Note: Most Internet Explorer 8 users encounter issues playing the presentation videos. Please update your browser or use a different one if available.

Symptomatic Presentation of Autism in Toddlers: What Can We See in 10 Minutes?

Thursday, 2 May 2013: 09:00-13:00
Banquet Hall (Kursaal Centre)
T. P. Gabrielsen1, M. E. Villalobos2, M. Farley3, L. A. Speer4, C. N. Baker5, J. Viskochil6 and J. Miller7, (1)Children's Hospital of Philadelphia, Center for Autism Research, Philadelphia, PA, (2)Children's Hospital of Philadelphia, Philadelphia, PA, (3)University of Utah, Salt Lake City, UT, (4)Autism Center, Shaker Campus, Cleveland Clinic, Cleveland, OH, (5)Department of Psychology, Tulane University, New Orleans, LA, (6)Utah Autism Research Program, Salt Lake City, UT, (7)Center for Autism Research, Philadelphia, PA

There have been recent emphases on ASD screening in health care settings and efforts to simplify identification of early signs of ASD.  The structure of a typical health care appointment, however, may not elicit an accurate sample of a child’s usual social behavior.  Individuals with autism spectrum disorders (ASD) can exhibit grossly appropriate social behaviors some of the time.   This may make it difficult for a clinician with limited autism expertise to detect subtle signs of ASD at very early ages by observation alone, which affects diagnostic impressions and subsequent referrals, and in turn can delay treatment during critical developmental periods.  In the United States, typical primary care patient contact time is 10-20 minutes, which may not be sufficient observation time to form an accurate clinical impression of need for referral.


This study aimed to characterize differences in symptom presentation during brief clinical observations among children with and without ASD.


Participants included 3 groups of children between the ages of 15-33 months: (1) children with early signs of ASD, (2) children with suspected language delay, and (3) typically developing children.  Participants were from a large community pediatric practice in which 80% of all children presenting for care were screened with standardized tools for early signs of ASD.  Only three children in the sample had previously been referred for an ASD evaluation and none were from high risk families.  Families spoke English or Spanish and had private insurance, Medicaid/CHIP, or were self-insured.  Participants likely represented an ecologically valid sample of community families.  A clinical evaluation for early signs of ASD (including the Autism Diagnostic Observation Schedule [ADOS]) was videotaped.   Social and communicative behaviors (e.g., initiating, responding, response to name, play and sounds) displayed by the children were measured across the first 10 minutes of the evaluation, and again 30 minutes into the evaluation for 10 minutes. Frequencies of grossly atypical and typical behaviors were compared among the 3 groups. Coders were licensed psychologists with ASD expertise who were research reliable on the ADOS and blinded to diagnostic group membership and screening results. 


Findings suggest very young children with ASD can show grossly typical behavior much of the time, and that some children with ASD may not manifest significant atypical behavior within a 10-minute observation window.   The ASD group showed statistically higher rates of atypical behavior and lower cognitive development and adaptive behaviors on standardized measures, but expert clinical impressions based on 10-minute observations were incorrect for children in the ASD group 39% of the time (false negatives).  Clinical impressions were incorrect 25% of the time for children in the suspected speech delay group and 11% of the time for typically developing children (false positives).


Brief observations likely do not provide enough of a behavioral sample to make a correct referral in all cases. Standardized screening tools identified more children as needing referral for ASD evaluation than clinical impression alone.

| More