To evaluate the diagnostic usefulness of AUDIT (Alcohol Use Disorders Identification Test) for detecting hazardous alcohol consumption in primary care settings, to assess the potential differences according to age or sex and to compare its diagnostic value with that of other conventionally used measures (CAGE questionnaire and biological markers). Patients and method: This is a descriptive cross-sectional study for evaluation of diagnostic tests with 500 participating patients randomly selected among those attending a family health care center. Data about alcohol intake were collected using a semistructured interview, considering as hazardous alcohol consumption a weekly intake of 35 SDUs (standard drink unit) or above for men and 21 for women. AUDIT and CAGE questionnaires were administered and blood levels of GGT, MCV, AST, ALT, alkaline phosphatase, platelets, tryglicerides and uric acid were determined. For each one of those sensitivity, specificity, global value, predictive values, positive and negative probability ratio were calculated as well as ROC analysis for the questionnaires. Results: Sensitivity of AUDIT in detecting current hazardous drinkers was 89%, specificity was 93% and area under ROC was 0.98 (95% Cl, 0.96-0.99). Its sensitivity was found to be lower in the female group (60 vs. 78%) as well as for age group under 60 years (86 vs. 100%). CAGE questionnaire showed a sensitivity of 48% and a specificity of 97%. Biological markers offered sensitivities lower than 40% and specificities over 90%. Conclusions: AUDIT is a useful tool for detecting hazardous alcohol consumption in ambulatory care. Its sensitivity being lower for women and its diagnostic usefulness significantly exceeding that of other screening measures commonly used.

581

576

16