There are a lot of reasons to be disturbed by the revelation, covered by the New York Times on Wednesday, that at least 26 published review articles in medical journals were ghostwritten by a medical communications company. But I’m not sure all of those reasons are obvious at first glance or fully addressed by the article. Sure, it’s ethically questionable for doctors to affix their names to a review article (which typically summarizes dozens or even hundreds of separate research articles into a cohesive statement about a medical or scientific topic) that they didn’t write. But it also raises serious questions about other, less overt ethical violations that beset the field.
The story has its origins in the controversy over hormone replacement therapy (HRT), which was found in 2002 to elevate risks of heart disease, stroke and some cancers. Wyeth Pharmaceuticals, as one of the leading sellers of hormone replacement therapy, has become one of the biggest targets for litigation: the Times piece counts roughly 8,400 lawsuits filed against the company by women claiming illness caused by HRT. As a result of those court cases, documents were released last week revealing that Wyeth paid DesignWrite, a medical communications company, about $25,000 per article for 26 articles advocating the benefits of HRT and published in prominent scientific journals. Though written by DesignWrite employees, who may or may not have medical or scientific training, the authors eventually listed for the articles were MDs.
Reprehensible? Sure. Unique? Hardly. Other companies, marketing other drugs, have also been forced to disclose similar arrangements in court. But the Times article and other online commentaries haven’t mentioned that this particular case is merely an especially egregious example of questionable practices that threaten scientific integrity.
1) Corporate entanglements are all too common. If a drug company commissions a review article that promotes their product and then finds a doctor to co-sign the article, that’s a blatant conflict of interest. But what about other, more subtle ways that doctor/corporation conflicts of interest can influence medical literature? Many doctors find themselves on medical advisory boards for pharmaceutical companies, obtain a million-dollar grant or two from those companies to do clinical or laboratory research or just take them up on the occasional offer for a free dinner or a night of open bar. Reputable journals rightly demand that the authors of journal articles reveal these potential sources of bias, and additional scrutiny by journal editors and readers should always follow such published admissions.
2) Review articles, like it or not, always contain an element of opinion. As described above, the articles paid for by Wyeth were all review articles that summarize available research. Only the purest-of-heart would be able to prevent bias from slipping into this process, and because scientific debates are typically closer to “your model vs. my model” than a clear-cut “right vs. wrong,” it’s natural for an author to focus on the research he or she agrees with, to the exclusion of papers that don’t match up with their own research or clinical experience.
Hormone replacement therapy is actually a great example for this practice, contentious as the field has become on the topic. In 2003, I attended the world’s largest conference of endocrinologists and was in the audience of a panel discussing the then-recent findings of HRT’s dangerous side effects. It was one of the most contentious scientific panels I’ve ever attended, as doctors argued loudly over whether the elevated risk for heart disease and stroke outweighed the very real benefit and relief from symptoms that it provided for menopausal women. So the perspective that Wyeth was pushing in their commissioned articles, while unethical in their creation, was not without its supporters in the medical community. The review articles, like most of their kind, had an element of bias, which I think is okay as long as conflicts are disclosed and the underlying science is correct. Which brings us to…
3) Where were the peer reviewers? Any reputable journal will send submissions to at least 2, often 3 reviewers with expert knowledge in the particular field that the article discusses. It is the job of these reviewers to ensure that the scientific facts of the article are sound, and that it makes no errors of omission, consciously or unconsciously. I have yet to see any accusations that the information in the Wyeth articles was outright false, and if valuable research was being omitted from the review article, then it should have been obvious to any qualified peer reviewer. That 26 articles made it past science’s traditional defenses into print means either that these commissioned articles did play by the rules, or that the peer review system didn’t do its job. If the latter is true, that’s far more disturbing than the practice of ghostwriting itself.
So what is to be done? Drug companies swearing off the practice of ghostwriting is a shallow gesture, for the reasons I outlined above about more subtle forms of influence. Journals bear a large responsibility to sniff out unsavory arrangements. The best journals now require an article’s authors to specify exactly what they did: experiments, or writing, or literature searches or an advisory role. More journals should take up that policy, and police it. Peer reviewers also need to step it up and shoot down review articles that fall outside the bounds of acceptable opinion or bias. Perhaps journals need a new kind of review article, one that is open about its bias/argument and is labeled differently – not quite an advertorial but more like a newspaper’s Op-Ed page.
And finally, readers of journal articles – be they doctors, scientists or laypeople – need to remember that the articles published in a journal are not infallible. Real people write journal articles, people who sometimes have ulterior motives or maybe just unconscious bias. Healthy skepticism is the best antidote to the poison of conflicts of interest.