Baylor College of Medicine

A doctor leads a child into an office to be screened with a Modified Checklist for Autism in Toddler, or M-CHAT.

Study improves understanding of pediatric autism screenings

Homa Shalchi


Houston, TX -
Media Component

The American Academy of Pediatrics recommends children to be screened for autism spectrum disorder (ASD) at both their 18-month and 24-month pediatric well-child visits. One of the most popular screening tools is the Modified Checklist for Autism in Toddlers (M-CHAT). A collaborative study led by researchers at Baylor College of Medicine explored providers’ diagnostic and referral practices for children who failed the M-CHAT at the 18-month and/or 24-month visits. Their report was published in Pediatrics.

The M-CHAT is a standardized questionnaire that screens for behaviors one would or would not expect to see in toddlers who may have ASD. It can be administered in a pediatrician’s office or accessed free online.

“If a child scores 3 or higher, he or she is considered to have failed the screening,” said Dr. Robin Kochel, associate professor of pediatrics – psychology at Baylor and associate director for research at the Texas Children’s Hospital Autism Center. “If the score ranges between 3 and 7, they are in the gray zone, and providers are encouraged to do a brief follow-up interview.”

The follow-up interview helps clarify the behaviors endorsed on failed items, such as the child failing to respond to his/her name or preferring isolated play. Although this interview helps to improve the specificity of the M-CHAT, it is not routinely used in clinical practice.

The researchers identified children who failed the M-CHAT at 18 and/or 24 months and continually reviewed their records over the course of 2 to 4 years. During this surveillance period, they examined what types of referrals were made following failed screens, whether families completed referrals, and what diagnoses were documented in the medical record.

Findings revealed high rates of screening with the M-CHAT at both the 18-month and 24-month visits. However, only 31 percent of providers placed autism-specialist referrals after failed screens, meaning that the vast majority did not take referral action and were not immediately evaluating the child for ASD. Only about half of the families that received referrals completed them.

“There needs to be action by pediatricians following that failed screening. Whether that action is immediately evaluating for autism themselves, or making those appropriate referrals if they are not sure a child meets the criteria for autism,” Kochel said.

The team’s findings concluded that ASD screening alone does not translate into earlier autism diagnoses or a younger age at diagnosis. Providers must act on failed screens—and families must follow through on referrals—in order for children to be diagnosed sooner and begin appropriate interventions, such as applied behavior analysis (ABA). ABA is one of the most well-researched behavior therapy programs for children with autism, with initiation at earlier ages being associated with better long-term outcomes. However, many children are not eligible to receive ABA without an autism diagnosis, which makes early identification of ASD that much more critical.

Kochel noted that educational tools and support must be placed in clinics so providers feel empowered to make ASD diagnoses when the symptoms are clear, as well as know where and when to refer families who need specialist evaluation and/or intervention.

“We already have developed a three-part educational webinar series that focuses on appropriate screening, diagnosis and referrals for autism. We are thinking of how we can use that as an educational tool in conjunction with in-clinic support, perhaps through a best practice alert: if a child fails an M-CHAT, a notification pops up in the clinical record to inform the provider of the next steps,” Kochel said.

Sonia Monteiro was the lead author on the study; other contributors to this work include Jack Dempsey, Leandra Berry and Robert Voigt. These authors are affiliated with Baylor College of Medicine, Texas Children’s Hospital, Children's Hospital Colorado and University of Colorado Denver School of Medicine.

This work was supported by the William Stamps Farish Foundation. Dr. Goin-Kochel is also partially supported through the Intellectual and Developmental Disabilities Research Center (1U54 HD083092) at Baylor College of Medicine.

Back to topback-to-top