Document Type : Original Article
1 Addiction Research Centre, Mashhad University of Medical Sciences, Mashhad, Iran
2 SYB Institute of Preventive Medicine, Mashhad, Iran
Background: Alcohol consumption is one of the main health problems in most countries. The purpose of the present study is to evaluate the reliability and validity of Alcohol Use Disorders Identification Test (AUDIT).
Methods: By employing the method of available sampling, 70 alcoholic patients were included as case group and 70 normal age, gender and education matched individuals were also included as the healthy group (controls). Both groups filled out the AUDIT questionnaire. Statistical tests, Cronbach’s alpha, correlation coefficients and exploratory factor analysis were done to analyze the data.
Results: The internal reliability of the questionnaire was estimated to be 0.84 by employing Cronbach’s alpha test, which indicates that approximately 16% of the total variance of the results obtained from the questionnaire was affected by the measurement error. This value was reported to be 0.85 for the cases and 0.84 for the controls. After omitting the second question of the questionnaire the Cronbach’s alpha increased, but the amount of the increase was rather limited and thus no change in the questionnaire was required. By omitting each of the questions from the questionnaire, the alpha coefficient changed ranging between 0.8 and 0.85. Moreover, using the method of questionnaire classification the coefficient of internal consistency was reported to be 0.78. The reliability of the questionnaire was determined to be 0.84 by employing Cronbach’s alpha. The coefficient of internal consistency for each question was estimated to be 0.18 to 0.74. The consistency between test and after-test results was found to be significant (P = 0.020 and r = 0.710). Using the ROC curve on cutoff point, sensitivity and trait of the test was determined to be 99% and 89% respectively. By employing exploratory factor analysis and varimax rotation, the test examines two key factors I and II that respectively accounted 35% and 34% and overall 56% of the total variance to them.
Conclusion: The AUDIT questionnaire has favorable psychometric properties and therefore can be used as a valid tool to carefully sift through patients for alcohol use.