Not enough sound evidence exists for routine screening of all preschoolers to determine whether they have autism, say Canadian researchers.

In a paper published in Monday's online edition of the journal Pediatrics, a team at McMaster University in Hamilton, Ont., said that screening and diagnostic tools are still being developed and revised, and no therapies have a "curative outcome or well-established efficacy to change the course of the condition."

More research is needed because none of the screening tests currently available has "high sensitivity, high specificity and high predictive value … in a population-wide screening program," wrote Dr. Jan Willem Gorter of the CanChild Centre for Childhood Disability Research, and colleagues.

But the American Academy of Pediatrics has taken a different stance, recommending that screening for autism be incorporated into routine medical practice. Gorter said in an interview that he hasn't heard of any similar advice from the Canadian counterpart organization.

"If you want to implement such a program you would like to know if you do more good than harm. That actually was our starting point, to go to the literature and find out more about screening," he said.

"Clearly autism is a developmental disability that has a huge impact on children, lives and families and society, so that would be a good reason for screening. Then we looked at what the tests are and what we know about the tests for screening for autism and found there was not much literature supporting that, and you wonder if you do more good than harm."

He expressed concern about the impact on parents and children if there is a false positive or false negative for autism spectrum disorder. The neurodevelopmental disorder has symptoms that can include differences in social and communication skills, motor skills and sometimes intellectual abilities.

"And with any test there is always a risk that you assume that the test will say you have autism when in fact you don't have autism."

Gorter is also quick to say that parents who have any concerns about their child's functioning should seek help, and "if they do have a diagnosis of autism, to be on the wait list and get the services."

Screening aging falls

"If a child is healthy, there are no parental concerns, is then screening better than not screening? That is the question."

In addition, his group looked at what happens once children are diagnosed with autism to see if they get the services they need, and also looked for evidence supporting interventions to help children with ASD. The researchers found there are wait lists often as long as a year in many centres providing therapies.

Diana Robins, chair of the neuropsychology and behavioural neurosciences program at Georgia State University, said that although there is still work to be done, a number of studies suggest that earlier intervention improves prognosis.

"Early intervention requires early identification, which triggers the referrals for intervention services," she wrote in an email reacting to the McMaster study.

"Although I agree that it is difficult to obtain intensive early intervention for reasons including high cost and long waiting lists, this does not seem to be a reasonable argument against screening.

Rather, establishing clear need through universal screening may propel change in the early intervention systems to meet the needs of our children."

Robins also noted that although screening age is dropping overall, disparities based on race, ethnicity, and socioeconomic status continue to be evident — and she cited a recent study in saying that methodical use of standardized screening tools may decrease these disparities.

"It will be a terrible disservice to children and their families if there is a reduction in screening efforts as a result of opinion pieces arguing that we do not have full empirical support for universal screening."

The study in Pediatrics says the number of children diagnosed with autism has grown from an estimated 0.8 cases per 1,000 in 1983 to 11 per 1,000 in school-aged children in 2007.