Prevalence and Predictors of Mental Health Service Receipt Among Cross-over Youth

Introduction: Over the past three decades, researchers stressed a call to action to disrupt the empirically supported linkages between child welfare and juvenile justice (CWJJ) involvement (Ryan et al., 2007). In their systematic review, Campie et al (2015) concluded that there were no CW and JJ studies that described formal service delivery systems to enhance well-being. To breakdown silos across child-serving systems of care, Congress enacted the Keeping Children and Families Safe Act (KCFSA) of 2003. While the Act is intended to enhance linkages between systems, no studies to date examined prevalence of mental health (MH) service receipt between then and now, and whether and under what conditions CWJJ involvement influences or serves as a “gateway” to receiving MH services. To that end, this study addressed three aims focusing on two cohorts of CWJJ involved youth in 2003 and approximately ten years later: 1) What is the prevalence of MH service receipt, 2) Which case characteristics (length of services, placement spells, placement type) predict MH service receipt, and 3) In light of documented racial disparities in CW and JJ systems, are there differences in MH service receipt by race/ethnicity among youth involved in both systems?   

Methods:  Data were collected from the Data Management Office (DMO) in a large mid-Atlantic City. DMO mined data from several systems of care to develop one integrated dataset of system involved youth in 2003-05 (Cohort 1; n=10,170) and 2012-14 (Cohort 2, n=5,787). For each cohort, the following data were provided: demographics, placement types, number of placement spells, and receipt of delinquency and MH services. This study focused on the subset of youth (µ=15.1 years) involved in CW and JJ systems (n=2,824). Most of the sub-sample included males (69%) and African Americans (77%), while 14% included Latinos and 8% were Caucasian. Logistic regressions were conducted to examine factors influencing MH service receipt after CWJJ involvement.

Results: Logistic regression models revealed that crossover youth in Cohort 2 were more likely than Cohort 1 to receive MH services (84% for Cohort 2 vs. 55% for Cohort 1), controlling for demographic and case characteristics. In addition, younger youth, females, and Latinos (vs. African Americans) were more likely to receive MH services. Length of time (in months) receiving CW and delinquency services, placement spells, not being placed in out-of-home care, type of crossover effects (CW to JJ and JJ to CW), and MH service receipt prior to CWJJ involvement were positively associated with MH service receipt. Interaction analyses showed the effects of race/ethnicity on service receipt did not differ between cohorts.

Discussion: Results suggest that the rate of MH service use among a cohort of CWJJ involved youth is higher in more recent years than it was in 2003 when KCFSA was enacted. Future survey and qualitative research may illuminate whether KCFSA played a pivotal role in the increase in MH service use. Moreover, future research is needed to examine whether MH services are effective in preventing CWJJ involvement, or if prolonged treatment before, during, and after involvement is warranted.