That may be easier said than done when the institution depends so heavily on those drug-company grants. Plus, while the researchers often disclose their own conflicts of interest, sponsoring academic institutions usually don’t reveal how many millions of dollars of its own endowment are invested in companies such as Aventis and Merck. Harvard is one of those schools that keeps its cards close to its chest. No wonder. In fact, a recent Harvard SEC report shows holdings of $16 million worth of Merck stock, $8 million of Bristol Myers Squibb, $34 million of Johnson & Johnson, and $33 million of Pfizer. (Aventis and other foreign-owned stocks would not appear on the report.) This is a small slice of the $26 billion endowment, perhaps, but nothing to sneeze at. “Institutional conflict of interest should matter,” Stafford says.
Ghost in the machine
Further obscuring the true participants in medical research, and their potential conflicts of interest, is a tremendous industry that thrives outside the halls of academia and contracts almost exclusively with drug and medical-device makers, and has been built up to create the content that fills hundreds of medical journals. Thanks to the wealth of medical expertise in the Boston area, much of this industry exists right here.
Among them are medical “ghostwriters,” paid by the pharmaceutical or biomed company, whose work ranges from simply touching up an article to writing the entire thing. No protocols are in place to standardize those arrangements: sometimes the academic researcher knows about the ghostwriter and has final approval, sometimes not.
A search of PubMed, the online resource of published medical research, shows no mention of Andrea Gwosdow since 1997. Not coincidentally, this is when she began her freelance medical-writing business. In fact, Gwosdow has written — or rather, ghostwritten — many medical-journal articles in the past decade from her home-based office in Arlington, Massachusetts, under contract to area companies such as Boston Scientific, Genzyme, Interleukin Genetics, and Anika Therapeutics. But her name does not appear on the articles.
Gwosdow is one of roughly 300 members of the American Medical Writers Association here in the Boston area. Some, like her, are freelancers. Others work as subcontractors through companies such as Rete Biomedical Communications, Cayuga Consulting, MedBio Publications, and Life Science Publishing.
No standard exists for the relationship among these writers, the company, and the academic researcher. The ghostwriter usually works for the company’s science department, but not always. “For Genzyme, I have in the past been hired directly for the marketing group,” Gwosdow says.
Other ghostwriters are employees of the company funding the study — people like Janice Schaap, who’s not listed as a co-author but who “assisted in the writing” of a January 2004 article in Cancer, according to an acknowledgment. That article, whose lead author was Paul J. Hesketh of St. Elizabeth’s Medical Center, in Brighton, reported that darbepoetin alfa effectively corrects anemia in patients with malignant disease. The article does not mention that Schaap was an employee of Amgen, whose sales of its darbepoetin-alfa drug Aranesp grew 32 percent — to $3 billion — the year after the article appeared.
