Screening for problem drinking in older people referred to a mental health service: a comparison of CAGE and AUDIT.

The aim of the study was to evaluate the validity of the Alcohol Use Disorders Identification Test (AUDIT), the five-item version (AUDIT-5) and the CAGE as screening tests for problem drinking in mentally ill older people. The study was of prospective cross-sectional design with questionnaire survey and interview and included all consecutive referrals to an old age psychiatry service fulfilling inclusion criteria. Sensitivity, specificity and positive predictive values and areas under the receiver operating characteristic curves (AUROC) for the AUDIT, AUDIT-5, and CAGE were the primary outcome measures. Using clinical criteria as the gold standard, the AUDIT, AUDIT-5 and CAGE had AUROCs of 0.961, 0.964, and 0.780 respectively. The AUDIT-5 performed best of the three scales with a sensitivity of 75.0%, specificity of 97.2% and positive predictive value of 83.3% when using a 4/5 cut-point. The AUDIT-5 performed as well as the AUDIT and better than the CAGE in identifying problem drinking in this sample. The AUDIT-5 may be a useful addition to the specialist mental health assessment of older people.