Threats to science-based medicine: Pharma ghostwriting

There is no doubt that the infiltration of quackademic medicine into medical schools in this country represents a profound threat to science-based medicine. By mixing mysticism, non-science, and pseudoscience along with science-based medicine, medical schools are in essence endorsing quackery and elevating it to the same level as science-based and science-tested modalities. Worse, they’re running the risk of training a generation of medical students accepting of this “integrating” woo with science, who can’t recognize highly implausible treatments or recognize obvious quackery. By letting pseudoscience take root, they’re risking the very foundations of medicine, the science that brought us such enormous progress over the last century. Indeed, over the last two or three years, this problem has been a frequent topic right here on this very blog.

But the dubious therapies, some of which are outright quackery, that make up CAM are, unfortunately, not the only threat to science-based medicine. They may not even be the most dangerous threat. Science-based medicine requires, as its name implies, a strong evidence base rooted in basic science and clinical trials. Without that, science-based medicine is not truly science-based; consequently, anything that threatens the integrity of the scientific literature supporting various therapies represents a threat to science-based medicine every bit as profound as the infiltration of reflexologists at the world-renowned and august M.D. Anderson Cancer Center. For instance, when the marketing division of a pharmaceutical company designs a clinical trial, it’s a threat to science-based medicine. When there’s publication bias in the trials used to support applications for FDA approval of drugs, it’s a threat to science-based medicine. When a drug company pays a publisher to publish a fake journal (six, actually) reprinting articles favorable to its products, it’s a threat to science-based medicine.

And when a drug company ghostwrites articles to be submitted to peer-reviewed journals for publication, it’s a threat to science-based medicine:

Newly unveiled court documents show that ghostwriters paid by a pharmaceutical company played a major role in producing 26 scientific papers backing the use of hormone replacement therapy in women, suggesting that the level of hidden industry influence on medical literature is broader than previously known.

The articles, published in medical journals between 1998 and 2005, emphasized the benefits and de-emphasized the risks of taking hormones to protect against maladies like aging skin, heart disease and dementia. That supposed medical consensus benefited Wyeth, the pharmaceutical company that paid a medical communications firm to draft the papers, as sales of its hormone drugs, called Premarin and Prempro, soared to nearly $2 billion in 2001.

But the seeming consensus fell apart in 2002 when a huge federal study on hormone therapy was stopped after researchers found that menopausal women who took certain hormones had an increased risk of invasive breast cancer, heart disease and stroke. A later study found that hormones increased the risk of dementia in older patients.

Hormone replacement therapy (HRT) is one of the better examples of both the success and failure of science-based medicine. It’s a failure in that HRT became widely used largely on the basis of retrospective and cohort studies that found cardiovascular benefits in addition to its obvious ability to relieve hot flashes and menopausal symptoms, something no other therapy does better. In retrospect, the evidence base supporting HRT did not justify such widespread use. Unfortunately for Wyeth (and patients), in 2002 an analysis of data from the Women’s Health Initiative showed that HRT with estrogen-progestin results in increased risk for stroke, myocardial infarction, and breast cancer among postmenopausal women. The triumph of science-based medicine is that this result showed that previous practice was subjecting women to a higher risk than had been thought and publication of these data resulted in a near-instantaneous change in practice in which routine recommendation of HRT for cardiovascular benefit as well as relief of menopausal symptoms. Again, as I’ve said time and time again, it may be messy and it may take more time than we like (not to mention encounter more resistance than is always desirable), but eventually science-based medicine, like science, is self-correcting. This is in marked contrast to “alternative” medicine, where, no matter how much evidence is accumulated against an “alternative” modality, its practitioners almost never abandon it.

Not that Wyeth apparently didn’t do its damnedest to prevent that correction and to promote its HRT product. It did it mainly by writing review articles surveying the state of the evidence on the topic of HRT and publishing them in journals such as The American Journal of Obstetrics and Gynecology and The International Journal of Cardiology. Most importantly, Wyeth did not disclose its role in writing the articles, as documents recently revealed in a product liability lawsuit against Wyeth by women who claim to have been injured by its HRT products show. Here’s what Wyeth did:

In 1997, for example, DesignWrite, a medical communications company in Princeton, N.J., proposed to Wyeth a two-year plan that would include the preparation of about 30 articles for publication in medical journals.

The development of an article on the treatment of menopausal hot flashes and night sweats illustrates DesignWrite’s methodology.

Sometime in 2003, a DesignWrite employee wrote a 14-page outline of the article; the author was listed as “TBD” — to be decided. In July 2003, DesignWrite sent the outline to Dr. Gloria Bachmann, a professor of obstetrics and gynecology at the Robert Wood Johnson Medical School in New Brunswick, N.J.

Dr. Bachmann responded in an e-mail message to DesignWrite: “Outline is excellent as written.” In September 2003, DesignWrite e-mailed Dr. Bachmann the first draft of the article. She also pronounced that “excellent” and added, “I only had one correction which I highlighted in red.”

The article, a nearly verbatim copy of the DesignWrite draft, appeared in 2005 in The Journal of Reproductive Medicine, with Dr. Bachmann listed as the primary author. It described hormone drugs as the “gold standard” for treating hot flashes and was less enthusiastic about other therapies.

