At the 1951 annual meeting of the American Association for the Advancement of Science, C.L. Emerson, a vice president of Georgia Institute of Technology delivered an ominous pronouncement:
"We may envision the time when industry and education are so closely related and so interwoven that it may be difficult to tell whether an individual is a worker in industry, or a faculty member in college, or whether he is a teacher or a student, and as a matter of sober fact, he may be all four over a year's time."
In a recent paper, Sheldon Krimsky, a professor of urban and environmental policy at Tufts University, delivered an unsettling progress report--one that suggests that Emerson's forecast may not have been far off the mark.
Krimsky and three of his colleagues examined 789 articles that appeared in 14 prominent scientific and medical journals in 1992. They were startled to find that about one-third of the papers had at least one author with some financial interest in the research published. (More details of the study are available at the end of this article.)
The Krimsky study, which was published last year in the journal Science and Engineering Ethics, ends by stating that, although most scientists probably reap no gain from a research publication, the "perception" exists that a financial tie could bias a manuscript. "At a time of diminishing research funds and institutional budgets, this is a perceptual problem worthy of consideration," the study concludes. The findings of the study add heat to the long-simmering issue of scientists' obligations to note links to industry when they publish a research paper. "We're talking about a sizable number of articles," said Krimsky. "If journals are serious about financial disclosure, they will have to do more than they're doing."
In 1992, the year surveyed, only two of the 14 journals (the New England Journal of Medicine and Science) demanded financial disclosure from authors. Since then, The Lancet and the Proceedings of the National Academy of Sciences have begun to require disclosure as well. In addition, the International Committee of Medical Journal Editors has passed resolutions recommending that authors detail financial interests.
The issue of disclosure exists in a much broader context. The Bayh-Dole Act of 1980 gave universities the right to patent or license their inventions that were created with federal money. Since then, academia has chosen to pursue vigorously its new-found freedom. In the 1993 federal fiscal year, royalty revenues for the top 10 universities or state college systems reached $170 million. The ties between academics and industry--faculty members are allowed a percentage of royalties--have grown. Benefits to society are widely acknowledged: new drugs to cure disease or electronics that lead to faster computers.
But the potential downside--the loss of the perhaps mythic notion of the university as a bastion of disinterested knowledge--has become the focus of increasingly intensive debate. In 1989, hearings by a subcommittee of the U.S. Committee on Government Operations resulted in guidelines for the disclosure of potential conflicts of interest. The issue took on a higher profile in 1995 when both the U.S. Public Health Service and the National Science Foundation issued guidelines to prevent financial conflicts of interest in federal grant applications.
Fraud is not necessarily the foremost concern, although one that certainly cannot be discounted. Perhaps the most pressing issue is whether involvement with a small biotechnology firm, for instance, can lead researchers to downplay relevant data--perhaps a potentially troubling side effect. Reviewers of a publication then need to know about ties with industry, as it may bear directly on the evaluation of research. Editors of a journal need to choose unbiased reviewers.
The need for openness suggests that the nature of scientific publishing must change. It must serve as a medium beyond just reporting research methods and evaluation of results to other scientists. In this respect, the media, as a pipeline to the general public, needs to be apprised of financial ties. And financial analysts may want to weigh information about the background of a researcher in making recommendations to shareholders.
Not everyone agrees that such disclosures are necessary, or even desirable. The difficulty with these arguments is that, in the absence of demonstrated wrongdoing, they are largely ones of perception. Such perceptions are important because they relate to how the public perceives--and supports--science. On a fundamental level, however, it is difficult to quantify how any weakness of the peer review process might compromise the integrity of scientific investigations.
In fact, the journal Nature, which is owned by the same corporate parent as Scientific American, published an editorial in its February 6th issue that responded to Krimsky's study by warning against what it called "financial correctness." Nature justified its current policy that foregoes the requirement that authors list financial affiliations. "It would be reasonable to assume, nowadays," the editorial stated, "that virtually every good paper with a conceivable biotechnological relevance emerging from the west and east coasts of the United States, as well as many European laboratories, has at least one author with a financial interest--but what of it? The measurements and conclusions are in principle unaffected, as is the requirement that uncertainties be made clear."
The Nature editorial also points out that the ethics journal study made no claims that the work linked to financial interests was fraudulent or deceptive. "[A]nd until there is evidence that there are serious risks of such malpractice," Nature concludes, "this journal will persist in its stubborn belief that research as we publish it is indeed research, not business." A related news story in Nature quotes Kenneth J. Rothman, a professor at Boston University School of Public Health, who sees a focus on financial conflicts as a distraction from, or even an impediment to, a journal's primary task--judging research on its intrinsic scientific merit.
Krimsky disagrees. Researchers' financial interest should be made known to editors, reviewers and readers, he argues: "I think that Nature is being quite naive in thinking that science operates outside of a social context." At a time when science, industry and public interest are increasingly interconnected, some critics believe that researchers have an obligation to address even the perception of bias.
Krimsky points to a recent example of a situation that has raised eyebrows. A study published last fall in the Annals of Internal Medicine showed that ingesting zinc lozenges shortened the length of a cold. One of the study's authors, Michael L. Macknin from the Cleveland Clinic Foundation, had reportedly bought stock in the company that made the lozenges before the article was published. When the company's stock soared after the journal circulated, the researcher sold some of his shares, netting a $145,000 profit. Journal editors knew of the investment; readers did not.
Financial incentives cannot help but exert pressure on at least the timing of scientific reports. David Blumenthal, the head of health policy research at Massachusetts General Hospital, published a study last year in the New England Journal of Medicine that documented one such effect. In a survey of 210 U.S. companies that sponsor or conduct research, Blumenthal found that commercial interests routinely lead to a delay in the publication of scientific findings, beyond even the time required for filing a patent. Fifty eight percent of the companies require that researchers refrain from publishing patentable work for at least six months; the National Institutes of Health, in contrast, normally allots one to two months of secrecy for the filing of patents.
Krimsky points to the Blumenthal study as a cautionary message about the ways in which financial considerations can affect the flow of scientific information. Profits may help feed research, but it is not necessarily a free lunch.
A Note on the Krimsky Study
In their study published in the journal Science and Engineering Ethics, Krimsky and his colleagues defined financial interest as being a member of a company's scientific advisory board, being listed as an inventor on a patent or a patent application for a product or process related to the published article, or serving as a director or major shareholder in a company related to a researchers' work. The study points out that other measures that could have been considered--consulting arrangements or personal stock holdings, for example--were not included because no data were available.
The researchers looked only at papers of scientists from Massachusetts who published in prominent journals such as the Proceedings of the National Academy of Sciences and the New England Journal of Medicine in 1992. Massachusetts was chosen because of the concentration of biomedical researchers. The study looked at the separate financial interests of each of the individual lead authors, as long as they were from Massachusetts. Of the 1,105 authors in the 14 journals surveyed, 15 percent had financial involvement with their research. Some authors with financial ties published more than one paper, each of which counted toward the 34 percent figure ( 267 of 789 articles) in which at least one lead researcher had some financial affiliation.