Can CBCL/1.5-5 Distinguish Between Autism and Attention Deficit Disorder?

Background:

Autistic Spectrum Disorders (ASD) and Attention-Deficit /Hyperactivity Disorder (ADHD) are common neuropsychiatry syndromes which begin in childhood. Attention problems have been described in autism. Some authors report  that 60% of children with Pervasive Developmental Disorder (PDD) meet criteria for ADHD. In contrast, autistic traits have also been identified in children with ADHD. Misdiagnosis is very common, in a recent study 33% of children classified as ADHD when assessed with the ADI-R were rediagnosed as PDD. Recent studies confirm that the most common diagnosis prior to autism is ADHD. Current DSM-IV exclusion criteria makes impossible to make a dual diagnosis (ADHD + Autism) in children.

The CBCL/1.5-5 is a parent-report checklist designed for assessing most frequent emotional and behavioral problems of children between 18 months through 5 years old. DSM-oriented scales have been added to empirically based scales which include: Affective, Anxiety, ADHD, Oppositional Defiant, and PDD Problems.

Objectives:

To examine if the CBCL/1.5-5/PDD and Attention-problems subscales can discriminate children with Autism versus ADHD and a healthy control group (HC).

Methods:

A total of 438 parents of preschool children with a mean age of 4.4±1.1 participated in the study from two different settings. The community sample consisted of 376 children recruited through 6 nurseries. The clinical sample consisted of 62 patients seeking attention in the Psychiatric Hospital Dr. Juan N. Navarro. Parents in both groups completed the CBCL/1.5-5, the response rate was 80%, the mother being the most frequent informant (77.6%). The clinical group was assessed through observation with appropriate play materials and their parents completed structured interviews to assign a DSM-IV diagnosis of Autism, Asperger, PDDNOS, ADHD and Other Psychiatric Diagnosis. ROC curves were used to determine criterion validity for the CBCL/1.5-5-PDD. Sensitivity and specificity for different cutoff points were calculated for assessing ASD group from either HC or ADHD. Mean differences were used to test if the CBCL/1.5-5/Attention problems and ADHD subscales could distinguish autism versus ADHD groups.

Results:

The ASD group had the highest ratings in the CBCL/1.5-5/PDD (p<.001) in contrast to the ADHD and the OPD group. However, the autism group had lower ratings than the Asperger group (11.6 vs.13.5) despite autism is considered as a more severe disorder. Elevated punctuations for the ASD group on the CBCL/1.5-5/ADHD and attention problems subscales could not differentiate them from the ADHD group. Roc curves show that the PDD subscale could effectively distinguish PDD vs. Healthy and ASD  vs. ADHD (AUC of .95 to .97). The CBCL/1.5-5-PDD showed excellent sensitivity and specificity through ROC curves in differentiating children with ASD from either healthy controls or with ADHD. The CBCL/1.5-5/Attention problems and ADHD- subscales fail to discriminate the groups.

Conclusions:

Given the high co-morbidity between autism and ADHD, resulting in common misdiagnosis, children should be evaluated with instruments that assess for both, autism and ADHD to avoid misclassification which leads to inaccurate and incomplete psychiatric and psychosocial treatment. The CBCL/1.5-5 can be used to identify children with developmental problems as a first screening measure for clinical or epidemiology studies.