Screening for Attention-Deficit/Hyperactivity Disorder and Comorbidities in a Diverse, Urban Primary Care Setting

Abstract

We tested the accuracy of 2 parent-report tools, the Pediatric Symptom Checklist (PSC-35) and Child Behavior Checklist (CBCL), to identify attention-deficit/hyperactivity disorder (ADHD) and distinguish complex (highly comorbid) cases in an urban, largely Latino pediatric practice. Spanish- and English-speaking parents of children aged 6 to 10 years completed a PSC-35 and CBCL at well visits. Those with CBCL Attention Problems Subscale (CBCL-APS) T scores ≥60 plus controls completed the diagnostic MINI-KID (Miniature International Neuropsychiatric Interview) for Children. Receiver operating characteristic (ROC) curves quantified accuracy of both scales to distinguish ADHD from non-ADHD, and complex from simple ADHD. Two hundred and nine children were screened, and 41 completed diagnostic interviews. Both the CBCL-APS and PSC Attention Scale (PSC-AS) accurately identified ADHD; the CBCL-APS performed best (AUROC CBCL_APS = 0.837; AUROC PSC_AS = 0.728). The PSC Total and Internalizing Scores and the number of CBCL subscale elevations accurately distinguished complex from simple ADHD; the PSC Internalizing Score performed best (AUROC PSC_TOTAL = 0.700; AUROC PSC_INT = 0.817; AUROC CBCL_SUBS = 0.762).