Effectiveness of a Wide-Scale Community Based Intervention for Preschoolers with Autism

Background: Evidence supports the efficacy of early intensive behavioural intervention (EIBI) based on principles of applied behaviour analysis (ABA) for improving the outcomes of children with autism spectrum disorders (ASD; e.g., Howard, Sparkman, Cohen, Green, & Stanislaw, 2005; Lovaas, 1987; McEachin, Smith, & Lovaas, 1993).  However, studies on the effectiveness of community-based programs are limited.  Perry and colleagues (2008) conducted the largest effectiveness study to date on adaptive behaviour, autism severity, and cognitive level in a sample of children enrolled in a community-based, publicly funded EIBI program in Ontario, Canada.  They found statistical and clinical improvement in all areas.  Smith and colleagues (2010) recently evaluated the effectiveness of an EIBI program in Nova Scotia, Canada and found accelerated growth across multiple domains, including adaptive behaviour, autism severity, behavioural problems, cognitive level, and language ability.

Objectives: This prospective study was a pilot investigation of the outcomes of nine preschoolers with ASD enrolled in a community-based, publicly funded EIBI program provided in New Brunswick, Canada.  The New Brunswick model differs from the Ontario and Nova Scotia models in that intervention is provided in for-profit agencies.  Further, measures of treatment fidelity are lacking in both Perry et al. (2008) and Smith et al. (2010); therefore, measures of treatment fidelity are provided.  

Methods: Nine boys (mean baseline age = 33.7 months) with ASD participated.  All children underwent a comprehensive assessment at intervention onset and at 6- and 12-months follow-up.  The assessment battery included measures of overall development, language ability, and adaptive behaviour.  To account for limitations in past research, fidelity measures were obtained from video-recorded samples of children’s therapy sessions. 

Results: Repeated measures analysis of variance (ANOVA) revealed significant increases in children’s overall development standard scores: F(2, 16) = 4.795, p < .05, overall development age equivalents: F(2, 16) = 14.699, p < .01, expressive language age equivalents: F(2, 16) = 18.237, p < .01, receptive language age equivalents: F(2, 16) = 24.544, p < .01, and adaptive behaviour age equivalents: F(2, 16) = 31.944, p < .01.  Analyses of learning rates at 12-months follow-up on the measure of overall development indicated that 22% of children had learning rates exceeding the normal rate of 1 year of development per year of age.  On the measure of adaptive behaviour, 44% of children had above-average learning rates, and on the measure of language ability, 56% of children had above-average learning rates.  The majority of children showed some improvement (67%), 22% of children achieved average functioning, while 11% of children showed no change.  No children regressed.  Therapists demonstrated correct use of ABA techniques 95% of the time.

Conclusions: The New Brunswick EIBI model showed promising results in improving the outcomes of nine preschoolers with ASD.  A range of outcomes was found given the heterogeneity of our sample; however, the majority of children showed some measurable improvement or progress.  This suggests that EIBI can be associated with significant child improvement, and can be administered with very high levels of treatment fidelity, in a community-based, publicly funded setting.