Detection of alcohol-related problems in general practice.

While primary care has considerable potential as a site for detecting and intervening for alcohol-related problems, few doctors currently identify these problems. The judgments of eight primary care physicians about alcohol-related problems in 371 of their patients were compared with the patients' responses to the Short Michigan Alcoholism Screening Test (SMAST) and the CAGE. The CAGE classified 11.4% of the patients as alcoholics and the SMAST identified 23.9% as probable alcoholics. However, the doctors identified only a small proportion (7.0%) of their patients as having any level of alcohol-related problem. The doctors did not identify 65.0% of CAGE-defined alcoholics and 82.3% of those patients classified by the SMAST as probable alcoholics. The discrepancy between primary care physician's judgments and the SMAST and CAGE may be attributable to the doctor's failure to identify patients with alcohol-related problems. An alternative explanation is that the SMAST and CAGE are inappropriate screening tools for use in Australian primary care. The findings are discussed in terms of the implications for training doctors and for the development of better measures of alcohol-related problems for use within a primary care context.