Evidence Behind Autism Drugs May be Biased: Study

(Reuters Health) - Doctors' belief that certain antidepressants can help to treat repetitive behaviors in kids with autism may be based on incomplete information, according to a new review of published and unpublished research.

The drugs, which include popular selective serotonin reuptake inhibitors (SSRIs), are sometimes used to treat repetitive behaviors in people with obsessive-compulsive disorder (OCD).

"The main issue to emphasize is that SSRIs are perhaps not as effective at treating repetitive behaviors as previously thought. Further research will help confirm these findings in the long run," said Melisa Carrasco, the study's lead author, in an email.

For their analysis, Carrasco, a researcher at the University of Michigan in Ann Arbor, and her colleagues examined PubMed and ClinicalTrials.gov for randomized, double-blind and placebo-controlled trials -- considered the gold standard in medical research -- supporting the use of SSRIs and similar antidepressants in children with autism.

Their search yielded 15 trials. Five studies were excluded because they did not meet the researchers' criteria. Another five were listed as completed but never published.

Carrasco's team emailed the researchers on the unpublished trials to request their data. One researcher complied and sent the findings.

Of the six studies, three showed a benefit from SSRIs and three -- including the unpublished study -- reported some or no benefit.

Overall, the 365 participants in the six studies showed a small response to the SSRIs, but that association disappeared when the researchers accounted for the studies that were completed but never published.

When only positive findings get published, and negative ones never see the light of day, the evidence on a topic is said to be subject to "publication bias."

As a result of including the unpublished data, "The research made it clear that the effects of (serotonin receptor inhibitors) treatment of (autism spectrum disorders) are considerably overrated," Carrasco and colleagues wrote in the journal Pediatrics.

She told Reuters Health, however, that the new study does not mean the drugs are not useful for treating other conditions related to autism spectrum disorders.

"There is compelling data available, for example, in regards to their use in treating anxiety in autism, and there may still be potential... in treating additional aspects of autism as well," she said.

The new study also had limitations, including there being limited data on the topic and a lack of a unified system to measure repetitive behaviors.

In an accompanying commentary, Dr. Scott Denne, at the Indianapolis University School of Medicine, wrote that the paucity of data "results in physicians being unable to make rational informed decisions" about the benefits and risks of using SSRIs to treat children with autism.

"You come to completely different conclusions based on the information you have," Denne told Reuters Health. "There is a mechanism to make those data available and that data should be available on ClinicalTrials.gov whether it's published or not."

As a result of the 1997 FDA Modernization Act, ClinicalTrials.gov was created as by the United States' National Institutes of Health and the U.S. Food and Drug Administration. The website serves as a listing of trials involving human subjects.

Several bodies including the NIH, Congress and an association of academic journal editors have passed guidelines and regulations encouraging -- and in some cases requiring -- researchers to register their clinical trials on the website.

In a separate study also published in the journal Pediatrics, researchers found the results for the majority of clinical trials involving children are unavailable.

Specifically, of 2,400 completed studies involving children and registered on ClinicalTrials.gov between 2000 and 2010, only 29 percent were ever published. On average, they were not published until two years after the study was completed.

Those researchers also found that only 53 percent of studies funded through the NIH were ever published.

"The problem is that existing clinical research policy does not guarantee availability of the results of all clinical studies and leads to biased health care decisions based on highly selected and incomplete information," said the study's lead author, Dr. Tatyana Shamiliyan of the Division of Health Policy and Management at the University of Minnesota in Minneapolis, in an email.

Dr. Sidney Wolfe, head of Public Citizen's Health Research Group, told Reuters Health that children are sometimes prescribed drugs for uses that are not approved by the FDA. It's called off-label prescribing.

Pharmaceutical companies cannot market a drug for an off-label use, but nothing stops a doctor from prescribing it.

Wolfe said, when children are prescribed medicines because of publication bias, it can be "very dangerous."

Regarding the people who went through clinical trials that were never published, Denne told Reuters Health, "I think the public ought to demand that the results of their data, of what they really sacrificed for, should be available so other people can benefit."