The Each Child Study: Early Autism Screening in a Pediatric Practice

Background: Systematic screening for autism in toddlers can lead to earlier identification and intervention. It may also improve our understanding of the epidemiology and clinical course of ASDs.  However, the feasibility and utility of comprehensive screening is not yet established.  

Objectives: We collaborated with a community-based pediatric practice to conduct autism screening for toddlers seen during a six-month period.  Objectives included outlining  methods for conducting screening in a pediatric practice, estimating the rate of significant ASD-related concerns in toddlers, and comparing screening results to parent and pediatrician concern levels. 

Methods: A large pediatric clinic with a diverse ethnic and socioeconomic patient base was approached for participation.  Researchers met with physicians and staff to determine an optimal method for incorporating screening questionnaires into their care service.  Screening took place over a six-month period, and included all children born in 2006 who were seen for sick or well-child visits.  Office staff distributed and collected screening questionnaires while parents waited for appointments.  Parents and providers were also asked to indicate their current level of concern about the child’s development (before knowing screening results).  Research staff obtained all necessary permissions to obtain the clinic’s daily schedule in order to determine how many eligible children were screened or missed.  Office staff received daily contact from researchers to maintain compliance, and were motivated with feedback about individual rates of completed screenings.  Both the Modified Checklist for Autism in Toddlers (M-CHAT; Robins et al., 2001) and Communication and Symbolic Behavior Scales Developmental Profiles Infant Toddler Checklist (CSBS-ITC; Wetherby & Prizant, 2002) were administered.  Researchers scored all questionnaires, and positive screens were followed up with phone interviews and in-person evaluations (ADOS; Lord et al., 2000 and Mullen Scales; Mullen, 1995) where appropriate.  All forms were also available in Spanish, and a Spanish-speaker administered phone follow-up and in-person evaluations when appropriate.      

Results: The clinic screened 81% of eligible children.  The group closely reflected ethnic and socioeconomic diversity statewide. Of the 817 children screened, 24% failed either the M-CHAT or CSBS-ITC or both.  Of those, 42% were eliminated as false positives after the phone interview; 38% could not be reached.  Of the remaining 20% (n=44),  57% (n=25) came for an in-person evaluation.  Two more were evaluated based solely on parent or pediatrician concern.  Thirteen children were identified as showing significant early signs of autism; 8 were already of concern to parents or pediatrician, including 3 who had already been diagnosed.  Fourteen children showed speech or other developmental delays (6 were already of concern to parents or pediatrician).  One appeared to be developing typically, and was not of concern to the parents or pediatrician.  Of the 27 children seen in person, 11 had been referred to early intervention services.   

Conclusions: Systematic screening identified possible autism in children that were not of  current concern to parents or pediatricians.  Our high rate of screening questionnaire compliance was likely due to removing the burden of scoring and interpretation from the pediatricians, as well as our frequent feedback to staff and physicians.