Diversion programs, often called mental health diversion or jail diversion programs, have received increased attention from clinicians, administrators, and researchers in the past decade. Indeed, the number of such programs in the United States increased from an estimated 52 in the mid-1990s to nearly 300 by 2005 (1,2). The number of people served has also increased. Fifteen years ago, even as the number of programs was increasing rapidly, limited empirical data were available to guide jurisdictions in developing and implementing them. Recent studies are more helpful in this regard.

Many trace the emergence and growth of mental health diversion programs to the process of deinstitutionalization in the United States that occurred mainly in the 1960s and 1970s. The assumption that incarceration is not good for citizens with mental illness is not new. Persons with severe mental illness or developmental disabilities have long faced incarceration in jails or prisons as the alternative to hospitalization. Indeed, the creation of asylums worldwide in the 18th century and of state hospitals in the United States in the 19th century was based on the work of many who wanted to promote more humane management of persons with mental illness.

As decisions were made by states across the United States to provide mental health services in their communities rather than in large, centralized institutions, observers began to raise concerns about the "criminalization" of persons with serious mental illness (3,4). Difficulties, such as "revolving door" hospital admissions, inadequate treatment in jails, unemployment, and homelessness, were compounded because persons with serious mental illness were unnecessarily arrested and incarcerated largely because of poor symptom control in the context of a fragmented network of support in the community (5). However, Junginger and colleagues (6) have cautioned against overstating the effects of criminalization.

The difficulties associated with criminalization have become clearer both to professionals in the mental health field and to those working in the criminal justice system (7,8,9). Persons with serious mental illness are sometimes seen by jail administrators as not fitting well within criminal justice settings, and those with a criminal history may be seen by mental health administrators as inappropriate for traditional treatment settings. As a result, organizational stresses have occurred in both mental health and criminal justice agencies. Agency missions and personnel skill sets have not always matched well with unusual demands, and resources have been stretched to serve individuals whose primary needs might lie outside an agency's usual domain. The clinical needs of individuals with serious mental illness in jails or prisons are often unmet or inadequately met (10,11).

Currently, perhaps surprisingly to some, the absolute number of persons with severe mental illness is greater in American jails and prisons than in our hospitals, making jails the largest de facto institution for the mentally ill population (3,12). As U.S. rates of incarceration surged in the 1980s (13), these issues became more acute for affected clients, for leadership in criminal justice and mental health systems, and for communities. Budgetary and fiscal pressures contributed to the necessity of finding better interventions and more efficient ways to meet the needs of criminal justice system clients with serious mental illness.

In response to concerns about criminalization and unmet treatment needs in jails and prisons, many U.S. communities have developed mental health diversion programs. Even though local needs and resources have resulted in a great diversity of programs (14,15), they are typically categorized as either prebooking or postbooking diversion (1).

Prebooking interventions occur before a client's processing into jail, are often police based, and usually involve reliance on some of the tenets of crisis intervention (16,17). Often, these interventions work best when a central facility is used as a setting to resolve crisis situations, which can be rather complex and time intensive for the multidisciplinary team members (18). Postbooking diversion can be divided into prearraignment and postarraignment types or mixed types, depending on when diversion occurs in the judicial process. Postbooking diversion may also include mandated treatment. Furthermore, the goals of postbooking diversion may be quite limited (for example, to reduce the frequency or duration of a person's incarceration with little attention to subsequent treatments) or quite encompassing (for example, to encourage and monitor a client's progress in an integrated system of individualized care).

The variety and complexity of diversion programs can make classification and study challenging. Even so, progress has been made in the past decade in clarifying some of the components that contribute to the effectiveness and success of both prebooking and postbooking diversion. This review provides a critical look at the effectiveness of postbooking jail diversion by examining the literature from the past decade. The focus is on postbooking programs because adequate study of the two types of diversion requires separate methodologies, and a satisfactory treatment of both would go beyond the constraints on this article's scope and length. Many of the thorny problems affecting those who work to care for persons with severe mental illness in jails and prisons also cannot be covered in depth here.

