Screening Instruments: Autism Spectrum Disorders


By Michael Reiff, MD, FAAP

michaelreiffWith the current prevalence of autism spectrum disorder (ASD) estimated at 1 in 68 children and rising, it is imperative that pediatricians take an active and informed role in identifying children at risk. Despite a great deal of effort regarding early identification, parents generally report initial concerns before their child is 18 months and still the average age of diagnosis remains at about 4-5 years of age. It is important that screening tools used by pediatricians be as evidence based as possible to avoid over- or under-referral for diagnosis.

The current AAP guidelines recommend ASD screening at 18 and 24 months. This recommendation is prudent in spite of some current controversy about the efficacy of screening children whose parents have not voiced concerns. We have evidence that we are already missing children at risk for ASD with present recommendations and procedures, and stopping screening would be likely to compound this and be a step backwards. Universal screening for ASD can streamline the early detection process and facilitate appropriate referrals.

General screening tools familiar to pediatricians such as the Ages and Stages Questionnaires (ASQ), Communication and Symbolic Behavior Scales (CSBS), and Parents’ Evaluation of developmental Status (PEDS) can all be helpful in identification of children at risk for developmental and language delays but are not specific to ASD screening.

Some of the more familiar ASD screening tools include the Modified Checklist for Autism in Toddlers (MCHAT), which is a parent completed questionnaire, and the Screening Tool for Autism in Toddlers and Young Children (STAT), an interactive screening tool designed for children in whom developmental delays are suspected.

The Modified Checklist for Autism in Toddlers, Revised (MCHAT-R), with Follow-up interview (M-CHAT-R/F) is one of the most used and researched screening tools. The original M-CHAT produced an unacceptably high false positive rate (the number of children who screened positive but were not found to have ASD on careful evaluation), even when a follow-up interview was added. The M-CHAT-R/F was developed in order to reduce the false positive rate and detect more ASD cases. The MCHAT-R scoring is simplified from the original MCHAT and contains an algorithm that separates low-risk, moderate-risk and high-risk scores. The M-CHAT-R follow-up interview is essential for moderate-risk scores because a it yields a dramatically increased positive predictive value (PPV) when both stages are implemented. In addition, using this new protocol, the PPV for any developmental disorder using the M-CHAT-R/F is .946, indicating that a high percentage of children screening positive by this procedure need evaluation and intervention. Scores on the M-CHAT-R/F can range from 0-20.

  • Those with low risk scores (0-2) can continue with developmental surveillance without the follow-up interview or referral.
  • The follow-up interview should be administered to those at moderate risk (3-7), and, if the interview yields positive scores (2+), children should be referred for evaluation and early intervention.
  • Those with high-risk scores (8-20) can bypass the follow-up interview and should be immediately referred.

One of the key issues in ASD screening is early detection to assure that identified children who receive the ASD diagnosis receive intensive early evidence-based interventions as soon as possible. In the past few years there have been many advances in detecting the early signs of ASD. These findings indicate that there is substantial heterogeneity in the presentation of ASD.

Some early markers of ASD between 12 and 24 months include reduced levels of social attention and social communication as well as repetitive behaviors with atypical body movements and motor development. Temperament observations around age 2 can include lower sensitivity to social rewards as well as negative affect and difficult behaviors. Helpful behavioral markers below 1 year of age have not been constantly identified. Language development and/or non-verbal cognitive gains may slow down after the first year. From 6-18 months gaze to faces, vocalizations directed toward people and social smiling may decline.

In light of this research we are now in a good position to start developing tools to measure these behaviors in clinical practice using observations and parent reports.

References available upon request.

Annual Sponsors

Children's Minnesota
Gillette Children's
Hennepin Healthcare
University of Minnesota Health
Essentia Health
Mayo Clinic
Shriners Healthcare for Children-Twin Cities