A screening focusing on alcohol consumption withdrawal effects in student’s population

Introduction: Students overestimate alcohol consumption of those around them and underestimate their own, so that quantitative approach may not be the most relevant to assess students’ drinking. The main objective was to provide an appropriate tool for screening students’ drinking. Methods: A multicentre cross-sectional survey was conducted by internet between February and June, 2013 in France. Thirteen questions explored alcohol consumption of which 8 concerned withdrawal effects (4 items of the AUDIT) and alcohol behaviour (CAGE test). A multiple correspondence analysis (MCA) was conducted to identify profiles of student’s alcohol consumption. Partitioning methods were used to group students according to their mode of alcohol use. The most relevant items included in the MCA were identified. Three questions were identified as most pertinent among the students with potential drinking problems and ranked by a decision tree with the Chi-square Automatic Interaction Detector method. Finally, the generalization of our model was assessed.Results: A total of 36,427 students participated in the survey: 25,679 females (70.5% of respondents), sex ratio 0.42 and mean aged 21.2 (sd 3.7 years). Among those who had experimented with alcohol (N=33,113), three consumption profiles were identified: “simple/non-use” (66.9%), “intermediate consumption” (25.9%) and “problem drinking” (7.2%) Instead, for the latter group, the three most relevant items were (Q20) "not able to stop drinking after starting”, (Q21) "failed to do what was normally expected”, and (Q23) "unable to remember what happened the night before”.Conclusions: Healthcare professionals have a screening tool of three items for students’ drinking