Evaluation of Early Intervention Outcome in Young Children with Risk for ASD

Background: Children with ASD often perform differently in varied contexts making assessment of overall functioning and prognosis challenging. Therefore, when determining overall child functioning, these children be require a comprehensive evaluation consisting of measures from multiple sources (Ozonoff, Goodlin-Jones, & Solomon, 2005). First, standardized assessments across developmental domains (cognition, communication, social skills, adaptive behavior, behavior challenges) and diagnostic assessments (e.g., ADOS) are necessary. In addition, a detailed assessment of a child’s response to treatment over time is necessary for an accurate picture of prognosis. This dual approach to evaluation provides data to examine possible correlates of differential responsivity to intervention, providing valuable information about child variables leading to best outcome.

Objectives: Our goals were to (1) create a comprehensive evaluation of child functioning based on standardized developmental and diagnostic assessments; (2) develop a measure of child responsivity to treatment; and (3) identify predictors of best outcome and treatment response for children with risk for ASD in an early intervention program.

Methods: Forty-nine children between 13 and 27 months of age (M=22.4) identified with risk for ASD participated in an early intervention program. Children received an average of 9.29 treatment hours per week until age 3. An empirically-based behavioral intervention utilizing Pivotal Response Training, Discrete Trial Training, developmental strategies and Functional Routines were used to teach a range of skills. .

Child outcome was quantified by ranking and combining scores from the Mullen Scales of Early Learning, the Vineland Adaptive Behavior Scales, and the Autism Diagnostic Observation Schedule at age 3.

Treatment responsiveness was measured every 3 months using an adapted Student Learning Profile (aSLP), a curriculum-based assessment measuring mastery of skills taught in intervention. Treatment responsiveness was quantified using changes in curriculum assessment scores from intake to age 3.  

Results: To date, outcome data have been analyzed for 17 children and treatment responsiveness has been analyzed for 28 children. Children were divided into four groups based upon child outcome scores on standardized assessments. A Pearson’s correlation revealed a positive relationship between child outcome groups and assessment scores at intake on the Mullen Early Learning Composite, and Mullen expressive and receptive language domains.

The aSLP revealed substantial heterogeneity in treatment responsivity as indicated by variability in the number of skills learned. Rate of learning after 3 months of treatment accounted for 52% of the variance on the final aSLP. Pearson correlations revealed intake scores on the Mullen Scales of Early Learning, Early Learning Composite were positively related to final aSLP scores. Final aSLP scores were not related to intake scores on the Vineland.

Conclusions:  Quantitative indices of level of treatment responsiveness and overall outcome for each child with risk for ASD in an early intervention program were developed that identify subgroups of children, including those who demonstrated good, moderate or limited progress. A combination of standardized language measures and early rates of learning may be predictive of overall prognosis.