The review was motivated primarily by an awareness of the unmet needs of criminal justice system clients with serious mental illness and the inadequate base of information related to outcomes of postbooking diversion programs. Four questions guide the discussion below. First, who are the persons whose service needs fall within both the mental health and the criminal justice systems? Second, what kinds of programs have evolved to meet their needs? Third, what progress has been made in substantiating the beneficial outcomes of these programs in their varied U.S. forms? And, finally, what factors might predict successful postbooking diversion? It is hoped that this article will help to clarify where the efforts of clinicians, researchers, and policy makers should be directed in order to contribute to the improved functioning and satisfaction of persons with serious mental illnesses involved in the criminal justice system.

The empirical studies reviewed were identified in a search of the literature on U.S. postbooking diversion programs published from 1999 to 2008. Approximately 130 articles were identified for initial review; the studies included adult criminal justice system clients with serious mental illness, either with or without substance abuse problems. The articles were found primarily by using the following keywords in PsycINFO and MEDLINE searches in January and November 2008: jail diversion, mental health diversion, serious mental illness, and criminal justice. Examination of reference lists in these articles also identified a limited number of additional relevant articles—a few from before 1999.

Persons with severe mental illness in the criminal justice system

The absolute number of clients currently served in either pre- or postbooking diversion programs is somewhat difficult to quantify, as is the number of persons who might be eligible for and thus benefit from such services. Studies have shown that persons with severe mental illness are greatly overrepresented in U.S. jails and prisons. The proportion of persons with serious mental illness in the adult U.S. population is estimated to be approximately 3% to 4%, whereas the proportion of such persons in the incarcerated population may be between 7% and 16% (19,20). It has also been estimated that between 38% and 52% of adults with serious mental illness in the United States have been arrested at least once (21,22,23). Much of the information about the population of criminal justice system clients with serious mental illness comes from published studies of jail diversion programs or reports on prison populations (10,11).

Even early studies of mental health diversion suggested that the population served is a diverse group of men and women (24,25). These clients vary in many ways—for example, in their psychiatric symptomatology, diagnoses, developmental and family experiences, and criminal histories (20). Although frequently single, many clients are married, and many have children (26). Differences in other demographic features are seen, such as in race-ethnicity, age, socioeconomic background, and urban or rural residence. These clients struggle with complex illnesses, including comorbid psychiatric, substance use, and general medical conditions, and their situations may include lack of resources or victimization (10). Although studies have typically focused on persons with diagnoses of schizophrenia, bipolar illness, or severe depression, there is good reason to consider the struggles that these clients also experience with substance abuse and chronic issues related to past traumas (27). It is estimated that as many as 59% of arrestees are under the influence of a substance when arrested; further, among persons with serious mental illness who are incarcerated, about three-quarters have a substance abuse problem (20). Finally, although the focus of this article is exclusively on adult clients, it appears that similar issues affect many juveniles (28,29,30,31).

Data illustrating the diversity of the population of criminal justice system clients with serious mental illness can be drawn from a study by Broner and colleagues (20) that was sponsored by the Substance Abuse and Mental Health Services Administration (SAMHSA). This project, which is the largest study of pre- and postbooking diversion completed to date, examined multiple diversion programs from eight states between October 1998 and May 2000. A total of 1,966 clients participated—971 in the diversion group and 995 who were not diverted. Among the participants, 69.1% were male and 30.9% were female; 44.4% were white, 40.9% were black, 9.4% were Hispanic, and 5.3% were from other racial-ethnic groups. Most (62.0%) had at least a high school diploma or GED, and 31.3% were employed in the 30 days before study entry.

Information from an earlier study (1995–1996) of a diversion program in an unnamed Midwestern city gives a similar indication of client diversity (25). Among 80 consecutive arrestees in the study (35 in the diversion group and 45 not diverted), 66.3% were black, 31.3% were white, and 2.5% were from other racial-ethnic groups. In the sample, 43.0% had less than a high school education, 31.6% had a high school diploma or equivalent, and 25.3% had at least some college. For 52.6% the most serious charge involved a crime against a person. For 15.8% it was a drug-related crime, for 23.7% a property-related crime, and for 7.9% another minor crime. Ten percent of participants were employed at the time of arrest. The mean±SD number of arrests for the sample was 17.3±19.5.

