Persons with severe mental disorders are overrepresented in jails and prisons in the United States. Studies by Teplin and colleagues (1,2), survey data from the Bureau of Justice Statistics (3), and a review by Lamb and Weinberger (4) suggest that the prevalence of severe mental disorders in correctional facilities ranges between 6 percent and 16 percent. These rates are significantly higher than the rate of 2.8 percent in the general population (5). The recent Criminal Justice/Mental Health Consensus Project (6), the Substance Abuse and Mental Health Services Administration (SAMHSA) (7,8), and national advocacy organizations (9,10) have expressed concern about the problem and have called for effective strategies to address it.
Jail diversion is currently a predominant approach to preventing unnecessary arrest and incarceration of persons with severe mental illness. This approach encompasses a wide range of strategies that are positioned primarily within the criminal justice system (11), including specialized police teams, mental health courts, and pretrial service agencies (12,13). Because these strategies are designed to prevent incarceration by diverting high-risk individuals to treatment, their effectiveness is likely to depend on the availability of appropriate services in the community (8). Despite the importance of access to treatment and support services, many diversion programs lack effective linkages to community-based care. In a national survey of jail diversion programs, few programs had specific procedures for following up diverted detainees or for ensuring that initial linkages to treatment were maintained (13). In addition, in a national survey of probation and parole agency directors, 82 percent of the directors indicated a need for improved access to mental health services and professionals (14).
Assertive community treatment was developed to help persons with severe mental illness who are at risk of homelessness and hospitalization become integrated into their communities (15,16,17). This treatment modality engages high-risk individuals in care by using mobile services that are available around the clock and by performing active outreach. Engagement is further promoted through delivery of comprehensive services, including mental health and addiction treatment, transportation, financial services, and vocational support. Although assertive community treatment has been shown to be effective at reducing hospital use and promoting community tenure, most studies have shown little effect on rates of arrest and incarceration (18). In a recent review of controlled studies examining assertive community treatment's impact on jail and arrest rates, Bond and colleagues (19) found that 70 percent of studies showed no effect, and 10 percent showed worsening.
If jail diversion requires access to treatment and assertive community treatment engages high-risk individuals in care, then combining these models should produce synergistic effects. An example of such a combined approach is Project Link, an assertive community treatment-based program established in 1995 to prevent arrest and incarceration of adults with severe mental illness in Rochester, New York (20,21). Described as a "comprehensive diversion approach" by the Bazelon Center for Mental Health Law (22), Project Link differs from typical assertive community treatment programs in a number of ways. These differences include its requirement of a history of arrest for admission, its use of jail as the primary referral source, its close partnership with multiple criminal justice agencies to divert clients from further involvement with the criminal justice system, and its incorporation of residentially based addiction treatment. Research has suggested that this program may be effective at reducing rates of arrest, incarceration, and hospitalization as well as improving community adjustment (23,24).
Studies have recently been published of other assertive community treatment programs that have been modified to treat mentally ill offenders (25,26,27,28,29,30). Despite the development of Project Link and other programs over the past decade, there has been a paucity of controlled research. Published reports have consistently shown these programs to be effective at reducing arrest and incarceration rates, but most of the studies have been naturalistic and were conducted without comparison groups or randomization. In addition, studies have not indicated the extent to which mean differences in service use may have been due to outliers in the study samples. Lack of information about outliers limits the ability to assess the effectiveness of particular programs or to compare outcomes. Also, basic descriptive studies that delineate the structural and functional elements of these programs have yet to be conducted.
The goal of the study reported here was to identify and describe assertive community treatment programs that have been modified to prevent recidivism among adults with severe mental illness who have been involved with the criminal justice system. On the basis of the results of a national survey, we propose that a new model for preventing recidivism called forensic assertive community treatment (FACT) is beginning to emerge.
