No abstract

The Autism Diagnostic Observation Schedule-Generic (ADOS-G) is a semistructured, standardized assessment of social interaction, communication, play, and imaginative use of materials for individuals suspected of having autism spectrum disorders. The observational schedule consists of four 30-minute modules, each designed to be administered to different individuals according to their level of expressive language. Psychometric data are presented for 223 children and adults with Autistic Disorder (autism), Pervasive Developmental Disorder Not Otherwise Specified (PDDNOS) or nonspectrum diagnoses. Within each module, diagnostic groups were equivalent on expressive language level. Results indicate substantial interrater and test-retest reliability for individual items, excellent interrater reliability within domains and excellent internal consistency. Comparisons of means indicated consistent differentiation of autism and PDDNOS from nonspectrum individuals, with some, but less consistent, differentiation of autism from PDDNOS. A priori operationalization of DSM-IV/ICD-10 criteria, factor analyses, and ROC curves were used to generate diagnostic algorithms with thresholds set for autism and broader autism spectrum/PDD. Algorithm sensitivities and specificities for autism and PDDNOS relative to nonspectrum disorders were excellent, with moderate differentiation of autism from PDDNOS.

Autism Diagnostic Observation Schedule (ADOS) Modules 1-3 item and domain total distributions were reviewed for 1,630 assessments of children aged 14 months to 16 years with an autism spectrum disorder (ASD) or with heterogeneous non-spectrum disorders. Children were divided by language level and age to yield more homogeneous cells. Items were chosen that best differentiated between diagnoses and were arranged into domains on the basis of multi-factor item-response analysis. Reflecting recent research, the revised algorithm now consists of two new domains, Social Affect and Restricted, Repetitive Behaviors (RRB), combined to one score to which thresholds are applied, resulting in generally improved predictive value.

Diagnostic stability at age 9 years was very high for autism at age 2 years and less strong for pervasive developmental disorder not otherwise specified. Judgment of experienced clinicians, trained on standard instruments, consistently added to information available from parent interview and standardized observation.