Scientific fraud comes in several varieties. Data can be fabricated, ideas swiped, experiments gamed. One dramatic version is the too-late revelation that an investigator was drinking at the watering hole of a pharmaceutical company when he or she published an article about the wonders of a certain drug, produced by the watering hole's operators.
To prevent such embarrassments, medical journals now run authors through the wringer to potentially illuminate their often-cozy relationships with pharmaceutical companies. We writers complete an extensive conflict-of-interest form describing any payments received for lectures or advice, research grants, and personal or family financial interests. This information is then included at the end of the published article, allowing the reader to judge whether an author might have been working under the influence.
Although onerous to coordinate (and annoying to complete), this approach is an important first step in preventing unseemly trysts between doctors and the pharmaceutical industry. But it does not go far enough because it ignores an even more conspicuous denizen of the watering hole, the medical journal itself.
Just as pharmaceuticals fund studies and pay doctors to give lectures, so too do they buy journal ads and reprints of favorable articles—lots of them. Often a drug company may find one of its products featured in a scientific article while another of its products is dolled up in a high-gloss ad a few pages later. Yet the journals keep quiet about these financial arrangements. When an article is published that shows a specific drug at great advantage, the reader may learn plenty about the author while nothing—absolutely nothing—is disclosed about the medical journal itself.
The public deserves to know about the extent to which every medical journal relies on pharmaceutical advertising revenue to run its business. In 2003, drug companies spent almost half a billion dollars on advertising in medical journals. The two lead general medical journals in the United States—the New England Journal of Medicine and JAMA, the Journal of the American Medical Association—receive about $18 million and $27 million each year, respectively, for display advertisements (as opposed to classified ads placed by individuals seeking jobs or institutions seeking qualified candidates). The display ads represent 10 percent to 21 percent of the journals' overall revenue, according to one study.
And display advertising is only one of the ways that dollars can move from a drug company to a medical journal. Many journals publish "supplements," typically a one-topic, free-standing volume that shares font and layout with the parent magazine—but not serious peer review. In other words, it looks and sounds like your favorite medical journal, but it's not quite the real thing. These supplements have sponsors (guess who?) that front the production and publication costs sometimes with a Producers-like overpayment.
Another trick is the sale of reprints of a specific article. Say a group of researchers has published a favorable article about a drug. Well, the pharmaceutical company can show its love by snapping up thousands and thousands of reprints to sprinkle around in the good name of education—and in the process, pay the journal tens of thousands of dollars. As Richard Smith, former editor of the British medical journal BMJ and a leader in trying to protect editorial integrity, once wrote about the effect of reprints: An editor in chief must "publish a trial that will bring US $100,000 of profit [in reprints] or meet the end-of-year budget by firing an editor."
The consequence of the pharma-journal relationship is far from abstract. A few years ago, Dialysis & Transplantation, a leading journal for kidney specialists, received an editorial that argued against a double dose of a pharmaceutical product, Epogen, which can ameliorate the anemia common in dialysis. The author claimed the big dose was twice as expensive (about $10,000 to $12,000 per patient annually, versus the typical $5,000 to $6,000) but no better than the regular dose—and perhaps less safe.
The piece was accepted through the standard peer-review process. And then the reviewing scientists were overruled by the marketing department. As the journal's editor explained when the situation came to light, the author's opinion that a lower dose was preferable "apparently went beyond what our marketing department is willing to accommodate." Though the journal eventually restored its original decision and decided to publish the article, the damage to its reputation had been done (and the author declined the invitation). Earlier this year, the FDA determined that the higher dose indeed was less safe than the standard dose because of an increase in the rate of stroke, blood clots, and heart attack.
How often do questionable episodes like this one take place? We don't know. But the examples that are known about, and the possibility that medical observations are influenced by anything other than cold hard facts, are sufficiently sobering to warrant additional disclosure. And so I have a modest suggestion: In addition to requiring authors to post conflict-of-interest statements when they publish an article, medical journals should tell readers how much revenue they themselves have received in the previous year from the company producing the drug or device under discussion. The total sum should include not just advertising pages purchased, but also the other ways that industry money can slip into journal pockets, by buying reprints and journal supplements. Show us the actual dollar (or euro or pounds sterling) amount. And if a professional society sponsors the journal, tell us about its financial dealings with the drug companies as well.
Such disclosures would take work, annoy scads of people (most of them honorable), and be completed under protest. But they're worth it, to help assure the integrity of medical literature. Just as compromised relationships are unusual among researchers, they are likely, in the end, to be unusual among medical journals. But it is naive to think that only authors are influenced by who is writing the checks.