Alan W. Leschied, Ph.D. (C.Psych.), Principal
Investigator
Associate Professor, University of Western Ontario
Senior Clinical
and Research Consultant, Centre for Children & Families in the Justice
System
Alison Cunningham, M.A.(Crim.), Research
Coordinator
Director of Research & Planning
Centre for Children & Families in the Justice
System
This report describes a research project that began in 1997 and involves the evaluation of a home-based intervention for high-risk young offenders called Multisystemic Therapy (MST). In agencies at four Ontario communities (London, Mississauga, Ottawa and Simcoe County) teams of therapists have been trained in the provision of MST and are cooperating with the research.
MST promises to be a cost-efficient, community-based option to keep high-risk youth out of residential placements such as custody without putting the community at risk. It is designed for youth at the high-need/high-risk end of the spectrum and is meant to be tailored to their individual needs. Its goal is to reduce future criminal behaviour. If successful, MST could contribute to public safety, in both the short and the long-term, which translates into cost savings in two important sectors: 1) state-paid justice services; and 2) losses experienced by victims of crime. The implementation of MST in Ontario is being evaluated against these goals using a scientifically rigorous methodology including random assignment, a control group, a large sample, and long-term follow-up.
Youths are considered candidates for MST if they have a history of criminal offences and are identified as having a high or very-high risk of criminal offending in the future. Most have a constellation of presenting problems that include school refusal, aggression, substance abuse, non-compliance, risk taking, and severe parent/child conflict. The research design involves randomly assigning half of qualifying cases to MST and half to the services currently available to youth in the local area. There is pre- and post-testing of both groups and all youths will be followed for up to three years, to gauge their subsequent offending and levels of correctional service utilization. The evaluation strategy focuses on process, outcome, and cost-efficiency.
This report presents available data on the short and long-term outcomes of MST and summarizes the activities that lie ahead in this last year of the study.
Ontario, our largest province, contributes almost four out of each 10 youths (38%) appearing before Canadian youth courts (Carrière, 2000). Figures from Statistics Canada for 1998/99 indicate that 4.5 of every 100 Ontario youths have contact with the youth courts (national average is 4.4). The previous year, 41% of cases that result in an adjudication of guilt were disposed of with custody terms, higher than the national average of 34% (Statistics Canada, 1999). This is the situation that prompted the present study.
The origin of the Ontario MST clinical trial dates back to the late 1980s when persuasive evidence began to emerge that this new technique was effective in reducing both criminal behaviour and correctional costs in the United States. MST was developed at the Family Services Research Center at the Medical University of South Carolina, where they observed that mental health services for serious young offenders were minimally effective at best, extremely expensive and not accountable for outcomes. They searched the research literature for interventions with documented success in shaping good outcomes for anti-social youth. They also noted which interventions, some quite popular, have no empirical support. This process of discarding ineffective techniques while gleaning those most effective means that MST is really more an amalgam of best practices than a brand new method.
Outcome data from the randomized studies in the United States made MST a stand-out among delinquency prevention efforts, a field not characterized by overwhelming success. While most meta-analyses of research were showing that some treatment programs can be effective with some youths, translation of "best practice" to field environments had proved challenging. Moreover, few were claiming success with the most hard-to-serve youth, whose anti-social behaviour seems intractable and who are on a trajectory that may well take them to the adult penal system. (For a recent summary of American research on MST, see Borduin (1999) and a 1999 special issue (vol. 38, no. 3) of the journal Family Process devoted to MST.)
Dr. Scott Henggeler, in conjunction with colleagues at the Family Services Research Centre, is the developer of the MST approach. Dr. Henggeler was invited to speak at a court-based clinical services conference organized by the London Family Court Clinic in 1993. His presentation generated sufficient interest for a group of practitioners and policy makers in the Ministry of Community and Social Services (MCSS) to begin a four-year effort to find both the funding and the community will to start MST teams in four areas of Ontario.
Devising successful alternatives to custody as a key part of MCSS's commitment to promote prevention strategies, community-based programs, and cost-effective placements for youth in Ontario as articulated in the 1997 document Young Offenders Framework 1997-2000. The Ministry has defined five goals for young offender programming: maintenance of community safety; cost-efficiency; responsiveness to individual needs; reduction of future recidivism; and, provision of a spectrum of services that addresses all young people according to their levels of risk and need. MST fit comfortably within all these priorities and the Ministry made a four year commitment to fund the implementation of MST, mostly with resource re-allocation.
At the same time, the federal Department of Justice was examining the Young Offenders Act with a view to its reform, a process they were to call youth justice renewal. First in a report of the Standing Committee on Justice and Legal Affairs (1997), and then in a proposed framework for youth justice reform (Department of Justice, 1998), the federal government articulated a determination to reduce the rates of custody sentences ordered in Canadian youth courts. This cannot be accomplished through law reform alone. Members of the public in general, and sentencing judges specifically, must be convinced of several things. First, incapacitation through custody may protect the public in the short term but not in the long term. Second, there are viable community-based alternatives to custody that can both protect the public in the short term and reduce recidivism in the long term. Third, the expensive option of custody will not 'purchase' as much reduction in offending as these other non-custodial sentencing options.
Providing empirical evidence of these three factors is the intent of the MST clinical trial. The federal Department of Justice (and later the National Crime Prevention Centre) provided research funding to determine if MST can lower recidivism among high-risk young offenders in Canada. The methodology necessary to make this determination -- the randomized clinical trial -- was not a popular approach in all circles, but its adoption was crucial. Preventing Crime: What Works? was a review of $3 billion worth of crime prevention programs sponsored by the U.S. Department of Justice. These reviewers determined that most so-called "evaluations" were little more than descriptions of the program under study. Few studies reviewed met their threshold test for scientific rigour and so added nothing to the debate about "what works." It was suggested that, at the minimum, evaluations should have these three characteristics:
These criteria were described as "extraordinarily difficult to achieve," but it was recommended that no evaluation study be funded unless the three were met. However, to definitively answer the question "Does this program work?", two further criteria are required:
This last point is referred to by them as the "gold standard." The clinical trial of MST in Ontario meets this standard because we cannot assume the success of MST in the United States will automatically be replicated in Ontario. Compared with the under-privileged, dysfunctional communities in which MST was tested, Ontario young offenders have access to a far superior array of services. There was a need to determine if the use of MST with high-risk youth can produce better outcomes than the services and interventions already available in this province. This study uses an experimental design because there is a need to answer these specific questions:
The study has high ecological validity in that the youths are identified by referral sources as being those clients in the local area who present the greatest challenge to current services. Unlike many programs, MST does not screen out treatment-resistant youth or those with serious criminal histories.
Community agencies in four parts of Ontario agreed to participate with the project. Early in 1997, a representative from MST Services Inc. travelled to all four sites to make community presentations about MST and to ensure that the requisite resources and personnel were in place. A key goal of that process was to ensure that the community service system was "on board," that everyone was in agreement about the target population, and that referral volume would be sufficient to make a program viable.
The MST teams themselves were ideally to be comprised of three full-time workers but a combination of part-time workers was assembled in some areas. Each team was to have an experienced and well-trained supervisor who was at a position in the agency hierarchy where he or she would have credible authority over the team and the ability to advocate for the needs of the program. Agencies were required to ensure that supervisors were able to devote at least 25% of their time to the MST team (with 50% preferable) or, ideally, to have a full-time appointment with a 50/50 split between supervision and a small MST caseload.
