Comparing Two Parent-Implemented Interventions on Direct Observation of Joint Attention and Language of Preschool Children with Autism: A Pilot Study

Background:  Acquiring joint attention (JA) and functional spoken language by the time a child with autism enters school around the age of 5 years old are critical. Involving parents in treatment maximizes the child’s opportunities to learn. The efficacy and effectiveness of two popular approaches to autism treatment have not yet been evaluated as parent-implemented treatments. Relationship Development Intervention® (RDI) is a relationship-based intervention focusing on recreating developmental milestones through meaningful interactions with parents with a primary goal of increasing JA. In contrast, an Applied Behavior Analysis model based on Skinner’s Analysis of Verbal Behavior (ABA/VB) increases spoken language directly when implemented by therapist and teachers; however, this approach has not been evaluated when implemented by parents. RDI and ABA/VB have different primary targets for intervention; however, both approaches assume that language and JA, respectively, will emerge as a corollary effect of the intervention.

Objectives:  The purpose of this pilot study is to compare the efficacy of two parent-implemented adjunctive interventions, RDI and ABA/VB on JA and language in children 2-6 year old with ASD in preparation for a larger randomized clinical trial.

Methods:  Participants were 12 children diagnosed with autism spectrum disorder (ASD) by community providers and the Autism Diagnostic Observation Schedule-2™.  Participants were randomly assigned to parent-implemented RDI or ABA/VB. To control for services received in the community, parents reported weekly on the type and extent of services delivered. The intervention consisted of 16 sessions over 12 weeks delivered in the home. Assessments were conducted pre-treatment and end of treatment (3 months). The primary outcome measure was direct observation of parents and children interacting under different scenarios designed to implement conditions most likely to produce the behavior targeted by the interventions (e.g., joint attention, specific language skills) including prompts that involved having the child ask the parent for preferred items (PC5) and engaging the child in a conversation about family photographs (PC6). The direct observations were coded by research assistants blind to treatment condition and trained to interobserver reliability levels of 80% or above on a coding taxonomy that included an array of parent and child behavior related to joint attention, language, and problem behavior.

Results: Post-treatment direct observation assessment revealed that both the parent-implemented RDI and parent-implemented ABA/VB conditions were followed by gains in joint attention and verbal behavior, as well as decreased levels of problem behavior. Although greater improvements as measured by the direct observation coding system were observed in the ABA/VB condition compared to the RDI condition, there was variability in individual response.

Conclusions:  Preliminary results from a pilot study comparing parent-implemented RDI and ABA/VB indicated that both interventions are able of producing gains in joint attention and verbal behavior, even though the two interventions differ in primary intervention targets. Further, these results support that parents can serve as effective agents of change in the implementation of empirically supported early interventions for ASD.