Children and Mental Health
This is a chapter from a Report of the Surgeon General on mental health.
This section examines the effectiveness of such treatment interventions as outpatient, partial hospitalization/day, residential, inpatient treatments, and medication. Much of the research on their effectiveness deals with children’s outcomes largely independent of diagnosis. As noted earlier in this chapter (see Treatment Strategies), practitioners and researchers previously shied away from diagnosis because of the inherent difficulty of making a diagnosis, concerns about labeling children, and the limited usefulness of DSM classifications for children. Each intervention was developed to treat a host of mental health conditions in children and adolescents. Each also was delivered in a wide range of settings. Over time, the combination of interventions and settings, with the exception of medication, became conceptualized as “treatments,” which stimulated research on their effectiveness (Goldman, 1998). They are not, however, treatments in the conventional sense of the term because they are less specific than other treatments with respect to indications, intensity (i.e., “dose”), and elements of the intervention. There is little research describing treatment in actual clinical settings.
The term “outpatient treatment” covers a large variety of therapeutic approaches, with most falling into the broad theoretical categories of the psychodynamic, interpersonal, and behavioral psychotherapy. Outpatient psychotherapy is the most common form of treatment for children and adolescents, utilized annually by an estimated 5 to 10 percent of children and their families in the United States (Burns et al., 1998). It is also the most extensively studied intervention and, with over 300 studies, has the strongest research base (Weisz et al., 1998). Outpatient therapy is offered to individuals, groups, or families, usually in a clinic or private office. The duration of treatment varies from 6 to 12 weekly sessions to a year or longer. Newer outpatient interventions (e.g., case management, home-based therapy) that were developed more recently for youth with severe disorders are provided with greater frequency (i.e., daily) in the home, school, or community. Those interventions are reviewed later in this chapter.
The strongest support for the effectiveness of outpatient treatment comes from a series of meta-analyses. Meta-analyses are an important type of research methodology, described in Chapter 1, that enable one to combine research findings from separate studies. Nine meta-analyses, published between 1985 and 1995, probed the effectiveness of research on individual, group, and family therapy for children and adolescents (Casey & Berman, 1985; Hazelrigg et al., 1987; Weisz et al., 1987; Kazdin et al., 1990; Baer & Nietzel, 1991; Grossman & Hughes 1992; Shadish et al., 1993; Weisz & Weiss, 1993; Weisz et al., 1995). Although these meta-analyses vary in time period, age groups, and meta-analytic approach, they were largely restricted to studies of treatment given in a research clinical setting, and their findings are relatively consistent. The major findings indicated that the improvements with outpatient therapy are greater than those achieved without treatment; the treatment is highly effective, as was found in meta-analyses of adults (Brown, 1987); and the effects of treatment are similar, whether applied to problems such as anxiety, depression, or withdrawal (internalizing problems) or to hyperactivity and aggression (externalizing problems) (Kazdin, 1996).
Given strong evidence of efficacy for outpatient treatment, the question of applicability to real-world settings has been examined. A meta-analysis was performed on studies of the effectiveness of various types of outpatient treatment, regardless of whether their efficacy had been established through research (Weisz et al., 1995). The researchers were able to identify only nine studies of treated children in nonresearch clinical settings where therapy was a regular service of the clinic and was carried out by practicing clinicians. Those nine studies demonstrated little or no effect. Clearly, real-world therapy was found to be less effective than that provided through a research protocol. A variety of factors may account for the gap, including less attention in real-world settings to careful matching of patients with treatments, less adherence to a treatment protocol, and less followup care.
Partial Hospitalization/Day Treatment
Partial hospitalization, also called day treatment and partial care, has been a growing treatment modality for youth with mental disorders. Research on partial hospitalization as an alternative to inpatient treatment generally finds benefit from a structured daily environment that allows youth to return home at night to be with their family and peers.
