The goal of this study is to construct a 16-week, 2-stage, adaptive intervention consisting of DTT (Discrete Trials Training, largely considered usual care for children with autism), JASP-EMT (a blended, naturalistic, developmental behavioral intervention involving JASPER [Joint Attention, Symbolic Play, Engagement and Regulation] and EMT [Enhanced Milieu Teaching]), and parent training (P) for improving spontaneous communicative utterances in school-aged, minimally verbal autistic children. Intervention was delivered both at school (DTT, JASP-EMT) and at home (P). This article reports results for the study’s primary aim and a closely related secondary aim.

The study used a 2-stage, sequential, multiple-assignment randomized trial design. In stage 1 (weeks 1-6), 194 minimally verbal (<20 functional words), 5- to 8-year-old autistic children were randomized initially to DTT vs JASP-EMT (stage 1, weeks 0-6). Early vs slower response status was determined at the end of stage 1. In stage 2 (weeks 7-16), early responders were re-randomized to stay the course vs P, whereas slower responders were re-randomized to stay the course vs combined DTT+JASP-EMT). The primary aim was to test whether there was a difference between starting with DTT vs starting with JASP-EMT on average change in socially communicative utterances (SCU; primary outcome) from baseline to week 16. A secondary aim was to estimate which of the 8 prespecified interventions was most favorable (ie, the largest average SCU at week 16). The secondary outcomes were total number of novel words, joint engagement, play diversity, requesting, and joint attention gestures from independent blinded assessments.

There was no evidence to reject the null hypothesis of no difference between starting with DTT or JASP-EMT on primary outcome ( p  = .41). The most favorable of the 8 interventions was the adaptive intervention, which starts with DTT, augments with P for early responders, and augments with JASP-EMT for slower responders. For this adaptive intervention, average change on SCU from baseline to week 16 for this intervention was estimated to be 7.68 (95% CI = 2.13-13.24).

The results showed no difference in treatment starting with JASP-EMT or DTT, and the differences among the 8 adaptive interventions of the secondary aim were modest. Based on these results, reflections on next steps are discussed.

There is limited research about the optimal strategy for interventions to improve spoken language outcomes among minimally verbal 5- to 8-year-old children with autism. The goal of this sequentially randomized trial of 194 children was to compare starting intervention with either a highly structured intervention (discrete trial training [DTT], considered usual care) or a blended, naturalistic, developmental and behavioral intervention (Joint Attention, Symbolic Play, Engagement and Regulation [JASPER] and Enhanced Milieu Teaching [EMT], JASP-EMT) and then changing the intervention depending on how the child responded to the initial intervention. Results at 16 weeks found that there was no difference in improvement in spoken language outcomes for children who began with either intervention first. However, an adaptive intervention was modestly better for having the greatest spoken language at the end of the study: begin with DTT, then add parent training in the home for early responders, and combine DTT and JASP-EMT for slow responders.

Adaptive Interventions for Minimally Verbal Children With ASD in the Community (AIM-ASD); NCT01751698 .