The Diagnostic Interview for Social and Communication Disorders: Algorithms for ICD-10 Childhood Autism and Wing & Gould Autistic Spectrum Disorder

Abstract

Background: The Diagnostic Interview for Social and Communication Disorders (DISCO) is an interviewer-based schedule for use with parents and carers. In addition to its primary clinical purpose of helping the clinician to obtain a developmental history and description of the child or adult concerned, it can also be used to assist in providing a formal diagnostic category. Method: In this study we compared two algorithms based on the ninth revision of the schedule (DISCO 9). The algorithm for ICD-10 childhood autism comprised 91 individual, operationally defined items covering the behaviour outlined in the ICD-10 research criteria. The algorithm for the autistic spectrum disorder, as defined by Wing and Gould (1979), was based on 5 DISCO items that represented overarching categories of behaviour crucial for the diagnosis of autistic disorders. The aim of the study was to examine the implications for clinical diagnosis of these two different approaches. Parents of 36 children with clinical diagnoses of autistic disorder, 17 children with learning disability and 14 children with language disorders were interviewed by two interviewers. Algorithm diagnoses were applied to interview items in order to analyse the relationship between clinical and algorithm diagnoses and the inter-rater reliability between interviewers. Results: Clinical diagnosis was significantly related to the diagnostic outputs for both algorithms. Inter-rater reliability was also high for both algorithms. The ICD childhood disorder algorithm produced more discrepant diagnoses than the Wing and Gould autistic spectrum algorithm. Analysis of the ICD-10 algorithm items and combination of items helped to explain the reason for these discrepancies. Conclusions: The results indicate that the DISCO is a reliable instrument for diagnosis when sources of information are used from the whole interview. It is particularly effective for diagnosing disorders of the broader autistic spectrum.