Partial hospitalization is a specialized and intensive form of treatment that is less restrictive than inpatient care but is more intensive than the usual types of outpatient care (i.e., individual, family, or group treatment). The most frequently used type of partial hospitalization is an integrated curriculum combining education, counseling, and family interventions. The setting, be it a hospital, school, or clinic, may be tied to the theoretical orientation of the treatment, which ranges from psychoanalytic to behavioral. Partial hospitalization has also been used as a transitional service after either psychiatric hospitalization or residential treatment, at the point when the child no longer needs 24-hour care but is not ready to be integrated into the school system. It also is used to prevent institutional placement.
Overall, the research literature points to positive gains from adolescent use of day treatment, but most of the studies are uncontrolled. Gains relate to academic and behavioral improvement; reduction in, or delay of, hospital and residential placement; and a return to regular school for about 75 percent of patients (Baenen et al., 1986; Gabel & Finn, 1986). Day treatment programs are not being used as frequently as they might be because third-party payers are reluctant to support this form of treatment. They claim that the modality is ambiguous, that it induces demand among those who would not otherwise seek treatment, and that its length, treatment outcomes, and costs are unpredictable (Kiser et al., 1986). Research is needed to address these issues.
To date, the only controlled study of partial hospitalization compared outcomes for young children (ages 5 to 12) with disruptive behavior disorders who received intensive day treatment with children who received traditional outpatient treatment services (in fact, a waiting list control) (Grizenko et al., 1993). The results at 6 months favored day treatment in reducing behavior problems, decreasing symptoms, and improving family functioning.
Findings from uncontrolled studies of partial hospitalization are informative, although not conclusive. Based on approximately 20 studies, multiple benefits have been reported even over the long term (see reviews by Kutash & Rivera, 1996; Grizenko, 1997). In general, child behavior and family functioning improve following partial hospitalization. Findings for improved academic achievement are mixed and possibly suggest that implementation of school-based models should be considered. About three-fourths of youth are reintegrated into regular school, often with the help of special education or other school- or community-based services. Several uncontrolled studies found that day treatment could prevent youth from entering other costly placements (particularly inpatient and residential treatment centers), which suggests that partial hospitalization may reduce overall costs of treatment (Kutash & Rivera, 1996). Finally, family participation during and following day treatment is essential to obtaining and maintaining results (Kutash & Rivera, 1996).
Residential Treatment Centers
Residential treatment centers are the second most restrictive form of care (next to inpatient hospitalization) for children with severe mental disorders. Although used by a relatively small percentage (8 percent) of treated children, nearly one-fourth of the national outlay on child mental health is spent on care in these settings (Burns et al., 1998). However, there is only weak evidence for their effectiveness.
A residential treatment center (RTC) is a licensed 24-hour facility (although not licensed as a hospital), which offers mental health treatment. The types of treatment vary widely; the major categories are psychoanalytic, psychoeducational, behavioral management, group therapies, medication management, and peer-cultural. Settings range from structured ones, resembling psychiatric hospitals, to those that are more like group homes or halfway houses. While formerly for long-term treatment (e.g., a year or more), RTCs under managed care are now serving more seriously disturbed youth for as briefly as 1 month for intensive evaluation and stabilization.
Concerns about residential care primarily relate to criteria for admission; inconsistency of community-based treatment established in the 1980s; the costliness of such services (Friedman & Street, 1985); the risks of treatment, including failure to learn behavior needed in the community; the possibility of trauma associated with the separation from the family; difficulty reentering the family or even abandonment by the family; victimization by RTC staff; and learning of antisocial or bizarre behavior from intensive exposure to other disturbed children (Barker, 1998). These concerns are discussed below.
