(PS11-C51) Examining Treatment Engagement for Children with Disruptive Behavior Multidimensionally: Does past Engagement Predict Future Engagement?

Treatment engagement has been shown to impact children’s outcomes in psychotherapy. Historically, engagement has been almost exclusively studied as the number of sessions attended. While attendance is an important part of treatment engagement, recent efforts have led to a broader conceptualization. Specifically, the REACH framework defines five domains of engagement: relationship with provider, expectancies for treatment, treatment attendance, clarity on treatment, and homework completion (Becker & Chorpita, 2016). This study examined whether domains of treatment engagement would show stability in the prediction of engagement one year later. Additional analyses will be run to understand if domains of treatment engagement demonstrate a cross-sectional relationship. The impact of children’s disruptive behavior disorder symptoms and family-level factors on engagement domains was also examined (n=19). Participants were drawn from a larger sample of a study of parental help seeking behavior which includes two cohorts of children (N = 149), one of children involved in juvenile court and the other of children seeking outpatient specialty mental health services. Only data from the second and third annual timepoints were used in this study because of the low levels of outpatient service usage in the juvenile justice cohort at time point 1. Detailed information on service usage including service type, attendance, expectancies for treatment, and relationship with provider were collected using the Child and Adolescent Services Assessment (CASA). Information on child symptoms and diagnoses were collected using the Diagnostic Interview Schedule for Children (DISC-IV). Demographic information, including family income, was completed for all families enrolled in the study. Regression models for longitudinal analyses showed that after controlling for all other variables in the model each domain of engagement was stable over time. Relationship at year 2 positively predicted relationship at year 3 (β = 1.24, p =.01), treatment expectancies at year 2 positively predicted treatment expectancies at year 3 (β =2.55, p =.03), and treatment attendance at year 2 positive predicted attendance at year 3 (β = .90, p =.03). Treatment expectancy at year 2 also negatively predicted treatment relationship at year 3 (β = -5.61, p =.03) and the number of conduct disorder symptoms in year 2 positively predicted relationship at year 3 (β = 5.05, p =.04). The child’s cohort also predicted treatment relationship, being a child in the juvenile justice cohort negatively predicted relationship with provider (β = -.891, p =.03). Taken together, the results of the analyses suggest that while not all longitudinally predictive of each other, each of the domains of engagement was stable over time. This suggests that active intervention may be needed to enhance domains of engagement, given their stability. Distinct populations (e.g., justice-involved youth, individuals with high DBD symptoms) may also require tailored interventions to enhance domains of engagement.