Mark W. Fraser
University of North Carolina at Chapel Hill
School of Social Work
CB #3550
301 Pittsboro St.
Chapel Hill, NC  27599-3550
(919) 962-6538
mfraser@email.unc.edu

Paper 1:  Mental Health Problems, Substance Abuse, and Recidivism among Delinquent Youth: A Field Test of the Massachusetts Youth Screening Instrument-2
Jeffrey M. Jenson
Cathryn C. Potter
William Dieterich

University of Denver
Graduate School of Social Work
2148 So. High Street
Denver, CO  80208

Paper 2:  Risk Assessment for Re-Offending in Juvenile Justice
Mark W. Fraser
Steven H. Day
Craig S. Schwalbe

University of North Carolina at Chapel Hill
School of Social Work
CB #3550
301 Pittsboro St.
Chapel Hill, NC  27599-3550

Paper 3:  Is “Ecstasy” (MDMA) Abuse a Risk Factor for Psychiatric Disorders and Suicidality among Antisocial Youth?
Matthew O. Howard
George Warren Brown School of Social Work
Washington University
One Brookings Drive
Campus Box 1196
St. Louis, MO  63130-4899
(314) 935-4957
howard@gwbmail.wustl.edu

The prediction of substance abuse and re-offending among juvenile offenders is fraught with challenges.  Knowledge of risk factors for substance abuse and recidivism is limited and risk assessment instruments designed to isolate factors associated with persistent antisocial behavior are often unavailable to practitioners.  Findings from three investigations that assess the relationship among risk factor domains, substance abuse, and recidivism are described in this symposium.  The first paper examines the concurrent and predictive validity of the Massachusetts Youth Screening Instrument-2 (MAYSI-2), an inventory that evaluates mental health and substance abuse problem symptoms among delinquent youth.  Analyses are based on a sample of 480 detained youth.  Exploratory factor analytic techniques are used to examine the underlying structure of the inventory.  Concurrent validity is assessed by comparing MAYSI-2 scale-scores to subjects’ scores on the Brief Symptom Inventory.  The utility of the MAYSI-2 in predicting recidivism at twelve-month follow-up is analyzed in a multivariate logistic regression model.  A 9-item risk assessment instrument developed to predict future offending among youth in the juvenile justice system is described in the second paper.  To estimate predictive validity, 400 juveniles adjudicated to community-based supervision were tracked for 12 months.  Survival models of the hazard for re-arrest are estimated on the basis of risk scores and the current offense, which is not included in the risk assessment instrument.  A classification table showing correct and incorrect predictions based on risk scores will be reviewed.  The experiences of 15 court workers who used the instrument are also described.  The third paper examines the prevalence of “ecstasy” (MDMA) abuse among 296 delinquent youth in juvenile justice and psychiatric settings.  Hierarchical multiple linear and logistic regression are conducted to examine the effect of light, moderate, and heavy MDMA use on current psychiatric symptoms and suicidality.

Findings in this presentation describe the concurrent and predictive validity of the MAYSI-2.  Factor analytic techniques are used to examine the underlying structure of the inventory; results illustrating differences between a 7-factor structure reported by Grisso and colleagues and a 4-factor model are shown.  Concurrent validity is assessed by comparing MAYSI-2 scale scores to the Brief Symptom Inventory, a validated measure of mental health problem symptoms among youth (Derogatis, 1993).  Findings indicate strong concordance between overall and corresponding scale scores on the two instruments.  The utility of the MAYSI-2 in predicting recidivism at twelve-month follow-up is analyzed in a multivariate logistic regression model.  The presentation concludes with a discussion of the need for risk assessment of co-occurring problem symptoms and argues for differential treatments that account for variation in patterns of mental health problems and substance abuse for delinquent youth.

References:  Grisso, T., Barnum, R., Fletcher, K.E., Cauffman, E., & Peuschold, D. (2001).  Massachusetts Youth Screening Instrument for Mental Health Needs of Juvenile Justice Youths.  Journal of the American Academy of Child and Adolescent Psychiatry, 40, 541-548.  Derogatis, L.R. (1993).  Brief Symptom Inventory.  Minneapolis: National Computer Systems.

To estimate predictive validity (Note: Reliability data have been presented at previous SSWR meetings), 400 juveniles adjudicated to community-based supervision were tracked for 12 months.  Survival models of the hazard for re-arrest are estimated on the basis of risk scores and the current offense.  Based on these survival models, the predictive validity of each item in the risk assessment instrument and the current offense will be discussed.  In addition, a classification table showing correct and incorrect predictions based on risk scores from the 400 youths with community dispositions will be reviewed.  Findings suggest that the instrument adequately predicts future offending and offers significant advantages over traditional idiographic methods of predicting risk.  Qualitative interviews with 15 court workers suggest that the instrument is easily used, but that the prediction of future general offending may not have as much practice relevance as the prediction of future violent offending.  Concurrent efforts to introduce risk assessment instruments in juvenile justice systems across the county will be reviewed.  We will conclude with a discussion of the strengths and limitations of risk assessment in juvenile justice.

All residents of each facility were interviewed; the overall response rate was 86.3%. Respondents completed the Brief Symptom Inventory (a measure of current psychiatric symptoms), a substance use assessment, and measures of school performance, vocational aspirations, bonding to family members, and self-esteem.  Lifetime MDMA use was reported by 41.3% and 27.3 % of youth at the 2 correctional settings, 9.2% of psychiatrically-disordered youth, and 17.9% of psychiatric hospital clients.  Lifetime MDMA users were evenly divided between light, moderate, and heavy users. Heavy MDMA users averaged 156.7 (?=278.2) occasions of use and evidenced significantly more symptoms of depression and anxiety than did lighter MDMA users/nonusers. Hierarchical multiple linear and logistic regression analyses were undertaken predicting current psychiatric symptoms/lifetime suicidality entering demographic variables in the 1st step, measures of family/social factors in the 2nd step, and measures of non-MDMA substance use and lifetime MDMA use in the 3rd and 4th steps, respectively.  Results suggest that MDMA abuse is prevalent among youth in correctional settings and an independent risk factor for psychiatric dysfunction among antisocial youth.