In the past, admission to an RTC has been justified on the basis of community protection, child protection, and benefits of residential treatment per se (Barker, 1982). However, none of these justifications have stood up to research scrutiny. In particular, youth who display seriously violent and aggressive behavior do not appear to improve in such settings, according to limited evidence (Joshi & Rosenberg, 1997). One possible reason is that association with delinquent or deviant peers is a major risk factor for later behavior problems (Loeber & Farrington, 1998). Moreover, community interventions that target change in peer associations have been found to be highly effective at breaking contact with violent peers and reducing aggressive behaviors (Henggeler et al., 1998). Although removal from the community for a time may be necessary for some, there is evidence that highly targeted behavioral interventions provided on an outpatient basis can ameliorate such behaviors (Brestan & Eyberg, 1998). For children in the second category (i.e., those needing protection from themselves because of suicide attempts, severe substance use, abuse, or persistent running away), it is possible that a brief hospitalization for an acute crisis or intensive community-based services may be more appropriate than an RTC. An intensive long-term program such as an RTC with a high staff to child ratio may be of benefit to some children, especially when sufficient supportive services are not available in their communities. In short, there is a compelling need to clarify criteria for admission to RTCs (Wells, 1991). Previous criteria have been replaced and strengthened (i.e., with an emphasis on resources needed after discharge) by the National Association of Psychiatric Treatment Centers for Children (1990).
The evidence for outcomes of residential treatment comes from research published largely in the 1970s and 1980s and, with three exceptions, consists of uncontrolled studies (see Curry, 1991).
Of the three controlled studies of RTCs, the first evaluated a program called Project Re-Education (Re-Ed). Project Re-Ed, a model of residential treatment developed in the 1960s, focuses on training teacher-counselors, who are backed up by consultant mental health specialists. Project Re-Ed schools are located within communities, facilitating therapeutic work with the family and allowing the child to go home on weekends. Camping also is an important component of the program, inspired by the Outward Bound Schools in England. The first published study of Project Re-Ed compared outcomes for adolescent males in Project Re-Ed with untreated disturbed adolescents and with nondisturbed adolescents. Treated adolescents improved in self-esteem, control of impulsiveness, and internal control compared with untreated adolescents, according to ratings by Project Re-Ed staff and by families (Weinstein, 1974). A 1988 followup study of Project Re-Ed found that when adjustment outcomes were maintained at 6 months after discharge from Project Re-Ed, those outcomes were predicted more by community factors at admission (e.g., condition of the family and school, supportiveness of the local community) than by client factors (e.g., diagnosis, school achievement, age, IQ). This suggested that interventions in the child’s community might be as effective as placement in the treatment setting (Lewis, 1988).
The only other controlled study compared an RTC with therapeutic foster care through the Parent Therapist Program. Both client groups shared comparable backgrounds and made similar progress in their respective treatment program. However, the residential treatment cost twice as much as therapeutic foster care (Rubenstein et al., 1978).
Despite strong caveats about the quality, sophistication, and import of uncontrolled studies, several consistent findings have emerged. For most children (60 to 80 percent), gains are reported in areas such as clinical status, academic skills, and peer relationships. Whether gains are sustained following treatment appears to depend on the supportiveness of the child’s post-discharge environment (Wells, 1991). Several studies of single institutions report maintenance of benefits from 1 to 5 years later (Blackman et al., 1991; Joshi & Rosenberg, 1997). In contrast, a large longitudinal six-state study of children in publicly funded RTCs found at the 7-year followup that 75 percent of youth treated at an RTC had been either readmitted to a mental health facility (about 45 percent) or incarcerated in a correctional setting (about 30 percent) (Greenbaum et al., 1998).
In summary, youth who are placed in RTCs clearly constitute a difficult population to treat effectively. The outcomes of not providing residential care are unknown. Transferring gains from a residential setting back into the community may be difficult without clear coordination between RTC staff and community services, particularly schools, medical care, or community clinics. Typically, this type of coordination or aftercare service is not available upon discharge. The research on RTCs is not very enlightening about the potential to substitute RTC care for other levels of care, as this requires comparisons with other interventions. Given the limitations of current research, it is premature to endorse the effectiveness of residential treatment for adolescents. Moreover, research is needed to identify those groups of children and adolescents for whom the benefits of residential care outweigh the potential risks.
