Effectiveness of Multilevel Implementation Strategies for Autism Interventions: Outcomes of Two Linked Implementation Trials.

Objective: This study tests the effectiveness of leader- and provider-level implementation strategies to implement evidence-based interventions (EBIs) in 2 of the service systems caring for autistic children. The TEAMS Leadership Institute (TLI) targets implementation leadership and climate, and TEAMS Individualized Provider Strategy (TIPS) targets provider motivation and engagement.

Method: A cluster randomized hybrid type 3 implementation-effectiveness trial tested the effects of the implementations strategies when paired with AIM HI (An Individualized Mental Health Intervention for Autism) in mental health programs (study 1) and CPRT (Classroom Pivotal Response Teaching) in classrooms (study 2). The combined sample included 65 programs/districts across 4 training cohorts (2018-2019 to 2020-2021). Organizations were randomized to receive a leader-level strategy, provider strategy, both strategies, or neither strategy (EBI provider training only). Leader and provider participants were recruited from enrolled programs/districts, and child participants were recruited from providers' caseloads or classrooms. Data from a total of 387 providers (mean age = 36.39 years; 91% female participants; 30% Latino/a/x participants) and 385 children (mean age = 9.13 years; 84% male participants; 60% Latino/a/x participants) were analyzed. Outcomes were assessed over 6 months. Provider outcome measures included provider EBI certification and observed EBI fidelity. Clinical outcome measures included the Eyberg Child Behavior Inventory (ECBI) (study 1) and the Pervasive Developmental Disorder Behavior Inventory (PDD-BI) (study 2). Outcomes were analyzed using intent-to-treat models.

Results: There was no significant effect of TLI on EBI Certification. TLI was associated with significantly higher EBI fidelity compared to non-TLI (B = 0.37, p = .04). Moreover, a statistically significant TLIX1Time interaction was found for child outcome T scores (B = -10.47, p = .03), with a significant reduction in T-scores across time only for those in the TLI condition. There were no significant effects of TIPS on any outcomes.

Conclusion: Findings support the effectiveness of leader-focused strategies to promote implementation and clinical outcomes of autism EBIs in multiple public service systems and for multiple EBIs.

Clinical trial registration information: Translating Evidence-based Interventions for ASD: Multi-Level Implementation Strategy (TEAMS); https://clinicaltrials.gov/; NCT03380078.