Reliability and validity of the Pervasive Developmental Disorders Assessment System

AIM: To test the reliability and validity of the Pervasive Developmental Disorders Assessment System (PDDAS), a Japanese semistructured interview system. METHODS: The PDDAS, consisting of 91 items including 12 major items corresponding to 12 items in criterion A of DSM-IV autistic disorder criteria, 36 items on autistic symptoms and three Asperger's disorder (AS) screening items for diagnosing pervasive developmental disorders (PDD) and their subtypes and 40 items for other information including early development and past/family histories, was administered to mothers of 77 PDD children and 64 non-PDD children. RESULTS: The PDDAS had satisfactory interrater reliability (ranges of kappa, r and raw agreement rate were 0.69-1.00 in 76 items, 1.00 in 11 items and 0.91-1.00 in four kappa un-calculable items, respectively). Thirty-three of the 36 items and all of the 12 major items scored significantly higher in the PDD than non-PDD groups to show satisfactory discriminant validity. PDDAS and consensus DSM-IV diagnoses agreed in the 77 children in PDD diagnosis and disagreed in only two children in subtype diagnoses of autistic disorder and PDD not otherwise specified. CONCLUSIONS: The PDDAS, which takes 1.5 h to administer, seems to have clinical and research utility, although further investigation is necessary.