Multisystemic Therapy

Study design and sample

The first study is a rigorously conducted RCT. 

This study involved random assignment of children to a MST or to a standard package of youth offending services involving individual and family support.  The trial involved 108 youths and their parents.  Treatment allocation was determined by a stochastic minimization program (MINIM) balancing for type of offending (violent vs. nonviolent), gender and ethnicity.

This study was conducted in the UK with a sample of children.  Families were eligible if they had a child between the ages of 13 and 17, who was living at home with at least one parent or caregiver, and was a recipient of a court order within the last three months.

Measures

Information about youth offending was gathered from police records at six months pre-treatment, six months post-baseline (typically just post-treatment), 12 months and 18 months.  Parents and youths also complete a battery of validated self-report measures at baseline and at six-months – i.e. immediately post-treatment.  These measures included:

• Self-report of Youth Behaviour (SRYB) (self-report)
• delinquency and aggression subscales of the Youth Self-Report (YSR) (self-report)
• Child Behavior Checklist (CBCL) (parent completed)
• Antisocial Beliefs and Attitudes Scale (ABAS) (self-report)
• Loeber et al.’s parent completed measure of positive parenting and disciplinary practices (PP) along with parent monitoring and supervision (parent-report)
• Subjective Family Image Test [SFIT]): A family measure completed by both the young person and primary caretaker looking at the quality of the emotional bond between adolescent and parent (emotional connectedness) (family-report)
• Antisocial Process Screening Device (APSD), a parent-completed measure of youth psychopathic traits (parent-report)
• a 16-item scale measuring the youth’s involvement with delinquent peers (IDP) adapted from the Youth in Transition Study’ (self-report).

Findings

This study identified statistically significant positive impact on a number of child and parent outcomes.

The study observed statistically significant reductions in MST youth’s offending behaviour. The study further observed significant reductions in MST parent- and youth-reported aggression and delinquent behaviour.

Study design and sample

The fourth study is a rigorously conducted RCT. 

This study involved random assignment of children to an MST or a treatment as usual group. Those assigned to the control group received individual therapy alongside a constellation of other youth offending services and MST participants received an average of six months of MST.

The study was conducted in Sweden with a sample of antisocial youths. Youths were an average of 15 years. 67% had one previous arrest and 32% had a previous out-of-home placement. Youths were eligible if they fulfilled the criteria for a DSM IV-TR diagnosis of conduct disorder and whose parent(s) or parent surrogate(s) were motivated to engage in an intervention. Youths were ineligible if there was a history of sexual offending, substance misuse, a serious cognitive difficulty or other mental health problem, or treatment in another facility. 

Measures

Social service and school attendance measures were collected for all participants. The following measures were also completed by the parent or youth. 

Youth:

• Youth self-report
• Sense of Coherence Scale
• Self-report Delinquency Scale
• Alcohol Use Disorder Identification Test (AUDIT )
• Drug Use Disorder Identification Test (DUDIT)
• ‘Bad Friends’ subscale from the Pittsburgh Youth Study
• Social competence with Peers Scales (SCPQ)
• Social Skills Rating System (SSRS)

Parent:

• CBCL (Caregiver – but about child)
• Parenting measure developed in Sweden
• Maternal mental health -- SCL-90.

Findings

There were no statistically significant differences between the groups. 

Study design and sample

The fifth study is a rigorously conducted RCT. 

This study involved random assignment of children to an MST group and a management as usual (MAU) group. Stochastic minimisation, stratifying for treatment centre, sex, age at enrolment to study and age at onset of antisocial behaviour was used. MAU involved the provision of best available local services for young people; the interventions were multicomponent and no less resource-intensive than MST.   

This study was conducted in the UK with a sample of children who were between 11 and 17 years old, with moderate-to-severe antisocial behaviour problems. 65% of the sample had persistent and enduring violent and aggressive interpersonal behaviour, more than 80% met DSM-IV criteria for any conduct disorder, and 26% had been permanently excluded from school for antisocial behaviour. 

