From the issue dated February 8, 2002
|
|
|
Conflicts of Interest Between the LinesWhy do scientists care who pays for research published in journals, but not in books?By LILA GUTERMAN The package that arrived on Griffith Edwards's desk last year made him angry. Addiction, the journal he edits, had received a copy of a book arguing that nicotine is not addictive. The publisher had sent the volume hoping that the journal would select it for a review. But the publisher got more than it wanted. Dr. Edwards discovered that the book's authors had received financial support from the tobacco industry and that nowhere in the book was the connection disclosed. He considered the book in violation of the journal's requirement that authors reveal financial links that constitute conflicts of interest. Although most journals apply such policies only to scientific papers, Dr. Edwards decided the policies extend to books as well. Instead of reviewing the nicotine book, Addiction published an angry editorial last month, insisting that the book had not met a standard expected of all scientific discourse, that of openness about financial interests. It wasn't the first time Addiction had received a book that angered Dr. Edwards. A year before that, a book covering alcohol-related research had arrived for review, and it had failed to disclose that its authors received money from the alcoholic-beverages industry. To prevent such problems in the future, the editors decided, the journal would review books only if publishers submitted a declaration of financial ties. And so, at a time when more journals are adopting disclosure policies, scientists are asking new questions about openness regarding conflicts of interest. Some scientists argue that books should meet the same ethical standards as journal articles and should therefore contain full disclosure. But critics of that approach say it only serves to tarnish the books -- and their authors -- and discourages readers from judging the actual content of scientific publications. "Most people would assume that failure to mention conflict of interest meant that no such conflict existed, for those are the rules of the game -- the rules by which we all trust our colleagues to play," the editors wrote about the nicotine book. The problem is, most scientists, editors, and publishers are only now starting to consider whether scientific books should play by those rules. Growing Criticism In the past decade, as connections between academic researchers and industry have become the norm, concern over conflicts has grown. Supporting those worries, bioethicists and scientists who survey scientific journals note an unmistakable trend between industrial sponsorship and pro-industry conclusions from experiments. For example, in a 1998 analysis of 15 years of review papers on secondhand smoke, Deborah E. Barnes and Lisa A. Bero, of the University of California at Berkeley and San Francisco, respectively, found that 94 percent of authors with some connection to the tobacco industry concluded that passive smoking was not harmful, while only 13 percent of authors without a link to the industry came to the same conclusion. Corporate ties received even more criticism when once-confidential documents showed that tobacco companies helped get tobacco-friendly scientists onto journal editorial boards and financed both bogus scientific conferences and studies that never received peer review. Many scientists, ethicists, and editors say that a partial solution is to make readers aware of scientists' ties to industry. "When it comes out later, it destroys your credibility," says Drummond Rennie, a deputy editor of The Journal of the American Medical Association and a professor of medicine at UC-San Francisco. If a prospective author reveals financial ties in a paper, he says, "I respect that, I allow a little bit for it, but I don't think of him as being dishonest. Whereas if I find out later, I think he was hiding something." Although many scientific journals do not require disclosure of financial ties, a majority of the most prominent ones have adopted policies, and a group of 27 addiction journals, including Addiction, adopted a joint policy in 1997. Last year, Nature, a holdout that had previously referred to disclosure policies as "financial correctness," finally adopted one. Still, the question of whether books should be held to the same standards is a new one that many important publications have yet to consider, according to 50 scientists, editors, and publishers contacted by The Chronicle. "That's an interesting question, one I never thought of," says Robert S. Schwartz, who edits book reviews at the New England Journal of Medicine. "I've never heard anybody mention this issue," says Michael L. Callaham, a professor and chief of emergency medicine at UC-San Francisco and the chairman of the World Association of Medical Editors' ethics committee. "I suspect that the issues and the principles that are getting established in periodicals are sooner or later going to have to spread to other kinds of publications." 