Other expectations that were made clear during this start-up period were:
Four clinical supervisors and 13 workers joined the MST project early in 1997 and began the training process with an intensive one week session in mid-April near Orillia. This was followed by a pilot period during which each worker was assigned a small caseload and the procedures for supervision and consultation were put into place. It was also necessary to finalize the referral process and harmonize it with the research protocol. Potential referral sources were then notified that cases were being accepted. Teams began intake of cases in May and weekly consultations with South Carolina began in early June, 1997. Depending upon the rate of referrals, which was slow at first, each therapist had three or four cases in the pilot phase. In all, there were 16 pilot cases in London, six in Mississauga, nine in Ottawa and nine in Simcoe County. The research protocol was suspended for these early cases. By the end of fiscal year 1997/98, 77 cases had been inducted into the study.
On-going support by MST Services Inc. took the form of weekly telephone consultations and monitoring of cases; quarterly booster training sessions; and, periodic consultations on broader issues such as referral screening and community liaison. Dr. Christine Hamel was the MST consultant assigned to the Ontario trials and it was she who conducted the weekly consultations and booster sessions. This system continued for two years after which it was necessary to begin developing greater independence from the South Carolina developers of MST. Moves were taken to foster the capacity of Ontario-trained MST supervisors to adopt much of the role that the MST consultant previously had played. This direction led to the appointment of a System Supervisor in April, 2000. As the last year of the clinical trial starts, over 300 youths have been inducted into the study and preliminary data indicate that the hypothesized benefits of MST may well be materializing.
MST adopts a social-ecological approach to understanding anti-social behaviour. The underlying premise of MST is that criminal conduct is multi-causal; therefore, effective interventions would recognize this fact and address the multiple sources of criminogenic influence. These sources are found not only in the youth (values and attitudes, social skills, organic factors, etc.) but in the youth's social ecology: the family, school, peer group and neighbourhood. Treating the youth in isolation of these other systems means that any gains are quickly eroded upon return to the family, school or neighbourhood. In addition, many common interventions can have an iatrogenic effect and foster higher levels of criminal behaviour than would have been the case with no intervention at all. It is a key premise of MST that community-based treatment informed by an understanding of the youth's ecology will be more effective than other methods including costly residential treatment. This is even true when you select as candidates for MST those youths who are bound for residential treatment or custodial placements because of the seriousness of their conduct or emotional problems.
MST differs from conventional approaches in several respects, above and beyond the small caseload and 24/7 availability. The work is done in the community, usually the family home, rather than an office. As needed, they will spend time at school and meet with the youth's peer group and extended family. Instead of weekly or monthly appointments, there is daily contact if needed, especially in the beginning. The therapist is completely responsible for engaging the family and is accountable for outcomes. Progress is measured daily in an objective and observable way and the case is closed when the family-defined goals have been attained.
MST is a highly individualized, flexible intervention tailored to each unique situation. In other words, there is no one recipe for success. Instead, there are nine principles which guide intervention:
The MST-specific training augments the education and experience therapists bring from their chosen fields (usually social work or psychology).
Collaboration with community agencies is a crucial part of MST. The school is a key player and workers may be in daily contact with teachers and administrators. MST therapists also work in close partnership with probation officers who in many cases are the referral source. There may be a need to involve the youth in substance abuse treatment or seek a psychiatric consultation about a parent, for example. While the initial MST involvement may be intensive, perhaps daily, the ultimate goal is to empower the family to take responsibility for making and maintaining gains. An important goal in this process is to foster in parents the ability to be good advocates for their children and themselves with social service agencies and to seek out supportive services and networks. In other words, parents are encouraged to develop the requisite skills to solve their own problems rather than rely on professionals.
The MST process begins with the identification of the problem behaviours, a task which involves the whole family. In other words, parents are key in identifying treatment targets. Examples of these behaviours include non-compliance with family rules, failure to attend school, failure to complete school work, substance use, disrespect to authority figures, and assaultive behaviour. While the focus is on elimination of problem behaviours, this is accomplished in great measure by building on strengths. So the assessment process also involves identifying the strengths in the youth and his or her family, which can include athletic ability, a trusting relationship with an extended family member or teacher, warmth and love among family members, or a hobby.
The next step is an assessment of the factors in the youth's ecology which support the continuation of the problem behaviours and the factors which operate as obstacles to their elimination. These factors may be a found in any sphere of the youth's ecology or the linkages among them so therapists go to the school, spend time with the peer group, or speak with members of the extended family. Examples of these factors might include poor discipline skills on the part of the parents or teachers, marital discord, parental substance use, lack of supervision, peer reinforcement of problem behaviours, neighbourhood culture which condones violence or encourages anti-social values, low commitment to education, chaotic school environment, poor parent-to-school communication, or financial stresses experienced by the family.
By identifying the "fit" between the problems and the broader systemic context, MST workers are defining both the targets of intervention and the indicators of whether the measures undertaken have been effective. A therapeutic strategy should produce observable results in the problem behaviour or else the strategy is revised. In other words, positive changes in the behaviour (e.g., school attendance) is used as indication that the intervention (e.g., parent contacting the school daily) is on the right track. Failure to achieve positive changes requires a reassessment of the "fit" and plainly indicates the need to try a new approach. The MST service providers are ultimately accountable for overcoming barriers to change. Blaming language such as "sabotage," "resistence," and "intractable problems" are not permitted. In fact, diagnostic labels of any type are discouraged in favour of a perspective that focuses on challenges and strengths.
MST is designed to be an intense but short-term involvement which can result in the generalization of treatment gains over the long-term. The frequency and duration of contacts will decrease over time, being intense in the beginning but lessening as improvements are observed. No social service intervention can last forever, so the ultimate goal is to empower the family or other care giver to continue with the strategies and interventions which were successful. The clearly articulated definition of success permits objective definition of when the case can be closed.
Since we began this project, the developers of MST have continued to test its efficacy, effectiveness and efficiency. High on the list of priorities is the need to learn more about its transportability, the real-world conditions under which MST be implemented with compromising its efficacy. Ours is one of ten randomized clinical trials now underway, including one in Norway. There are also five non-randomized studies in the United States. Several of these projects are testing MST for new populations of youth, such as sex offenders, substance abusers, emergency psychiatric patients, abused and neglected children, and school-referred youth. The first international MST conference is scheduled for October of 2000.
In addition, since we began this project, MST has been recognized and promoted as an effective technique by several prominent bodies in the United States.
Strengthening America's Families, a project funded by the Office for Juvenile Justice and Delinquency Prevention, identified MST as an exemplary program for delinquency prevention. Overseen by the Department of Health Promotion and Education at the University of Utah, this on-going project began with a literature review that examined 500 interventions in terms of evaluated results and ease of dissemination. They initially identified 25 model programs, 11 of which were elevated to the category of "exemplary" in the year 2000. Key among the findings is that family-focused interventions have greater empirical support than youth-focused interventions in preventing or reducing youthful criminal behaviour. Also important are matching programs to family needs, being developmentally appropriate, paying attention to recruitment and retention to make the program attractive for family members to attend, developing strategies to overcome barriers and measuring program effectiveness.