In summary, the needs of the population of criminal justice system clients with serious mental illness, which are great, are driven by the diversity of the individuals in this group, and these needs have become more obvious over time. Although these descriptive statistics provide some information about this population, they fail to capture all the dimensions along which a sample of such clients may differ. We still have an imperfect picture of the diversity of clients currently served in U.S. postbooking diversion programs.

Postbooking diversion programs

Significant efforts have been made in the United States to promote the diversion of people with serious mental illnesses from jails and prisons toward more appropriate services in their communities (32,33). The severe problems facing persons with mental illness are best managed not in jail but in the community. Incarceration may sometimes be seen as serving the goal of rehabilitation, but more typically it is seen as serving as a deterrent and punishment. In this light, many consumers, family members, clinicians and administrators, and policy makers advocate for substantial improvement in the quality of services and systems designed to meet the needs of this population.

Two basic approaches have been proposed and utilized to improve treatment outcomes and quality of life for persons with severe mental illness who come into contact with the criminal justice system. The two strategies currently used in the United States are essentially either to provide mental health services to incarcerated persons in jails and prisons or to divert persons with serious mental illness from the criminal justice system to mental health services in the community. The rationale for much of the research on diversion in the past decade is the assumption that in-jail treatment services are often inadequate and have failed to meet the needs of many clients. As a result, this research has centered on the second strategy—mental health diversion—which is the strategy discussed in the remainder of this article, specifically postbooking diversion. This focus does not minimize the important work of psychiatrists and other mental health providers in correctional settings or the challenges faced there.

Even among postbooking diversion programs, the means for achieving objectives vary. A number of specific clinical interventions have been used either alone or in combinations, and they have been studied to some degree. Traditional outpatient mental health treatment has usually included a combination of counseling and visits to a psychiatrist, which are often focused on medication prescribing and monitoring. Several other clinical tools utilized in a diversion strategy are briefly mentioned below.

Assertive community treatment. Assertive community treatment is an approach developed in the 1970s to better care for persons with serious mental illness in their communities. A team-based model, assertive community treatment generally targets objectives of helping clients with continuity of treatment, decreasing hospitalizations, and enhancing community and social functioning. Teams are interdisciplinary, try to provide for all the psychiatric and social needs of clients in their homes or workplaces, and offer "assertive outreach" (34,35,36). Marshall and Lockwood (34) conducted a thorough examination of the effectiveness of assertive community treatment. The authors documented assertive community treatment as a "clinically effective approach to managing the care of severely mentally ill people in the community." It should be noted, however, that although studies examined by these authors included participants with serious mental illness (alone or in combination with a substance use disorder), the studies did not focus specifically on the population of criminal justice system clients with serious mental illness or the effectiveness of assertive community treatment specifically for this group.

Intensive case management. Intensive case management is a community-based intervention often used within diversion programs and has some objectives in common with assertive community treatment: it is intended to enhance linkage to services, community integration, and successful adaptation to living without frequent hospitalization. In contrast to assertive community treatment, intensive case management is based on the responsibility of one person for a caseload of clients and is focused less on fidelity to a specific model than on general tenets of practice (37). A recent review by Loveland and Boyle (37) looked at intensive case management as utilized in the process of jail diversion (not solely postbooking); the authors distinguished between "general intensive case management programs" and intensive case management interventions specifically used as part of diversion programs. They concluded that although an intensive case management component appears to be essential in the diversion process, intensive case management, alone, may not be sufficient to ensure successful diversion.

Intensive psychiatric probation and parole. Intensive psychiatric probation and parole are elements of supervision and monitoring that are primarily the responsibility of individuals in the criminal justice system, typically a probation or parole office. In maintaining contact with a person after detention or incarceration, the officer is responsible for monitoring the individual's compliance with conditions of the court. The officer does not have a clinical role per se, but in fulfilling his or her role for the court, the officer may influence the likelihood of a client's hospitalization or incarceration. This influence has been examined for clients involved in mental health services. Solomon and colleagues (38) provided evidence that some elements of mental health services can help to protect a person with serious mental illness from incarceration for technical violations.

