Computerized screening for DSM Classifications using CBCL/YSR Extended Checklists: A clinical try-out

Abstract

In this paper, we present two studies on the development and evaluation of a computerized aid to screen for DSM classifications with the use of CBCL/YSR/TRF input data. In the first study we examined the screening capacity of the computerized aid by administering the Achenbach extended questionnaires CBCL and YSR in a clinical sample and undertaking a DSM-III-R based interview (the ADIKA, the Dutch version of the DICA). Computer generated and checklist based classifications were compared to a criterion diagnosis based on the structured interview. Analysis of the agreement (sensitivity and specificity) between the classifications showed that the computer generated diagnoses did not always agree with classifications generated by the standard DSM-III-R interview. A detailed analysis of the symptom patterns revealed at least five extraneous factors that appear to restrict the screening capacity of the computerized aid. Corrective measures were undertaken to adjust for the influence of some of these additional factors. Next, we ported the program to Windows and adapted the content to DSM IV. In the second study the usefulness of the program for clinical psychodiagnostic practice was examined in eight clinical settings and results showed a positive evaluation. It was reported that the program served a supportive function during the diagnostic process.