Psychometric evaluation of a DSM-5-based semi-structured parent interview for externalizing disorders in preschool children.

Abstract

Reliable and valid diagnostic interviews for preschool-age children are needed to identify children with externalizing behavior problems requiring treatment and thus prevent adverse outcomes. This study evaluated the psychometric properties of the semi-structured Clinical Parent Interview for Externalizing Disorders (ILF-EXTERNAL). Data were collected from 211 children (3;0-6;11 years, 62% boys) who were screened for participation in a multicenter study on evidence-based stepped care treatments for preschool children with externalizing symptoms (ESCApreschool). The factor structure, interrater reliability, internal consistency, convergent validity, and divergent validity of the ILF-EXTERNAL were examined from a multi-rater perspective (clinicians, parents, preschool teachers). A first-order factor model with four correlated factors showed acceptable model fit and allowed for straightforward interpretability. Interrater reliability estimates were fair or moderate to good [intraclass correlation coefficients(1,1) =  .52-.74]. Categorical diagnostic agreement was moderate (κ = .44-.52). The ILF-EXTERNAL scales exhibited acceptable to good internal consistencies (α ≥ .69, ω ≥ .72), except for the ADHD Functional Impairment scale (α = .65, ω = .68). Correlations between the ILF-EXTERNAL scale scores and the Externalizing Problems scale of the Child Behavior Checklist (CBCL/1½-5) ranged from  .38 to  .56, whereas correlations with the Internalizing Problems scale ranged from  .09 to  .37. Except for one subscale, all ILF-EXTERNAL subscales correlated more strongly with the Externalizing Problems scale than with the Internalizing Problems scale, providing support for convergent and divergent validity. Cross-informant correlations between clinician and parent ratings were higher (r = .46-.65) than between clinician and preschool teacher ratings (r = .05-.33). The ILF-EXTERNAL shows sound psychometric properties in clinically referred preschool children with ADHD and/or ODD.