May 1991

Characterization of DSM-III-R Criteria for Uncomplicated Alcohol Withdrawal Provides an Empirical Basis for DSM-IV

Author Affiliations

From the Clinical Pharmacology Program, Addiction Research Foundation, Toronto, Ontario (Drs Sellers and Sullivan and Mss Somer and Sykora); and the Departments of Pharmacology and Medicine (Dr Sellers), University of Toronto.

Arch Gen Psychiatry. 1991;48(5):442-447. doi:10.1001/archpsyc.1991.01810290054011

• The DSM-III-R criteria for uncomplicated alcohol withdrawal require the presence of coarse tremor of the hands, tongue, or eyelids plus one of a number of other clinical features. We examined the validity and other characteristics of these items in 137 patients in pure alcohol withdrawal using the reliable and valid Clinical Institute Withdrawal Assessment for Alcohol. The DSM-III-R items of hand tremor amplitude, nausea or vomiting, headache, transient hallucinations, autonomic hyperactivity (increased pulse or sweating), and anxiety correlated significantly with total score and significantly indicated clinical severity. Addition of an "agitation" item improved the correlation. The diagnostic accuracy is greater than 95% if any two or more items are present. The number of positive items, of which tremor can be one, to grade clinical severity shows that a score of 2 indicates "very mild"; 3, "mild"; 4, "moderate"; and 5, "severe." We propose that an Alcohol Withdrawal Diagnostic Inventory and a DSM-III-R—compatible brief Clinical Institute Withdrawal Assessment for Alcohol are useful for clinical research, where graded symptom characterization is needed. Our data may be helpful in the development of criteria for DSM-IV.