The Therapeutic Alliance In a Social Skills Intervention for Adolescents with ASDs

Background: While social skills interventions (SSIs) for youth with autism spectrum disorders (ASDs) have recently accumulated sufficient evidence to be considered “empirically-supported” (Reichow & Volkmar, 2009), the essential therapeutic ingredients remain unknown. The relationship with those administering such interventions may represent such an ingredient (Gutstein & Sheely, 2002). That the relationship is a mechanism of therapeutic change is consonant with established literature on the therapeutic alliance. Alliance relates to outcomes in child psychotherapy (Shirk & Karver, 2003), is influenced by therapist experience with target populations (Ackerman & Hilsenroth, 2003), and may play a role in friendship-making in SSIs for other populations (Lerner, Mikami, & McLeod, in press). However, alliance has never been examined in SSIs for youth with ASDs. Identifying the role of alliance in this population could provide direction in unpacking essential components of SSIs for this population.

Objectives: 1. To examine baseline therapist characteristics related to alliance. 2. To assess the contribution of alliance to change in treatment outcomes.

Methods: Thirty-four youth (9 – 16 years; 23 male), primarily with ASDs, in a 6-week summer SSI with an explicit focus on fostering child-therapist relationships (Lerner, Mikami, & Levine, 2011) completed reports of their social skills (Social Skills Rating System; Gresham & Elliott, 1990) and social anxiety (Social Anxiety Scales; La Greca, 1999) immediately before and after the SSI. Additionally, they completed blinded sociometric ratings of their peers in their SSI group (Coie, Dodge, & Coppotelli, 1982), and alliance with group therapist (Therapeutic Alliance Scale for Children; Shirk & Saiz, 1992) after the first (early) and last (endpoint) week of the program. Their parents completed a baseline measure of child autism symptoms (Social Responsiveness Scale; Constantino & Gruber, 2005).

Results: Hierarchical Linear Modeling was used to control for nesting within therapists, and past experience with the rated therapist was controlled in all models. To assess Objective 1, after controlling for therapist gender and education, years of therapist experience with ASDs predicted higher early alliance (β₀₄ = .11, p <.01), and therapist age predicted lower early alliance (β₀₂ = -.02, p = .03).  To assess Objective 2, after controlling for child autistic symptoms and early sociometric ratings, early alliance predicted increased friendship (β₁₀ = .21, p < .01) and reciprocated friendship nominations (β₁₀ = .15, p = .01), but not social preference (β₁₀ = -0.01, p = .91).  Additional analyses for Objective 2 were unavailable at time of submission, but will be available for conference presentation.

Conclusions: Therapist experience with ASDs predicted higher early child alliance, while therapist age predicted lower early alliance. Additionally, youth reporting higher early alliance increased in the number of reciprocated friends they made in the SSI relative to those with lower early alliance. These results suggest that alliance may play a role in friendship-making in SSIs for ASD populations, but vary based on therapist characteristics. Additional analyses predicting therapeutic change may augment these initial promising findings. Future research should expand upon this initial investigation into this underexplored treatment mechanism in SSIs for ASD populations.