Physician, Reveal Thyself

A report that 72% of journal articles on heart stents in 2006 failed to note who funded the research draws sharp criticism of disclosure practices

by Arlene Weintraub

A group of researchers at Duke University scoured more than 700 studies on heart stents published in medical journals over the course of a year, and were shocked to discover two huge omissions. First, 83% of the papers failed to disclose whether the authors were paid consultants for the companies that made the products they wrote about.

Perhaps more surprising, 72% of articles describing clinical trials and other research-related matters didn't identify who funded the research. The Duke paper, published May 7 in the online journal PLoS One, raises fresh questions about the ever-growing financial ties between doctors and companies, and the adequacy with which those relationships are communicated to the public.

Doctors who study drugs and medical devices often get financial support from the companies that make those products. It's a symbiotic and important partnership, especially as federal funding for medical research dries up (BusinessWeek, 5/8/08). But in the past few years, the medical community has become increasingly concerned that some physicians who author studies about products could be biased by the money they get from industry. That may prompt them to downplay dangerous side effects, critics say, or simply not publish studies that come to negative conclusions about experimental products.

Tougher Policies After Vioxx

In the wake of the controversy surrounding Vioxx—the Merck (MRK) arthritis pill that was pulled from the market in 2004 after it was linked to heart attacks and strokes—publications such as the New England Journal of Medicine and The Journal of the American Medical Assn. (JAMA) toughened policies requiring authors to disclose financial ties to industry. The idea was that such transparency would allow doctors, patients, regulators, and insurance companies to see the studies in light of their backers, and decide for themselves how serious the potential for bias might be.

But the Duke study suggests the quest for transparency has failed. Aside from the revelation that most authors are ignoring journals' disclosure rules, the study suggests the disclosures themselves may be suspect.

The researchers zeroed in on the 170 or so authors, out of a total of 2,985, who fulfilled their duty to disclose in the journal articles, which were published in 2006. By doing informal Internet searches, they discovered that some of the physicians who said they had no conflicts of interest in fact served on the advisory boards of companies that make stents. One person co-founded a stent company. (The Duke paper doesn't name individual doctors.) Of the 75 authors who disclosed relationships with companies, only two did so consistently; the rest named their corporate sponsors in some journals but not others.

Lead author Kevin Weinfurt of the Duke Clinical Research Institute says journal editors are as much to blame as authors are for the lack of transparency. "These journals should be doing better" at policing their disclosure rules, he says. "We hope this is a wake-up call."

The Companies Respond

Disclosure guidelines are set by the journals, and they're completely voluntary. Some federal legislators have proposed mandating a national database that would list all doctors' financial ties to companies, but so far those efforts have stalled.

The study also reveals just how deeply corporate interests have infiltrated medical research. The top funding sources named in the disclosures that the Duke researchers found were stentmakers Johnson & Johnson (JNJ), Boston Scientific (BSX), and Medtronic (MDT), as well as drugmakers Bristol-Myers Squibb (BMY) and Sanofi-Aventis (SNY). Those last two companies co-market the blood thinner Plavix, which is widely prescribed to stent patients.