A two-phase survey was conducted between July 2002 and December 2003 in the department of psychiatry of the University of Rochester Medical Center. The study was approved by the university's research subject review board. Phase 1 consisted of an electronic survey of members of the National Association of County Behavioral Health Directors (NACBHD), a nonprofit membership organization. NACBHD encompasses states where county authorities are mandated to oversee the planning and delivery of mental health, developmental disability, and substance abuse services.
All 314 NACBHD members from 28 states and the District of Columbia, including all organizational, state association, and associate members, were e-mailed a Web-based survey. The survey asked for contact information about assertive community treatment programs in the members' regions that met two broad screening criteria: the programs served severely mentally ill adults with histories of arrest and incarceration, and the programs worked in close coordination with the criminal justice system. NACBHD members who did not respond to the initial e-mail request were sent two reminder e-mails, followed by telephone contact.
During the second phase, the person who was identified as being in charge of each program was contacted by telephone and administered a survey requesting detailed information about the design and operation of the program. Program fidelity was briefly assessed by using five criteria from the Dartmouth Assertive Community Treatment Scale (DACTS) (31): in vivo service delivery, a staff-to-client ratio of at least 1:10, a psychiatrist-to-client ratio of at least 1:100, 24-hour availability for crises, and time-unlimited services. Programs were required to meet at least four of the five DACTS criteria to qualify as assertive community treatment programs. Telephone interviews required approximately 45 minutes to complete and were conducted Monday through Friday between 9 a.m. and 5 p.m., eastern time. The telephone survey instrument is available from the authors on request.
Assertive community treatment programs were selected that met three criteria. First, client history of involvement with the criminal justice system was an admission requirement. Second, a criminal justice agency was the primary source of referrals. Third, the program worked in close partnership with a criminal justice agency to perform jail diversion. Identified representatives from each selected program were subsequently asked to review written summaries of information gathered about their respective programs to ensure accuracy. Verified survey data were numerically coded and entered into a Microsoft Excel database for analysis.
A total of 291 NACBHD members (93 percent) responded to the survey. Of these, 98 (34 percent) identified programs that met phase 1 screening criteria. Of the 98 programs identified, 16 programs in nine states subsequently met the DACTS criteria and the study inclusion criteria and are listed in t1. Approximately two-thirds of all programs had begun operations since 1999. Although all the programs received funding through Medicaid or other sources of billable revenues, all received additional funding through grants, contract sources, or both. Major funding sources were the California Board of Corrections Mentally Ill Offender Crime Reduction Grant (MIOCRG) program (eight of the 16 programs) and other state health authorities (five programs). Private foundations and SAMHSA's Center for Mental Health Services (CMHS) each funded two programs. The primary client referral sources for 13 programs (81 percent) were local jails. The most common secondary sources of referrals involved various parts of the court system (five programs, or 31 percent). Eight programs (50 percent) accepted clients under involuntary outpatient treatment statutes. Referral sources, admission requirements, and program capacities are listed in t2. Program partnerships with criminal justice agencies and additional program characteristics are summarized in t3 and t4, respectively.
Eight programs (50 percent) had supervised residential components that were incorporated either as part of their programs or through special service contracts with residential providers. Five of these programs provided residentially based addiction treatment. Eleven programs (69 percent) incorporated probation officers as members of their assertive community treatment teams. These officers provided probation services to all enrollees served by the assertive community treatment teams who were on probation, and they actively collaborated with team members around management of those individuals. Among the 16 programs, a mean±SD of 32±25.7 percent of all team members providing direct care were African American, Hispanic, or from other racial or ethnic minority groups. The most common deviation among the measures of assertive community treatment fidelity was inadequate availability of a psychiatrist, noted for five programs (31 percent). Twelve programs (75 percent) reported having an advisory or oversight board with mental health and criminal justice representatives.
