Harvard Researcher Develops New Method for Autism Diagnosis

Researchers at Harvard Medical School have developed a new web-based diagnostic procedure to identify autism more quickly, an advance that they hope will allow clinicians to provide better care to their patients.

Dennis P. Wall, the lead researcher on the project, said that previous approaches to diagnosing autism were often flawed and took hours to perform. Wall said he sought to develop a new method that would be more efficient and easier for families.

“Clinical observation requires a clinical setting and the room can be artificial, which affects the behavior of the child,” Wall said.

Wall’s method requires parents to answer seven targeted questions and to record a home video of the child that is then examined by an analyst. According to Wall, the home environment is a more natural setting for the child, and videotaping enables families who live in rural areas or may not have easy access to clinical facilities to receive a diagnosis more quickly.

Wall said that many of his colleagues agree that the stress placed on families during the diagnostic process is unnecessary and harmful to the child.

“Some people have to wait months for diagnosis and their children miss the chance for better treatment,” Wall said. “Also it is not until after diagnosis that the families receive compensation for the care which can amount to thousands of dollars.”

Wall used the Autism Genetic Resource Exchange’s database, one of the largest data sets of autism-related behaviors and genetic information, to generate his questionnaire.

Using the database allowed Wall to craft questions based on thousands of previous cases, reducing the possibility of bias and variability between clinicians, he said.

While the project has received some criticism from others in the field, Wall emphasized that the test is not meant for parents to self-diagnose their child. The test is intended to be used to support clinicians in their practice, and should not be used by non-professionals.

“It would be a dangerous idea to put something online as a definite diagnostic test,” Wall said. “It is meant to be a clinical tool and not [strictly] diagnostic.”

The next phase of the project includes making the test widely available, which will require further testing and additional funding.

“We want to work with clinical colleagues to validate it and help design a better mobilized framework,” said Wall. “We hope it will give valuable information to clinicians.”

—Staff writer Armaghan N. Behlum can be reached at behlum@college.harvard.edu.