18 FEB 2005
Drug ads: Data-free zone?
With the pain-killer Vioxx just yanked from the shelves, with Merck facing billions in lawsuits, and with Congress investigating the Food and Drug Administration, the last thing the pharmaceutical industry needs is more bad news. Now, a study in the Canadian Medical Journal shows that many drug advertisements in medical journals don't have references to back up their claims.
And many of the references that do appear mention industry-funded studies -- or data secret that the drug companies won't release.
In a study of 438 ads published in 10 medical journals in 1999, two UCLA emergency medicine professors found that 29 percent contained no references. Assistant professor Richelle Cooper and professor David Schriger found that 19 percent of the ads that did have references cited "data-on-file," unpublished information held by the drug manufacturer.
It's in my file, somewhere
The researchers asked for the data-on-file, but it "was rarely available" says Cooper, "and it took several weeks to come, or to get a letter saying it was proprietary." Only 20 percent of those requests for data produced any document. (Drug companies are required to release this information to the FDA, but not to physicians or researchers.)
In contrast, all of the research articles in the same journals contained references, and 88 percent of the citations were to articles in medical journals. The researchers were able to find 99 percent of those references, but only 84 percent of references the drug-ad refs, largely because it was data-on-file.
The comparison between ads and research articles was suggested by a study reviewer, Cooper acknowledges. "It's not made to be a comparison, in the sense that one is better, and one is worse, but it's an effort to put it in context. If a physician saw an article in a journal and read the claims and wanted to look at the evidence, how easily could they obtain it, versus something cited in a pharmaceutical ad?"
Follow the money
Many references in the drug company ads cited research that was sponsored by the company or performed by company employees, Cooper and Schriger wrote. Among 294 ads that cited original research, "58% indicated that the research was sponsored by or had authors affiliated with the product's manufacturer.... "This was true of only 8 percent of the research-article references.
Now, before you jump to the conclusion that data-free ads are reckless ads, remember that the FDA is supposed to be monitoring drug ads. And, as Cooper says, it is logical for a company to cite its own research... "In part, it's not surprising that they will cite the research they have sponsored about their drugs, but it's definitely in contrast to what you would see in the medical articles."
So why worry about the absence of references? First, doctors may need to know who, exactly, was in the study group, when the study was done, or whether it found limitations or cautions. To answer these questions, you need to read the original report.
Second, drug ads sell drugs. A survey sponsored by medical publishers found that these ads had the highest "return on investment" (payback) of four major drug industry marketing tactics.
Third, footnotes are an essential part of the scientific self-regulation. If I say the sun comes up in the East, you can check it out by looking at my references.
Is silence golden?
The Why Files sought comment from the American Medical Association, publisher of JAMA, one of the journals studied by Cooper and Schriger, but Jann Ingmire, manager media relations for JAMA and other AMA journals, emailed to tell us, "The JAMA editors are not available for comment on the study in the Canadian Medical Journal. The data is really old - 1999 and Dr. DeAngelis our current editor-in-chief was not here then."
The Principles Governing Advertising in Publications of the AMA do not mention references, although they do include these requirements:
Although it's hard to know exactly how much drug advertising appears in the thousands of medical journals, industry critic Marcia Angell estimates, based on industry numbers, that the drug biz spends nearly $54 billion per year on all types of marketing.
And page after page of drug ads appears medical journals, where the research articles often contain 50 or more references. Should ads and articles be held to the same standard when they appear in the same context? Perhaps, says Cooper. "The ideal is that if you make a claim, you should provide evidence for the claim. If there's a reference, somebody should be able to look at the data, and decide if the claim is appropriate, if it refers to my patient population, and make an informed decision.
"When data is collected, it should be available. If it's information
that affects the safe prescribing of medicine, all the information should
be available to everyone."
-- David Tenenbaum
- The Availability of References and the Sponsorship of Original Research Cited in Pharmaceutical Advertisements, R.J. Cooper, D.L. Schriger Canadian Medical Journal, Feb. 15, 2005.
- Drug ad return on investment The Truth About the Drug Companies, Marcia Angell, Random House, 2004.