Objective: The present study examines the agreement of structured child and parent interviews as well as clinical diagnosis for depressive episodes in children and adolescents. Moreover, it compares the accuracy and optimal cutoff scores of self-report questionnaires with reference to each of these diagnostic assessments.
Method: 81 children (9-12 years) and 88 adolescents (13-16 years) in psychiatric care and their parents completed the structured diagnostic interview Kinder-DIPS. The children answered the German Children's Depression Inventory (CDI), and the adolescents answered the German Center for Epidemiologic Studies Depression Scale (CES-D). Kappa coefficients quantified the agreement. Receiver operating characteristic (ROC) curves were used to determine optimal cutoff scores, sensitivity, specificity, as well as positive and negative predictive values.
Results: The agreement between the child and parent interviews as well as between the interviews and clinical diagnosis was low to moderate. Clinicians diagnosed depressive episodes more frequently than the interviews. Cutoff scores and measures of accuracy varied between the reference standards, with less favorable results for clinical diagnosis.
Conclusions: Clinicians may profit from conducting structured interviews. Strategies for dealing with conflicting information from children and parents should be tested empirically and described in detail.