Diagnostic Accuracy of the Drug Use Disorder Identification Test and Its Short Form, the DUDIT-C, in German Adolescent Psychiatric Patients

Background A common screening instrument for substance use disorders (SUDs) is the Drug Use Disorders Identification Test (DUDIT) which includes a short form regarding only drug consumption (DUDIT-C). We aim to assess if a German version of the DUDIT, adapted for adolescents, is a suitable screening instrument in a sample of adolescent psychiatric patients. Methods N = 124 (54 female) German adolescent (M = 15.6 + 1.5 years) psychiatric patients completed the DUDIT and received a diagnostic interview (MINI-KID) assessing DSM-5 SUD criteria. A confirmatory factor analysis (CFA), receiver operating characteristic (ROC) curves, the area under the curve (AUC), and Youden's Index were calculated. Results A two-factor model of the DUDIT shows the best model fit (CFI = 0.995, SRMR = 0.055, RMSEA = 0.059, WRMR = 0.603). The DUDIT as well as the DUDIT-C show high diagnostic accuracy, with AUC = 0.95 and AUC = 0.88, respectively. For the DUDIT a cut-off value of 8.5 was optimal (sensitivity = 0.93, specificity = 0.91, J = 0.84), while for the DUDIT-C the optimal cut-off value was at 1.5 (sensitivity = 0.86, specificity = 0.84, J = 0.70). Conclusion This is the first psychometric evaluation of the DUDIT in German, adolescent psychiatric outpatients, using the DSM-5 diagnostic criteria. The DUDIT as well as the DUDIT-C are well suited for use in this population. Since in our sample only few patients presented with a mild or moderate SUD, our results need to be replicated in a sample of adolescents with mild SUD.