Objectives: 1) To determine the operating characteristics of the Alcohol Use Disorders Identification Test (AUDIT) in an inner-city population and 2) to compare AUDIT performance with physician recognition of alcohol-related problems.
Design: New patients were administered a health habits questionnaire, which included the AUDIT and the alcohol portion of the Structured Clinical Interview for DSM-III-R (used as the "gold standard" for alcohol abuse or dependence). The findings were compared with physician recognition based on chart review.
Setting: Inner-city general medicine clinic staffed by resident physicians with faculty supervision.
Patients/participants: Consecutive new patients over a ten-week period. Of 166 consecutive new patients, 23 were not contacted, 17 refused to participate, two were excluded for gross cognitive deficits, and 124 completed the study.
Results: Of the 124 patients, 41 (33%) met criteria for past or present alcohol abuse or dependence. The AUDIT correctly identified 26/27 (sensitivity of 96%) of patients with current problems and 0/14 of patients with past problems only. Physician rates of recognition of current and past alcohol-related problems based on chart review were 12/27 (sensitivity of 44%) and 1/14 (7%), respectively. The specificities of both the AUDIT and physician recognition were high, 96% and 99%, respectively.
Conclusions: 1) Alcohol-related problems were common in this setting, 2) The AUDIT had a high sensitivity and a high specificity for detection of current alcohol problems in this setting, but it failed to detect patients with only past histories of alcohol problems. 3) The AUDIT performed significantly better than did the physicians in detecting alcohol problems. 4) The addition of an assessment of past alcohol use to supplement information from the AUDIT would appear to represent a promising screening strategy worthy of further investigation.