ABSTRACT
To estimate prospectively the concordance between diagnoses of alcoholism by a physician and by use of a questionnaire-type screening test in patients in a coronary care unit, we studied 608 patients admitted to a coronary care unit in a 5 1/2-month period. The results from the Self-Administered Alcoholism Screening Test (SAAST) were compared with the diagnoses by physicians recorded in the medical records for the same patients. A strong association (P < 0.001) existed between the clinicians' identification of an alcohol problem and the SAAST scores indicative of such a problem. However, 23 (4.2%) patients were identified as alcoholics by the SAAST but not by clinicians, and 16 (2.9%) patients were identified as alcoholics by clinicians but not by the SAAST. The SAAST was well accepted in this coronary care unit and can complement and extend the clinicians' impression, although each approach may identify some patients that the other will miss. Routine use of the SAAST in the coronary care unit or other hospital settings is feasible and, with appropriate clinician assessment and physician involvement, should yield more complete evaluation and case finding for alcoholism.
