence from all substances of abuse, adherence to prescribed medications, and participation in mental health and substance abuse treatment groups. These objective conditions are often written into terms and conditions of release, and the failure to achieve these standards can be used as a justification for reincarceration. Second, for an individual who has not been in treatment, incarceration represents an opportunity to engage that person, mandate treatment interventions that can reduce high-risk behavior, optimize other treatment strategies (for example, medication management) in a controlled setting, and closely monitor the person's behavior. The effectiveness of coercive strategies for persons with co-occurring disorders who are involved in the criminal justice system is essentially unknown ( 27 ).

From the perspective of the person involved in the criminal justice system, the perception of coercion may be more important than the extent to which the coercion is applied ( 28 ). The Institute of Medicine suggests that when coercion is legally authorized, patient-centered care is still applicable and client decision making should be maximized by involvement in the selection of treatments and providers ( 29 ).

If desired public health and public safety outcomes are to be achieved, these clinical, social, and legal features of persons with mental illness involved in the criminal justice system must be addressed. What evidence exists to show that current practices can achieve these objectives? Are the evidence-based practices that have been developed to generate positive behavioral health outcomes also associated with reductions in criminal activity, rearrest, and reincarceration?

The documented high prevalence rates of co-occurring mental and substance use disorders among persons involved in the criminal justice system demands an inspection of effective interventions to reduce these illicit and destabilizing behaviors. For persons with co-occurring disorders who are not involved in the criminal justice system, integrated treatment has been identified as an evidence-based practice, and its core components have been articulated ( 30, 31, 32 ).

For persons involved in the criminal justice system, the hypothesis underpinning effective interventions for co-occurring disorders can be stated as the following: interventions (at the program or provider level) that reduce substance use (licit and illicit) and improve levels of functioning among persons with co-occurring disorders will reduce both the frequency of involvement with the justice system and the time spent in justice settings or under correctional supervision. The outcomes sought are reduced criminal activity (specifically the use of illegal drugs and violent behavior), fewer persons with co-occurring disorders at all points in the justice system, and improved reintegration of offenders with co-occurring disorders into community settings. Specific treatments within integrated programs include psychopharmacologic strategies ( 33 ), motivational interventions ( 34 ), and cognitive-behavioral interventions ( 32 ). Several specific program models that use integrated treatments have been applied to persons with co-occurring disorders who are involved in the criminal justice system.

First, the integrated dual disorder treatment model is an evidence-based practice that combines program components and treatment elements to ensure that persons with co-occurring disorders receive combined treatment for substance abuse and mental illness from the same team of providers ( 35 ). Although routinely applied to persons with serious mental illness involved in the criminal justice system, the model has not been specifically studied for its impact on criminal justice outcomes.

Second, the modified therapeutic community is a residential treatment program for persons with co-occurring disorders that uses integrated strategies that have been studied in populations involved in the criminal justice system ( 36 ). Modified therapeutic communities use the "community as method" as the basis for both the program and treatment integration. Compared with nonintegrated services, modified therapeutic communities have been shown to significantly lower reincarceration rates and reduce harmful substance use among persons with co-occurring disorders ( 37 ).

Finally, ACT is reviewed below, but it should be noted that when the model is implemented with high fidelity, it achieves positive outcomes for persons with co-occurring disorders, although its impact on public safety outcomes remains to be established.

Although some services integration models have data to support their effectiveness as evidence-based practices for persons with co-occurring disorders who are involved in the criminal justice system, more specific tests are needed of integrated treatment for cohorts involved in the criminal justice system.

The housing situations for persons with mental illness who have contact with the criminal justice system range from owning their own homes or living in independent rental units to institutional care. It is also the case that housing type for any individual will vary over time. There has been a concerted focus on the utility of supportive housing for individuals with a history of mental illness and housing instability. Supportive housing includes a variety of permanent housing settings coupled with on-site or easily accessible services. Services include case management, counseling, medical care, mental health and substance abuse treatment, vocational training, cognitive skills groups, and assistance in obtaining income supports and entitlements. These models have proven effective in achieving positive residential outcomes ( 38 ).

No single model or package of services has emerged specifically to persons with mental illness involved in the criminal justice system. Furthermore, evaluation of the effectiveness of different approaches has not employed rigorous methodology, and measured outcomes vary widely. Culhane and colleagues ( 39 ) found that individuals with mental illness at risk of homelessness who were placed in supportive housing options in New York City had reductions in shelter use and hospitalizations and spent fewer days in jail in the year after attaining housing, compared with the year before attaining housing. In addition, compared with a matched group of New York residents without supportive housing, they had fewer episodes of incarceration. The modified therapeutic community model mentioned above is another example of a supportive housing model producing positive behavioral health and criminal justice outcomes.

