When physician Adriane Fugh-Berman of Georgetown University School of Medicine in Washington, D.C., was asked to write a review article on interactions between herbs and warfarin, she said maybe. A clinician and expert on herb-drug interactions, Fugh-Berman thought the information could be useful to clinicians who prescribe warfarin as an anticoagulant.
A few months later, a finished manuscript arrived on her desk. All she needed to do was read and approve it. She was intrigued and more than a little suspicious. She did a little research and learned that the company was working on a related drug that had not yet come up for approval by the U.S. Food and Drug Administration. By planting this paper in the medical literature, the company was preparing for the approval process by highlighting in the medical literature the deficiencies of the best-selling generic that the new drug would compete with.
'Some of it is just naïveté on the physicians' part,' she says. 'But there's also the fact that this is so common that it's not considered unusual. There's no shame attached to it.'
Fugh-Berman withdrew, but some time later it arrived on her desk again, this time for review. She notified the journal's editors and urged them to not publish it. Instead of the ghostwritten article, Fugh-Berman published a different article, "The Corporate Coauthor," which appeared in the same journal -- the Journal of General Internal Medicine (JGIM) -- in June 2005, alongside comments from the journal's co-editors and a statement from the World Association of Medical Editors (WAME).
This wasn't the first time a manuscript had been rejected because it wasn't written by the putative author. But the event helped bring to light the role that ghostwriters, and the medical-education companies (MECs) that employ them, have played -- and continue to play -- in shaping the medical-science literature. The event initiated a dialogue on the topic in the community of peer-reviewed journal editors and led to guidelines on dealing with ghostwritten articles that are now embraced by many -- though not all -- scientific journals.
Haunting the literature
Ghost authorship occurs when an unacknowledged author writes, or makes substantial contributions to, an article published in the peer-reviewed science literature. Typically, the author is a professional writer with scientific expertise, hired by medical-education and communications companies. These articles are destined to appear under the names of scientists who contribute little to their writing. They are backed by for-profit companies -- often pharmaceutical companies -- which, like most scientific authors but for different reasons, hope to influence scientific opinion. Such activities skew the medical literature by injecting a marketing spin that benefits the sponsor, Fugh-Berman says.
"People tend to think of marketing messages as 'buy drug A,' but that's never the message imbedded in such articles," Fugh-Berman says. "The message may be, disease A is underdiagnosed or far more serious than previously believed, and currently used drugs for this disease are highly problematic." Most academics know that it's unethical to sign their names to work they didn't do, but it's just as important not to hide contributions from others, says Fugh-Berman, who currently writes about how drug companies inject marketing messages into the medical literature on her Web site, Pharmed Out.
The cynical injection of marketing messages into the scientific literature sounds outrageous, but abuses common in academic publishing set the stage. When a lab chief insists on being an author for every article emerging from his lab, or when a prominent scientist is invited to be an "honorary author" on a paper to boost prestige, lines that should be bright become fuzzy, Fugh-Berman says. "It's such a mixed message because we throw students out of school for plagiarizing. And then we, as scientists, plagiarize their work, and then that's supposed to be acceptable."
It's common for researchers -- especially those who are thought leaders in a particular field -- to receive invitations to serve as "authors" for industry-sponsored, journal-bound, ghostwritten manuscripts, she says. "Often, the invitation is presented under the guise of 'editorial assistance' offered by another scientist with an advanced degree, so people can get fooled into thinking it's a collaborative effort," she says.
The selected author often has some connection to the sponsoring company, but not always. "In most cases, [the signing author is] not being paid directly for the article, but they might have other financial arrangements with the company," Fugh-Berman says. A physician may be paid as a consultant researcher or speaker, for example. "Some of it is just naïveté on the physicians' part," she says. "But there's also the fact that this is so common that it's not considered unusual. There's no shame attached to it."
In a recent case documented in court records, the pharmaceutical company Wyeth paid ghostwriters to write dozens of papers on hormone replacement therapy. The company had researchers sign the papers to make it look like they were written independently. Published in medical journals between 1998 and 2005, the articles detailed the benefits of the therapy while minimizing its risks. Fugh-Berman is serving as an expert witness for the plaintiffs in a lawsuit brought by patients against the company.
Fugh-Berman says that, according to court documents, most of Wyeth's ghostwriters worked at MECs. MECs -- or more precisely, some of them -- often target pre-tenure faculty members who are eager to publish, says Martha Gerrity, a clinician-educator and health-services researcher at Oregon Health & Science University in Portland.
Gerrity, who served as co-editor of the Journal of General Internal Medicine from 2004 to 2009, was sorting through manuscripts when that herb-warfarin review was submitted. Noting that the article would need some revisions, she passed it on to a deputy editor, who sent it to Fugh-Berman for review.