The acknowledgments thanked several medical writers for their “editorial assistance,” not disclosing that those writers worked for DesignWrite, which charged Wyeth $25,000 to generate the article.

Here’s a graphical representation of how DesignWrite did it.

So, in essence, Dr. Bachman appears to have signed on to a previously drafted article, suggested some changes, and then submitted it for publication under her own name. This sort of thing disgusts me. Besides being wholly deceptive and quite dishonest, it provides ammunition to the quacks to dismiss SBM as being hopelessly corrupt. Not surprisingly, the anti-vaccine quack apologists over at Age of Autism have already jumped all over this as being somehow “vindication” and evidence that the studies failing to find a link behind vaccines and autism are similarly tainted. Never mind that the HRT papers were review articles and not primary studies.

Look, I can understand as much as any academic the pressure to publish. It truly is “publish or perish,” and I probably haven’t published as much as I should have by this point in my career, at least in terms of sheer numbers of publication. If a drug company came up to me with a mostly written review article, I can see how it would be highly tempting to say yes. Hopefully my sense of honor would prevent me from saying yes, but, admittedly, also helping me to stand tall would be the knowledge that review articles don’t really count for much when it comes to career advancement or tenure, at least not among those of us who do lab research.

I believe that treatments based on the best available science represent the best medicine. I really do. Anything that contaminates the best available science contaminates the evidentiary basis behind what I believe to be the best medicine. it’s like weakening the pillar of a structure. It may stand with the weakened support, but it takes a lot less stress for the whole edifice to come tumbling down. In this particular case, Wyeth firmly asserts that everything it published was correct and based on good science. Even if that is absolutely, positively, 100% true, it’s also absolutely, positively, 100% beside the point. Failure to disclose that Wyeth (or, more specifically, a company hired by Wyeth to promote its products in the medical literature) had a major hand in writing a review article, if it didn’t write it completely itself, deprives the reader of a critical piece of information necessary to judge the content of that article. Review articles are different from primary research articles in that there is no new data. What they do is to review the existing literature, synthesize it, and come to conclusions. Thus, decisions about what articles to include, which ones to emphasize, and how to synthesize the totality of the literature become very important. The “spin” put on the article is everything, and you can bet that Wyeth tried to spin the data in the most favorable possible manner to its interests. Moreover, if a reader (or a peer reviewer) doesn’t know who was really behind the article, he assumes that it represents the analysis, conclusion, and opinion of the professor whose name is on the article. The practice of ghostwriting review articles is thus, at its heart, completely corrupt.

I don’t know what’s worse, that Wyeth would pull such a stunt or that medical faculty would willingly sign on to this deception. As Dr. Roy Poses pointed out, as hard as it is for those of us in academia to believe, we hold trusted positions. Our opinions matter, both to patients and to community physicians who read what we write for guidance regarding what the latest science- and evidence-based practice is. Whoring one’s good name to a pharmaceutical company for an extra line on one’s CV or for a “consulting fee” destroys that trust and poisons the scientific literature. It deceives physicians and patients into thinking that a therapy is more efficacious and/or less risky than it really is. It would be different if it were disclosed that a drug company employee or hired gun co-authored the paper, because then at least the reader would be aware of the article’s source and could take that into account when analyzing its conclusions. As Adriane Fugh-Berman put it:

While ghostwriters for celebrity autobiographies may do readers a service by rendering a non-writer’s story readable, the ghostwriters who haunt medical journals represent the views of product manufacturers rather than the academic “authors” whose names decorate the articles.

Riffing on the baseball analogy in the NYT article, Dr. Daniel Carlat puts it well:

As with baseball players on steroids, when companies pour marketing money into ghostwriting campaigns, they change the rules of the academic game. The playing field is no longer level; the drug company’s version of the truth gains the upper hand. Sometimes, their truth really is the truth, but sometimes it’s a carefully crafted lie. Sorting it out is difficult even for physicians who specialize in the area being written about. It’s essentially impossible for the average generalist physician, to say nothing of patients who did not have the advantage of attending medical school.

Exactly, which is, of course, the whole reason why pharmaceutical companies do this.

I used to complain about the increasingly detailed reporting about which author did what demanded by journals to which I’ve submitted manuscripts over the last few years. Sometimes the requests for exactly which author did what in a manuscript, from designing and carrying out the experiments to analyzing the data to writing the manuscript seemed to be reaching ridiculous heights. (Or should I say depths?) Maybe it’s not so ridiculous. On the other hand, I’m not sure what journals can do. After all, as I pointed out before, even articles with ostensibly very strict disclosure rules for authors often don’t catch omissions when authors fail to disclose pharmaceutical company ties. Journal staffs aren’t large enough to investigate the veracity of every author when it comes to reporting conflicts of interest. What is really needed is to attack the problem at the most vulnerable point, namely the faculty who lend their names to ghostwritten papers. Universities need to make it very clear that such activity will have a major negative impact on the careers of its faculty and that ghostwritten articles absolutely, positively will not count towards tenure, promotion, or pay raises. In fact, they should impact negatively on these career milestones.

Actions must have consequences.