A study by the Watching the Bottom Line. Each program followed youths until the age of 25. None eliminated offending but 15 of the 16 documented lower rates of recidivism among program participants compared with control youth. After subtracting the cost of the MST intervention itself, MST saved taxpayers on average $7,881 (U.S.) per youth for services associated with criminal behaviour, such as incarceration. The cost of the intervention was recouped after two years. In addition, the reduction in crime was associated with $13,982 in savings to potential victims of crime. Five of the programs reviewed, including the Perry Pre-School Project, did not reduce crime enough to pay for themselves and none generated the level of crime-related savings linked to the MST intervention.
With state and federal funding, the Center for the Study and Prevention of Violence at the University of Colorado reviewed 450 violence prevention programs and identified ten model programs for violence prevention to promote as exemplary. Each "blueprint" is a practical description of an effective program and MST was one of the ten (Henggeler et al., 1998).
The National Institute of Drug Abuse identified MST as one of 12 scientifically based approaches to drug addiction treatment in a 1999 document called Principles of Drug Addiction Treatment. Among the principles of effective treatment they identified for substance abuse are that the treatment must be matched to each individual's particular problems and needs; treatment must attend to the multiple needs of the individual, not just the drug use; the treatment and services plan must be assessed continually and modified as necessary as the individual's needs change; and treatment duration depends on his or her problems and needs.
The Surgeon General reviewed over 3,000 research articles on mental health and mental illness to separate the practical and effective approaches from those that constitute little more than "good intentions" (U.S. Department of Health and Human Service, 1999). In the chapter on Children and Mental Health, they addressed both the causes and the treatments for serious emotional and behavioural problems among children and adolescents. A key conclusion was that the multiple problems associated with serious emotional disturbance among children and adolescents are best addressed with a systems approach. Multiple service sectors should work collaboratively and the participation of families is essential. Home-based services have a particularly "strong record" of effectiveness and MST was identified as an exemplary approach among both anti-social youths and those at risk for hospitalization.
Generally, it was concluded that community-based interventions, perhaps augmented with brief and targeted in-patient stays for period of acute crisis, had the best empirical support. In contrast, residential treatment evidenced only "weak evidence" of effectiveness from largely uncontrolled studies. When compared to community-based options, they were not more effective, were more expensive, and depended on the quality of post-discharge environment to maintain any gains that were made. Seriously aggressive and violent youths may be unlikely to improve at all. Moreover, several potential risks for residents were noted including failure to learn behavior needed in the community; the possibility of trauma associated with separation from the family; difficulty reentering the family or even abandonment by the family; and possible victimization by residence staff. In addition, proximity to deviant and disturbed peers increases the risk that anti-social and other negative behaviours will be learned or fostered. In conclusion, the Surgeon General believed it was "premature to endorse the effectiveness of residential treatment for adolescents." It was also necessary to learn for whom the benefits of residential care outweigh the potential risks.
While the Centre for Children and Families in the Justice System is undertaking the evaluation, the MST services are delivered through the collaboration of eight community agencies in four parts of Ontario:
For the first three years of the project, Ontario's Ministry of Community and Social Services has funded the provision of MST to qualifying youths and young offenders who fall within their mandate, specifically those under the age of 16. Beginning in April of 2000, referrals of Phase II young offenders will also be welcome from probation/parole officers of the Ministry of Correctional Services. This latter group is joining the clinical trial for the remaining year of the study.
MST services in this area are provided through Associated Youth Services of Peel. This team accepts referrals from the catchment area of the Mississauga Area Office of the Ministry of Community and Social Services, specifically Mississauga, Brampton, Oakville and Halton. Referrals are accepted only from the probation offices in this area.
The clinical supervisor for the first three years, Kelly McDonnell, has recently been appointed to the position of System Supervisor, as discussed below. To fill her position, one of the therapists, Lisa Bachmeier, was elevated to the position of clinical supervisor. As with Kelly before her, she has a half-time MST caseload and half-time supervisory responsibilities. The remainder of the team are John Choi, who has been with the project from the beginning, and Jennifer Watson who has just joined the team.
MST services are available to referred probationers across Simcoe County through the cooperation of two agencies: Kinark Child and Family Services and New Path Youth and Family Counselling Services of Simcoe County (formerly Catupla Tamarac Child and Family Services and Robert Thompson Youth and Family Centre). Both are children's mental health centres that offer a complete spectrum of community-based and residential services to families and youth. The catchment area for the Simcoe team includes Midland, Cookstown, Orillia and Bradford. This team is co-supervised by Stan Spicker of New Path and Barbara Curwen of Kinark. The three team members are housed in separate agencies and meet centrally for consultation. They are John Fiddes of Kinark in Midland, Laurie Teed of New Path (formerly Catupla Tamarac) in Barrie, and Brad DeLong of Kinark in Barrie. This site was chosen for inclusion in the MST project to test the viability of an MST team in a rural area with wide geographical dispersion. The presence of a large Aboriginal community in Simcoe County was also a factor.
In London, MST services are delivered through Craigwood Youth Services, a children's mental health centre that provides both community-based and residential services to youth, many of whom are in conflict with the law. The London MST team is composed of a clinical supervisor, Jon DeActis, and three full-time MST therapists: Michelle Edwards, Roger Houghton and Jody Lake. Penny Loube conducts the initial assessments to determine the suitability of a family for MST. All have been with the project from the beginning.
In London, MST is one of a range of services available through the Safer Community Program (SCP), to which families self-refer. The SCP services, that also include a W.D. Sutton school placement or a range of treatment groups (Choices, enhanced Choices, substance abuse, anger management, etc.), are overseen by an advisory committee made up of representatives from the agencies which participate in the SCP: Craigwood, the St. Leonard's Society of London and Madame Vanier Children's Services as well as the Children's Aid Society of London and Middlesex. To qualify for the MST program, the youth must have engaged in criminal behaviour but many of the referred youth are not on probation and some have never even been charged.
All referred families participate in an intake assessment to determine suitability for the various SCP services. The case is then presented to the Youth Access Committee (YAC). This committee is composed of representatives from the St. Leonard Society, the Children's Aid Society, the local probation office and, when required, a school board representative. The intake worker and the manager of the Safer Community Program (who is also the clinical supervisor of the MST team), also sit on YAC. The YAC members review the decision to assign the case to MST and also develop a contingency plan should the case draw the "usual services" option. Typically, that contingency option is selected from one of the other SCP services. If the case is deemed appropriate for MST, the family is approached about participating in the clinical trial. The case is referred back to YAC for the random assignment after the family agrees and signs the consent forms.
In Ottawa, the MST services are delivered through Eastern Ontario Young Offender Services with the cooperation of Crossroads Children's Centre and the William E. Hay Centre, a closed custody facility for Phase I young offenders. Crossroads and the Hay Centre have each seconded a staff member to the team since the on-set of the project. They are Alain Corriveau from Crossroads and Heather Jenkin of the Hay Centre. Also on the team are the clinical supervisor, Vickie Jennings, and MST therapist Ned Jackson, who are both with EOYOS and came to the project from that agency's Community Support Team for young offenders.
Referrals are accepted from a variety of sources that include the youth division of the Ottawa Carleton Regional Police Service. Crossroads Children Centre is also a large source of referrals. In this site, referrals for youth under 12 years old are accepted for MST, in an attempt to see if this approach is useful for youngsters who are manifesting serious criminal behaviour at an early age.