Measures

Data was collected at post-intervention (6 months after randomisation), approximately 6-months follow-up (12 months after randomisation) and approximately 12-month follow-up (18 months after randomisation):

• Out-of-home placements were assessed using administrative data.
• Time to first offense was assessed using police administrative data.
• Proportion free of offending behaviour was assessed using police administrative data.
• Exclusion from school was assessed using administrative data (National Pupil Database).
• Behaviour problems were assessed using the Strengths and Difficulties Questionnaire (youth self-report and parent report versions).
• Callous and unemotional traits were assessed using the Inventory of Callous/Unemotional traits (youth self-report and parent report versions).
• Delinquency was assessed using the Self-Report Delinquency Measure (youth self-report).
• Antisocial beliefs and attitudes were assessed using the Antisocial Beliefs and Attitudes Scale (youth self-report).
• Materialistic attitudes were assessed using the Youth Materialism Scale (youth self-report).
• Young person wellbeing was assessed using the Moods and Feeling Questionnaire (youth self-report).
• ADHD symptoms were assessed using the Conners Comprehensive Behaviour Rating Scales – ADHD subscale (parent and teacher report).
• Behaviour disorders were assessed using the Development and Well-being Assessment measure (clinician report).
• Parent wellbeing was assessed using the General Health Questionnaire (parent report).
• Problems of monitoring and supervision in parenting were assessed using the Alabama Parenting Questionnaire (youth report and parent report versions).
• Parental support was assessed using the Loeber Parental Support measure (youth report and parent report versions).
• Family satisfaction, cohesion and communication were assessed using the Family Adaptability and Cohesion Evaluation Scale-IV (parent report).

Findings

This study identified statistically significant positive impact on a number of child and parent outcomes. This includes:

Post-intervention

• Improved SDQ – total score (parent report)
• Improved SDQ – impact score (parent report)
• Improved SDQ – conduct problems (parent report)
• Improved SDQ – emotional problems score (parent report)
• Improved SDQ – hyperactivity or inattention score (parent report)
• Improved SDQ – prosocial score (parent report)
• Reduced inventory of callous/unemotional traits (parent report)
• Reduced self-report delinquency measure scores – variety of substance misuse (young person self-report)
• Reduced self-report delinquency measure scores – volume of substance misuse (young person self-report)
• Improved Moods and Feelings questionnaire scores (young person self-report)
• Improved Conners Comprehensive Behavour Rating Scales – ASDHD scores (parent report)
• Improved Alabama Parenting Questionnaire – problems of monitoring and supervision (parent report).
• Improved Loeber parental support score (parent report)
• Improved FACES family satisfaction (parent report)
• Improved FACES family cohesion (parent report)
• Improved FACES family communication (parent report)
• Improved General Health Questionnaire scores (parent report)

6-months

• Improved SDQ – total score (young person self-report)
• Improved SDQ – emotional problems score (young person self-report)
• Improved SDQ – total score (parent report)
• Improved SDQ – emotional problems score (parent report)
• Improved SDQ – hyperactivity or inattention score (parent report)
• Improved Moods and Feelings questionnaire scores (young person self-report)
• Improved FACES family satisfaction (parent report)
• Improved General Health Questionnaire scores (parent report)

12-months

• Improved inventory of callous/unemotional traits scores (young person self-report)
• Improved General Health Questionnaire scores (parent report)

A statistically significant negative impact was identified on:

• Increased criminal offenses at 12-month follow-up. It is worth noting that MST participants’ outcomes did not worsen over time, but rather that MAU participants’ outcomes improved more relative to MST participants. 

This study receives a rating of NE (no effect) on a balance of considerations, including:

• The positive effects which are identified tend to occur early and not be sustained, with the MAU group catching up over time.
• The highest quality results in terms of lowest-attrition suggest that there was no effect (out of home placements), or that MAU participants’ outcomes improved more relative to MST participants (arrests). The results suggesting positive impact (child- and parent-reports) are less robust due to higher attrition (18% at post-test, 29% at 6-month follow-up, and 31% at 12-month follow-up).
• The highest quality results in terms of objective measurement (versus self/parent-report) suggest that there was no effect (out-of-home placements), or that MAU participants’ outcomes improved more relative to MST participants (arrests).   
• There are negative results that suggest that MAU participants’ outcomes improved more relative to MST participants.