Mental health courts. Mental health courts are specialized judicial entities that have become a prominent element in many diversion programs and are widespread in the United States. Frequently utilized in combination with some of the other elements described here, mental health courts have followed drug courts in their development and are similar in function. One judge or a very limited number of judges take responsibility for clients with serious mental illness who enter the court, and a team of criminal justice staff and treatment providers collaborate in a problem-solving approach for the benefit of these individuals. An emerging body of literature has offered some empirical support for the effectiveness of the mental health court model over the past decade (12,39,40).

Residential support. Although residential support is not always included in diversion programs, it has received attention as a useful element. The association between serious mental illness and homelessness serves as one basis for offering residential support to criminal justice system clients with serious mental illness. There is evidence that specialized housing programs may be necessary for these clients, rather than housing programs available for the general population of persons with serious mental illness (41). The value of housing support as a well-integrated element in an array of diversion services, such as those provided by Project Link in Rochester, New York (42,43), has been shown.

The tools described above have been successfully used to help address the needs of persons with severe psychiatric illness. As discussed below, when these tools are used as part of a comprehensive diversion program, they may help to prevent problems for clients with serious mental illness that may otherwise arise from recurrent, excessive, or inappropriate contacts with the criminal justice system.

Postbooking diversion outcomes

Previous studies of jail diversion programs describe some of the benefits of postbooking programs as they currently operate in the United States. It appears, however, that research has not kept up with practice in the area of postbooking diversion. The outcome evidence below is organized in terms of the effects of diversion programs on individual clients, on systems of mental health or criminal justice services, on communities in which diversion programs exist, and on society as a whole.

Individual outcomes. Studies indicate that postbooking diversion programs benefit enrolled clients by reducing criminal justice contacts, such as number of arrests and the frequency and length of incarcerations (2,20,24,44,45). Clients sometimes experience increased life satisfaction and reduced psychiatric symptoms (12,45). Diversion programs appear to reduce the number of psychiatric hospitalizations for a client over the longer term but may increase them in the shorter term and may increase use of services at lower levels of care (20). Some studies have shown reduced alcohol consumption among clients with mental illness who have comorbid substance use problems (12). Some clients experience increased work stability and increased residential stability, depending on the components offered in the diversion program and their coordination (42,43).

System outcomes. Although diversion programs may benefit individual clients, the effects of the programs on the mental health and criminal justice systems may vary, depending on the particular diversion program and the management of its resources. In general, criminal justice organizations may see lower levels of rearrest or jail recidivism, depending on the severity of the initial charge, as noted above. For mental health agencies, the levels of services used and the distribution of use may be affected; that is, utilization of mental health care at various levels of care (outpatient, day hospital, or inpatient) may increase or decrease depending on the specific diversion program (26,39,46). Such shifts in the use of various community services may reflect more appropriate utilization of mental health services by clients; for example, intensive case management may help direct an individual client to an optimal combination of services. However, data on the issue of optimal redistribution of service use is scant or nonexistent. There is some evidence that criminal justice system clients with serious mental illness may gain improved access to mental health services (26,47). However, little evidence exists about whether clients receive appropriate treatment services at the most beneficial intensity.

The economic benefits of diversion programs depend on the specific program and the balance between costs for the mental health system and for the criminal justice system in a particular location (26,48,49). Generally, diversion programs appear likely to reduce some of the costs of providing correctional services for enrolled clients while increasing costs of providing mental health treatment services, at least in the short term. However, data are limited. Clearly, more information is needed about economic costs from a community or taxpayer perspective. The effects of separate funding streams from mental health and criminal justice agencies within a state are not well studied.

Finally, it seems reasonable to hypothesize that setting up and implementing a diversion program will enhance the overall interactions between the mental health and criminal justice systems involved. Although research in this area appears sparse, there is some evidence that positive system-level changes may occur (50). Indeed, such system-level and intersystem benefits may be seen as highly beneficial for the community as a whole.