'Years Behind' The reasons for requiring book authors to reveal conflicts of interest are simple and the same as those for journal-article authors, say the editors of Addiction. Disclosing monetary support and any other financial interests -- including consultancy to companies, stock ownership, or other such ties -- helps readers decide whether bias is present and protects against industrial manipulation of the scientific literature, says Dr. Edwards, who is affiliated with the National Addiction Centre at King's College London. "If a special-interest group like, say, the tobacco industry or the pharmaceutical industry, wishes to shape a scientific field," he says, "one well-known way of doing it is to shape the production of books." Such disclosures shouldn't mean readers will automatically assume bias exists, he adds. It just alerts them to look out for it. In fact, says Wayne Hall, a professorial research fellow at the University of Queensland's Institute for Molecular Bioscience and a regional editor of Addiction, the industry support wasn't the problem with the book about alcohol. Instead, he says, Dr. Edwards discovered "a deliberate attempt to obscure the source of funding and cover it up, which encourages the view that there's some sort of conspiracy here to present the book as something it isn't: authoritative, independent, impartial." Scientists have not generally worried about conflicts in books because researchers focus on professional journals, where they publish results of new experiments or reviews of recent developments. But some warn that the power of books should not be underestimated. "Books can be an important source of scholarly information and enormously influential," says Mr. Hall, adding that their contributions to the literature can subsequently affect policy decisions. The idea of requiring disclosure has received some support. Why should publishers be concerned about scientists' conflicts in some instances but not in others? asks Ronald Collins, the former head of the Integrity in Science project at the Center for Science in the Public Interest. Sheldon Krimsky, a professor of urban and environmental policy at Tufts University, says that scientific books aimed at expert readers, such as monographs, proceedings of conferences, or compilations of review papers, should "have the same standard of disclosure as review articles in journals if they are intended to be a contribution to the scientific literature." Yet book publishers, despite their influence on science, don't require disclosure. "It's a funny thing about books," says Dr. Edwards. "They're years behind." Looking for Bias The book that triggered the whole debate had an innocuous title, Health Issues Related to Alcohol Consumption, (Blackwell Science, 1999). The foreword raised Dr. Edwards's suspicions. It described the sponsoring organization, the International Life Sciences Institute (ILSI), as a nonprofit foundation that "receives its funding from member company dues, governmental projects, and publication sales." Curious about the member companies, Susan Savva, Addiction's managing editor, sent a letter to Ian Macdonald, the book's editor, asking who had financed the book. Two weeks later, Berry Danse, the executive director of the life-science institute's European branch, replied. "Funding of ILSI Europe scientific activities is from our members (food companies)," he wrote, adding that support came from international organizations such as the World Health Organization, too. In the editorial about the alcohol book, Ms. Savva and Dr. Edwards recounted that she wrote to ILSI again, asking specific questions about who paid for the book, how much they paid, and whether the alcohol industry was involved in choosing authors or topics for the book. More than a month later, Mr. Danse responded that the book was a project of ILSI Europe's Alcohol Task Force, which has both industry and nonindustry members. That group's projects "are mainly paid for by the Member Companies in the Task Force," he wrote. "The members of the Task Force also decide upon the choice of authors." He also explained that task-force members had asked to mention the alcohol industry's role in financing the book in its foreword, but that the institute's policy prevented singling out specific companies as financial sources. After an exchange of letters with Dr. Macdonald and another letter to Mr. Danse -- whose only response was a brochure and other general information about ILSI -- Dr. Edwards and Ms. Savva published their first editorial on disclosure policy. "Put simply, the public was offered a book which derived from a project commissioned by the drinks industry, which had the drinks industry in a position to influence the choice of authors and possibly other aspects of the editorial process, and with the drinks industry involvement in the project entirely undeclared in the published volume," they wrote. " ... In our view, the community of addiction science has been insulted." The editors received letters both supporting and criticizing their view. "Regardless of the legitimacy of an organization's mandate, it has a duty to disclose any funding information," wrote