In This Section

What We Need to Know About Screening for Autism Spectrum Disorder

Published on April 27, 2021 in Cornerstone Blog · Last updated 5 months 3 weeks ago


Subscribe to be notified of changes or updates to this page.

5 + 10 =
Solve this simple math problem and enter the result. E.g. for 1+3, enter 4.

Addressing the role of screening in promoting the equitable identification of and provision of supports for children with autism spectrum disorder is sorely needed.

By Kate E. Wallis, MD, MPH; Amanda Bennett, MD, MPH; Marsha Gerdes, PhD; Whitney Guthrie, PhD; Susan E. Levy, MD, MPH; Judith S. Miller, PhD; Juhi Pandey, PhD

Editor’s Note: A group of clinicians and researchers from Children’s Hospital of Philadelphia with expertise in autism spectrum disorder (ASD) commented on draft plans by the U.S. Preventive Services Task Force (USPSTF) regarding their upcoming review of the scientific evidence on screening for ASD. The USPSTF makes evidence-based recommendations by weighing the risks and benefits for preventive practices, such as screening children for early signs and symptoms of ASD. This guest blog sums up prior research findings from CHOP that are important factors to consider in the UPSTF’s review.

The American Academy of Pediatrics recommends universal screening for autism spectrum disorder (ASD) in primary care pediatrics, and the CHOP Care Network has been screening all young children for ASD in accordance with these recommendations. However, when the U.S. Preventive Services Task Force (USPSTF) last published its review on screening for ASD in 2016, they concluded that there was insufficient evidence to "assess the balance of benefits and harms of screening" for ASD in children without specific ASD concerns, and therefore insufficient evidence to recommend universal screening in the general population.

When the USPSTF published a draft of their planned approach to revisit this screening data and reassess their recommendation on universal screening, we welcomed to opportunity to share our feedback. We quickly identified several areas that our prior research highlights as important factors to consider in this updated review. The comments that we shared are the following:

  1. Racial/ethnic disparities in screening practices, screening accuracy, referral decisions after a positive screen, and access to autism diagnostic evaluations and interventions may greatly impact diagnosis and functional outcomes. Our prior work has demonstrated racial, ethnic, gender-based, and socio-economic disparities in the accuracy of the most commonly used ASD screening tool, and in who is referred after a positive ASD screen. These differences undoubtedly contribute to disparities in rates and age of ASD diagnosis and intervention, which will affect the ultimate outcomes of interest. Not accounting for these disparities in the primary analysis and analytic framework will lead to inaccurate estimates of effects. Because disparities currently exist, they should be treated as components of the analytic framework, not secondary to it, and thus should be included as moderators of the primary effects.
  2. Developmental surveillance also contributes to autism identification and diagnosis. Our prior work found that 61% of children who completed universal screening in pediatric primary care who were ultimately diagnosed with ASD (n=454) had a negative ASD screen. Many of these children were thus identified by other means, likely in part through ongoing surveillance efforts in primary care. The American Academy of Pediatrics recently reissued clinical practice guidelines for ASD and emphasized the importance of pairing developmental surveillance with screening. Therefore, inclusion of developmental surveillance in the analytic framework may more accurately reflect real-world practice.
  3. If screening is not done universally, it is likely driven in large part by level of caregiver concern, and thus may under-identify children whose parents don't have early concerns about their development. But little is known about the accuracy of caregiver observations and reporting of ASD-related and general developmental concerns in the general population. Prior work (currently unpublished) in which we scrutinized electronic medical records to identify documentation of parental concerns found that nearly 15% of children who screened positive and were later diagnosed with ASD had no reported parental concerns by the time of their positive screen. Without universal screening, these children would otherwise have had delayed or missed diagnoses. Additional evaluation of alternative approaches to universal screening, including relying on caregiver-reported concerns to drive assessment referrals, is an important question to consider.
  4. Referral decisions by clinicians (both aided by screening results and independent of screening results) rely on clinical observations and judgment. If screening is not completed universally, it will be driven in part by clinical decisions, which are shown to be less accurate than when formal or standardized measures are used. Our group has found that clinicians frequently rely on other clinical information when making decisions about when to refer a child for intervention and evaluation, even after a positive screen. Furthermore, prior work has found that brief observations (such as those available to a general pediatrician) are insufficient for making good referral decisions. Additional information about the role and accuracy of clinical decision-making in ASD identification is needed to determine the value of universal screening.
  5. Screening for ASD may facilitate the identification of other developmental conditions of concern. The benefit of ASD screening may therefore have benefits beyond those appearing in the proposed analytic framework. Our group has conducted and contributed to research that has found that the majority of children (72% to 100%) who screen positive for ASD end up with another developmental diagnosis. While not the primary goal of ASD screening, the secondary benefits accrued to children who are identified with other developmental conditions as a result of a positive ASD screen is an important factor to consider when evaluating the usefulness of ASD screening.

We are hopeful that our comments and prior research contribute in a meaningful way to the review and development of evidence-based guidelines by the USPSTF. Now, more than ever, it is critical that public health agencies, such as USPSTF, put equity issues front and center. Addressing the role of screening in promoting the equitable identification of and provision of supports for children with ASD is sorely needed.