Diagnostic accuracy of the alcohol use disorders identification test (AUDIT) in Swedish young adult outpatients in general psychiatry.

AIMS: The Alcohol Use Disorders Identification Test (AUDIT) is a widely used screening instrument for harmful alcohol use. Comorbidity between alcohol use disorders and other mental disorders is common, but there is a shortage of studies investigating AUDIT's diagnostic accuracy in young adult psychiatric patients. The aim of this study was to examine AUDIT's diagnostic accuracy and psychometric properties in young adult psychiatric outpatients.

METHODS

In data from two samples of young adults (18-25 years) from general psychiatric outpatient clinics in Sweden, recruited in 2002-2003 (n = 197) and 2012-2016 (n = 283) psychiatric diagnoses were assessed using the Structured Clinical Interview for DSM-IV, axis I disorders, clinical version (SCID-I-CV) or the Mini International Neuropsychiatric Interview (MINI 6.0). Internal consistency of AUDIT, sensitivity, specificity, positive and negative predictive values (PPV and NPV), receiver operator characteristic (ROC) curves, and area under the curve (AUC) were calculated.

RESULTS

The internal consistency of AUDIT was acceptable (Cronbach's alpha = 0.76). For the samples combined the optimal cut-off was 10, sensitivity 87% and specificity 82%, PPV 0.28 and NPV 0.99. For Sample 2002-2003, the optimal cut-off point was nine For Sample 2012-2016, the optimal cut-off point was 10. ROC curves are presented, with an AUC of 0.88 (95% CI= 0.81-0.95) for both samples combined, 81.1% for Sample 2002-2003 (95% CI = 0.63-0.99) and an AUC of 90.6% for Sample 2012-2016 (95% Cl = 0.85-0.96).

CONCLUSION

The full version of AUDIT is a reliable and valid screening instrument for alcohol use disorders, as shown in two samples of young adult psychiatric outpatients.