As noted above, the MST training begins with a one-week course provided by MST Services Inc. However, only through application to actual cases can even an experienced therapist achieve a satisfactory level of MST adherence. The training process is on-going, involving didactic instruction on case formulation, weekly feedback on clinical interventions and case progress, and boosters to consolidate skills. Each therapist completes a weekly summary of all active cases for a telephone consultation. For the first two years of the project, these weekly summaries were faxed to the MST consultant in South Carolina, Dr. Christine Hamel. During conference calls that averaged two hours in duration, Dr. Hamel provided case-specific feedback and guidance to help the teams gain proficiency in the MST method. Now these consultations are being conducted by the System Supervisor, with a greater emphasis on peer supervision.
Quarterly booster sessions are opportunities for the four teams to meet at a central location to receive continuing training on MST and related issues. The System Supervisor arranges the agenda, based upon an assessment of need areas. Twice each year, the new MST consultant, Dr. Dan Edwards, will travel to Canada to assist with the booster training. Topics addressed have included worker stress, marital therapy, substance abuse treatment, suicide assessment, and how all of these issues are responded to within the MST framework. The members of the research team are also in attendance at booster sessions, as are interested parties from the local area who want to learn more about MST.
As consultation with MST Services in South Carolina wound down after two years, there was a need to develop the internal capacity of the Ontario teams to provide case supervision. The position of System Supervisor was filled by one of the four clinical supervisors. This move is part of a strategy to cultivate, within Ontario, the expertise to train and supervise MST workers with less reliance on the developers of MST in South Carolina. Throughout this process as it evolves, the intention is to support the teams in maintaining a high level of adherence to the MST model. Local expertise will also provide a base for the expansion of MST to other areas of the province. The new System Supervisor, formerly the Clinical Supervisor of the Mississauga Area Office Team, officially assumed her duties on April 1, 2000. She now directly supervises the four Clinical Supervisors and is assisted in her new role by Dr. Edwards of MST Services Inc. in South Carolina.
The methodology employed here accommodates three different information needs. First, the evaluation charts outcomes. Put simply, an evaluation should be able to document the degree of success in achieving stated goals. The benefits of outcome evaluation include accountability to funders, consumers and the public. This information also contributes to the knowledge base in the area of prevention. Outcomes need to be comprehensive and long-lasting. That is, the benefits of the program should not only be observed in the short term but also sustainable over time. Another goal of MST is to decrease the services utilized by such youths. It is here that program outcomes can be related to cost-effectiveness and service utilization rates.
Second, the evaluation will monitor program delivery to ensure treatment fidelity. This is sometimes called process evaluation. Program integrity is crucial to any test of a program, to be able to unambiguously relate outcomes to the program as defined. It is also important to be attentive to the possibility of program drift and intervene when it is observed. Especially with a best practice model compiled from the literature, as with MST, drifting from that practice may dilute the success of the program overall.
Third, the design will accommodate the need for comparative information, specifically the portability or transferability of the program components to any community and for use with any defined group. Comparative information is best gathered by implementing the same program in several areas. All programs, even those with demonstrated positive outcomes, do not work equally well in all communities. The four participating sites vary in terms of population size and density, urbanism, ethno-cultural profile, proximity to major centres, and sophistication of social service infrastructure.
MST supervisors and therapists have responsibility for on-site research tasks. Supervisors screen cases against the exclusionary criteria and maintain files with referral information. The therapists themselves seek consent from families in their homes, administer pre-tests, debrief control group members, contact the school regarding the control group, and administer post-tests to MST and control cases. They also review control cases for abuse and suicide risk and take appropriate action when necessary.
Referrals to MST come from the community, typically probation officers or other case managers. The ideal MST referral is a youth who presents a challenge to existing community- based interventions and who is facing the prospect of an out-of-home placement in custody or a residential setting. Each site has devised a local variation for securing referrals and processing cases. The common goal is to select youths who are at high risk for committing criminal offences in the future, particularly those likely to be incarcerated as young offenders. The goal is to have a referral strategy that limits referrals to the most appropriate cases and that refers potential candidates in such a way that no MST waiting list is created.
Prediction of future offending among the youths under 16 is made with the Risk/Need Assessment (RNA) instrument developed by Drs. Robert Hoge, Don Andrews and Alan Leschied. This tool has been used by the Ministry of Community and Social Services for many years. Youths who qualify for MST will fall into one of two categories: high (scores of 27 to 34); or, very high (35 to 42). If insufficient number of these cases are available, clinical supervisors have the discretion to include "high moderates" (scoring 21 to 26). In addition, some youths who score lower than 27 will have been overridden into a higher category by a probation officer. They will also qualify for MST. As noted above, referrals are accepted for youths under 12 years of age in Ottawa, meaning they will have no official criminal record and no RNA scores. A modified version of the RNA is used to determine their risk for future offending.
For the Phase II young now being referred to the project, the prediction of criminal risk is made with the youth version of the Level of Services Inventory, a tool which is widely used in the probation/parole offices of the Ministry of Correctional Services. Potential referral agents have been told to restrict applications for MST to the cases that score in the "high" risk category.
In summary, youths become eligible for MST when it has been determined they have a high risk for future criminal conduct. That is the first and most important criterion, but there are other factors that must be taken into account in the referral process. The timing of referrals is sometimes such that waiting for a while is prudent. If a youth is before the court facing a high probability of a lengthy custody term, the referral is best put on hold until later.
In addition, some referrals will not be appropriate for the MST trial. Cases will not be screened for treatment amenability or excluded because of poor prognosis for success. However, there are two categories of exclusionary criteria that used to determine when the youth does not qualify for MST (at least at this point but potentially later if the situation changes):
The first category of exclusionary criteria requires consideration of these four factors:
1. Requisite Level of "Family" Involvement
MST being a family-based intervention, a youth must have at least one adult caregiver. This may be a parent but could also be an older sibling, grand-parent, aunt, uncle or friend of the family. A Crown ward in a stable foster placement could qualify. However, a CAS client in a new placement may not qualify, as there is no way to determine if the placement will break down. Typically, youths in group homes or other residential settings will not be suitable MST candidates unless a family reunification is imminent or a substitute caregiver can be identified.
2. Current Family Therapy
If the family is already engaged with a therapist and making gains, the intervention of a MST worker would be neither needed nor appropriate. Should the arrangement break down, however, a referral could be made.
3. Safety of Youth and Family
MST uses a family preservation model but some families cannot be preserved safely. When assessing the appropriateness of an MST referral, safety concerns override all others, whether that involves youths who are at risk of abuse, at risk of suicide, or at risk of harming other members of the family. MST is not a substitute for CAS involvement, in-patient hospitalization, or community safety through custody/detention.
4. Risk of Injury to Worker
Clinical supervisors, perhaps in consultation with the police or probation officers, have the discretion to disqualify a case from the clinical trial because of a risk of injury or harm to the MST worker while in the family home. This situation is NOT indicated merely by family violence or assault convictions.
The second category of exclusionary criteria pertains to the types of cases with which MST has been demonstrated effective. It has been tested on youths with many presenting problems, all of whom have one thing in common: criminal behaviour. Based upon clear direction from MST Services Inc. South Carolina, two groups are ineligible for MST at this point in time:
5. Sex Offenders
Sex offenders must be excluded because MST has not yet been demonstrated as effective with this group (although a project is under way to adapt MST to this purpose).
