Skip to main content

Medical studies about drugs may be victims of spin, says report

This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American


Does the Food and Drug Administration (FDA) have access to the data it needs to make informed decisions on approving drugs, or are drug companies cherry-picking the studies they publish to make their drugs look better than they actually are?

A new report in Monday's PLoS Medicine questions whether doctors and patients are getting objective information about whether a medicine works. That's because more than half of studies on government-approved medications—presumably the ones that show a drug doesn't work—are never published, and those that are show disproportionately positive results.

Just 43 percent of trials behind 90 drugs approved by the FDA were published in the medical journals most commonly consulted by doctors, according to the report by researchers at the University of California, San Francisco. Of those, 66 percent show the drugs are beneficial. The results are based on 909 clinical trials of medicines approved between 1998 and 2000.


On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.


"When trials are selectively published … it will skew the efficacy of the drug and make it look like it works better than it does," says Erick Turner, who co-authored another study on the issue published in January. "It's going to create a lot more enthusiasm among consumers of that information or in the words of Alan Greenspan, 'irrational exuberance.'"

So-called "positive publication bias" has been described before. Why this happens is unclear—optimism is a natural human instinct, so perhaps the desire to publish good news may be innate. (Except among us journalists, of course.) The theory is that drug company funding can ramp up that bias significantly. And previous findings by Turner—notably an analysis of information submitted to the FDA for approval of antidepressants—suggests there may be something to the theory.

His analysis published in the New England Journal of Medicine, for example, found that 69 percent of studies of commercially available antidepressants were published; of those, 94 percent conveyed positive results, compared to half of the trials the FDA considered in green-lighting the drugs for marketing. A third of the negative or questionable results of those trials were never published, that analysis found.

The researchers didn't identify the 90 drugs whose data they looked at in the current study. One of the authors, Ida Sim, says the selective publication practice is likely to be just as common for today's drugs as it was for those approved by a decade ago.

But a spokesman for a pharmaceutical industry group says that publishing the results of drug trials in medical journals isn't as important as the time the FDA spends reviewing new drug applications. Doctors and patients know everything they need to from the labels on approved meds, says Ken Johnson, senior vice president of the Pharmaceutical Research and Manufacturers of America (PhRMA).

There may be hope for transparency on the horizon, says FDA spokeswoman Susan Cruzan: A 1-year-old law, the FDA Amendments Act of 2007 (FDAAA). The law, signed by President Bush last year, requires that all trials backing FDA-approved drugs and devices be registered on ClinicalTrials.gov, a Web site of the National Institutes of Health. Information including the demographics of trial participants, the number of people who drop out and outcomes of the measurements scientists say at registration that they'll be looking for in the study also must be made public on the site starting Saturday.

FDAAA was a Congressional response to hearings in which it became clear that companies were less likely to make results of studies that showed significant side effects public. The idea is that the public, and the FDA, can go back and check how many studies were started, the assumption being that if a given one wasn't published, the findings might have been negative, arguing against approving the drug. (PhRMA supports the requirements of FDAAA and also sponsors another informational Web site, Clinicalstudyresults.org, Johnson says in the statement.)

But while the FDA releases a summary of information it uses in approving new drugs, it doesn't specify the trials it considered, notes Sim, an associate professor of medicine at the

University of California San Francisco. FDAAA doesn't affect what information the agency releases when it gives its stamp of approval to the meds.

"It's critically important that we know trials exist and that we get the summary results, positive and negative, into the public domain—that's a huge step and more than any country is doing now," says Sim, citing the World Health Organization's International Clinical Trials Registry Platform that suggests guidelines for transparency in drug studies.

Updated at 1:40 p.m. with comment from PhRMA.

(Image from iStockphoto/Marcelo Wain)