Predictive validity of an observer-rated adherence protocol for multisystemic therapy with juvenile drug offenders

Multisystemic therapy (MST) is perhaps the best validated treatment for youth who engage in serious and chronic antisocial behavior (Henggeler, Schoenwald, Borduin, Rowland, & Cunningham, 2009). Despite evidence suggesting that high treatment adherence is needed to achieve optimal MST outcomes, this research is limited because past studies have relied on adherence reports derived solely from treatment participants (i.e., caregivers, youth, and therapists). To address this gap in the literature, the present study assessed the reliability and predictive validity of an observational protocol for rating adherence to MST.

The sample was drawn from a randomized clinical trial of juvenile drug offenders (77.5% male, 65% African American) referred to one of four treatment conditions (Henggeler et al., 2006). Audiotaped sessions of youth and their families were selected from the first month of MST and trained undergraduate students independently rated therapist adherence to the nine MST treatment principles. We assessed the validity of MST adherence in predicting outcomes at post-recruitment and 12-month follow-up.

Good interrater reliability (ICC=0.642) was found across all raters for our composite index of adherence. High adherence to MST during the first month of therapy predicted decreases in externalizing behavior at post-recruitment and decreases in youth alcohol consumption at 12-month follow-up.

These results provide independent support for the link between treatment fidelity and behavioral outcomes in the context of MST. Further, this study demonstrates the feasibility of using novice, undergraduate judges to reliably code therapist adherence.