6. Acute Psychosis
Youths who are acutely psychotic are not candidates for the MST clinical trial. However, a psychiatric diagnosis is not a disqualifying factor in itself.
The frequency and rationale of case exclusions are being tracked.
When a youth is referred to MST and meets the inclusionary criteria, the next step is to approach the family about participation in the clinical trial. In most cases, the option of MST will already have been discussed with the family by the person who made the referral. The family will have signed a consent form agreeing to the release of their names to the MST team. Signing such a consent does not oblige the family to participate in the study and they are free to refuse consent once they are told about the study.
The next step is for the MST team to contact the family and explain to them both how MST works and the contingencies of the clinical trial. Consistent with the MST philosophy, this meeting will almost always occur in the family home. It is not uncommon that the youth is living elsewhere at the time, perhaps in custody, which necessitates two sessions at different places.
Strict observation of ethical principles is crucial in this process. The characteristics of an ethical study are that:
For the MST clinical trials, the implications of these standards are these. Potential participants must understand:
A Letter of Information for potential participants provides a written explanation but the worker must be assured that each person fully understands MST and what the study involves. This usually entails giving a verbal explanation after the letter is read and answering any questions. (There is a French language version used in Ottawa when the need arises.) Those who agree to participate, both the youth and a parent/guardian, will sign two consent forms: one consenting to participation in the clinical trial and one granting permission to access the police and correctional data bases used in the follow-up.
The frequency and rationale of decliners is being tracked. The rate of refusal must be low enough to avoid the possibility of a volunteer bias and to keep the generalizability of the results high. It is important not to systematically screen out treatment-refusing youth. The MST intervention can proceed without the active participation of the youth, but he or she must sign the consent form.
Preliminary observation suggests that the parents are more likely to decline participation than are the youths. Some parents are reluctant to have a therapist in the home or the intensive nature of the intervention does not appeal to them. Some families fail to see the need for counselling while others are satisfied with the services already being provided by other agencies. Finally, there is a tendency of some parents to believe that the youth should be the "identified client" rather than the whole family. Families initially categorized as refusers may be re-referred if they still meet the inclusionary criteria and later indicate a willingness to participate.
The pre-testing is administered once the family has given consent and before the random assignment is made. Part of the consent process involves an understanding on the family's part that they are required to complete these forms. All the forms are self-administered, some completed by the youth and some by a parent. However, where literacy is an issue, the worker will assist in their completion. This process usually takes place in the family home immediately after they have signed the consent forms, except in London because of the different referral procedures.
The instruments have been chosen to reflect several key domains which MST might impact as well as to gather some socio-historical information on the family. Five instruments are used, all of which are re-administered at the end of the MST intervention. Two of them have teacher versions. The families are asked to sign a consent to give permission for their completion by the school.
1. Standard Client Information System (SCIS)
These forms, developed by Offord and Boyle as part of the Ontario Child Health Survey, are used already by many members agencies of Children's Mental Health Ontario (formerly the Ontario Association of Children's Mental Health Centres). Permission for their use has been secured. At intake, these four forms are used:
2. Beliefs and Attitudes Scale
Developed by Steven Butler and Alan Leschied, this is an instrument similar to the Criminal Sentiments Scale but designed for use with adolescents. The Self-Report of Youth Behaviour is a corollary test which is administered at the same time.
3. Family Adaptability and Cohesion Scale - II (FACES)
FACES-II is a 30-item scale that measures family adaptability (negotiation style, roles, assertiveness, leadership, discipline, child control, rules) and family cohesion (emotional bonding, coalitions, space, family boundaries, shared time/friends, decision-making, and shared activities). Permission to use this instrument was secured from the authors, who recommended the second version, rather than the third of fourth one, as being best for research. It is to be administered to both parents and to the youth.
4. Social Skills Rating System
There are three forms, one each for youth, parent and teacher with separate versions for elementary school (grades 3 to 6) and secondary school (grades 7 to 12). Norms are available for learning disabled youth.
5. Parental Supervision Index
A two-item parental supervision index (Jang & Smith, 1997) asks the youth to answer two questions: 1) during the course of a day, how often (do/does) your parent(s) know where you are; and 2) how often would your parent(s) know who you are with when you are away from home?
Random assignment creates two groups of equal size that are identical, especially in terms of variables which might impact future offending (criminal history, etc.). The key tenet of random assignment is that each youth has an equal chance (i.e., 50/50) of being assigned to the treatment condition. The other half continue to receive the services available to them in their communities and they constitute the control group. When the assignment to groups is random, you can assume the two groups differ only in the intervention they receive.
Random assignment occurs in the family's home with the MST worker present, so they know immediately into which group they have been placed. In London, where the random assignment is conducted by the Youth Access Committee, members of the control group are immediately assigned to another service within the Safer Communities Program, chosen to match the youth's individual needs.
If the family is assigned to the control group, the case manager, who is usually the probation officer, is notified and the youth and family carry on with the original case management plan. An ethical obligation does attach to the MST team, however, in that information may be revealed in the testing that needs to be addressed. Specifically, the responses to two items in the Standard Client Information System could indicate that a youth is at risk for suicide. If either or both of these items are endorsed by a youth, the therapist will pursue the issue further with the youth and, when age-appropriate, with a parent. This assessment may involve consultation with the clinical supervisor and another visit to the home. Therapists and their clinical supervisors are able to consult with Principal Investigator Dr. Alan Leschied in London, a clinical psychologist, if the need arises. When warranted, a referral to the appropriate agency will be made.
The process of securing consent and administering the intake battery is a time consuming one for the MST therapists, but it is desirable that an experienced clinician do so because of the need to assess risk. Another situation that arises in these sessions is the disclosure of child abuse. A parent may endorse the use of physical discipline with their children. Where this occurs, the MST therapists review the matter with the parent, asking question to ascertain level and immediacy of risk. When necessary, where both the control and the MST groups are concerned, the team will contact local child protection officials. This move has been necessary on several occasions.
Treatment fidelity to the MST method is important for two reasons. First, it has been demonstrated that level of adherence to the MST method is correlated with outcome (Henggeler et al., 1997). Second, because this is so explicitly a test of MST, we cannot draw conclusions about the impact of MST without knowing that it was, in fact, MST we were evaluating. The MST Therapist Adherence Measure was developed at the Family Services Research Center in South Carolina. It is based upon the nine MST principles outlined above. In our study, the family is asked to complete this form at two points during the intervention and again when the case is closed.
The MST teams are now cooperating with a 27-site dissemination study being conducted by the Family Services Research Center with funding from the National Institute of Mental Health in Washington. In consequence, the administration of the adherence form is now being accomplished via telephone by researchers at the Medical University of South Carolina. A Supervisor Adherence Measure has also been developed but it not being used as part of the Ontario study.
When the case is closed, the instruments from the intake battery are readministered. The members of the control group are contacted five months after intake and asked to complete the testing again.
Follow-up data on recidivism and correctional service utilization are collected from three sources:
While all participants sign consent forms granting such access, we have also taken the step of securing a court order. In addition, access to OMS records from the Ministry of Correctional Services was granted by that Ministry's research committee. Records are checked six months and after one two and three years post-discharge.
