*Objective*: Juveniles under community supervision have a higher prevalence
of substance use compared to the general population. Previous research has
examined factors that predict treatment outcomes once treatment has been
initiated; however, there is a gap in the literature regarding the factors
that affect initiation of treatment, particularly after referral by the
juvenile justice system. This study analyzes individual-level factors with
site and state-level variation predicting initiation of treatment following
referral from juvenile probation agencies.
*Methods*: Using the Behavioral Health Services Cascade framework of the
Juvenile Justice-Translational Research on Interventions for Adolescents in
the Legal System (JJ-TRIALS) multisite project, a series of mixed effects logistic
regression models were estimated using two methods. Referral and treatment
initiation data from two analytic methods addressing missing data from
3,312 youth records in Method 1 and 5,325 records in Method 2 collected
from 22 counties across six states were analyzed. We examined between site
variation and assessed the impacts of individual- and state-level factors
on initiation to treatment after referral.
*Results*: Results from the mixed effects model indicate that being
referred and initiating treatment varied significantly across sites in both
Methods. A mixed model in Method 1 with individual-level characteristics
indicated that youth with a higher substance use treatment need and youth
with a higher supervision level are 1.98 and 1.74 times more likely to
initiate treatment after referral, respectively. In Model 3 for Method 2
analyses, when controlling for state-level differences, youth with a higher
level of supervision were 2.26 times more likely to initiate treatment
following referral compared to those with a lower level.
*Conclusion*: Findings suggest that individual factors matter in predicting
initiation, independent of state or site. Level of supervision is the most
salient factor in a youth’s initiate of treatment in both missing data
methods, however, need for treatment is also important although it was
significant in Method 1 only. Understanding the differences in predictors
allows partnering behavioral health and juvenile justice agencies to
develop strategies to increase the likelihood that referred youth initiate
treatment.