Stan Shatenstein, the editor of Tobacco News Online. One author of a chapter in the alcohol book, Boris Tabakoff of the University of Colorado Health Sciences Center, commended the journal "for its commitment to promoting ethical safeguards regarding conflict of interest." But another, Sir Richard Doll, an emeritus professor of epidemiology at the University of Oxford, wrote, "At no time did I have any reason to suspect that the scientific discussions [regarding the book] were influenced by the interests of any particular section of the industries that financed ILSI." Critics Speak Out A spirited discussion also began on an Internet mailing list of addiction scientists. Kaye M. Fillmore, a professor of social and behavioral science at the University of California at San Francisco, was one of the most outspoken critics of Addiction's editorial. "It seems to me that the publication of a book is well outside the interventionist responsibilities of academic journals," she wrote. Later, in an e-mail interview with The Chronicle, she said that demanding that books adopt journals' conflict-of-interest-disclosure policies was akin to telling Germans to heed the American speed limit in Germany. What's more, she said, the book should be judged by the quality of its science, not by who paid for it -- a point Dr. Macdonald echoes. Dr. Edwards is "condemning the book without reading it," he says. "If he read it and said it was biased, then his case would have been stronger." (Dr. Edwards says he did read the book.) Some readers have quibbles with some of the chapters' conclusions. For instance, Robin Room, a prominent alcohol researcher at Stockholm University, says he thought the book was "very conservative" about the relationship between drinking and breast cancer. "My reading of the literature [is that] there is a link, and it's causal. That's not what the chapter in the book concludes," he says. But others say the book is fair, and two other journals gave it positive reviews. Ms. Fillmore calls the volume a "pretty good state-of-the-art summary of the health consequences of heavy alcohol use" and says she could find no evidence of bias. James Stanley, the president of the life-sciences institute, confirmed in a letter to The Chronicle that the alcohol industry financed the book. He wrote that the institute had avoided publishing the names of companies for fear they would try to use ILSI to strengthen their public images, but that, "in retrospect, further elaboration about the sponsors would clearly have been more appropriate." He added that a committee is now creating a policy that will encourage "ILSI branches to list committee or task force members in the publication when space permits (as it would in a book)." Ms. Fillmore, Dr. Macdonald, and others suggest that the Addiction editorial is part of an anti-industry campaign by Dr. Edwards and the other editors. "Do I think this is an attempt to paint the alcohol industry and any research associated with it as bad? Absolutely," says Ms. Fillmore. When the nicotine book arrived, Addiction's editors realized that the life-sciences institute wasn't alone in failing to disclose connections to industry. The foreword of A Critique of Nicotine Addiction (Kluwer Academic Publishers, 2000) made no mention of payment to the authors, Hanan Frenk and Reuven Dar. Suspicious, Dr. Edwards wrote an e-mail message to Mr. Frenk asking if the tobacco industry was involved. Gold Standard? In response, Mr. Frenk, a professor of psychology at Tel-Aviv University, wrote, "I do not believe that the answer is relevant for scientists who read our book," arguing, as others had about the alcohol book, that his book be judged on its merits. He explained that he had been paid by a law firm to review the literature, and that the lawyer had refused to disclose his client, but "it seems obvious that client is from the tobacco industry." Dr. Edwards then sent Mr. Frenk 11 questions, among them: "Have you or your colleague ever previously accepted tobacco industry support in terms of grants, fees, travel support or hospitality?" Mr. Frenk did not answer any of the questions once he gleaned from them that Dr. Edwards might write about conflict of interest, rather than about the book itself. In the editorial on the nicotine book, Dr. Edwards and three other Addiction editors wrote, "We take it as a done deal that declaration of potential or actual conflict of interest is a gold standard in scientific publishing. We regard any attempt to detract from that position to be profoundly damaging to the integrity of science and, we believe, indefensible." But Mr. Frenk sees bias in conflict-of-interest disclosure even within Addiction. In a response to the editorial, he and Mr. Dar pointed out that several authors with connections to companies that produce smoking-cessation aids had published papers in the journal without full disclosure of what those companies made. "The 'gold standard' of disclosure is applied in a manner that turns it into a double standard," they wrote. Disclosure should be voluntary, whether in books or in journals, says Mr. Frenk in an interview. Mandatory