At the outset of the meeting, it was acknowledged that trauma interventions were not at the level of an evidence-based practice, although they are a promising practice for which there is the beginning of a persuasive database. The topic was included in the evidence-based practice series because of the exceedingly high relevance of this issue for persons with mental illness involved in the criminal justice system.

The high rates of trauma exposure among persons, particularly women, with mental illness who are involved in the criminal justice system require a systemic response that views trauma history as the norm rather than an exception. As an evidence-based practice for persons involved in the criminal justice system, a trauma-specific intervention assumes that effectively dealing with symptoms of trauma will reduce the likelihood of subsequent criminal activity. Although a number of effective trauma interventions for the general population with posttraumatic stress disorder (PTSD) have been described, applications for persons involved in the criminal justice system are not common. Zlotnik and colleagues ( 40 ) applied a well-established cognitive-behavioral model for the treatment of PTSD ( 41 ) to a cohort of incarcerated women with PTSD and substance dependence and found positive outcomes consistent with its application among nonincarcerated women. Valentine and Smith ( 42 ) used traumatic incident reduction in a sample of female prisoners and found significant reductions in trauma-related symptoms. However, although these and other small-scale studies suggest that trauma-specific interventions for persons involved in the criminal justice system can positively affect trauma symptoms, no evidence exists that these interventions reduce rates of rearrest and jail utilization.

If some criminal activity is driven by the need for money, successful employment may mitigate subsequent contact with the criminal justice system. Over the past 15 years supported employment has emerged as an evidence-based practice that can improve the success of persons with serious mental illness in competitive employment circumstances ( 43, 44 ). In a multisite, randomized clinical trial, Cook and colleagues ( 45 ) showed that participants in a supported employment program were twice as likely as persons who were enrolled in traditional vocational programming to achieve competitive employment. Although there are limited data that suggest supported employment is effective among persons with mental illness, regardless of whether or not they are involved in the criminal justice system ( 46 ), there are no data on the impact of supported employment on criminal justice outcomes.

There is face validity to the concept that improving employment outcomes for persons involved in the criminal justice system can reduce criminal justice contact. Programs may need to be modified to accommodate conditions of release, which can require court or community correction monitoring concurrent with work hours.

Illness management and recovery are evidence-based practices that teach persons with serious mental illness the required skills to manage their own illnesses in collaboration with health care professionals and other natural supports. The application of these evidence-based practices has been demonstrated to prevent relapse and rehospitalization in addition to reducing the disabling effects of mental health symptoms ( 47 ). There are several illness management and recovery program types—for example, the wellness recovery and action plan and the social and independent living skills program—that share practice principles while employing different approaches. These programs have been implemented for persons with mental illness within correctional settings and appear to produce the expected gains in social skills. There is no evidence concerning the effect of illness management and recovery on public safety outcomes.

The common application of psychoeducational and cognitive components within illness management and recovery programs makes them well suited for adaptations that could address unique aspects of persons involved in the criminal justice system. Addressing criminogenic thinking or antisocial tendencies within an illness management and recovery framework could specifically target behaviors that lead to criminal justice contact. Alternatively, using motivational enhancements for consumers in illness management and recovery programs could reduce perceived coercion and may offset any potential negative impact of mandated treatment.

ACT is a well-documented evidence-based practice that combines treatment, rehabilitation, and support services within a multidisciplinary team ( 43, 44, 48 ). It is a high-intensity, high-cost service that is typically reserved for persons most disabled by mental illness and those who use multiple community acute and emergency services. The ability of ACT to consistently reduce arrests and jail times among persons with serious mental illness has not been established ( 49 ). As a result, there has been recent interest in augmenting ACT services by specifically focusing on populations involved with the criminal justice system and training team members to be responsive to criminal justice partners ( 10 ). These forensic ACT teams, or FACT teams, have been implemented in many communities around the country. Some demonstration studies have found significant reduction in jail days and arrests ( 50, 51 ), but these studies had no control groups and used FACT teams that made many adaptations to the original model. Adaptations to intensive case management models have also been applied to populations involved with the criminal justice system, although outcomes have been mixed ( 49 ).

The bottom line for persons with mental illness who are involved in the criminal justice system is that FACT teams are relatively new adaptations of the ACT model and come in many forms. When adhering to the core ACT model, they show promise for reducing inpatient hospitalizations. With their "criminal justice savvy" ( 49 ), they can be expected to reduce recidivism and maintain certain clients in the community. Nonetheless, they are a high-intensity, high-cost intervention that fits the most disabled segment, perhaps 20%, of the persons being diverted or released from the criminal justice system. The case management models of choice for the other 80% or so of less disabled individuals rely on brokering services from mainstream providers rather than providing all services via a FACT team. Although less costly to operate, brokered case management models are still a challenge for many communities with limited services.