A Naturalistic Social Communication Intervention for Minimally Verbal Children with Autism

Background: Acquisition of spoken language before age 5 is an important predictor of long-term academic and social outcomes for children with autism (Rutter, 1978). For children with autism who are not using verbal communication in social interactions even after early intervention, it is imperative to identify an effective treatment.  Two interventions have been shown to impact early social communication during relatively short term treatments. Enhanced Milieu Teaching (EMT) has been found effective for increasing spoken communication in young children with autism (Hancock & Kaiser, 2011). The JASPER (Joint-Attention, Structured-Play, Emotion Regulation) intervention has been shown to increase not only joint attention and play, but also communication (Kasari, 2008). The current study examined the effects of a blended intervention (EMT-JASPER) targeting social attention, play and spoken language for minimally verbal children with autism who have not been responsive to previous communication interventions. 

Objectives: 1) Does the blended EMT-JASPER intervention improve social communication in minimally verbal children? 2) Will the effects of an initial EMT-JASPER intervention generalize to other adults who do not use the EMT-JASPER intervention strategies, and (3), those who do?

Methods: The participants included three boys with autism (MN age 51.6 months) each of whom was minimally verbal and had participated in more than two years of intervention. A multiple probe across participants single subject design was used to evaluate the intervention.  A secondary multiple probe design was implemented across three therapists for each participant in order to examine generalization during baseline and treatment conditions with new partners.  All interventionists were trained to fidelity on the EMT-JASPER intervention. Intervention occurred during 30-minute play sessions in a clinic setting. Every session was video recorded transcribed and coded for social communicative use of specific language target forms. Interobserver agreement averaged 96% (range 83-100%) for child communication. 

Results: Each child demonstrated a large increase and an accelerating trend in spontaneous target use within six intervention sessions. The participants did not generalize to new therapists when they did not use intervention techniques (baseline condition); however, after 4-5 sessions of intervention, all children responded to the EMT-JASPER intervention with new therapists at levels similar to those observed with their first therapist. Spontaneous use of social communicative targets for each participant averaged 4.75 (SD=6.29), 4.92 (SD=2.50), and 3.79 (SD=5.62) during baseline, respectively. Spontaneous target use increased during intervention for each participant to 36.48 (SD=18.13), 30.96 (SD=15.17), and 13.20 (SD=7.04) respectively. 

Conclusions: The EMT-JASPER intervention was effective for increasing spontaneous targets for these minimally students however, generalization was not observed until the intervention was implemented by the additional therapists. These findings suggest that at least initially, children with a history of minimal verbal communication may need systematic support across partners in order to use newly learned social communication skills. This is consistent with the proposal that that joint attention is a state that is dyadic and unique between conversation partners (Tasker & Schmit, 2008). Future research should investigate the effectiveness of introducing multiple interventionists including familiar communication partners during blended EMT-JASPER sessions to promote faster generalization.