For Autism, No Such Thing as a Quick Diagnosis

In a presentation at the 12th Annual International Meeting for Autism Research (IMFAR), The Center for Autism Research (CAR) concluded that a 10-minute office visit is not sufficient to accurately diagnose autism, even when administered by a group of autism experts. The implications of this study are very important for clinicians and families and show the need for CNNH’s “Specialty Care Medical Home”™ model of care in diagnosing autism.

Diagnosing ASD is a complicated process that requires expertise supported by objective testing results. For those health care providers that are often the first to encounter a child with ASD, relying on observations during a brief office visit is insufficient to determine a diagnosis. Thus, pediatricians, family physicians, and other providers should listen carefully to parental concerns and should refer readily when any “red flags” of autism are raised. For those involved in performing autism evaluations, it is also insufficient to rely solely upon testing results (such as the ADOS) or subjective clinical impressions, but to make a diagnosis based upon a comprehensive evaluation that combines historical information, developmental and examination findings, behavioral assessments and objective testing results.

In their study, CAR researchers took 10-minute video clips of children (ages 15 to 33 months) divided into three groups: early signs of an autism spectrum disorder (ASD); suspected language delay; and typically developing children. Video clips were taken at different points during administration of the Autism Diagnostic Observation Schedule (ADOS) assessing social interaction, communication and play skills. When the experts were asked to rate the video clips based on the question “Would you refer this child for an autism evaluation?”, the researchers found, surprisingly, that the experts were often wrong in their diagnostic conclusions. For 39% of the children with ASD, the experts stated there was not evidence of an ASD diagnosis. There were also high rates of incorrect diagnoses for predicting language delays, and even diagnosing typical children as having ASD 11% of the time.
By Sarah Woldoff