Community outcomes. When criminal justice system clients with serious mental illness are well served in a diversion program in a particular community (municipality or county), there may be advantages not only to those clients and to criminal justice and mental health systems but also to other citizens in the community. Evidence of improved individual outcomes for these clients, such as reduced arrests, reduced substance use, and improved community adjustment, is encouraging. However, some of the spillover benefits have not been explicitly examined. A reduction in even minor crimes, for example, may enhance perceptions of safety in particular neighborhoods. Fewer arrests of individuals with violent histories may result in fewer injuries to police officers and others. Increases in residential stability and decreases in the number of homeless criminal justice system clients with serious mental illness may enhance the satisfaction of both the clients and other citizens. Increased work stability may mean greater economic gains not only for clients but also for affected family members and possibly for the community. For example, successful and efficient programs may allow funds to be allocated to other public programs in need of support.

However, empirical research on community outcomes of postbooking diversion is almost nonexistent. Information on costs or comprehensive benefits from the perspective of the overall community is quite limited (26,49). A more complete accounting of such benefits is a necessary part of an accurate cost-effectiveness analysis of postbooking diversion from the perspective of community leaders. Mauser and colleagues (51) attempted to take such a broader viewpoint in a study of services for substance-abusing offenders.

In summary, much of the available information on the outcomes of postbooking diversion programs is for individual clients. However, data for outcomes at other levels are scarce. We know much less about positive effects for agencies that offer diversion services or for communities. Benefits might even be considered on a broader societal level. For example, a well-functioning and widespread strategy for meeting the needs of criminal justice system clients with serious mental illness allows the program to serve more clients. In addition, the practices or elements of diversion programs that are found to contribute to the cost-effectiveness of the program can be disseminated to other states. The possibilities for economies of scale or economies of scope as effective and efficient diversion programs become more widespread in a geographic area suggest other benefits related to resource allocation. For example, a greater absolute number of clients in a particular state could be served without a proportional increase in the societal resources devoted to such programs. However, it is again unfortunate that adequate information is not available to answer questions at this level.

Variables that may influence successful diversion outcomes

For administrators, clinicians, and particularly for clients involved in postbooking diversion programs, it is critical to have some understanding not only of how to provide services but also to whom they will be most useful. Factors that appear to be related to positive outcomes are discussed below. The information is organized somewhat chronologically—that is, according to the points at which the variables appear to influence the diversion process.

Prediversion factors. Variables that come into play before a client enters a diversion program may be considered prediversion. They primarily comprise characteristics of the individuals with serious mental illness; however, they may also include aspects of a program's development or an agency's functioning that exist before an individual's entry. For example, compared with groups not diverted from the usual judicial process after arrest, diverted groups have been found to include more females, older clients, clients with less severe charges for their index arrest, and clients with no alcohol problems (52). In the large samples in the SAMHSA-sponsored study by Broner and colleagues (20), the diverted groups were also more likely than the nondiverted groups to have indicators of better mental health and higher life satisfaction, and they were less likely to have been previously arrested or to have spent prior time in jail (20). Such selection parameters or biases appear to be in operation or "built into" a program before an individual is accepted or rejected (53). In research studies, attempts are made to statistically account for such group differences; however, it is still difficult to reach conclusions about whether some of these characteristics (such as gender, age, and substance use) may predict more successful diversion outcomes.

Once in a diversion program, it appears that individuals with an index arrest for minor crimes that are more serious—that is, higher-level misdemeanors and low-level felonies—may have better outcomes in terms of fewer days of subsequent incarceration (24,44). Also in terms of index arrest, it appears that those with charges for violent crimes fare no worse than those charged with nonviolent crimes, although persons in the latter group are apparently less often approved for participation in diversion (54). Comorbid substance use disorders are extremely common in populations of criminal justice system clients with serious mental illness, and it appears that these clients also have positive outcomes. When analyses controlled for differences between diverted and nondiverted samples, findings have suggested that diversion "may be one option for increasing access to services, increasing time in the community, and reducing jail days, without a concomitant increase in arrests, substance abuse, or psychiatric symptoms" (54). However, evidence is still limited on other prediversion factors that may be associated with positive outcomes.

