Using the VABS-II Item Sets As Predictors of Outcome in Preschool-Based Esdm Vs. Preschool-Based Eclectic Intervention

Background: Identification of child related factors that may serve as predictors of outcome in early intervention plays a key role in the personalization of treatments in ASD (Sherer, Schreibman & Cunningham, 2011).

Objectives: The current study aimed to identify these differential factors while comparing two types of preschool-based intervention models: The Early Start Denver Model (ESDM), a manualized early intervention, integrated into ASD community preschools in Israel, and an eclectic intervention commonly implemented in the country. Child factors were assessed utilizing content analysis of specific Vineland Adaptive Behavior Scales (VABS-II) item-sets related to play and communication skills.

Methods: Fifty-eight children, aged 35-57 months, from eight ASD preschools, participated in the current study. Thirty-three children attended preschools in which ESDM was integrated, and twenty-five children attended eclectic preschools. Groups were matched on age, developmental level and autism severity, both receiving equal numbers of therapy hours per week. Changes in children's adaptive behaviors (teacher reported VABS-II) and cognitive ability (MSEL) were measured pre- and post- 8 months of intervention. Four item-sets reflecting early behaviors that may affect treatment outcome (Vivanti et al., 2013) and may be impaired in children with ASD (Balboni et al., 2015) were chosen as predictive variables to intervention related changes in child MSEL scores. These item-sets include 'beginning to talk', 'following instructions', 'playing' and 'imitation'.

Results: The 'following instructions' item-set predicted developmental gains in the areas of expressive language, visual reception and fine motor across intervention groups. The 'imitation' item set predicted improvement for both groups in receptive language and was also found to predict greater gains in fine motor scores in the ESDM group, but not in the group receiving the eclectic intervention. However, the 'play' item-set yielded an opposite trend, predicting higher fine motor scores in the eclectic intervention group in contrast to the ESDM group. Finally, the 'beginning to talk' item-set predicted significantly higher gains in expressive language scores in the ESDM group but not in the eclectic intervention group. A similar trend was marginally significant in predicting receptive language scores (p=0.058).

Conclusions: Results indicate that both imitation and following instructions skills are important learning mechanisms that underlie children's ability to gain from both interventions. However, differential gains in fine motor and expressive language abilities reflect the notion that different children might gain more from different intervention models. In the ESDM model, imitation and spoken language abilities are marked as central early social learning abilities. Put together, these results further broaden current knowledge regarding child-related intervention outcome predictors. The current study also offers a possible use of the VABS-II item-sets in community settings, in order to identify children who might benefit more from the ESDM.