Decision Factors in Referrals for Autism Spectrum Disorder Evaluations

Background: Children with autism spectrum disorder (ASD) have better outcomes with early identification and early intervention. Early screening and diagnosis can help identify children with ASD. Professionals have the responsibility to refer children, of any age, for early identification in the presence of concerns or positive screening results. Currently, however, less than half of children with ASD received their first comprehensive evaluation by a professional before age 3, which is late, given the stability of diagnoses at age 2. If signs are evident in many children at age 2, it is unknown why more children are not referred at earlier ages.

Objectives:  This study looked at the factors that professionals may use in order to make referral decisions.

Methods:  Over 360 interdisciplinary clinicians and educators attended a professional development workshop, where they watched brief clips of children with typical development, language delays, or ASD. Professions included in the sample were health care, speech language and hearing, early intervention, educators, therapists and family service professionals. Professionals were also asked to make a decision about referral, as well as state the factors they used to make this decision.

Results:  The most commonly used factors to make referral decisions were interactions/engagement, verbal/non-verbal cutes, and appropriate play. Across disciplines, referral accuracy was consistent (75%). A Chi-square test showed no significant effect of profession on accuracy of referral decision (χ²=0.719). A more detailed analysis of the Chi-square residuals for the collapsed professional categories revealed that speech and hearing professionals were less likely to refer children for autism evaluations in our sample. All other professions showed similar and higher rates of referral. The sensitivity and specificity in referral decisions for this group of professionals was fairly well balanced at .74 and .75 respectively. In our sample, although professionals were accurate in their referrals for a majority of children, they still missed 26% of children with subtle early signs.

Conclusions:  This result is encouraging, because it tells us that the broad range of disciplines represented in “frontline” early childhood personnel can notice early signs of autism and refer for further evaluation equally, but that this method alone is not foolproof and that referrals are not happening at the same level as shown in this demonstration, suggesting further exploration into factors affecting actual referrals and follow through. These findings also support the need for adding universal screening for autism to detect early signs that might be missed in brief observations.