Inpatient hospitalization is the most restrictive type of care in the continuum of mental health services for children and adolescents. Questions about excessive and inappropriate use of hospitals were raised in the early 1980s (Knitzer, 1982) and clearly documented thereafter in rising admission rates from the 1980s into the mid-1990s, without evidence of increased social or clinical need for such treatment (Weller et al., 1995). Inpatient care consumes about half of child mental health resources, based on the latest estimate available (Burns, 1991), but it is the clinical intervention with the weakest research support. Nevertheless, because some children with severe disorders do require a highly restrictive treatment environment, hospitals are expected to remain an integral component of mental health care (Singh et al., 1994). More concerted attention to the risks and benefits of hospital use is critical, however, along with development of community-based alternative services.
Research on inpatient treatment mostly consists of uncontrolled studies (Curry, 1991). Factors that are likely to predict benefit have been identified from such studies. Beneficial factors were found to include higher child intelligence; the quality of family functioning and family involvement in treatment; specific characteristics of treatment (e.g., completion of treatment program and planned discharge); and the use of aftercare services. Neither age nor gender affected prognosis after hospitalization. The prognosis was poor for several clinical characteristics, including children with a psychotic diagnosis and antisocial features with conduct disorder (Kutash & Rivera, 1996).
Only three controlled studies evaluated the effectiveness of inpatient treatment: one that randomized antisocial children to specific interventions on an inpatient unit (Kazdin et al., 1987a, 1987b) and two older clinical trials (Flomenhaft, 1974; Winsberg et al., 1980). All three studies demonstrated that community care was at least as effective as inpatient treatment.
More recently there have been preliminary favorable findings from a randomized trial of inpatient treatment versus multisystemic therapy (MST), an intensive home-based intervention. For example, MST was more effective than psychiatric hospitalization in reducing antisocial behavior, improving family structure and cohesion, improving social relationships, and keeping children in school and out of institutions (after the initial period when the control group was in the hospital). Hospitalized youth reported improved self-esteem, and youth in both treatment conditions showed comparable decreases in emotional distress (Henggeler et al., 1998). A great deal more research is needed on inpatient hospitalization, as it is by far the costliest and most restrictive form of care. Recent changes in health care management have resulted in short lengths of stay for children and adolescents. Preliminary results from the study of MST indicate that intensive home-based services may be a viable alternative to hospitalization. However, even when such services are available, there may be a need for brief 24-hour stabilization units for handling crises (see Crisis Services).
Since the 1980s, the field of children’s mental health has witnessed a shift from institutional to community-based interventions. The forces behind this transformation are presented in a subsequent section, Service Delivery. This section attempts to answer the question of whether community-based interventions are effective. It covers a range of comprehensive community-based interventions, including case management, home-based services, therapeutic foster care, therapeutic group homes, and crisis services. Although the evidence for the benefits of some of these services is uneven at best, even uncontrolled studies offer a starting point for studying the effectiveness and feasibility of their implementation. Many of the evaluations to date offer a first glimpse into the benefits of these services and the extent to which they may be valuable for further examination. Of these inter- ventions, the most convincing evidence of effectiveness is for home-based services and therapeutic foster care, as discussed below.
There is a special emphasis throughout this section on “children with serious emotional disturbances,” as many of these community-based services are targeted to this population of the most serious severely affected children. The term serious emotional disturbance refers to a diagnosed mental health problem that substantially disrupts a child’s ability to function socially, academically, and emotionally. It is not a formal DSM-IV diagnosis but rather a term that has been used both within states and at the Federal level to identify a population of children with significant functional impairment due to mental, emotional, and behavioral problems who have a high need for services. The official definition of children with serious emotional disturbance adopted by the Substance Abuse and Mental Health Services Administration is “persons from birth up to age 18 who currently or at any time during the past year had a diagnosable mental, behavioral, or emotional disorder of sufficient duration to meet diagnostic criteria specified within the DSM-III-R, and that resulted in functional impairment which substantially interferes with or limits the child’s role or functioning in family, school, or community activities” (SAMHSA, 1993, p. 29425). The term is used in a variety of Federal statutes in reference to children fitting that description and does not signify any particular diagnosis per se; rather, it is a legal term that triggers a host of mandated services to meet the needs of these children (see Service Delivery section).
