Detecting alcohol consumption as a cause of emergency general medical admissions.

In a general medical unit 27% of 104 admissions and 17% of bed occupancy were attributed to alcohol consumption, although only 10 of these 28 patients had classical alcohol-related conditions. Questioning on the amount of alcohol consumed was the most accurate method of screening for these patients. The brief MAST questionnaire, mean red cell volume, gamma-glutamyl transferase, aspartate transaminase and urate were all inadequate as screening tests. Only 13% of the admissions were defined as 'problem drinkers' on the brief MAST questionnaire, suggesting that harmful effects of alcohol are not confined to 'alcoholics'. Clinical suspicion and questions on the level of alcohol consumption are more efficient than questionnaire, biochemical or haematological screening tests in detecting alcohol-related medical problems.