The Effectiveness and Feasibility of the PEERS® School-Based Social Skills Intervention in a Transdiagnostic Group of Adolescents

Background: Social skills difficulties are common in adolescents with a range of neurodevelopmental (NDD) and mental health disorders (MHD). There are few evidence-based social skills interventions for adolescents, and access to these programs is limited. Social skills interventions are generally evaluated in tightly-controlled research settings, which differ substantially from community settings in ways that may limit the generalizability of program effectiveness. For example, social skills programs are generally studied within a single diagnostic group, while in community settings youth with a variety of NDDs/MHDs may be included together. School-based social skills interventions have potential to overcome barriers to care. Given the limited research on these interventions in school settings, considerations of social validity and feasibility are warranted.

Objectives: The present study evaluated the feasibility, acceptability, and initial effectiveness of the PEERS® social skills intervention for a transdiagnostic group of adolescents with social difficulties in the public high school setting.

Methods: Twenty-four high school students were assigned to the PEERS® intervention or a waitlist control group (WLC). The majority of students (71%) had IEPs. Twenty-nine percent had ASD, 29% had other diagnoses, and 42% had no diagnoses. We assessed variables related to feasibility including attendance, dropout, and scheduling obstacles. We also assessed parents’ and adolescents’ ratings of program acceptability. We evaluated the impact of the PEERS® program on participants’ social skills, ASD symptom severity, social anxiety, and explicit knowledge of social skills, as rated by parents, teachers, and the adolescents themselves. Further, we directly evaluated participants’ conversational abilities with the Contextual Assessment of Social Skills (CASS).

Results: Both parents and adolescents rated the intervention acceptable overall. Drop out was low, and the remaining students attended most sessions, although scheduling difficulties arose. Results of ANCOVA indicated adolescents in the intervention group displayed significantly improved social skills knowledge on a measure of skills taught in the program (Cohen’s d = 3.26) and overall involvement in conversation on the CASS (Cohen’s d = 1.00) compared to adolescents in the WLC. There were small-to-large effect sizes favoring the intervention group for most other intervention outcome measures, although these differences did not reach statistical significance. Reliable change indices revealed that almost half of adolescents (46%) in the intervention group displayed reliable improvement in at least one conversational skill domain that was observable to an independent rater blinded to treatment status and time point, compared to 18% of adolescents in the WLC.

Conclusions: Our findings provide preliminary support for the effectiveness of the PEERS® intervention in transdiagnostic groups. They also highlight the importance of continued research into factors that facilitate implementation of interventions in community-based settings. Future research should evaluate moderators of treatment response to provide guidance for which students are most likely to benefit from the intervention and identify modifications or alternative treatments for those who are less likely to benefit. As this work progresses it is also critical to consider factors that increase the reach and sustainability of interventions, such as training school-based providers and integrating programs into the structure of existing services.