Two groups of volunteers were recruited for this study: the clinical sample encompassed 70 alcoholics from alcoholism treatment units (mean age 43.22 ± 9.91; M/F: 57/13; years of education 11.96 ± 3.91) and healthy control group contained 74 subjects from general population (mean age 38.33 ± 11.58; M/F: 43/31; years of education: 13.77 ± 2.96).
All subjects were assessed by AUDIT Core (the first 10 questions) and AUDADIS (Alcohol Use Disorders and Associated Disabilities Interview Schedule) developed by Grant & Hasin (1991) from National Institute of Alcoholism and Alcohol Abuse-Rockville USA.
The discriminant validity of the AUDIT was performed determining the ability of this instrument to differentiate alcoholics vs non-alcoholics and various alcoholism diagnostic criteria vs non-criteria. The cutoff point of AUDIT score is the true threshold between the categories above mentioned.
The AUDIT score of 10 is a good cutoff point discriminating between alcoholics vs non-alcoholics diagnosticated according to AUDADIS algorythm (kappa = 0.81, sensitivity/specificity: 90.6/90.6) and between non-criteria and harmful (kappa = 0.65, sensit./specif.: 83.5/81.0), tolerance (kappa = 0.40, sensit./specif. 67.0/72.9), neglecting interests (kappa = 0.51, sensit./specif. 63.5/91.5), and withdrawal/relief (kappa = 0.71, sensit./specif. 88.2/83.0) criteria.
This study highlights that AUDIT is a suitable instrument that is able to distinguish between alcoholics and non-alcoholics and fournishes confident thresholds outlining diagnostic criteria. Contrary, AUDIT is not able to differentiate between harmful and dependency, the diagnostic categories from ICD-10.
Incorporating the AUDIT into a general population telep... Drug and Alcohol Dependence |
Incorporating the AUDIT into a general population telephone survey: a methodological experiment Drug and Alcohol Dependence, Volume 60, Issue 1, 1 July 2000, Pages 97-104 Frank J. Ivis, Edward M. Adlaf, Jürgen Rehm Abstract This study assessed potential ordering and wording effects of the alcohol use disorders identification test (AUDIT). In total, 688 respondents were randomly assigned to one of four experimental conditions: Intact order/original wording (n = 148), intact order/revised wording (n = 183), split order/original wording (n = 192), split order/revised wording (n = 166). Changes to question order and wording had no discernable impact on the scores of the AUDIT. Our results suggest that alterations to the AUDIT can be made in order to integrate it within a larger survey without adversely affecting its measurement properties. PDF (937 K) |
View More Related Articles |
An epidemiological survey of alcohol use disorders in a Tibetan population