Validation of a screening questionnaire for alcoholism (MAST) in an Italian sample

One hundred fifty subjects (75 index cases and 75 controls) completed the Italian version of the Michigan Alcoholism Screening Test (MAST). Using Relative Operating Characteristic (ROC) analysis, the 5/6 cut-off point was found to be the best as for discriminating between "cases" and "non-cases" in terms of balance between sensitivity (96%) and specificity (89%). The positive predictive value, the negative predictive value, and the overall misclassification rate were 90%, 95%, and 7% respectively. Using item by item assessment of discriminating capacity, we developed a brief ten item Italian version (2/3 cut-off point). This discriminated between alcoholics and non-alcoholics as well as the longer version, and is thus superior in situations where a brief evaluation is necessary.