To facilitate the analysis of research data, five data bases have been created:
1. Referral Information, Criminal History and Risk Need Data
This statistical data base contains referral information including the risk/need scores. The RNA is always completed by the probation officers except those cases in two sites (London and Ottawa) where a portion of referrals comes from other sources. In these cases, the RNA is completed as part of the referral or intake package, by the referral agent, intake worker or clinical supervisor. Young Offender Strategic Information System (YOSIS) data are being provided by the Ministry of Community and Social Services as an indication of prior history with the youth correctional system. CPIC data are also useful in this regard.
2. Pre/Post Testing Data
One data base contains the summary score results of the five psychometric instruments completed by the families at the time of intake, before the random assignment is made. These instruments are re-administered at discharge. A complete set of parent/youth intake testing is available from 87% while one or more test is missing in the case of 19, or 13%. Teacher intake testing is available in 60% of cases, the shortfall due to non-responding by teachers or the fact that the youths were not attending school at the time.
3. MST Process Data
For the half of the sample that receives the MST intervention, information on that process is collected, including the name of the MST worker, the length of time the case was open, the number of sessions undertaken, and the reason the case was terminated if it ended prematurely. In addition, the MST adherence scores are recorded here for later use, to correlate with recidivism data when available.
4. Case-specific Demographic Information
This data base is used to keep track of the key dates in the project, specifically the closure of cases and the times at which criminal record checks need to be conducted. The file contains the names and birth dates of the youths so that lists of names can be generated for secure transmission to the police and correctional agencies which will conduct the checks.
5. Recidivism Data
As noted above, information on reconvictions is being collected from the Canadian Police Information Centre, the Ministry of Community and Social Services (YOSIS) and the Ministry of Correctional Services (OMS). This data base is structured to permit the generation of survival curves (months without convictions and days before first incarceration) as well as information on the number, severity, timing and frequency of convictions after intake into the MST study. Information on custody stays in five types of facilities (Phase I open custody, Phase I secure custody, Phase II open custody, Phase II secure custody and adult facilities) will be collected to examine the cost effectiveness of MST. Recidivism data will also be available for correlational analyses with factors such as MST adherence, gender, age, risk score, etc.
The data available for this report will focus on a client profile, preliminary results of the post-testing, and the six month post-discharge reconvictions of 145 youths. These data are taken from the first 147 cases inducted into the study. This includes 46 cases from Simcoe County (31.3%), 34 cases from London (23.1%), 42 cases from Mississauga/Peel (28.6) and 25 cases from Ottawa (17%). Half have received MST (73) and the other half constitute the control group (74). Data on more recent cases are now being assembled.
One quarter of these referrals (26%) were young women. The youngest youth was
10 at intake (in Ottawa which takes under 12 referrals) while the average age
was 14.3 for boys and 15 for girls. The youth referred in Mississauga/Peel were
older on average (15.7) while the youth in Ottawa tended to be younger (13.6).
The age breakdown for all 147 youth is found in Figure 1.
Using the risk/need assessment conducted at intake, about one third (36%)
were rated as moderate risk to re-offend, 57% as high and seven percent as very
high. None of the MST referrals was rated as a low risk. The override was
applied in 30% of these cases, almost always to place the youth in a higher risk
category than did the straight summation of sub-scale scores. Both boys and
girls had the same average score of 24 (see Table 1), which approaches the top
of the moderate category (9 to 26). Girls had significantly higher average
scores in two areas: substance abuse and leisure/recreation. They were also
higher on family and attitudes but the differences with the boys' scores were
not significant. Boys tended to score higher on education/employment.
TABLE 1
Average Risk/Need Scores on Eight RNA Subscales by Sex of
Youth
| Boys | Girls | |
| Prior/Current Offences or Dispositions (0 to 5) | 1.44 | 1.03 |
| Family Circumstances (0 to 6) | 4.20 | 4.49 |
| Education/Employment (0 to 7) | 4.71 | 4.21 |
| Peer Relations (0 to 4) | 2.93 | 2.87 |
| Substance Abuse (0 to 5) | 1.22 | 1.69 |
| Leisure Recreation (0 to 3) | 2.07 | 2.44 |
| Personality/Behaviour (0 to 7) | 4.75 | 4.59 |
| Attitudes/Orientation (0 to 5) | 2.75 | 2.87 |
| TOTAL (0 to 42) | 24.44 | 24.11 |
Overall risk/need score was not correlated with family adaptability or cohesion, parent reports of internalizing symptoms, teacher reports of externalizing symptoms, teacher ratings of social skills, problems behaviours or academic competence or parent reports of social skills. However, there were small correlations between RNA score and parent reports of conduct problems (r=.24, p=.006), oppositional behaviour (r=.22, p=.012), and general behaviour problems (r=.24, p=.01). In addition, teacher reports of depressive symptoms were negatively correlated with RNA score (r=-.30, p=.01).
One quarter (26.5%) of the parents were born outside Canada, while almost all (91.5%) of the youths were born in Canada. Youths born outside of Canada typically emigrated here when young but two of them came to Canada only five years prior to referral. English was the language in the home for 92% of families while one family was French speaking. Twelve percent of the referred youths were described by their parents as being a status or non-status Indian, Métis or Inuit.
Many of the referred youths are the youngest in the sibline (34%) or are only children (13%). Only one was described as being adopted, a much lower number than expected. The referred youth was the only child in the home in 37.4% of cases while the remainder of families had between one and five other children. In 29% of the families, there were five or more people living in the home.
The income distribution is skewed toward the lower levels with one half of the families (47.7%) earning less than $20,000 per year. Fifty-one percent are on social assistance and a further 13.5% are on a pension or drawing worker's compensation. In addition, 10 percent of the families derived at least part of their income in the previous year from employment insurance. One fifth of the parents (19.4%) said they were unemployed or had lost a job at some point in the previous twelve months and almost two thirds (63.9%) indicated they had "financial problems." However, almost two thirds (62.6%) were employed at least some of the time in the previous year. Probations officers who completed the risk need assessment believed that 36% of the families had financial or accommodation problems.
When asked to describe their child's general health, 90% of the parents indicated that it was good or excellent. Only one youth was described as having poor health but one tenth (11.1%) had a health problem that limits participation at school or age appropriate activities. It is clear that this sample of youth have had more than their share of medical problems, beginning before birth. When pregnant with the youth, one fifth of the mothers were hospitalized for complications of the pregnancy. However, the rate of extremely low birth weights (less that 2,500 grams) was low at only 5.2%. One tenth (11.1%) spent at least three consecutive months away from their parents during the first year of life. Seventeen percent of parents indicated that their children had at some point been so sick that they thought he or she might die. One quarter had suffered a broken bone. More then one third (36%) had stayed overnight in a hospital, not including at birth. Ten of them had a hospital stay during the previous ten months. Twenty-four percent had been treated in an emergency room in the past six months. More than one quarter of the youths were on prescribed medication, including 20% who were taking medication such as Ritalin for hyperactivity or behaviour control.
The MST approach assumes that "anti-social behavior is multi-determined by the reciprocal interplay of characteristics of the individual youth and the key social systems in which youths are embedded." These systems include the family, peers, school and the neighbourhood.
Characteristics of families that are correlated with delinquency include lack of monitoring, lax and ineffective discipline, low warmth, high conflict, and parental difficulties such as drug abuse, psychological problems and criminality. We found that the families in our sample of cases referred to MST correspond to many of these factors.