The mean level of enrollment in the programs at the time of interview was 53±30 clients; maximum capacity averaged 63±29 clients. Of clients in all programs, a mean of 69±11 percent were men; 56±22 percent had a diagnosis of schizophrenia or schizoaffective disorder, and 21±10 percent had a diagnosis of bipolar disorder. A mean of 89±12 percent of clients had co-occurring substance use disorders, and 52±35 percent of all clients were homeless at the time of enrollment. A mean of 49±29 percent of clients were African American, Hispanic, or from other racial or ethnic minority groups. A mean of 64±32 percent had previous felony convictions, and 37±26 percent had histories of committing violent crimes. A mean of 55±39 percent of clients were on probation at the time of enrollment in the program. Five of the 16 programs accepted patients who were on parole at the time of enrollment.
Three programs reported that they had published outcome data in academic journals. Cimino and Jennings (25) reported on the first 18 patients treated in the Arkansas Partnership Program. Seventeen patients had remained arrest free and without substance abuse while living in the community an average of 508 days. In a study comparing outcomes among 41 patients during the year before and after enrollment in Project Link, the mean number of jail days per patient dropped from 107.7±133.5 to 46.4±83.7 (p<.01, two-tailed Wilcoxon test) (23). Significant reductions were also noted in the number of arrests and hospitalizations, along with improved community functioning as measured with the Multnomah Community Ability Scale (MCAS) (32,33). The mean MCAS scores improved from 51.5±7.6 to 61.5±8.6 (p<.001, two-tailed Wilcoxon test) during the first year in the program. In a study of the first 30 patients enrolled in the Thresholds Jail Project, the total number of jail days dropped from 2,741 in the previous year to 469 during the first year of enrollment (26). The total number of hospital days dropped from 2,153 to 321 for the group. Total savings in jail costs during the one-year study period was $157,000, and total savings in hospital costs was $917,000.
This study identified a group of 16 promising programs with similar target populations, system coordination, and service elements that have been implemented since 1991. Although a well-defined and replicable model has yet to be developed, the observed convergence in approach can be understood as representing an early stage of the development process. The proposal that a new approach to care called forensic assertive community treatment (FACT) is emerging raises several basic questions that are only partially addressed by data from this study. Does FACT differ from assertive community treatment in substantive and meaningful ways? What are the core elements of the FACT approach? Who should FACT programs treat?
On the basis of study findings and experience with Project Link, we suggest that the primary distinction between FACT and standard assertive community treatment programs lies in the extent to which the goals of preventing arrest and incarceration determine program structure and function. Although assertive community treatment teams often treat patients who have criminal histories and interface with criminal justice agencies, these activities are undertaken more by necessity than by design. FACT prioritizes the treatment of mentally ill offenders, as evidenced by its requirement that clients have a criminal history and the predominance of criminal justice agencies as primary referral sources. Beyond interfacing with criminal justice agencies on an as-needed basis, the 16 programs identified in this study are developing integrated mental health and criminal justice service systems. An example is the incorporation of probation officers as team members by 69 percent of identified programs. In addition to promoting effective communication, such integration may be strategically important in preventing unnecessary incarceration, because it can facilitate the use of legal leverage to promote treatment adherence when necessary.
Assertive community treatment's client-centered nature and FACT's use of legal leverage may appear contradictory in terms of treatment philosophy. However, assertive outreach and legal leverage can be understood as existing on a continuum of interventions that may be necessary to promote engagement. Decisions about the relative use of these approaches may remain client centered if guided by an understanding of each client's needs, strengths, and goals and by the principle of using the least restrictive alternative. Of note, close treatment monitoring by probation personnel has been associated with increased incarceration rates in other treatment venues (34,35). To prevent this outcome, FACT should be implemented as an intensive treatment and support program rather than a coercive extension of probation as much as possible (36).
An additional distinction between FACT and assertive community treatment relates to the type of housing provided. Half the programs identified in this study had a supervised residential component, with most providing addiction treatment services. Although assertive community treatment programs routinely link patients to existing housing (17), the development and incorporation of a residential treatment component is not part of the assertive community treatment model. Such development may be critical for persons with mental illness in correctional facilities, especially those with felony convictions, histories of violence, and active psychosis.
