A Little Goes a Long Way: A Randomized Controlled Implementation Trial of a Social Engagement Intervention for Children with Autism Spectrum Disorder in Public Schools

Background: Children with autism have poorer social inclusion than do their classmates throughout elementary school. While several interventions have been developed and tested to address this challenge, few evidence-based social engagement interventions for children with autism are successfully implemented in schools. This disconnect may occur because these interventions are not feasible to implement in many public schools. To address this challenge, we partnered with public school personnel to adapt existing social inclusion interventions to increase their feasibility and acceptability, and to develop a set of implementation strategies that can support school personnel in using the intervention. 

Objectives: To examine the implementation, child outcomes and sustainment of an evidence-based social engagement intervention for children with autism in public schools.

Methods: Thirty-one elementary-aged children with autism (87% male; mean age=8.8 (SD=2.0) years; 48% white) and 28 school personnel (86% female, mean age=39.5 years; 61% white) from 12 public schools in five districts participated. A stepped-wedge randomized controlled design was used in which data were taken at baseline, second baseline, exit and follow-up. Schools were randomized to: 1) training in Remaking Recess, a social engagement intervention for children with autism; or 2) training in Remaking Recess plus implementation support. Intervention implementation occurred during recess (30-45 minutes) for 12 sessions over six weeks. Implementation support was provided to school administrators three times over 12 weeks and entailed consultation in implementation strategies. School personnel rated their attitudes about evidence-based practices, acceptability, and implementation climate at exit. Children completed sociometric ratings while blind observers recorded playground peer engagement and implementation fidelity at each time point. 

Results: There were no significant differences in attitudes about evidence-based practices, acceptability, or implementation climate between school personnel in either condition. Fidelity of implementation was low in both conditions. On average, school personnel in both conditions improved in some components of Remaking Recess (assessment (p=.02), set up (p=.045), accuracy (p<.001), and understanding (p<.001)) but not others (transition, participation and feasibility). There were no significant differences in the rate of improvement between the two groups of school personnel from entry to exit.  Multilevel models where time was nested within children and children were nested within schools were used to test differences in social outcomes between schools that were randomized to Remaking Recess or Remaking Recess plus implementation support. There was a significant main effect of time for solitary (p<.001) and joint engagement (p<.001), where children in both conditions spent significantly less time in solitary engagement and more time in joint engagement after the intervention and at follow-up. There was a significant time by condition interaction (p=.01) where children with autism in the Remaking Recess plus implementation support condition had significantly more improvement in social network salience than in the Remaking Recess alone condition after the intervention (p=.002) and at follow-up (p=.05).

Conclusions: These data suggest that implementing some components of an evidence-based intervention with fidelity may lead to improved child outcomes; however, organizational-level implementation support may be necessary to improve child outcomes in public school settings.