Abstract

OBJECTIVE:

To examine the psychometric properties of the German version of the Autism Diagnostic Observation Schedule (ADOS).

METHODS:

Interrater and retest reliability, internal consistency, convergent and diagnostic validity were determined in a total sample of 137 subjects with autism, 23 with atypical autism or pervasive developmental disorder not otherwise specified, 16 with Asperger-syndrome and 13 with other psychiatric disorders.

RESULTS:

Interrater and retest reliability on the level of diagnosis (kappa w = 1.00 and .62) and raw-scores (ru = .84 and .79) were good. Likewise, the internal consistency of the algorithm scale communication and social interaction of modules 1 to 4 was fair (ru = .78 to .89). The categorical convergence for autism between the ADOS and the Autism Diagnostic Interview-Revised (ADI-R) reached 79% (kappa = .23), with their corresponding subscales correlating moderately (rtc = .31 to .45). The concordance of the ADOS judgment and the clinical consensus diagnosis was 77% (kappa w = .37), with a sensitivity of the ADOS of 90.4% and a specificity of 48.1% regarding the discrimination of autism and other autistic disorders.

CONCLUSIONS:

The ADOS is a reliable and sufficiently sensitive diagnostic tool in the assessment of autistic disorders. For ICD-10/DSM-IV classification and to ensure a high specificity of diagnosis additional information concerning repetitive, stereotyped behavior and early development (e.g. taken from the ADI-R) has to be collected.