Diversion factors. Diversion factors are variables that exist during a client's participation in diversion. They may comprise characteristics of the client or the program, or they may refer to the interaction or "fit" between client and program. One such variable with some empirical support is whether the diversion process includes mandated treatment services and court monitoring; a related variable is the degree of structure or even "coercion" that is inherent in treatment interventions, such as residential substance abuse treatments versus outpatient treatment (26). Broner and colleagues (26) have suggested that qualities such as perceived choice and level of insight deserve consideration as predictors of successful outcomes. Their findings on the interactions between diversion and race-ethnicity led them to suggest that diversion interventions need to be "culturally informed consistent with the established practice guidelines." Of interest, diverted and nondiverted clients often appear to receive similar amounts of treatment (26). It is unclear which clinical interventions contribute substantially or predominantly to positive diversion outcomes. Unfortunately, data are insufficient to help determine whether the quality and components of treatment are appropriate, even within diversion programs generally. In fact, it has been suggested that in far too many programs the "array of community-based services clinically indicated is seldom provided" (2).

Postdiversion factors. To ensure that all variables that might contribute to positive diversion outcomes are considered, factors that exist after a client participates in or graduates from a diversion program should be considered. Some elements of diversion or of follow-up monitoring may help criminal justice system clients with serious mental illness solidify or even build upon the gains made during their participation in diversion programs. Such variables may be more relevant in longitudinal studies examining long-term effects of diversion programs. However, few studies of diversion programs extended longer than about 12 months, and almost none extended to 18 months. Even within these time frames, researchers could examine whether clients who successfully completed or "graduated" from a diversion program had different longer-term outcomes than clients who did not complete the program. Little information is available on how gains made during the diversion period are best sustained over two to three years or more.

In summary, the ability to answer the question of how to predict successful diversion is limited. We can say that persons from diverse demographic groups (for example, gender, race-ethnicity, and age groups) have successfully participated in postbooking diversion. Individuals with charges for nonviolent crimes and those with charges for certain violent crimes appear to be able to successfully complete and benefit from diversion. An individual's perceived degree of choice and level of insight may be related to positive outcomes. Finally, mandated treatment within a diversion program may confer a higher probability of successful completion than nonmandated treatment.

Although researchers have made significant progress in gathering evidence for some of the general benefits of diversion programs, the evidence so far is still quite limited and sometimes equivocal. Additional information and better quality of information are needed in a number of areas. For instance, which clients should be accepted into diversion programs? This basic question about clients for whom diversion is most appropriate deserves further examination. Strategies other than diversion for criminal justice system clients with serious mental illness may deserve more study. Alternatives or additions to current diversion programs include, for example, jail mental health services (55), an increased focus on vocational development (56,57), and other legal mechanisms to reduce the rate of incarceration for people with serious mental illness, such as the insanity defense or involuntary emergency evaluations (58,59).

Once diversion is deemed appropriate for a particular person or group of clients, which individuals are likely to derive the most benefit? Published studies have provided evidence of the effects of diversion for clients with a range of diagnoses, with a range of legal charges, and in several demographic categories. Empirical data suggest that administrators and policy makers who make decisions about diversion in their communities should take a rather inclusive view of who may benefit. Because no evidence indicates that certain demographic characteristics predict better outcomes, diversion programs appear suited for individuals of any demographic background. Because no evidence indicates that offenders with more serious index charges—including some with violent charges—have poorer outcomes, diversion should be considered for all criminal justice system clients with serious mental illness, without necessarily excluding those with felony charges or a violent history. Because no evidence indicates that people with substance use disorders have poorer outcomes, diversion should also be offered to them.

This is not to say, however, that the process of diversion can or should be a "one size fits all" endeavor. Evidence remains limited about specific groups of people who benefit most and through what interventions. Evidence about the predictors of success, including prediversion, diversion, and postdiversion factors, is still quite limited. Research is needed to determine which types of programs and program elements are most useful at the individual level so that administrators and clinicians can better tailor the process for individual clients. The importance of client priorities and satisfaction deserves more attention. Considerable information is available about treatments for clients with serious mental illness who have comorbid substance use disorders. However, little can be said specifically about clients who have issues related to past traumas, certain personality disorders, developmental disabilities, or specific medical conditions. Interactions between professionals from the mental health and criminal justice systems who serve the same clients require further study. For example, not enough is known about the interactions between probation officers and mental health teams. Although not covered here, the literature on youths with serious mental illness who are involved with the juvenile justice system is likewise limited, and further research here is sorely needed.