According to the youths, levels of parental supervision at the time of referral were low. As illustrated in Figure 2, only 29% of the youths indicated that their parents often knew where they were and 23% said their parents often knew who they were with. The profile was similar for males and females. Probation officers also had concerns about supervision (Table 2).
TABLE 2
Family Circumstances/Parenting Rating from Risk/Need
Assessment
| Boys | Girls | Total | |
| Inadequate Supervision | 67.6% | 60.6% | 65.9% |
| Difficulty Controlling Behaviour | 95.2% | 100% | 96.4% |
| Inappropriate Discipline | 49.5% | 63.6% | 52.9% |
| Inconsistent Parenting | 81.9% | 75.8% | 80.4% |
| Poor Father/child Relations | 72.4% | 75.8% | 73.2% |
| Poor Mother/child Relations | 47.6% | 69.7% | 52.9% |
It was also clear that these families were experiencing a high degree of conflict and problems with discipline. Raters using the Risk Need Assessment, under the category of Family Circumstances and Parenting, placed half of the youths in the "high" risk category, more likely true for girls (54%) than for boys (45%). The most commonly observed parenting problem was a difficulty controlling the youth's behaviour, true for 96% of the cases (see Table 2). Indeed, the parents acknowledge this themselves in that 88% felt in need of outside help in disciplining their child. Moreover, one third of the parents reported that their child's behaviour made it often or always difficult to take them shopping or visiting. Another 35% said this was sometimes true. Similarly, about one third of parents reporting foregoing a vacation in the past year because of the child's behaviour. Almost two thirds (62%) had quarrelled with a spouse about the child's behaviour and virtually all the parents (97%) had worried about their child's chances for doing well in the future.
According to probation ratings on the Risk/Need Assessment, these youths were generally beyond the control of their parents, the parents were often inconsistent and, in half of the cases, inappropriate discipline was used (see Table 2). Also according to the probation officers, the families were often characterized by marital conflict (58%) and parental drug or alcohol abuse affected the referred youth in 41% of cases. This latter figure contrasts with parental self-reports where there were infrequent acknowledgements of problems related to drinking. However, 28% of primary caregivers reported having felt they should cut down on their drinking. Probation officers reported parental histories of chronic offending in 11% of cases (16.8% of parents acknowledge a previous arrest). They also felt that at least one parent suffered from emotional distress or psychiatric problems in 32% of cases. Moreover, 31% of youths were described as having an abusive father and 12% as having an abusive mother. In one in ten cases (12%) the parents were described as being uncooperative with intervention efforts.
One third (32.4%) of the caretakers were in the clinical range for depression. Over one third (36%) acknowledged when asked that they are not too happy these days. Health wise, the parents reported a high rate (34.2%) of chronic health problems of either a permanent or long-term (more than six months) nature. Over one fifth (22%) indicated that a medical problem limited their ability to carry out normal activities at home, school or at a job. Moreover, 31% replied in the affirmative to this question: "have you ever been treated for 'nerves' or a nervous condition. Half of them had been treated in the previous six months.
Primary caregivers are asked to rate some aspects of their family functioning. Using the norms available through the Standard Client Information System, the average t-score on overall family functioning was 65 while the average t-score for householder depression was 64, both scores that approach the clinical range. Twenty-nine percent of the families were in the clinical range on family functioning.
Peer factors correlated with propensity for criminality include association with deviant peers, peer relationship skills and low associations with pro-social peers. More information on this topic will be available in subsequent analyses. At present, we can observe that teachers rated the youths as falling on the 12th percentile on average for social skills and the 84th percentile for problems behaviours in school (see Figure 3). Also note that in Table 3, probation officers rated 78% of the youths as having problems with peer relations.
School factors correlated with criminality in youth include low achievement, early school leaving, low commitment to education, and a weak and chaotic school environment. Our understanding of school performance comes from the reports of parents, teachers and probation officers. About one quarter of this sample was in elementary school and the remainder in secondary school.
At referral, about 17% of the youths were no longer in school, most commonly because they had dropped out but almost as likely because they were excluded by the school system. Many had histories of multiple school placements, with an average of 4.9 different schools. Indeed, almost half had been in five or more different schools and ten percent had been in nine or more schools. In addition:
The parents of almost half of these youth (46.4%) described their children as
having difficulty learning. As noted above, 20% were taking Ritalin or similar
medications. With reference to the past six months, few parents (10.5%)
described their children as good students. Two thirds (63.4%) of the youth were
described by their parents as below average students. Teacher ratings were
similar in that 76% were described as performing below average. Again according
to teacher descriptions, this sample of youth was described as falling at the
16th percentile, on average, for academic competence (see Figure 4).
TABLE 3
Education/Employment Ratings from Risk/Need
Assessment
| Boys | Girls | Total | |
| Disruptive Classroom Behaviour | 76.4% | 51.5% | 70.5% |
| Disruptive Schoolyard Behaviour | 70.8% | 45.5% | 64.7% |
| Low Achievement | 88.7% | 87.9% | 88.5% |
| Problems with Peer Relations | 77.4% | 78.8% | 77.7% |
| Problems with Teacher Relations | 86.8% | 69.7% | 82.7% |
| Truancy | 67.0% | 81.8% | 70.5% |
| Unemployed/Not Seeking Employment | 7.5% | 15.2% | 9.4% |
The parents' commitment to education may have been low. More than one third of the primary caregivers (37.5%) had not finished high school while 15.8% had completed a college diploma and 2.6% had graduated from university. The profile was similar among their spouses where 39.4% had stopped short of high-school graduation.
Characteristics of neighbourhoods associated with elevated risk of youth crime include high mobility, low support available from neighbours, high disorganization and criminal subculture. Little information is now available on these factors but it is our goal to collect crime rate data by family postal code in the future.
It is the assumption of the MST approach that the above-listed system factors interact, in a reciprocal or iterative way, with youth characteristics. Key variables of youth with elevated propensity to criminality include low verbal skills, favourable attitudes toward anti-social behaviour, psychiatric symptomatology and a cognitive bias to attribute hostile intentions to others. Again, we find that the youths referred to the MST program present with a range of individual-level problem behaviour that has already manifested in criminal conduct.
TABLE 4
Substance Abuse Ratings from Risk/Need Assessment
| Boys | Girls | Total | |
| Occasional Drugs Use | 60.4% | 75.8% | 64.0% |
| Chronic Drug Use | 19.8% | 33.3% | 23.0% |
| Chronic Alcohol Use | 11.3% | 27.3% | 15.1% |
| Substance Use Interferes w. Functioning | 20.8% | 33.3% | 23.7% |
| Substance Use Linked to Offence(s) | 12.3% | 6.1% | 10.8% |
The majority of youths were known to use drugs occasionally. Probation
officers saw a link between criminal offending and substance use in 11 percent
of cases in this sample (see Table 4). Two thirds of the youths (68%) admitted
at least once having three or more drinks at one time. On average, they were
12.4 years old the first time that happened and a few were in the primary
grades. Almost half said that they had been drunk at least once in the last 12
months. The youths were also asked about their drug use and 55% reported using
cannabis products during the last six months. Less commonly reported was the use
of psychedelics (17%), stimulants (10%), glue or gasoline (7.7%), cocaine or
crack (7.1%), and narcotic depressants (5.8%). As noted above, probation
officers were more likely to see substance abuse as a problem among the girls
than the boys. One finding of note was that 81% of the youths said that they had
smoked cigarettes every day for a month or longer. On average, they were 11.6
years old when first this statement was true.