These factors have been found to predict failed community placements shortly after release despite the involvement of a transition team that links clients to community housing (37). In addition, the reluctance of housing providers to serve high-risk individuals can be a significant barrier to obtaining existing community housing (20). The incorporation of a supervised residential treatment component in FACT programs suggests that structured housing may be necessary to promote safety and residential stability among certain mentally ill offenders. Further research is needed to clarify the purpose, target subpopulation, level of care, and types of residential programming that are most effective as FACT programs continue to develop.
The core elements of the FACT approach have yet to be fully determined. We chose three critical elements as the criteria for this study because they were integral to determining the structure and function of Project Link and because they could be reliably measured. Although these criteria enabled identification of a group of similar programs, it is important to note that the programs differed on several dimensions. A key dimension is how diversion is accomplished. Some programs are "pre-booking" in nature (engaging clients at the point of arrest), whereas others are "post-booking" (engaging clients on release from the courts or jail). Most programs have developed partnerships with multiple criminal justice agencies, but others use a small number of partnerships as their basis for diversion, as can be seen in t3. The characteristics and needs of clients who are typically served in these different diversion strategies as well as the relative effectiveness of these approaches require further study.
The programs also varied in terms of the services provided. As can be seen in t4, programs vary in the scope of residential and addiction treatment services as well as in the racial or ethnic composition of the service providers. The degree to which the demographic characteristics of service providers approximates that of service recipients may be important given that minority groups are overrepresented in correctional facilities (38,39) and because cultural and language differences represent major barriers to treatment (39). It became apparent during the interviews that the programs also varied in their level of fidelity to the assertive community treatment model on dimensions beyond those formally assessed, such as whether they included vocational specialists. Structural and organizational DACTS items were chosen to screen programs to ensure some measurable degree of consistency between programs. A small number of items were used to limit the duration of the telephone interview, although a full DACTS assessment would have enabled more detailed program descriptions. Whether particular fidelity deviations are more common among FACT teams than among assertive community treatment teams is not known.
The question of who FACT programs should treat is difficult to answer in the absence of a clearly defined model of intervention. It is noteworthy that admission criteria vary substantially among current programs. Although most programs target persons with schizophrenia and bipolar disorder, others admit clients who have a wide range of diagnoses, including primary personality disorders. Because assertive community treatment may not be effective for persons with personality disorders (40), the inclusion of these individuals in FACT programs is potentially problematic, especially when drug use and criminal behaviors are present.
The programs also varied substantially in their criminal justice admission criteria. Understanding the level of risk of criminal recidivism among clients who are admitted to FACT programs is critical to determining the model's effectiveness at preventing recidivism. In a meta-analysis of studies of criminal and violent recidivism among mentally ill offenders, Bonta and colleagues (41) found that criminal history variables were the best predictors of recidivism. These variables include adult criminal history, history of juvenile delinquency, association with criminal companions, criminal use of weapons, and presence of an antisocial personality. In a recent study of 802 adults with severe mental illness, Swanson and associates (42) found that past violent victimization, violence in the surrounding environment, and substance abuse had a cumulative association with risk of violent behavior.
Despite the likely importance of criminal history and social-environmental variables in determining the risk of criminal recidivism, none of the 16 programs we surveyed assessed these variables by using a standardized measure. One method for assessing the risk of recidivism attributable to such variables is to incorporate the use of standardized assessment tools such as the Level of Service Inventory-Revised in screening potential clients (43,44). This instrument has achieved the highest predictive validity with recidivism in the general population among available instruments (45). Additional instruments that have good predictive power include the Psychopathy Checklist-Revised (46) and the Lifestyle Criminality Screening Form (47). Although not validated and normed for use among adults with severe mental illness, such tools may help us understand which clients are most appropriate for FACT and to evaluate the effectiveness of this approach.