Furthermore, what program models and components are most efficacious within a given community? How can the best practices be optimally implemented, given the resources and capacities available to a particular community's mental health and criminal justice systems? Further evidence should be sought on the usefulness of specific tools (for example, mental health courts, intensive case management, and assertive community treatment) for individuals with specific issues (for example, a trauma history or certain medical conditions) and in particular community systems. Information on the cost-effectiveness of such elements by themselves and in various combinations is very much needed. The impact of having funding streams that originate from two separate agencies—the mental health system and the criminal justice system—must be understood better. Issues of management decision making, quality control, and service integration deserve closer examination within both systems. The proper role of information technology in the implementation and operation of diversion programs has also received inadequate research attention.

At the community or societal level, what variations in diversion programs must be considered when implementation occurs across several communities, perhaps statewide or regionwide? In such cases, the needs and resources of particular communities and states must be considered. Do differences in the availability of informal family or other community supports have an impact on the effectiveness of diversion programs? Are differences among clients living in urban, suburban, and primarily rural areas significant, and how might these differences affect the likelihood of successful diversion? Rural-urban comparisons of diversion participants or programs are difficult, if not impossible, to find. Are some interventions better for clients living in rural areas than for those in urban areas? Differences among various racial-ethnic or cultural groups may be significant, and program designs based on these differences may improve outcomes. Funding mechanisms across counties or statewide may be critical, as may issues of adequacy of health insurance or managed care in a community. Available studies focus on clients and programs in only a limited number of sites. Better and additional information is needed to help guide policy makers and administrators in many other settings.

Finally, some questions about diversion programs go beyond the individual, system, or even community outcomes discussed here. For example, a number of research and policy issues deserve further study. Methodological improvements in research on diversion are needed, such as using more sophisticated designs and measures, pushing beyond descriptive studies toward more explanatory models, and helping providers and managers to better predict how an individual may benefit from a combination of program elements or tools. A policy-related question concerns how diverse points of view can best be utilized by policy makers and managers in their work. Perhaps an optimal policy or administrative decision can be made only after the needs of the client population and available models of intervention are considered from a variety of viewpoints. Examination of such questions through a mental health lens may be complementary to perspectives gained through other lenses, such as criminal justice, developmental, multicultural, or economic. For example, clinical and public safety outcomes may be weighed differently according to how the diversion programs are administered and funded—through a criminal justice or a mental health agency. Future research should take into account such diversity of viewpoints and organizational cultures.

Several issues are of particular interest or concern to decision makers. The cost-effectiveness of various diversion program models deserves more rigorous study. Using an economic lens, researchers must look at efficiency and at optimal ways of implementing a diversion program. To what degree is human capital being underutilized in systems that fail to adequately promote employment of clients with serious mental illness? How many clients can realistically be served in a given state or across the United States with current diversion models? Are there economies of scale or of scope to be realized? To what extent can programs increase their effectiveness with more attention to up-to-date information technology? Do state or regional differences in funding or in managed-care penetration influence the kinds of programs best suited to particular geographic areas in the United States?

As is clear from this discussion, numerous paths are open for clinicians, administrators, researchers, and clients who are interested in improving the outcomes of diversion programs. Clearly there is more work to be done in the direction of successfully diverting criminal justice system clients with serious mental illness from incarceration and to optimal treatment interventions. Efforts to improve our understanding of diversion programs and their implementation may lead to additional individual, system, community, and societal benefits not yet realized or even imagined.

Problems facing persons who are involved in both the mental health and criminal justice systems are often extremely complicated and challenging. Substantial progress has been made over the past decade in developing strategies for meeting the needs of clients with serious mental illness involved in the criminal justice system. This review has attempted to provide an overview of some of these strategies, with a focus on the emergence and growth of postbooking diversion programs in the United States. It is hoped that discussion of these issues will contribute to continued progress in this important arena.

The authors report no competing interests.