Efficacy of the LEAP and TEACCH Comprehensive Treatment Models for Preschoolers with ASD

Thursday, May 17, 2012: 2:45 PM
Grand Ballroom East (Sheraton Centre Toronto)
2:00 PM
B. Boyd1, K. Hume2, M. Alessandri3, A. Gutierrez3, L. D. Johnson4, L. A. Sperry5 and S. Odom6, (1)Occupational Science, University of North Carolina at Chapel Hill, Chapel Hill, NC, (2)Frank Porter Graham Child Development Institute, University of North Carolina, Chapel Hill, Chapel Hill, NC, (3)Psychology, University of Miami, Coral Gables, FL, (4)Educational Psychology, University of Minnesota, Minneapolis, MN, (5)Arts, Education and Law, Griffith University, Brisbane, Australia, (6)University of North Carolina, Chapel Hill, NC, United States
Background: TEACCH and LEAP are two comprehensive treatment models (CTMs) with a long-standing history in the field of autism.  Yet, researchers are only now beginning to examine their programmatic efficacy (Strain & Bovey, 2011).  There is still a critical need to directly compare these CTMs to determine any differential effects as well as for whom and under what circumstances these interventions best work.

Objectives: (1) To compare child and family outcomes for participants in TEACCH, LEAP, and “business-as-usual” (BAU) control classrooms, and (2) To describe child and family characteristics that moderate participant outcomes. 

Methods: A rigorous, quasi-experimental study involving n = 75 classrooms (28 BAU, 22 LEAP, 25 TEACCH) and n = 205 children was conducted across four study locations.  All study classrooms had to meet stringent inclusion criteria to ensure they were “high quality” classrooms and, for TEACCH and LEAP, they exhibit a high degree of fidelity  A battery of measures was collected on children and families across three time points to examine change over time.  Prior to data analysis, seven composite variables were derived from the measures to serve as study outcomes.  The composite variables were generated using a combination of exploratory and confirmatory factor analytic methods. In addition, longitudinal invariance analysis was used to ensure the composite variables exhibited stable measurement properties across time points.  Following empirical confirmation of the composite variables, a series of three data analytic models were fit to the data.  The three analysis models were (1) gain score analysis, (2) repeated measures model, and (3) repeated measures model with adaptive centering.  Because of the quasi-experimental study design, the data analytic models are of increasing complexity to act as sensitivity checks.  Multiple imputation methods were used to address missing data.

Results: Data analysis is still ongoing. However, our current findings based on the gain score analysis only, which reflect change from pre to post-test, reveal no group differences on any composite variables with the exception that children in TEACCH classrooms showed less improvement in sensory and repetitive behaviors (based on teacher report) than children in BAU classrooms (p=0.0267).  In general, irrespective of program type, children made gains across the school year.  However, there are variables that appear to moderate participant response. For example, for the autism severity composite, children with initial lower IQ scores and parents with higher levels of stress made less improvement. Additional discussion of child characteristics that may interact with program type to differentially predict response to intervention will be discussed.

Conclusions: In this study, two historical, widely implemented CTMs and a high quality business as usual program all produced positive results across time. The absence of differential treatment effects suggests that all models were having a significant impact, although one cannot rule out the possibility of maturation.  This study raises the issues of the replication of effects for CTMs when the model developer is not involved in conducting the research, and whether having access to a high quality, early intervention program is as beneficial as access to a specific CTM.

| More