Validity of the CAGE questionnaire in hospital

OBJECTIVE: The CAGE questionnaire is considered a useful screening and case-finding tool for alcohol use disorders in clinical populations. Our objectives were to validate the French version of the CAGE against DSM-IV criteria and to assess performance of each item of the scale. METHOD: Data were extracted from a hospital morbidity study conducted in central France. It concerned 5452 patients-48.5% men-in short and medium-stay units. Patients answered the CAGE questionnaire as a past-year assessment. The alcohol use disorders were diagnosed by the physicians using DSM-IV alcohol abuse or dependency criteria. RESULTS: The CAGE questionnaire for a cut-off of 2 had a sensitivity of 77% and a specificity of 94%. The CAGE test was more sensitive for patients diagnosed as alcohol-dependent than for alcohol abusers (61% vs. 84%) with the same specificity (94%). These values are close to those for the English-language CAGE. The first three items (CAG) were very similar, with sensitivity 70% and specificity 94%. The eye-opening question (E) differentiated sharply between abuse and dependency, with sensitivities of 18% and 46%, respectively. A questionnaire comprising only the CAG questions of the CAGE had properties similar to the full questionnaire. CONCLUSION: CAGE is a good screening tool for alcohol abuse or alcohol dependency. Given the frequent-and insufficiently diagnosed-alcohol problems among inpatients, CAGE is indicated as a first-line tool for screening for the most severe alcohol use disorders in hospital. It should ideally be used systematically. A positive reply to any of the first three items should alert the clinician and prompt further investigation.