Case management is an important and widespread component of mental health services, especially for children with serious emotional disturbances. The main purpose of case management is to coordinate the provision of services for individual children and their families who require services from multiple service providers. Case managers take on roles ranging from brokers of services to providers of clinical services. There is a considerable amount of variation in models of case management. In one important model, called “wraparound,” case managers involve families in a participatory process of developing an individualized plan focusing on individual and family strengths in multiple life domains. Research on wraparound is still in its early stages (Burns & Goldman, 1999).
There have been controlled studies of three programs that used case managers who work individually rather than as part of an interdisciplinary team (discussed later). In one study of the Partner’s Project in Oregon, case management was compared with “usual services,” which did not include case management (Gratton et al., 1995). The authors found at 1-year followup that children in the Partner’s Project scored significantly higher on measures of social competence and had received more individualized, comprehensive services, and a greater degree of service coordination.
The second study compared the outcomes of intensive case management and regular case management for mentally ill homeless children in Seattle (Cauce et al., 1994). The case managers in the intensive condition had lower caseloads, were required to spend more hours supervising the youth, had flexible funds (for clothing, transportation, etc.) at their disposal, spent more hours in consultation with psychologists, and were of higher educational status. After 1 year, the study found that both groups showed substantial yet similar improvement in mental health and social adjustment.
A model known as Children and Youth Intensive Case Management (CYICM) was evaluated in two controlled studies. The program has been described as an Expanded Broker Model, which means that the case manager, in addition to brokering services, is responsible for assessment, planning, linking, and advocating on behalf of the youth and family. Case managers, with caseloads of 10 children, are given $2,000 of flexible funds per child each year to purchase treatment and ancillary services (e.g., transportation and educational aids). In the first study, the authors found that children in the program spent significantly more days in the community between episodes of psychiatric hospitalization and were hospitalized for fewer days than before enrollment (Evans et al., 1994). A subsequent study evaluated a random sample of 199 children enrolled in CYICM (Evans et al., 1996b). Findings at 3-year followup indicated significant behavioral improvements and decreases in unmet medical, recreational, and educational needs compared with findings at enrollment. As in the previous study, children who had been in CYICM for 2 years had spent fewer days in psychiatric hospitals and more days in community settings during the intervals between hospitalizations. This study went further to compare their hospital utilization with that by children not enrolled in the program. Although CYICM clients spent more days in psychiatric hospitals before enrollment, they used inpatient services after enrollment significantly less than did non-enrollees. CYICM clients’ hospital admissions declined fivefold after enrollment whereas among non-enrollees the decline in admission rates was less than half that value. This difference translated into a savings of almost $8,000,000 for New York State, where the project took place.
Some research has investigated the effects of extending case management on children with a dual diagnosis of a mental disorder and a substance abuse problem. Within the CYICM program, researchers looked at whether adolescents with mental disorders and substance abuse problems derived comparable benefits from the program as did those without substance abuse problems (Evans et al., 1992). No significant differences were found in the average number of inpatient admissions both before and after enrollment. There was also no significant difference between groups in the average decrease from pre- to postenrollment in the number of days spent in hospitals. These results indicate that case management can be as effective for youth presenting with substance abuse problems as for youth presenting with other psychiatric disorders.
