On the usefulness of the DSM-III-R versus the DSM-III for child psychiatrists.

The purpose of this paper is to compare and contrast the disorders of infancy, childhood and adolescence in the DSM-III-R with those of its predecessor, the DSM-III. Design features of the child psychiatry sections of the DSM-III-R are described, with comparisons of reliability and validity assessments in the two classifications. Categorical and dimensional systems of psychiatric nosology are described; the DSM-III-R has features of both systems. To be most useful for child psychiatrists in ordinary clinical practice, DSM-III-R symptom criteria should be available in a standardized but brief fashion to ensure adequate data gathering from both child and parent. This avoids problems inherent in lengthy standardized interviews are based on DSM-III-R criteria; although these interviews are excellent for research purposes, clinicians tend to avoid them as clinically constraining. The commonly used alternate of clinicians' overall evaluations is of uncertain reliability and validity, since it is unclear whether all symptoms have been asked for. A symptom checklist approach is therefore suggested as a intermediate procedure to ensure that appropriate questions are asked from the parent and child, while allowing fuller exploration by the clinician. This approach also indicates parent-child variance, and allows for rank ordering of diagnoses which may indicate priorities for treatment of child psychiatric disorders. Overall, the DSM-III-R is a positive step towards more adequate diagnosis and treatment for child and adolescent psychiatric disorders, which will lead to further improvement in the future DSM-IV.