Aims: The Alcohol Use Disorders Identification Test (AUDIT) is a well-established and widely used screening instrument. It has been shown that AUDIT has good criterion validity in relation to alcohol abuse and dependence according to DSM-IV, but it has not yet been validated following the introduction of the DSM-5 diagnostic system. The aim of this study was to evaluate concurrent validity for the AUDIT in relation to self-reported DSM-5 severity levels for Alcohol Use Disorder (AUD) in a Swedish general population sample. Methods: A postal questionnaire, containing the AUDIT and the 13-item brief DSM-5 AUD diagnostic assessment screener, was sent to a random sample of 1,500 persons drawn from the Swedish population, aged between 17 and 80 years and having a public residence address in Sweden. To evaluate the concurrent validity of AUDIT in relation to DSM-5 severity criteria for AUD, a Receiver Operating Characteristics (ROC) curve analysis was conducted. Results: Area under the curve (AUROC) showed excellent differentiation between AUD or not, mild (.93), moderate (.92) and severe (.99). Higher individual AUDIT scores were associated with more severe levels of AUD according to the DSM-5 screener. The optimal cutoff scores approximate earlier research on the DSM-IV and were identified as 5, 7 and 13 points, respectively, for mild, moderate and severe AUD. Conclusions: Our findings indicate that AUDIT is a valid screener for detecting concurrent AUD at three severity levels in the Swedish general population.
Keywords: Screening; alcohol use disorders; optimal cutoff; receiver operating characteristics (ROC).