"The thing that bothered me is that the author who submitted the manuscript to the journal was an assistant professor at a fairly high-powered academic university," she says. "This suggests that these medical-education companies understand that junior faculty members at academic institutions need to demonstrate academic productivity, and one of the ways of doing it is to have first-author publications in high-quality journals."
"It's very well thought out, the use of the medical literature as a marketing tool," Gerrity says. "But it's very subtle."
Many of the ghostwritten manuscripts are review articles allowing writers to selectively point to studies that highlight the shortcomings of a specific product, push specific diseases, or promote off-label uses of a drug. Fugh-Berman says review articles are also popular because busy physicians rely on experts to synthesize the medical literature and provide a clinically relevant summary.
No one knows exactly how common the practice is; it's a hard thing to study. In a 2006 study published in the Journal of the American Medical Association, the authors reviewed 1000 research articles from 10 high-ranking, international peer-review journals and found that up to 11% of articles had authors who weren't named.
And those surveys reveal the percentage of only the people who admit to working with an unacknowledged writer, Fugh-Berman says; the practice of ghostwriting likely stretches far beyond these high-profile cases. "The whole point of ghostwriting is that the ghost is invisible. So we have no idea how many articles are ghostwritten."
An interesting conflict?
When the ghostwritten article surfaced at JGIM, the journal's guidelines didn't specify what steps should be taken. Gerrity informed the author's institution but didn't do anything more. "It was clear in my e-mail interaction with the author that this person was frightened and didn't realize that they had gotten themselves into such hot water that it could have potential implications for their position," she says. "We decided not to push it at that point until there was very clear guidance from a group like WAME," an international group of editors of peer-reviewed medical journals.
The section on ghostwriting in WAME's code of ethics was "fairly thin and not very explicit," she says. "As I read through it, it was unclear that it would cover this type of authorship. There wasn't anything that specifically addressed the relationship between pharmaceutical industry's medical-education companies and ghost authorship."
The group has since rewritten its guidelines. "Examples of financial ties to industry include payment for research, ownership of stock and stock options, as well as honoraria for advice or public speaking, consultation, service on advisory boards or medical education companies, and receipt of patents or patents pending," the new guidelines say.
After that, Gerrity and her co-editor, William Tierney of the Indiana University School of Medicine, began scrutinizing manuscripts that came across their desks. They sent notices to authors informing them that anyone who had contributed to the writing of a manuscript must be listed as an author. If they learned that an MEC was involved -- often this information came from an Internet search of the listed affiliates -- they became especially diligent in tracing the funding, to see if it could be tied to a particular product.
Most studies tied to an MEC were rejected outright, she says. She and Tierney inserted a stock paragraph into the rejection letters discussing the relationship between scientists and MECs. "Over the course of the next year, the numbers of these manuscripts dropped off precipitously," Gerrity says.
Some academic medical centers, such as the one at Yale University, are cracking down on ghostwriting by barring faculty members from being listed as authors unless they make a substantive contribution. And top-tier medical journals, such as The Journal of the American Medical Association, PLoS Medicine, the British Medical Journal, and the Journal of General Internal Medicine, have developed stringent authorship guidelines to ascertain the contributions of each author and obtain full disclosure of authors' funding sources and financial ties.
Fugh-Berman highlights the conflict-of-interest requirements of the journal American Family Physician. Instead of just asking if there is a conflict of interest, the new guidelines include a checklist of questions with scenarios spelled out. Faced with such specific language, authors would have to "actively lie" to cover up the involvement of an unacknowledged author, Fugh-Berman says. "You can lie, but you can't say that you didn't understand the question. And I think that that's important. ... You don't want people deciding for themselves what are relevant conflicts of interest because nobody thinks their conflicts of interest are relevant."
But not all medical journals have taken such measures, and many still publish ghostwritten works, Gerrity says. Last year, it came to light that Merck had paid the publishing company Elsevier to produce a journal, Australasian Journal of Bone & Joint Medicine, which looked like a peer-reviewed medical journal but was filled with articles and reviews from MECs, including articles favorable to Merck products for osteoporosis.
Creating enforceable, specific publication guidelines prohibiting ghostwriting can be tricky, Fugh-Berman says, because researchers may not recognize conflicts of interest obscured by the intermediation of MECs. And the practice of not paying authors helps to maintain a "pseudo-academic air" about the enterprise, Fugh-Berman says.
The key, says Gerrity, is for researchers and institutions to be specific. "Saying that people can't participate in ghostwriting is not useful if people don't know exactly what ghostwriting is," she says. "You have to spell things out so that it mentions the euphemism that industry uses, and that is editorial assistance."