Study
Schurer Coldiron and colleagues (2019) conducted a randomized controlled trial assessing the impact of a Wraparound Program on youth arrests. The evaluation period was 20 months and took place in a large urban county in southeastern Florida. The evaluation compared arrest outcomes of youth in foster care and involved in the juvenile justice system receiving Wraparound Program services with youths in foster care and involved in the juvenile justice system receiving community treatment as usual. Eligibility criteria included 14- to 19-year-old youths, who were involved in both the Florida Department of Children and Families (DCF) and the Department of Juvenile Justice (DJJ) services, but not residing in a DJJ facility, and living within the Florida county.
Youth in the Wraparound Program received treatment as usual with the addition of a Wraparound care coordinator. Because of their involvement in both juvenile justice and child welfare, youths in the Wraparound Program were provided a range of services from local community-based organizations and state agencies before and during the study period. The Wraparound care coordination being evaluated supplemented this preexisting mix of supports. Treatment as usual consisted of a combination of DCF services and DJJ probation. Child welfare?related services included near-monthly contact, either in person or by phone, from a child advocate; an assigned guardian ad litem, if needed; an assigned aftercare specialist; and life skills coaching and/or transition services. Probation services typically were mandated for 1 year and included routine contact with a probation officer, monitoring and/or restricting of the given youth?s behavior, and requirements to attend treatment services. Noncompliance of the requirements would result in youth referrals to secure detention for up to a month. About half of all the study-involved youths received juvenile probation services from the local DJJ?s Probation and Community Intervention office during the evaluation period. About two thirds of all study-involved youths were held in secure detention for noncompliance or reoffending, during the intervention follow-up period.
The final sample included 24 youth participants of the Wraparound Program and 23 youths receiving community treatment as usual (TAU). There were no statistically significant differences between the groups at baseline. Youths participating in the Wraparound Program were an average age of 15.9 years; consisted of 62.5 percent males and 37.5 percent females; consisted of 58.3 percent Blacks, 29.2 percent whites, and 12.5 percent Hispanics; had an average of 2.5 placements since first entering DCF care; and had an average of 3.5 arrests before 2015. Youths receiving TAU averaged age 16; were 69.6 percent male and 30.4 percent female; consisted of 58.3 percent Blacks, 21.7 percent whites, and 8.7 percent Hispanics; had an average of 2.3 placements since first entering DCF care; and had an average of 2.8 arrests before 2015.
Youth outcome measures of interest for the study authors included outcomes relating to juvenile justice (number, type, and level of offenses committed, and the occurrence and timing of rearrests during the study period), child welfare (the change and difference in the placement stability, level of living restrictiveness, and runaway behavior), and educational achievement (graduated with diploma, on track for graduation, enrolled in GED program, number of years behind, and not engaged). The juvenile justice outcome of percentage of youth arrests were the only outcome eligible for during the CrimeSolutions review of this study because the remaining analyses of outcomes were based on a subset of individuals and were not presented as the overall participants of Wraparound program versus treatment as usual.
Change in youth arrests was examined by evaluating change from 20 months before and 20 months after the index date of January 2015. Florida DJJ?s Bureau of Research and Data Integrity provided de-identified offense, arrest, and DJJ services records for all youths from October 2005 through August 2016. Analyses included chi-square, paired-samples t-tests, and survival analyses. Because of the small sample sizes?and subsequent lack of ability to detect small to medium-sized effects?effect sizes (odds ratios or Cohen?s d) were also calculated for each comparison, including point-in-time differences between groups and changes over time. The study did not conduct subgroup analyses.
