More and more doctors are being urged to base their patient decisions on "evidence-based medicine," in which a doctor allows the results of clinical studies to guide his treatment choices.
But obviously, if doctors are to base their treatment decisions on the results of clinical trials, then they must have confidence that those trials were carried out with scrupulous care by researchers who did not have a financial stake in the trial's results.
Because any financial reward to the authors may bias the validity of their reports (either knowingly or unknowingly), most medical journals now require that all authors who submit manuscripts declare any financial relationships that may have a bearing on their research.
And yet, a spate of medical and media publications has recently highlighted possible conflicts of interest in a number of clinical trials. These revelations undermine our trust in the results of studies, even some of those reported in the leading journals.
During a two-month period in 2006, for example, authors of 3 different articles in JAMA on the treatment, respectively, of arthritis, cancer, and migraine, failed to acknowledge funding and consulting relationships with firms selling drugs to treat those very conditions.
An even more egregious cluster of incomplete and unsatisfactory financial disclosures was described in a recent article in The New York Times, about a 2006 New England Journal of Medicine paper concerning the use of computed tomography (CT) to screen for early-stage lung cancer. Author Claudia I. Henschke, M.D., Ph.D., of Weill Cornell Medical College in New York, failed to disclose to the journal that the tobacco industry had provided much of the funding for her study.
According to the Times article, $3.6 million from the tobacco companies was funneled through a charity called the "Foundation for Lung Cancer: Early Detection, Prevention & Treatment." And it seems that Henschke, the principle investigator of the research study, was also the president of this foundation. What's more, one of the study's co-investigators served as the foundation's secretary-treasurer, while both the dean of the medical college and the vice-chairman of its board of overseers were listed as directors.
Dr. Henschke said that the results of the study, which involved people at high risk for lung cancer, suggested that annual low-dose CT scans could prevent 80 percent of lung cancer deaths in such individuals.
Such a startling conclusion might encourage physicians to use CT screening in their patients, without those physicians knowing that this strategy would provide financial benefits to Dr. Henschke and to a Cornell colleague who was a co-author of the article. Dr. Henschke and this colleague, it turns out, held royalty-paying patents on CT-based, tumor-detection technology.
Dr. Henschke says that she disclosed these patent holdings to the journal, which inexplicably failed to reveal them at the time the article was published. It is evident that both the authors of the article and the editors of the journal didn't properly inform their readers.
The real problem here is the conflicts of interest themselves, rather than the improper disclosure of conflicts. Investigations into how monetary payments affect the way scientists interpret their results have shown that a study sponsored by a company is more likely to present that company's product in a favorable light.
