Detecting and Assessing Preschool Children with Autism Spectrum Disorder Using Child Behavior Checklist.

Background: Children with autism spectrum disorder (ASD) usually accompanied emotional and behavioral problems. Children with emotional and behavioral problems could result in more impairments on their social and communication, thus make stress on their caregivers.

Objectives: Children Behavior Checklist 1.5-5 (CBCL/1.5-5) is a tool for evaluating emotional and behavioral problems of preschool children and is used to screen ASD children. However, there is no research in Taiwan. In this study, the CBCL/1.5-5 is used to detect and assess preschool children with ASD.

Methods: 328 children under 48 months old, including 130 children with ASD, 135 children with developmental delay (DD), and 63 children with typical development (TD). The CBCL/1.5-5 was completed by the caregivers for measuring emotional and behavioral problems of all children.

Results: ASD group is higher than DD group and TD group on internalizing problems of broadband scales, withdrawn subscale of syndrome scales, and pervasive developmental problems subscales of DSM-direct scale. In addition, ASD group is higher than TD group on both externalizing and total problems of main scales, and emotionally reactive of syndrome scales. All of withdrawn, pervasive developmental problems and internalizing problems are better index for distinguishing children with ASD from children with DD. Using signal detection theory, cut-off scores of each subscales are as below: T-score of 65 in the withdrawn subscale, the sensitivity and specificity were 69% and 68%, respectively; T-score of 69 in the pervasive developmental problems subscale, the sensitivity and specificity were 70% and 64%, respectively; T-score of 62 in the internalizing problems subscale as the cut-off, the specificity were 63% and 60%, respectively. All of withdrawn, pervasive developmental problems, attention problems, internalizing problems and total problems are better index for distinguishing children with ASD from children with TD. Using signal detection theory, cut-off scores of each subscales are as below: T-score of 59 in the withdrawn subscale, the sensitivity and specificity were 87% and 75%, respectively; T-score of 64 in the pervasive developmental problems subscale, the sensitivity and specificity were 82% and 81%, respectively; T-score of 56 in the attention problems subscale, the sensitivity and specificity were 73% and 64%, respectively; T-score of 59 in the internalizing problems subscale, the sensitivity and specificity were 75% and 72%, respectively; T-score of 58 in the total problems subscale, the sensitivity and specificity were 79% and 68%, respectively. The results revealed that using the withdrawn and pervasive developmental problems subscale as a screening index for detecting children with ASD, it showed accepted validity.

Conclusions: Used the CBCL/1.5-5 as a screening tool for children with ASD, the withdrawn and pervasive developmental problems subscale have moderate validity, suggest the two subscales can be screening tool in the clinical setting.