Aalto Mauri, Tuunanen Mervi, Sillanaukee Pekka, Seppä Kaija
Department of Mental Health and Alcohol Research, National Public Health Institute, FIN-00251 Helsinki, Finland.
Alcohol Clin Exp Res. 2006 Nov;30(11):1884-8. doi: 10.1111/j.1530-0277.2006.00233.x.
There is a need for an effective and feasible alcohol screening instrument. The aim of the study was to evaluate how the abbreviated versions of the Alcohol Use Disorders Identification Test (AUDIT) questionnaire perform in comparison with the original AUDIT and what the optimal cutoffs are when screening for heavy drinking among women.
All the 40-year-old women in the city of Tampere, Finland, are invited yearly for a health screening. From 1 year, data from 894 women (response rate 68.2%) invited for a health screening were utilized in the study. The original 10-item AUDIT, AUDIT-C, Five Shot, AUDIT-PC, AUDIT-3, AUDIT-QF, and CAGE were evaluated against the Timeline Followback. Consumption of at least 140 g of absolute ethanol per week on average during the past month was considered heavy drinking.
In the Timeline Followback, the mean+/-SD weekly reported alcohol consumption was 45+/-67 g (range 0-936 g) of absolute ethanol. Of the women, 6.2% (55/894) were heavy drinkers. The optimal combination of sensitivity and specificity was reached for the AUDIT with cutoff > or =6, for the AUDIT-C with cutoff > or =5, for the Five Shot with cutoff > or =2.0, for the AUDIT-PC with cutoff > or =4, and for the AUDIT-QF with cutoff > or =4. When choosing the optimal cutoffs, the AUDIT-C, the Five Shot, the AUDIT-PC, and the AUDIT-QF performed as well as the 10-item AUDIT. With these cutoffs, sensitivities were 0.84 to 0.93 and specificities were 0.83 to 0.90. The AUDIT-3 and the CAGE did not perform as well as the other questionnaires.
The 10-item AUDIT, AUDIT-C, Five Shot, AUDIT-PC, and AUDIT-QF seem to be equally effective tools in screening for heavy drinking among middle-aged women. However, their applicability is achieved only if the cutoffs are tailored according to gender.
需要一种有效且可行的酒精筛查工具。本研究的目的是评估酒精使用障碍识别测试（AUDIT）问卷的简化版本与原始AUDIT相比的表现，以及在筛查女性重度饮酒时的最佳临界值是多少。
芬兰坦佩雷市所有40岁的女性每年都会被邀请参加健康筛查。本研究使用了894名受邀参加健康筛查的女性（应答率68.2%）从1年起的数据。将原始的10项AUDIT、AUDIT-C、五题法、AUDIT-PC、AUDIT-3、AUDIT-QF和CAGE与时间线追溯法进行比较评估。过去一个月平均每周饮用至少140克纯乙醇被视为重度饮酒。
在时间线追溯法中，每周报告的纯乙醇平均摄入量为45±67克（范围0 - 936克）。这些女性中，6.2%（55/894）为重度饮酒者。AUDIT临界值≥6、AUDIT-C临界值≥5、五题法临界值≥2.0、AUDIT-PC临界值≥4以及AUDIT-QF临界值≥4时，达到了敏感性和特异性的最佳组合。在选择最佳临界值时，AUDIT-C、五题法、AUDIT-PC和AUDIT-QF的表现与10项AUDIT相当。采用这些临界值时，敏感性为0.84至0.93，特异性为0.83至0.90。AUDIT-3和CAGE的表现不如其他问卷。
10项AUDIT、AUDIT-C、五题法、AUDIT-PC和AUDIT-QF似乎是筛查中年女性重度饮酒的同等有效工具。然而，只有根据性别调整临界值，它们的适用性才能实现。
