Journal Editor Again Misled

Catherine DeAngelis, the editor of the Journal of the American Medical Association, says that for the third time in as many months researchers have ignored JAMA's policy requiring disclosure of conflicts of interest in published articles. The excuse by the six physicians who studied the link between migraines and cardiovascular disease was that they didn't think their relationships with drug companies were relevant since the study didn't involve a test of any company's drug.

Some people just don't get it. The study authors, led by Dr. Tobias Kurth of Harvard, had relationships with numerous firms that sell drugs for treating migraines. Yet the authors, after being confronted by their failure to disclose those relationships, claimed those financial ties were not "relevant to our describing a biological link between migraine and cardiovascular disease."

DeAngelis snapped back: "Financial interests and relationships with manufacturers of products that are used in the management of migraine or cardiovascular disease certainly are relevant and should be disclosed." The authors' letter contained the disclosure.

This situation graphically points out why conflicts of interest must be taken into account when judging the results of "basic science" studies conducted by physicians or scientists with financial dealings with the drug industry. Those ties can subtly influence the interpretation of results in many ways. After all, if I'm committed to a treatment regimen for treating migraines, wouldn't it be nice to think that it was also helping my patients prevent heart disease?

Many researchers, like Dr. Kurth and colleagues, continue to think that their attitude toward a field of research has no impact on the way they conduct that research. Yet study after study has shown that the financial ties of researchers do in fact correlate with positive outcomes for their sponsors. This can only be explained by almost impossible-to-detect influences like how they framed the questions for study, how they chose patients for their sample, and how they interpreted the data.

For years, critics of the growing influence of private corporations over scientific research have fought for conflict-of-interest disclosure policies as one way to alert readers to these potential biases. Yet most of them now recognize that disclosure doesn't go far enough. They say it is time to reduce the influence industry exerts over research by setting up independent institutes for the conduct of clinical trials and basic health research. Attitudes like those exhibited in this most recent exchange by Dr. Kurth and colleagues graphically demonstrate why such changes are sorely needed.