Two in five biomedical journals have no declared policy on how to separate editorial and commercial matters, according to the preliminary results of a survey of editors.

Some journal editors say they find the results disturbing, given the commercial pressure that journals have to deal with. Many say they have experienced cases where companies have sought, to varying degrees, to influence journal content.

The survey asked the 350 members of the Committee on Publication Ethics (COPE), a London-based association of journal editors and publishers, whether their employers had a “declared policy on how to ensure a separation between editorial and commercial decisions”. The 118 responses received by 11 March indicate that 39% of respondents' journals had no such policy.

Reprints are one area in which editorial and financial issues can become entangled, says Alex Williamson, publishing director at the BMJ (formerly British Medical Journal) Publishing Group in London. She has seen submission letters in which authors claimed that, should the paper involved be accepted, a pharmaceutical company would buy large numbers of reprints. “That can be very lucrative for the journal,” Williamson says.

Some editors argue that formal policies are required to help them fend off more subtle attempts from companies to influence their content. “It doesn't have to be direct,” agrees Merrill Goozner, director of the Integrity in Science project at the Center for Science in the Public Interest in Washington DC. “It can be implied.”

Book worms: advertisers can put editors under pressure. Credit: A. MACDONALD

And editors warn of the dangers of approaching firms with offers to sell adverts next to papers about the product being advertised, a practice that many publications have policies against. “I know other journals do this,” Williamson says. “That can put editors under pressure.”

Some journals, including the medical giants such as the BMJ and The Lancet, check the factual content of the adverts that they run. But smaller journals often lack the resources to make such checks, say COPE members. A senior staff member at one journal that does check — who asked not to be named — says that he asks for changes to be made to one in every three adverts submitted to his publication. Monitoring often uncovers the use of references that do not back up the claims made in the advert, he says, adding that the process is “very labour intensive”.

Public rows between sponsors and journals may also put pressure on editors. In 2000, the house journal of the Hastings Center, a bioethics research institute in Garrison, New York, published a critical article on antidepressant drugs by David Healy, a director of psychological medicine at the University of Cardiff, Wales. A few months after publication, pharmaceutical manufacturer Eli Lilly withdrew its annual $25,000 grant to the centre, saying that the Healy article was “inaccurate and unbalanced”.

Some editors fear that such incidents could make journals think twice before publishing material attacking pharmaceutical or other industries. To provide editors with support if they feel under pressure, COPE members suggest that journals produce policies that cover relationships with advertisers, sponsors and firms involved in any other revenue streams. The policies would stress that issues such as reprint sales should not influence the peer-review process and that adverts should be checked for accuracy.

But some publishing insiders suggest that the 39% figure may overstate the problem. “The policies may not be explicit, but they are almost a given at most organizations,” says Andrea Powell, chair of the Association of Learned and Professional Society Publishers, based in Worthing, West Sussex. “Editors are not likely to be swayed, as they don't want their objectivity thrown into doubt.”