Autism Screening and Identification: From Paper to Electronic Medical Record

Parker Webster and John W. Harrington, MD

The American Psychiatric Association’s recent release of the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) describes autism spectrum disorder (ASD) as a range of complex neurodevelopmental delays characterized by marked impairment in social and communicative functioning with repetitive or restrictive behaviors and interests.1

A recent report from the Centers for Disease Control and Prevention’s Autism Developmental Disabilities Monitoring Network (ADDMN) notes that the U.S. prevalence of ASD is estimated at 1 in 68 (14.7 per 1,000) children, with a male to female ratio of approximately 4 to 1.2

While signs and symptoms of ASD can be detected as early as a child’s first birthday, the average age of diagnosis is approximately 4 to 5 years in most U.S. states (50 months, according to ADDMN).2 Although the report shows improvements in the recognition of autism in children of all socioeconomic and ethnic groups, differences still exist in the rates of diagnosis among the different communities, with fourfold differences among the ADDMN’s reporting sites (5.7 per 1,000 in Alabama compared with 21.9 per 1,000 in New Jersey).2

Experts agree, however, that the best prognoses for ASD are found among children with an early diagnosis and subsequent early intensive behavioral intervention.3 No biomedical tests are now available to diagnose or screen for ASD. Rather, clinicians must rely on the clinical presentation and historical report of parents at primary care visits. The challenge for the medical community is this: How can we reliably diagnose a condition in children for which no objective tests are available?

The American Academy of Pediatrics has endorsed the use of a 2-stage developmental screening tool, the Modified Checklist for Autism in Toddlers, Revised With Follow-up (M-CHAT-R/F).4 This revision of the original M-CHAT screening tool has been whittled down to a 20-item, yes/no questionnaire that assesses a child’s risk of ASD by focusing generally on joint communicative behaviors and supplying examples for parents to help qualify their answers. 

The M-CHAT-R/F, when administered and scored appropriately, is widely considered to be sufficiently sensitive and reliable; however, barriers remain to its consistent and standard use. Improper scoring, parental misunderstanding of questions, and improper use of the new follow-up interview questions have been shown to significantly jeopardize the positive predictive value of the screening tool.5

When 3 to 7 M-CHAT-R/F answers are scored as failed, the child is said to be at risk for autism, and further screening is indicated. This further screening consists of administering the follow-up interview developed by Robins and colleagues,4 which is designed to address the previously identified barriers to consistent screening.

Problems arise, however, when health care providers, rather than administering the standardized, algorithmic follow-up interview questionnaire, conduct their own ad hoc interview with the assumption that the child seems fine, and that the parent likely misinterpreted one or more of the questions. Furthermore, when providers review the answers with parents without strictly adhering to the follow-up interview, they may add a layer of subjectivity to the screening process, thereby reducing the accuracy and reliability of the M-CHAT-R/F.

Moreover, as many practices struggle to implement the meaningful use of electronic medical records (EMRs), they face the additional challenge of developing a workflow or interface that allows results from the M-CHAT-R/F follow-up interview, which may have been collected and recorded using an online tool, to be downloaded, scanned, or otherwise entered into a specific patient’s record.

For example, equipping each examination room with a tablet or computer that parents can use to complete a screening questionnaire may be beyond the financial means of many practices; instead, they will need to create alternative ways to reliably screen patients for autism in a busy clinical space without succumbing to the use of shortcuts that can undermine the integrity of the screening tool.

The following is a suggested approach that might work if your practice uses paper forms for charting:

1. The parent completes a paper copy of the M-CHAT.

2. Either the M-CHAT is hand-scored, or responses may be transferred into an EMR version that features built-in scoring.

3. If the screen is negative, the form can be scanned into the patient’s chart if it is not already part of an EMR.

4. If the screen is scored as having 3 or more failed items, the parent is rescreened: He or she completes the M-CHAT-R/F follow-up questionnaire using a tablet or computer that allows the use of the skip-logic algorithmic survey software (such as that at required to screen reliably.

Given the variability of EMR systems and clinical practices, it is likely that your current usual workflow is very different from this suggested approach. However, it is important to realize that using an automated system will improve the objectivity of your autism screening and thus its sensitivity.

While the M-CHAT-R/F is not infallible, and a significant number of ASD diagnoses in high-functioning children will be missed,4 it is hoped that if autism screening is performed well, the children with the highest needs will be identified as early as possible.


Parker Webster is a clinical research assistant at the Virginia Commonwealth University School of Nursing in Richmond.


John W. Harrington, MD, is the division director of General Academic Pediatrics at Children’s Hospital of The King’s Daughters and a professor of pediatrics at Eastern Virginia Medical School in Norfolk.


1. Harrington JW, Allen K. The clinician’s guide to autism. Pediatr Rev. 2014; 35(2):62-78.

2. Autism and Developmental Disabilities Monitoring Network Surveillance Year 2010 Principal Investigators. Prevalence of autism spectrum disorder among children aged 8 years — Autism and Developmental Disabilities Monitoring Network, 11 sites, United States, 2010. MMWR Surveill Summ. 2014;63(SS02):1-21.

3. Myers SM, Johnson CP; American Academy of Pediatrics Council on Children With Disabilities.  Management of children with autism spectrum disorders. Pediatrics. 2007;120(5):1162-1182.

4. Robins DL, Casagrande K, Barton M, Chen CM, Dumont-Mathieu T, Fein D. Validation of the Modified Checklist for Autism in Toddlers, Revised With Follow-up (M-CHAT-R/F). Pediatrics. 2014;133(1):37-45.

5. Harrington JW, Bai R, Perkins AM. Screening children for autism using an electronic form of the M-CHAT in an urban clinic setting. Clin Pediatr. 2013;52(1):35-41.