Detection of alcoholism in three populations by the brief-MAST.

The brief MAST (B-MAST) questionnaire was used to detect alcoholism in three populations, namely, alcoholics in treatment, clinical outpatients, and the general population. Nearly all the alcoholics in treatment tested positive in the B-MAST (99.2%), for lifetime as well as for current (past year) alcohol-related problems. Among the clinical outpatients and general population, 35.6% and 20.2% respectively, tested positive for lifetime alcohol problems, but only 19.6% and 7.6%, respectively, tested positive for past-year alcohol problems. The sensitivity/specificity of the B-MAST (lifetime) was 64.7%/90.0% and 48.3%/96.4% in the clinical outpatients and general population, respectively. Even lower sensitivity (28.2%) was obtained when the B-MAST was used to detect past-year alcohol problems in the general population sample. The poor sensitivity of the B-MAST in the general population was probably because most of the B-MAST questions deal with severe alcohol problems, and they are not sufficiently sensitive to detect those who drank heavily but who had not yet developed these alcohol problems. Anomalous responses to the two questions about "normal drinker" and to the question about "attendance in AA meeting" also contributed (< 3%) to inflated positive B-MAST scores. There were gender differences, and to a lesser extent, racial differences (Whites versus Blacks), in the responses to individual B-MAST questions in the three subject populations, with more males than females in each group reporting more alcohol-related problems. However, there were no gender and racial differences in the average B-MAST scores among those clinical outpatients and general population subjects who tested positive.