The American Academy of Pediatrics recently recommended that screening for autism be incorporated into routine
physician check-ups, even if no concern has been raised by the parents.

Such routine screening of all children for autism gets a thumbs down from researchers at McMaster University in a Pediatrics study.  The researchers say there is "not enough sound evidence to support the implementation of a routine population-based screening program for autism."

Not only are good screening tools and effective treatments lacking but there is no evidence yet that routine screening does more good than harm, said Dr. Jan Willem Gorter, a researcher in McMaster's CanChild Centre for Childhood Disability Research and associate professor of pediatrics.

Autism, or the autism spectrum disorders (ASDs), is a group of neurodevelopmental disorders with symptoms like differences and disabilities in social and/or communication skills, fine and gross motor skills, and sometimes intellectual skills.

During the past three decades, autism diagnoses have risen dramatically to 11 cases per 1,000 school-aged children from 0.8 cases per 1,000. Reasons for this increase are debated and may  include improved detection, over-diagnosing the disorder or an actual increase. Autism is more common in males with a 4:1 male-to-female ratio. 

For their analysis, the researchers conducted a literature search to assess the effectiveness of community screening programs for autism.

"None of the autism screening tests currently available has been shown to be able to fulfill the properties of accuracy, namely high sensitivity, high specificity, and high predictive value (proportion of patients with positive test results who are diagnosed correctly) in a population-wide screening program," researchers said.

Gorter said that unlike breast cancer screening, no autism screening programs have been studied in randomized controlled trials. "There is no solid evidence on which to base the recommendations of the American Academy of Pediatrics."

"None of the autism screening tests for the general population that we have today have proven accuracy," said Gorter. "That is, they aren't good enough to accurately detect children who have autism or to accurately detect those who don't."

Gorter said the study is a "call for action."

At this time, the researchers called community screening of all preschoolers premature. Alternatively, they recommend careful surveillance and assessment of all preschoolers who show signs of language, social and cognitive problems.