THE VALIDITY OF THE MICHIGAN ALCOHOLISM SCREENING-TEST (MAST)

This review examines the validity of the Michigan Alcoholism Screening Test (MAST) as a screening instrument for alcohol problems. Studies t hat compare the MAST-questionnaire with other defined diagnostic crite ria of alcohol problems were retrieved through MEDLINE and a cross-bib liographic check. A total of 20 validity studies were included. The st udies varied considerably regarding the prevalence of alcohol problems , the diagnostic criteria, and the examined patient categories. The MA ST compared with other diagnostic criteria of alcohol problems gave va lidity measures with the following span: predictive positive value (PV pos) 0.24-0.96, predictive negative value (PVneg) 0.78- similar to 1, sensitivity 0.36- similar to 1, and specificity 0.36-0.96. It appears that the PVneg are high. Except from one study, the sensitivities (0.5 7- similar to 1) are also high. The PVpos and the specificities show s ubstantial variations. The variables that seem to have the largest inf luence on the PVpos seem to be the prevalence of alcohol problems, the diagnostic method against which the MAST-questionnaire is validated, and the populations on which the MAST is applied. The MAST should in t he future be validated against internationally accepted diagnostic met hods on populations with representative prevalences of alcohol problem s.