Emergency Exorcism Part 2: Five ways to bust pharma ghostwriting in medical, science journals
Antidote proposed on Friday that medical and science journal editors do more than talk tough about conflicts of interest in their journals. Investigative journalists uncovered a series of scandals in recent years that have revealed the huge role that drug and device companies play in manipulating journal publications and continuing medical education programs through ghostwriting, speaking fees and other gun-for-hire arrangements.
Journal editors, though, have been slow to respond. No journal has stepped up to say, "We have investigated the extent of industry influence on the research in our journal and here is what we found." Absent that, patients worldwide are left to wonder whether the drug they are being prescribed, the device being implanted in their body or the advice they are being given by their physician is based on strong science.
Here are some concrete steps journal editors, and reporters, could take to restore confidence in scientific publications.
1. Read the documents. Journals could have one intern spend one week searching the Drug Industry Document Archive for all references to that journal. JAMA, for example, turns up 415 hits. Many of these are literature citations that have nothing to do with ghostwriting. But there is also evidence in here that needs more investigation. A study that showed how estrogen might help fight Alzheimer's disease has ghostwriting fingerprints on it. And much could be made of these minutes from a meeting of staff from Wyeth and DesignWrite, Wyeth's ghostwriting company, in which they discussed how they were going to try to shape the discussion of the study through an editorial in JAMA.
2. Check the outbox. Journals should go back into their archives and see how many of the articles by the same authors favored drugs or devices. They should look at submissions and rejections, too, and see who has been trying to sneak tainted research into their pages. For example, here's something that might make the JAMA editors smile. Dr. Rhoda Sperling wrote this memo to Merck executives in 2001 about a Merck-crafted paper being submitted as a piece of independent research.
The paper was submitted to JAMA-Express and rejected, Dr. Konstam spoke to the editor who expressed a great deal of interest in having us submit to JAMA through the usual route. I did not consider JAMA a viable alternative given (in the best scenario) a 6-8 month delay to publication and an anticipated rigorous peer review (given the recent CLASS publicity, Merck's experience with the VACT manuscript and the quote from the JAMA editor-in-chief in Wednesday's Wall Street Journal). If I understood Barry's MVX, he thinks we should not discount the JAMA option.
Where did this paper end up? And how many studies did Sperling have published that should be reexamined? Frederic Curtiss, editor-in-chief of the Journal of Managed Care Pharmacy, told Reuters in 2009 that he had begun combing through the histories and metadata in Microsoft Word documents to look for the involvement of outside writers or pharma executives. It's a simple tool that more editors could use. Reporters should be doing this type of digging, too and following up on whether the researchers who worked with pharma-sponsored ghostwriters have continued to produce scientific results that make drugs and devices look good.
3. Ask for help. The International Committee of Medical Journal Editors should set up an independent review board comprised of researchers, editors without pharmaceutical industry or device industry ties and members of the public. After all, the research that is being published in these journals determines the treatment options offered to patients. If they need suggestions for how to staff such a panel, ask Dr. Adriane Fugh-Berman, Dr. Daniel J. Carlat or Dr. Charles Rosen. Give the review board the evidence and let them decide what should be done. Should certain articles be retracted? Should the researchers be banned from publishing in journals? Should a letter be published in the journal listing their names?
4. Call their bosses. Reporters should follow in the footsteps of the Milwaukee Journal-Sentinel's John Fauber and Susanne Rust and start contacting their local universities and asking for conflict of interest statements or "outside activity reports." Fauber found one doctor who was making millions a year from medical device makers. Find out whether there are any rules barring these researchers from working with companies or ghostwriters. If not, why not? Where has the research been published? Ask the journals how these studies made it through the peer review process. Perhaps the peer reviewers have their own ghosts working behind the scenes.
5. Don't hunt witches. There is a difference between trying to find the truth and trying to punish people. There are also gradations of shadiness in what these different researchers and journals have done. No case is quite the same and one remedy won't work for everyone. That's why some of the best reporting on ghostwriting thus far has carefully walked readers through the ethical, legal and medical implications of industry influence on a particular study or whole field of research. Natasha Singer and Duff Wilson provided an Exhibit A for how to do this in The New York Times in December.
As Dr. Catherine DeAngelis, JAMA's executive editor, said at the recent Association of Health Care Journalists conference, there are great scientists working for drug companies and device companies, but she also pointed out that the scientific side of the industry is now, in many cases, being driven by the marketing side. It is up to journal editors and reporters to make sure that patients have the information they need to sort the good science from the bad.