Evaluation of clinical assessments of social abilities for use in autism clinical trials by the autism biomarkers consortium for clinical trials

Abstract

Clinical trials in autism spectrum disorder (ASD) often rely on clinician rating scales and parent surveys to measure autism‐related features and social behaviors. To aid in the selection of these assessments for future clinical trials, the Autism Biomarkers Consortium for Clinical Trials (ABC‐CT) directly compared eight common instruments with respect to acquisition rates, sensitivity to group differences, equivalence across demographic sub‐groups, convergent validity, and stability over a 6‐week period. The sample included 280 children diagnosed with ASD (65 girls) and 119 neurotypical children (36 girls) aged from 6 to 11 years. Full scale IQ for ASD ranged from 60 to 150 and for neurotypical ranged from 86 to 150. Instruments measured clinician global assessment and autism‐related behaviors, social communication abilities, adaptive function, and social withdrawal behavior. For each instrument, we examined only the scales that measured social or communication functioning. Data acquisition rates were at least 97.5% at T1 and 95.7% at T2. All scales distinguished diagnostic groups. Some scales significantly differed by participant and/or family demographic characteristics. Within the ASD group, most clinical instruments exhibited weak (≥ |0.1|) to moderate (≥ |0.4|) intercorrelations. Short‐term stability was moderate (ICC: 0.5–0.75) to excellent (ICC: >0.9) within the ASD group. Variations in the degree of stability may inform viability for different contexts of use, such as identifying clinical subgroups for trials versus serving as a modifiable clinical outcome. All instruments were evaluated in terms of their advantages and potential concerns for use in clinical trials.

Lay Summary

Evaluating interventions for autism often relies on clinician rating scales and parent surveys to measure results. We compared several common measurement instruments to evaluate which were most likely to be completed, distinguished children with and without ASD, and were impacted by participant and family characteristics. We also examined whether the instruments measured similar aspects of social ability and provided consistent scores across a short period of time. Results may be used to guide better selection of tests for future studies of autism interventions.