A Pragmatic Comparative Trial of Educational Service Delivery Models for Children with ASD

Background:  Different approaches to educational service delivery for children with ASD in regular schools are available, based on different conceptual and theoretical underpinnings. Nevertheless, there has been little comparative research examining outcomes and predictors of success under different models. Two approaches to the support of children with ASD in Australian primary schools are the Consultative model used by Autism SA in South Australia (SA) and the Satellite Class model used by Autism Spectrum Australia (Aspect) Aspect in New South Wales (NSW). The Satellite model provides placement into an autism specific class in regular school, with phased transition and support to mainstream when the child is adequately prepared. The Consultative model provides on-demand support in mainstream classes from the point of school entry.

Objectives:  The objective of this research was to examine comparative outcomes and predictors of success across the two models of service delivery.

Methods:  A four-year comparative pragmatic trial of the models was conducted with 85 children. Inclusion criteria for the project were that the child must have a formal diagnosis of Asperger’s disorder or autistic disorder (DSM-IV), be functioning within or above the mild range of intellectual disability, be in their first four years of schooling at the project commencement. Data were collected twice yearly. Primary child outcomes identified prior to the commencement of the project related to continuity of placement, social behaviour, school engagement and adjustment. Primary outcomes for school staff and parents related to satisfaction with service delivery and perception of the success of placement. A multilevel model approach to data analysis was taken. Predictors of each outcome variable, including program model, were examined with data structured as rounds (repeated measure), nested within child, and nested within school (i.e., three-levels).

Results:  Continuity of placement was high in both models. Satellite class transitions were lower than expected with many families electing to remain in satellite classes. With regard to child social skills and school engagement and adjustment, there was no significant difference between the models of service delivery, and the only significant predictor of both was teacher-rated academic competence. Mean level of rated placement success was high in both states across all respondents, typically scoring between 4 and 5 on a 5-point scale. Principal and teacher-rated success of placement was not different between states, and adaptive behaviour at pre-test was a significant predictor of successful outcome for both. Parents in NSW rated school placement in the satellite classes as more successful. Both parent and teacher ratings of success trended down across the study, and children with greater problem behaviours were rated as less successful by both. Both principals and parents rated support as being better in NSW.

Conclusions:  Child outcomes were not significantly different but some differences were evident in perceived success of placement and satisfaction with support. Adaptive behaviour at pre-test was a predictor of rated success of placement. While intensity of support differed across the models of support, both included several basic features that are considered best practice in support of children with ASD.