Report Faults Scientific Journals on Financial Disclosure
CSPI Says Authors Fail to Disclose Financial Conflicts of Interest; Journals Fail to Enforce Disclosure Policies
July 12, 2004
Several leading medical and science journals fail to enforce their own policies for disclosing financial conflicts of interest among contributing authors, according to a study released today by the nonprofit Center for Science in the Public Interest (CSPI). The study examined 163 articles in the New England Journal of Medicine (NEJM), the Journal of the American Medical Association (JAMA), Environmental Health Perspectives (EHP), and Toxicology and Applied Pharmacology (TAP). It identified at least 13 articles where authors did not disclose relevant conflicts of interest that should have been disclosed according to the journals’ policies. CSPI found another 11 articles where there were undisclosed conflicts of interest that might not have directly related to the subject at hand, but should have been disclosed nevertheless.
Some of the unpublished conflicts of interest include:
- a University of Arkansas College of Medicine professor, Dr. John Shaughnessy, published a NEJM article outlining the potential efficacy of a treatment for multiple myeloma, but did not disclose that he intended to apply for a patent on the underlying technology. He also failed to disclose that he is a paid consultant for drug companies developing vaccines for the condition.
- a Procter & Gamble scientist, William Owens, was only identified in EHP as an official of the Organization for Economic Co-operation and Development in an article that validated a toxicity test that would likely be used on various P&G products. There was no disclosure of Owens’ employment with Procter & Gamble in this article, even though it was known to EHP editors.
- two scientists at the U.S. Armed Forces Institute of Pathology, Frank D. Kolodgie and Renu Virmani, published an article in NEJM about the formation of plaque in coronary arteries, but did not disclose their consulting relationships with over 20 companies in the heart disease treatment field, including Medtronic, Guidant, Boston Scientific, and Novartis.
- a National Institutes of Health senior scientist published a study in JAMA on predictors of kidney disease, but did not disclose his consulting relationships with Merck, Bristol-Myers Squibb, GlaxoSmithKline, and Pfizer, all of which sell products whose marketing could benefit from the insights gleaned from that study.
“Published research that fails to disclose authors’ ties to drug companies threatens the credibility of scientific journals and rightly undermines public confidence in studies about the safety or efficacy of various drugs or chemicals,” said Merrill Goozner, director of the Integrity in Science Project at CSPI and the author of the study.
Nondisclosure of financial conflicts of interest was a problem at all four journals, but JAMA had the highest rate of nondisclosure of conflicts at 11.3 percent (six out of 53 articles). The undisclosed conflicts in JAMA ranged from consulting fees from companies immediately involved in the subject of the study to authors holding patents on technologies that may one day prove valuable because of information contained in the study. EHP had a nondisclosure rate of 8.6 percent (three of 35 articles), TAP had a nondisclosure rate of 6.1 percent (two of 33 articles), and NEJM had the lowest rate of nondisclosure at 4.8 percent (two of 42 studies examined). CSPI typically researched only the first and last of the authors cited for each article, and only when no disclosure statement was published for either author, so there are likely to be undisclosed conflicts among the other authors not researched.
CSPI recommends that journal editors require authors to disclose any financial arrangements they have had with private firms within the past three years, regardless of whether those arrangements relate to the subject of the article, and that the conflicts be published if they are in any way related to the article’s subject. CSPI also says that authors should be required to disclose any patent applications, or intentions to apply for any patents. To encourage authors to comply with journals’ policies, CSPI also recommends that editors adopt strong sanctions for failing to disclose conflicts of interest, such as a three-year ban on publication imposed on authors who fail to make complete disclosures.
“Some of the blame for the failure to disclose these conflicts rests with the individual scientists, who clearly feel comfortable withholding fairly glaring conflicts,” Goozner said. “But much of the blame must rest with the journal editors themselves, who, for the most part, have created disclosure policies that too narrowly define what conflicts are relevant.”