Objectives: The CAGE questionnaire is considered a useful tool for detecting alcohol problems, especially in clinical populations. Its sensitivity exceeds 75% and its specificity 90%. Our objective was to assess performance of the French version of the CAGE (DETA) in identifying patients who meet the DSM-IV criteria and to discuss its usefulness in hospitals.
Methods: Data come from study of hospital morbidity in the Auvergne district of central France in 1998. Alcohol problems were diagnosed by physicians using the DSM-IV alcohol abuse or dependency criteria. For each diagnostic category, we calculated the performance of the CAGE questionnaire at a cut-off of 2 (CAGE 2+) and of 1 (CAGE 1+).
Results: The study included 5,452 patients--48.5% of them men--in acute- and intermediate-care facilities. CAGE 2+ had a sensitivity of 77% and a specificity of 94% among the diagnosed patients. It was more sensitive for patients diagnosed as alcohol-dependent than for alcohol abusers (61% vs 84%), but specificity was identical for both groups (94%). These results are similar to the properties of the English version. For a cut-off of 1 the CAGE had a sensitivity of 88% and a specificity of 87%.
Conclusion: The French version of CAGE is a useful screening tool for alcohol abuse and dependance. Its systematic use in hospitals is desirable: a positive response to any item should attract the physician's attention and lead to further investigation.