The publishers of the Nature family of journals are considering extending their policy on disclosure of interests to all the types of articles they publish, after the author of a review article published in Nature Neuroscience was criticized for failing to reveal a potential conflict of interest.

Robert T. Rubin, director of the Center for Neurosciences Research, Allegheny General Hospital, and Bernard J. Carroll, director of the Pacific Behavioral Research Foundation, criticized the review of treatments for mood disorders—published in Nature Neuroscience in October 2002—for not disclosing significant financial interests of one of the authors in three therapies that were reviewed favorably.

Rubin and Carroll wrote to the Nature publishing group in February 2003, asking that it publish a letter in which they described the review author's conflicts. Charles G. Jennings, executive editor of Nature research journals, replied to them with a letter explaining that the company had no conflict of interest policy for review articles.

Rubin told The Scientist, "We wanted them to publish our letter and their response—and let the scientific community decide. They stalled and still haven't given us a response." Last week, Rubin and Carroll took the story to the New York Times.

Rubin said, "There should be complete disclosure on all articles, including review articles, in a similar way to that required by most other medical journals—to make it absolutely clear if there are any real or apparent conflicts of interests."

The author of the review, Charles B. Nemeroff, chairman of the department of psychiatry and behavioral sciences at the Emory School of Medicine, said that he would have reported the conflicts of interest, which included owning the patent on one of the treatments mentioned, if the journal had asked him to. In an interview with the New York Times, he commented, "I have always been totally compliant, probably gone overboard, with disclosure. If there is a fault here, it is with the journal's policy."

Jennings told The Scientist in an e-mail, "We are currently reviewing our policy, in particular the question of whether to extend the disclosure requirement to reviews. We have not taken a final decision on this." He continued, "Nature journals have taken a lead among basic science journals by requiring every author to make a statement on the record—this is common among clinical journals but I don't know of many other basic science journals that do this."

The case illustrates the ongoing need for journals to deal appropriately with disclosure of potential conflicts of interests. At the same time, experience in this area indicates that policies should allow for the associations that a growing number of researchers have with companies. The New England Journal of Medicine (NEJM) had to relax its rules in June 2002, to allow authors of reviews and editorials to receive up to $10,000 per year from interested companies, because it was unable to find writers willing to contribute to the journal when the amount was lower. The editor-in-chief, Jeffrey Drazen, found that everyone he approached to write for the journal had financial ties of some kind to companies that made drugs they were being asked to write about. Over the past year, 10% of editorials and reviews published in the NEJM have carried a disclosure.

"These authors would have been unable to contribute to the journal under the previous policy that allowed no associations at all," Drazen told The Scientist. "Under the new arrangements, we have had to exclude only about 10 to 15% of people who we were considering writing for us because of conflict of interest. Even though they could have written brilliant articles, the appearance of conflict of interest would have lessened the impact of the voice of the journal."

Writing in the NEJM, Drazen stated, "Not all financial associations are the same. Some, such as the receipt of honorariums for occasional educational lectures sponsored by biomedical companies, may be appropriately viewed as minor and unlikely to influence an author's judgment. Others, such as ownership of substantial equity in a company, are of greater concern. It is our intent to focus on the financial relationships that, in our judgment, could produce bias, or the perception of bias, in an article." He added that the journal's definition of significant financial interest was the same as that used by the National Institutes of Health and the Association of American Medical Colleges.

Richard Smith, editor of the British Medical Journal (BMJ), argues that rather than trying to introduce "a form of McCarthyism," journals should be aiming for transparency, providing sufficient information for readers to assess research as it is reported, taking into account the possible impact of outside influences. He considers that one reason authors may be wary of reporting conflicts of interests is they think that it implies potential wrongdoing. As a result, the BMJ has replaced the term "conflict of interest" with "competing interests."

"This will, we hope, reduce the sense of wrongdoing and encourage people to disclose competing interests," Smith explained to The Scientist.

The BMJ has also restricted its declaration requirement to financial interests, considering that this will make it more likely that authors will declare competing interests. The journal now requires disclosure of competing financial interests by authors of all original papers, editorials, and reviews. Reviewers and members of the editorial board are also asked about competing interests.

Jennings agreed that there is nothing wrong with researchers receiving funding or fees from commercial organizations. "The problem arises (in my view) when those interests are not disclosed. There is evidence from the clinical literature that industry funding is associated with pro-industry conclusions, but I'm not aware of any corresponding studies of the basic science literature, and unless disclosure is common practice, it's very difficult to evaluate the extent to which it might be a problem."

Transparency is the key, argued Smith: "It is difficult for researchers to assess if their research has been influenced in any way by potential conflicts of interest. A declaration of competing interests allows readers to judge for themselves."

Jennings suggested that it was also important to maintain public trust in research. "If the public comes to believe that basic science is systematically tainted by researchers' financial interests, that is not good for science, regardless of whether the public perception is accurate or not. If journals can do something to counteract this perception and to increase public trust in the research we publish, that's good for everyone."