Team Approaches to Case Management
Several studies assessed the value of case management as part of a treatment team. In a randomized trial in North Carolina (Burns et al., 1996), youth served by an interdisciplinary treatment team led by a case manager were compared with a control group of youth served by a treatment team led by their primary clinician in the role of case manager (also called clinician case manager). At 1-year followup, case managers in the experimental group reported spending significantly more time with their clients, as well as significantly more time on the core functions of case management (e.g., outreach; assessment of strengths, needs, and resources; service planning and monitoring; linking, referral, and advocacy; and crisis intervention). The experimental group also remained in the case-managed program longer, spent fewer days in psychiatric hospitals, and received more community-based services and a more comprehensive array of services. Although both groups showed similar clinical and functional improvements, parents of youth in the experimental group reported more satisfaction with the service system. The study concluded that traditional case managers, rather than clinician case managers, provide a more cost-effective method for attaining positive behavioral outcomes and access to mental health services.
Another example of a team approach to case management is the Family Centered Intensive Case Management (FCICM) program. This was originally created as a variation of Child and Youth Intensive Case Management in New York, with the later addition of a wraparound approach. The wraparound approach is based on a belief that the child and family should be placed at the center of an array of coordinated health and mental health, educational, and other social welfare services and resources, which a case manager wraps around the patient and family. In a randomized trial, children were assigned to either FCICM or Family-Based Treatment (Evans et al., 1996a). Family-Based Treatment included training, support, and respite care for foster families but did not include case managers.
The findings at 18 months (or at discharge) indicated that children in FCICM had significantly fewer behavioral symptoms and significantly greater improvements in overall functioning than those in Family-Based Treatment. In addition, the average annual cost of FCICM was less than half that of Family-Based Treatment.
The Fostering Individualized Assistance Program (FIAP) is an example of case management provided through a wraparound approach. The effectiveness of this model, which used clinical case managers, was compared with standard foster care in a randomized trial involving 131 children and their families (Clark et al., 1998). The most important duty of the FIAP case managers was to arrange monthly team meetings for the monitoring of individualized service plans. Although both groups showed significant improvement in their behavioral adjustment over a 3_-year period, children in the FIAP group were less likely to change placements, and boys in the group reported better social adjustment and fewer delinquencies. Older youth in the group were more likely to maintain placements in homes of relatives and less likely to run away. Youth in FIAP were also absent from school less often and spent fewer days suspended from school. Overall, youth in the FIAP group showed more improvement than did youth in standard foster care. Multiple uncontrolled studies of case management using a wraparound approach were summarized in a recent monograph focusing on the wraparound process (Burns & Goldman, 1999). Overall, the reviewed studies, although using uncontrolled methods, offer emerging evidence of the potential effectiveness of case management using a wraparound process.
While evidence is limited and many of the positive outcomes focus on service use rather than clinical status, there is some indication that case management is an effective intervention for youth with serious emotional disturbances. Studies in this area are difficult to conduct because of resource limitations and of varying approaches to case management. Agreement on standards for specific case management models is
needed in order to proceed with efficient and reliable controlled research in this area. In addition, future research needs to address the issue of cost-effectiveness, as some evidence presented above has shown savings from less utilization of institutional care.
This section describes the strong record of effectiveness for home-based services, which provide very intensive services within the homes of children and youth with serious emotional disturbances. A major goal is to prevent an out-of-home placement (i.e., in foster care, residential, or inpatient treatment). Home-based services are usually provided through the child welfare, juvenile justice, and/or mental health systems. They are also referred to as in-home services, family preservation services, family-centered services, family-based services, or intensive family services.
Stroul (1988) identified three major goals of home-based services: to preserve the family’s integrity and prevent unnecessary out-of-home placements; to put adolescents and their families in touch with community agencies and individuals, thus creating an outside support system; and to strengthen the family’s coping skills and capacity to function effectively in the community after crisis treatment is completed. The specific services provided most often include evaluation, assessment, counseling, skills training, and coordination of services. The historical evolution of home-based services is discussed further under Support and Assistance for Families in Service Delivery.