disclosure "allows people to judge a book by its cover and throw it away." What's more, conflict comes in many flavors. Religious or moral convictions could influence some authors, while others might use their writing to seek fame or promotion. "Your conflict of interest, financial or not financial, is always there," Mr. Frenk says. "The only way to cope with that is not having a certain selection of people declare it but to be extra careful with everything you read." He calls Dr. Edwards's question-laden e-mail message "scientific McCarthyism." "I really expected the first question to be, Are you or have you ever been a member of the Communist party?" Some other scientists agree that mandatory disclosure can have troubling consequences. "We've all got conflicts of interest, we've all got all sorts of agendas. Let's stop pretending it's other than that," says Doug-las Cameron, a senior lecturer in psychiatry at the University of Leicester and an associate editor of Addiction Research and Theory, which did not sign on to the joint ethics policy of the 27 other addiction journals. "I worry that these conflict-of-interest [policies] can be used to invalidate perfectly good research by innuendo." Despite such objections, the trend in recent years has shifted toward more disclosure in journal articles. Yet the fashion seems unlikely to spread to books anytime soon. None of 14 scientific book publishers contacted by The Chronicle currently requires disclosure and none intends to institute a policy. Many publishers note that most scientists voluntarily acknowledge their source of funds in books' prefaces. Most also hope that their review processes will catch biases. Addiction, it seems, may be waging a lonely battle. Although Stockholm University's Mr. Room, who is the editor of Contemporary Drug Problems, defended Dr. Edwards and Addiction in an online discussion following the first editorial, he has "no intention" of instituting a disclosure policy for books under review in his journal, he says, because most books sponsored by industry or lobbying groups disclose their connections voluntarily. Book-review editors at several major medical journals also say they don't need ethical declarations from the authors of books submitted for review. Several already consider the book's source and sometimes investigate possible conflicts when choosing books for review. About a mandatory disclosure policy, Harriet S. Meyer, the book-review editor at The Journal of the American Medical Association, says, "I don't know if there's enough of a problem to warrant it." But Dr. Rennie, of the same journal, disagrees: "Nobody thinks it's an issue until you get caught." Journal editors didn't consider disclosure important either until they had to because of studies that linked conflicts to industry-friendly results, he says. "We didn't know what a huge problem it was." But now the link between corporate ties and potential bias is clear to all, he says. "If publishers are saying it doesn't matter, they are putting their heads in the sand. And it means that they don't care about credibility." He and others who back disclosure policies for books say such policies are not meant to prejudice readers against industry-related research. "The fact that conflict or potential conflict is openly declared does not of course at all prevent discussion of a book 'on the basis of its arguments' but merely provides the reader with background information which he or she may want to take into account when evaluating the arguments," the Addiction editors wrote in the editorial about the nicotine book. Of all the possible conflicts, financial interests most damage credibility when they are discovered after publication, Dr. Rennie adds. "It isn't just a thought that we are influenced by money. It's a fact." Dr. Edwards says Mr. Frenk and Mr. Dar's argument that Addiction had not been applying its disclosure policy uniformly is "dubious." "Everyone who publishes in Addiction is asked to declare, and I look at that very carefully," he says. The counterexamples Mr. Frenk and Mr. Dar gave were "at worst, errors of a very different magnitude" than leaving out the source of funds for an entire book. In the examples, the authors mentioned their connections to companies but failed to disclose that those companies produced nicotine replacements. Dr. Rennie also defends Dr. Edwards from charges of instituting an anti-industry campaign. "That's nonsense. When people don't have a good argument, they attack the person," he says. In practical terms, Dr. Edwards suggests that if a group of journals were to require conflict declarations of books submitted for review, then book publishers would begin to pay more heed. But he admits he doesn't have all the answers. "Evolution is going on, which will be slow," he says. "I find it rather cheering to get the debate going." Dr. Rennie agrees. "I think it's something that will be done," he says. "And I think in 20 years' time, this will be a very old conversation." http://chronicle.com Section: Research & Publishing Page: A14
|
Copyright © 2002 by The Chronicle of Higher Education