TABLE 5
Prior and Current Offences/Dispositions from Risk/Need
Assessment
| Boys | Girls | Total | |
| Three or More Prior Convictions | 30.5% | 18.2% | 27.5% |
| Two or More Prior Failure to Comply | 21.4% | 15.2% | 19.9% |
| Prior Probation | 35.2% | 24.2% | 32.6% |
| Prior Custody | 37.1% | 33.3% | 36.2% |
| Three or More Current Convictions | 30.5% | 27.3% | 29.7% |
Because of the different referral streams, not all youths in this sample have a formal criminal record. However, all must have a history of offending to qualify for MST. Some information about prior record, as recorded in the RNA, can be found in Table 5.
As part of the intake battery, a number of psychometric tests are
administered to the youth and the parents complete several tests about their
children's behaviour. In Table 6, the parent responses to the Stand Client
Information System are summarized. Compared to Ontario norms, where a group of
average youth would have an average score of 50, this sample has a high level of
externalized behaviour problems. In fact, 81% fell above the clinical cutoff for
conduct problems. This means that only two youths out of 100 would be expected
to score at that level or above. Also of note was the high rate of internalizing
problems such as anxiety and depression. This is clearly a sample of youth with
extraordinarily high levels of emotional and behavioural problems.
TABLE 6
Parent Ratings on Standard Client Information
System
Low |
High |
Mean |
SD |
%age over clinical cutoff | |
| Conduct Problems | 42.6 | 165.0 | 96.6 | 26.4 | 81.4% |
| Oppositional Problems | 39.1 | 89.3 | 72.1 | 12.6 | 58.6% |
| Attention Deficit | 37.0 | 95.5 | 67.7 | 13.4 | 50.0% |
| EXTERNALIZING TOTAL | 42.1 | 105.0 | 77.3 | 15.3 | 70.0% |
| Over Anxious | 37.1 | 103.0 | 65.8 | 12.6 | 17.1% |
| Separation Anxiety | 36.1 | 90.8 | 58.3 | 11.7 | 23.9% |
| Depressive | 39.4 | 112.0 | 62.5 | 14.6 | 49.6% |
| Social Relations | 36.6 | 102.0 | 67.8 | 13.0 | 52.7% |
| INTERNALIZING TOTAL | 37.9 | 91.8 | 70.3 | 11.2 | 34.3% |
This sample of youths had already accessed a wide range of social service and mental health interventions: According to parent report:
Also, as noted in Table 5, about one third had previously spent periods in youth custody.
According to parent report, one quarter of the youths (24.85) had been physically abused and 14% had been sexually abused but we have no information on whether these incidents were perpetrated by family or non-family members. When the youths were asked about prior abuse, 36% said they had been physically abused and 16% said they had been "sexually molested." Self- report of prior physical abuse was more common among girls (43%) than boys (34%), as was self- report of prior sexual abuse (31% and 12% respectively).
Post-testing is administered to the MST group by the therapist assigned to the case. Control group post-testing is gathered by intake workers in some sites and through the mail in others. Post-testing data are not available for 21 MST cases and 23 control cases, representing a response rate of 70%. With any response rate less than 100%, the issue of response bias must be investigated. Specifically, there is the possibility that families who do not respond are different in an important way from those who do. Responders and non-responders were compared in terms of their intake testing scores and this analysis revealed the two groups were significantly different on six of the 33 scores of the Standard Client Information System. Non-responders scored higher on parent report of overall internalizing symptoms as well as two individual sub-scales: over anxious and depressive. They were also higher in teacher reports of conduct problems, oppositional behaviour, and problems in social relations. On the Social Skills Rating System, non-responders scored higher on average on teacher reports of overall problem behaviours. However, non-responders were not different in terms of family adaptability or cohesion.
Teacher post-testing data are not available for 43 MST cases and 43 control cases representing an overall response rate of 41%. This low response rate makes analysis of pre/post scores unwise.
Comparison of pre and post scores has revealed that both the MST and control groups, as groups, show improvement in many areas, especially externalizing problems. This observation validates the need for a control group to be able to distinguish changes brought about by MST from changes that would have occurred anyway as youths mature and/or are exposed to other interventions. At this point, with a small sample, the short-term impact of MST on the variables measured psychometrically is difficult to determine. For example, with the Family Adaptability and Cohesion Scale (FACES-II), youth report of family cohesion (t=-2.1, df=38, p<.05) and family adaptability (t=-2.0, df=38, p=.05) increased significantly in the MST group. On the other hand, youth reports of family cohesion and adaptability did not change in the control group. However, there were significant differences observed in parental reports for both groups. Parent reports of family cohesion (t=-2.6, df=38, p=.01) and family adaptability (t=-3.3, df=38, p<.01) increased in the MST group but the families in the control group also saw improvements by parent report (t=-2.5, df=43, p,.05; t=-2.5, df=43, p<.05).
It is interesting to note that neither group evidenced improvements in internalizing symptoms, not an MST treatment target and not likely to be the focus of criminal justice-based interventions such as probation. The MST group did evidence significant improvements in general family functioning (t=2.0, df=48, p=.05) and parental depression (t=2.9, df=47, p<.01), as measured by the Standard Client Information System. On these variables, the control group did not change at all. It is possible to speculate that the MST intervention has succeeded in changing the youths' home ecology thereby, perhaps, setting the stage for more enduring changes or more lasting impact on youth behaviour. It remains to be seen how these patterns of short-term changes in symptomatology and family functioning will translate into long-term changes in criminal behaviour. The next phase of the study will involve collecting and analysing recidivism data which will permit a survival analysis and cost efficiency assessment of MST.
At three years into the study, we are starting to see how MST is affecting the target behaviour of crime. At the close of each MST case, and six months post-intake for the control group, we are tracking re-involvement in the justice system as measured by re-conviction. Unfortunately, some of these convictions are for offences that pre-dated the intake and we cannot isolate them, in the absence of any reliable source of information on the offence date. However, this factor will affect the MST and control groups equally. It is also important to note that our measure of "recidivism" will no doubt be an underestimate of the actual amount of crime committed by these young people. Again, the factors that contribute to this error -- including failure of victims to report, police inability to apprehend the perpetrator, clearance by other than charge, and pre-adjudication termination of the case in the courts -- will affect both groups equally.
Figure 5 illustrates a survival curve of the first 145 cases who reached six months post- discharge. After six months, 21% of the MST group had been re-convicted of a criminal offence (excluding breach of probation) compared with 28% of the usual services group. For youths who were in custody at discharge from MST or at administration of post tests, the zero point for the survival analysis was the day of release. This group has now reached one year post discharge and the CPIC data are being collected.
In the last year of the clinical trial, emphasis will be on amassing the data set to permit the answering of the research questions posed at the beginning of this document. Specifically, focus will be on:
1. Process evaluation to determine the challenges and barriers to the successful implementation of MST likely to be faced by those who seek to adopt this method in other parts of Canada
2. Continued monitoring of the short and long-term outcomes of the sample youth that now numbers over 300
3. Explaining variations in outcomes using statistical prediction models and theoretically derived variables such as MST adherence levels
4. Cost-efficiency analysis that monetizes the cost savings associated with MST
Preliminary results will be presented at the MST conference in October.