Validity, reliability, and usability of the Pediatric Evaluation of Disability Inventory-Computer Adaptive Test for autism spectrum disorders

Abstract

Aim: This study aimed to explore the test-retest reliability of the Pediatric Evaluation of Disability Inventory-Computer Adaptive Test for autism spectrum disorders (PEDI-CAT [ASD]), the concurrent validity of this test with the Vineland Adaptive Behavior Scales (VABS-II), and parents' perceptions of usability.

Method: A convenience sample of participants (n=39) was recruited nationally through disability organizations. Parents of young people aged 10 to 18 years (mean age 14y 10mo, SD 2y 8mo; 34 males, five females) who reported a diagnosis of autism were eligible to participate. Parents completed the VABS-II questionnaire once and the PEDI-CAT (ASD) twice (n=29) no more than 3 weeks apart (mean 12d) using computer-simulated administration. Parents also answered questions about the usability of these instruments. We examined score reliability using intraclass correlation coefficients (ICCs) and we explored the relationship between instruments using Spearman's rank correlation coefficients. Parent responses were grouped by common content; content categories were triangulated by an additional reviewer.

Results: Intraclass correlation coefficients indicate excellent reliability for all PEDI-CAT (ASD) domain scores (ICC ≥ 0.86). PEDI-CAT (ASD) and VABS-II domain scores correlated as expected or stronger than expected (0.57-0.81). Parents reported that the computer-based PEDI-CAT (ASD) was easy to use and included fewer irrelevant questions than the VABS-II instrument.

Interpretation: These findings suggest that the PEDI-CAT (ASD) is a reliable assessment that parents can easily use. The PEDI-CAT (ASD) operationalizes the International Classification of Function, Disability and Health for Children and Youth constructs of 'activity' and 'participation', and this preliminary research suggests that the instrument's constructs are related to those of VABS-II.