Diagnostic Aid Offers New Approach to Detecting Early Autism
The FDA has authorized the first aid for diagnosing Autism Spectrum Disorder in children 18 months through 5 years of age called Canvas Dx. It was developed by Cognoa, a behavioral health company.
“Currently, pediatricians refer most children with suspected developmental delay to specialists to diagnose and prescribe treatment, which can result in families waiting months or even years before their child receives an initial autism diagnosis and can start potentially life-changing early interventions,” said Dr. Sharief Taraman, division chief at Cognoa.
The senior medical director of Clinical Adoption at Cognoa, Dr. Colleen Kraft, said that one of the reasons primary care physicians often aren’t able to detect autism early is because they have been positioned to identify developmental delay and behavioral conditions, while already managing children with behavioral health conditions such as ADHD, anxiety, and depression.
Autism spectrum disorder is a complex neurodevelopmental condition that affects communication, learning, and behavior, but there is a lack of established biomarkers for diagnosis. Typically, it is diagnosed by a primary care physician reviewing a child’s behavioral and developmental history and based on a direct observation of symptoms.
Canvas Dx can give primary care physicians a broader neurodevelopmental picture of a child. It works by having a parent or caregiver complete a questionnaire in the Canvas Dx caregiver-facing app. Two short videos of the child are then uploaded into the app and the videos are reviewed by analysts who complete a questionnaire based on their analysis of the child’s behavior. A doctor then completes a visit and questionnaire with the parent or caregiver and child.
Canvas Dx’s AI algorithm evaluates all the inputs and provides a positive or negative result for autism that the pediatrician utilizes in combination with the clinical presentation of the child to provide a diagnosis and direct appropriate next steps.
“In order to reduce the risk of false classifications, the algorithm has also been designed to provide an indeterminate output as a safety feature. Based on this information and their own clinical evaluation of the child, the doctor will diagnose the child with autism or direct to other next steps in care,” said Taraman.
The diagnostic aid for autism received FDA Breakthrough Device Designation in late 2018, and in 2020 the company completed a study in which the results exceeded the study endpoints and submitted an application for FDA De Novo marketing authorization.
The De Novo premarket review pathway is a regulatory pathway for low-to moderate-risk devices of a new type.
“Along with this authorization, the FDA is establishing special controls for devices of this type, including requirements related to labeling and performance testing,” said Taraman.
“The marketing authorization of the Cognoa ASD Diagnosis Aid provides a legally marketed device that other sponsors may reference as a predicate device when submitting a 510(k) notification for a device of the same type,” said a spokesperson from the FDA.
While the device has been authorized by the FDA, the agency acknowledged this was done with limitations.
“The risks associated with the use of the device include misdiagnosis and delayed diagnosis of ASD, based on a false positive result, which was observed in 15 out of 303 study subjects without ASD, and a false negative result, which was observed in one out of 122 study subjects with ASD, or when no result was generated. Both misdiagnosis and missed diagnosis can result in delayed treatment of ASD or delivery of treatment not appropriate for ASD,” said the FDA spokesperson.
Research demonstrates that with autism specific early intervention, children can benefit from greater improvement in cognitive, language, and social skills and decreases in problem behavior and autism symptoms.
The aid may also help to combat diagnostic disparities in detecting ASD.
A study from the Health Resources and Services Administration shows that children with racial or ethnic minority backgrounds, children from disadvantaged socio-economic backgrounds, and non-native English-speaking children are less likely to be diagnosed with ASD or are diagnosed later than White children.
While the device will aid in early diagnosis, according to Taraman it will also help primary care physicians avoid biases and disparities existing in the current autism diagnostic process, as most pediatric autism research has focused on White males from an above median socioeconomic background, which isn’t representative of all children.
In fact, Taraman said that girls are typically diagnosed 1.5 years later than boys, and one in four children under the age of 8 living with autism, most of whom are Black or Hispanic, are not being diagnosed at all.
“We therefore consciously built our AI algorithm to account for differences in gender, racial, ethnic, and socio-economic backgrounds,” said Taraman.
While Cognoa doesn’t expect to make Canvas Dx available in the U.S. until later in 2021, the company is working directly with insurers and Medicaid to ensure that Canvas Dx can be reimbursed by health insurance plans.
The company is also developing other solutions as separate products which could potentially use google-glass and AI to further extend the capabilities of the device.
“It is clear that AI has immense potential to make therapeutic solutions more accessible, and available to children with behavioral conditions,” said Taraman.