Implicit in FACT's design to promote engagement of clients in psychopharmacology, addiction treatment, and community support services is the notion that such interventions will reduce criminal recidivism. Although intensive services may reduce recidivism among persons who are arrested as a result of untreated psychosis, drug addiction, or homelessness, such services are probably not sufficient for everybody. Individuals with co-occurring psychopathy may also benefit from additional interventions that directly target antisocial attitudes, skills, and cognitions (48,49). Research in populations of persons who do not have a mental illness has suggested that the most effective approaches to individuals who are at a high risk of criminal recidivism incorporate highly structured cognitive-behavioral interventions (45,50,51). The FACT approach may benefit from inclusion of such strategies in managing persons who have both severe mental illness and psychopathic traits. One example of a program that is currently using these strategies is the Monterey County Supervised Treatment After Release (MCSTAR) program. In addition to standard treatments, the program incorporates cognitive-behavioral interventions, including specialized groups and courses designed to target and restructure criminal thinking.
Is "forensic assertive community treatment" an appropriate name for this approach to care? The acronym FACT is currently used for other health care models, including family-assisted assertive community treatment (52). In addition, the term "forensic" has a connotation that relates to persons who have been found not guilty by reason of insanity, guilty but mentally ill, or incompetent to stand trial. Such individuals are often treated in specialized forensic hospitals by forensic psychiatrists and other forensic specialists. By contrast, forensic assertive community treatment is designed to engage adults with severe mental illness in community-based care.
Forensic specialists can play a key role as FACT clinicians because of their special knowledge of the criminal justice system and relevant case law. However, although clinicians in the surveyed programs were knowledgeable about criminal justice and mental health systems, few were forensic specialists. Formal forensic training is helpful in bridging these systems, but it may not be necessary as long as clinicians are comfortable and familiar with both systems. Forensic assertive community treatment was chosen as a name for this approach to care because of the criminal justice system involvement of the target population and the criminal justice identity of its referral sources and partnerships. In addition, some programs surveyed have already begun using the FACT designation to identify themselves.
Several limitations of this study must be recognized. Because the survey covered only the 28 states represented by NACBHD, the actual number of such programs operating across the country is probably higher than indicated here. However, the goal of this study was to describe the early emergence of a promising model of care rather than to create a directory of existing programs. Although county behavioral health directors are generally knowledgeable about programs in their jurisdictions, the NACBHD members we surveyed had a high degree of discretion in determining which programs to identify in phase 1 of the survey. Also, the use of DACTS items to screen programs in phase 2 raises significant methodologic limitations. The selected items fail to capture important aspects of assertive community treatment programs, such as their team approach and vocational interventions.
In addition, the construct validity of the DACTS rests on direct review of records and interviews of staff members that were not conducted. The reliability of survey data collected in phase 2 was also limited, because only one senior representative from each program was interviewed, and the roles and responsibilities of these respondents varied between programs. Respondents sometimes provided rough estimates in response to interview questions, and data were not validated beyond being reviewed by the respondents. Also, many of these data points will change as programs continue developing. Given the likelihood of change over time, this study is best understood as providing a snapshot of an evolving approach to care.
FACT is an emerging approach to the prevention of recidivism that incorporates both assertive community treatment and criminal justice components. Although the extent and nature of incorporation varies among programs, the blend has created a foundation for new interventions that offer enhanced community treatment as an alternative to involvement with the criminal justice system. The combination of intensive service delivery and legal leverage represents a critical balance for persons who would otherwise be left at the mercy of untreated illness, streets, and jails. Sadly, 69 percent of the program representatives surveyed were uncertain whether their programs would continue after their current grants and contracts expired. Further work is necessary to develop funding streams to support the continued development of this promising model of care.
Dr. Lamberti and Dr. Weisman are affiliated with the department of psychiatry of the University of Rochester Medical Center, 300 Crittenden Boulevard, Rochester, New York 14642 (e-mail, firstname.lastname@example.org). Ms. Faden is with the National Institutes of Health in Bethesda, Maryland.
Forensic assertive community treatment (FACT) programs that met the FACT study criteria
Program referral sources, criminal justice admission requirements, and capacity of forensic assertive community treatment(FACT) programs
Criminal justice system partnerships
Additional characteristics of forensic assertive community treatment (FACT) programs