The evidence for the benefits of home-based services was recently evaluated in a meta-analysis of controlled studies only (Fraser et al., 1997). The analysis referred to home-based services as “family preservation services”; these were sponsored either by the child welfare or juvenile justice systems. For 22 studies the authors analyzed specific measures such as out-of-home placement, family reunification, arrest, incarceration, and hospitalization, with the control group defined as youth receiving “usual” or “routine” services. While a majority of the studies demonstrated marginal gains in effectiveness, other services appeared to be significantly more effective than usual services. The findings are presented below according to their organizational sponsorship by either child welfare or juvenile justice system.
Family Preservation Programs Under the Child Welfare System
Within the child welfare system, particularly effective family reunification programs were the Homebuilders Program in Tacoma, Washington, which was designed to reunify abused and neglected children with their families by providing family-based services (Fraser et al., 1996), and the family reunification programs in Washington State and in Utah (Pecora et al., 1991). Studies suggested that 75 to 90 percent of the children and adolescents who participated in such programs subsequently did not require placement outside the home. The youths’ verbal and physical aggression decreased, and cost of services was reduced (Hinckley & Ellis, 1985). The success of these family preservation programs is based on the following: services are delivered in a home and community setting; family members are viewed as colleagues in defining a service plan; back-up services are available 24 hours a day; skills are built according to the individual needs of family members; marital and family interventions are offered; community services are efficiently coordinated; and assistance with basic needs such as food, housing, and clothing is given (Fraser et al., 1997).
Multisystemic therapy programs within the juvenile justice system have demonstrated effectiveness. MST is an intensive, short-term, home- and family-focused treatment approach for youth with severe emotional disturbances. MST was originally based on risk factors that were identified in the published literature and was designed for delinquents. MST intervenes directly in the youth’s family, peer group, school, and neighborhood by identifying and targeting factors that contribute to the youth’s problem behaviors. The main goal of MST is to develop skills in both parents and community organizations affecting the youth that will endure after brief (3 to 4 months) and intensive treatment. MST was constructed around a set of principles that were put into practice and then expanded upon in a manual (Henggeler et al., 1998). Elaborate training, supervision, and monitoring for treatment adherence make this an exemplary approach. Furthermore, publication of an MST manual and the high level of clinical training in MST distinguish this model from other types of family preservation services.
The efficacy of MST has been established in three randomized clinical trials for delinquents within the juvenile justice system. The first of these studies took place in Memphis, Tennessee, and revealed that MST was more effective than usual community services in decreasing adolescent behavioral problems and in improving family relations (Henggeler et al., 1986). The second was conducted in Simpsonville, South Carolina, and compared outcomes for 84 juvenile offenders randomly assigned to either MST or usual services. At 59 weeks after referral, youth who had received MST had fewer arrests and self-reported offenses and had spent an average of 10 fewer weeks incarcerated than did the youth in usual services. In addition, families served by MST reported increased family cohesion and decreased youth aggression in peer relations (Henggeler et al., 1992). In the third study, MST was compared with individual therapy in Columbia, Missouri, and was found to be more effective in ameliorating adjustment problems in individual family members. A 4-year followup of rearrest data indicated that MST was more effective than individual therapy in preventing future criminal behavior, including violent offenses (Borduin et al., 1995). Studies found improved behavior, fewer arrests, and lower costs. These findings encouraged the investigators to test the effectiveness of MST in other organizational settings (e.g., child welfare and mental health), allowing them to target other clinical populations, including youthful sex offenders (Borduin et al., 1990), abused and neglected youth (Brunk et al., 1987), and child psychiatric inpatients (see Inpatient Treatment section). Initial results are promising for youth receiving MST instead of psychiatric hospitalizations (Henggeler et al., 1998). As expected, some adjustments to MST are required to handle children who are dangerous to themselves and who do not respond as quickly to treatment as the delinquent youth in previous studies. The efficacy of MST was demonstrated in real-world settings but only by one group of investigators; thus, the results need to be reproduced by others and future effectiveness research needs to determine whether the same benefits can be demonstrated with less support from experts.