Diagnoses of autism and ASDs have become increasingly common. Clinical evaluations often involve extensive observations, psychological testing, medical procedures, interviews with family members, and consultations with multiple professionals. Such evaluations are typically instigated when concerns are raised by pediatricians, preschool teachers, welfare workers, daycare providers, or family members. Because concerns may be raised about many kinds of early childhood problems, it is essential to distinguish effectively between problems that do vs. do not warrant extensive clinical evaluations for autism. It is also essential for all stages of assessment to include a broad spectrum of characteristics in order to avoid neglecting other aspects of children’s functioning.
A team at the Oregon Health and Science University (Sikora et al., 2008) tested the degree to which clinical conclusions about autism were associated with scores on the CBCL/1.5-5 and the Gilliam Autism Rating Scale (GARS), which is specifically designed to identify autism. The Oregon team studied 147 36- to 71-month-old children evaluated via tests, observational procedures, diagnostic interviews with caregivers, and forms completed by caregivers. The Autism Diagnostic Observation Scale-Generic (ADOS-G) was used to classify children as having autism, ASD, or non-autistic spectrum disorders. CBCL/1.5-5 scores were more strongly associated with the ADOS-G classifications than was the GARS score for autism. In fact, a MANOVA showed no significant association between the three ADOS-G categories and the GARS, whereas the CBCL/1.5-5 Anxious/Depressed, Withdrawn, and Aggressive Behavior syndromes, as well as the DSM-oriented Pervasive Developmental Problems (PDP) scale, were all significantly associated with the three ADOS-G categories. Furthermore, for all children classified as autism-positive vs. autism-negative and also for children grouped by gender and by cognitive level, the CBCL/1.5-5 Withdrawn syndrome and PDP scale had better sensitivity than the GARS. For every group comparison, at least one of these two CBCL/1.5-5 scales also had better specificity than the GARS. Moreover, CBCL/1.5-5 scores were significantly associated with GARS scores, as reflected in a Pearson correlation of .66 with the PDP scale and .58 with the Withdrawn syndrome. The authors concluded that “The efficiency of the CBCL in screening for ASDs as well as other behavioral problems is thus appealing for primary care and educational settings where efficiency, cost, and response burden must be taken into consideration.”
Reference: Sikora, D.M., et al. (2008). Does parent report of behavior differ across ADOS-G classifications: Analysis of scores form the CBCL and GARS. Journal of Autism and Developmental Disorders, 38, 440-448.