A conflict that might interest the ethicists

When publishing findings, many researchers tend to omit some of the more pertinent facts, reports Stephen Phillips.

A poll published this week of 1,400 science and biomedical journals has renewed US academics' concern about the way university ties with industry are skewing research findings. The poll, published in Science and Engineering Ethics , shows that in 1997, when the data were collected, barely 16 per cent of the journals had mechanisms for policing authors' financial interests in research subjects and just 0.5 per cent of the 61,000 articles on scientific and biomedical topics published in journals with conflict-of-interest policies carried financial disclosures.

The report, co-authored by Sheldon Krimsky, professor of urban and environmental policy and planning at Tufts University, and L. S. Rothenberg, an associate professor of clinical medicine at the University of California, Los Angeles, probed journals' policies for dealing with financial interests from patents, consulting work for manufacturers, seats on firms' scientific advisory committees and share ownership.

Krimsky, who was involved in a previous study that showed that 34 per cent of articles had at least one lead author with a vested financial interest in the outcome of the research being documented, believes that "poor compliance is the likely explanation for low disclosure rates in journals with policies".

Industry funding of research and development at US universities is steadily increasing - up from 6.8 per cent of research budgets in 1992 to 7.5 per cent in 1999. According to the National Science Foundation, it now contributes $2.05 billion (£1.42 billion) to higher education coffers. Topping the table of industry-funded institutions is North Carolina's Duke University. Of the $348.3 million it spent on research in 1999, 34.9 per cent came from private sources - a 280 per cent hike from seven years earlier.

Corporate funding is particularly strong in life sciences faculties, where 28 per cent of respondents said they received industry support in a 1994-95 poll published by Harvard Medical School professor David Blumenthal in the New England Journal of Medicine .

Growing awareness of the toll such ties may be taking on cherished academic canons of intellectual freedom, the open interplay of ideas and research for the public good has sparked intense soul searching in the US about the prevailing laissez-faire attitudes highlighted by Krimsky's research.

But a more immediate catalyst has been the public outcry stirred by the death, in December 1999, of 18-year-old Jesse Gelsinger, who had volunteered to undergo experimental gene therapy for a chronic genetic illness at the University of Pennsylvania. Subsequent revelations that researchers failed to disclose adverse patient reactions from previous trials, and that a company founded by the head researcher partly funded the research, have galvanised efforts to address conflicts of interest.

The US Department of Health and Human Services has since established a Federal Office for Human Research Protection as a watchdog over commercially funded research that involves human subjects. And the US Food and Drug Administration is due to publish guidelines for universities dealing with potential financial conflicts of interest where human subjects are involved.

Such guidelines will attempt to plug a vacuum. In the absence of central guidance, US universities have fashioned their own policies, resulting in variable standards. Research published by Mildred Cho, senior research scholar at Stanford University's Center for Biomedical Ethics, in the November 2000 issue of the Journal of the American Medical Association , uncovered disparate regimes and policy omissions at 89 of America's leading medical research universities. Some 55 per cent of policies required disclosures from all faculty members, while the remainder sought them only from principal investigators. Only 19 per cent put limits on academics' financial interests in research sponsors.

Cho believes efforts to rectify this situation must take place in tandem with those to overcome the apparent editorial complacency unearthed by Krimsky. "If research institutions are to be effective in enforcing disclosure policies, they need cooperation from the publications," she says.

But resistance from academic authors frequently makes policies difficult to enforce. "It all comes down to people saying, 'How dare you!'" says Drummond Rennie, deputy editor of Jama and an assistant professor at the University of California, San Francisco's Institute for Health Policy Studies.

Cho adds: "Scientists are taught that they can achieve objectivity, that factors such as money do not impinge on their thinking and that the tools they use to eliminate bias in their work are effective."

However, there is rising concern that peer reviews may not guard against research being slanted by sponsors and that readers may benefit from knowing the provenance of findings. Krimsky favours policies that spell out what salient information academics need to disclose. "Many journals leave it to the author to decide."

But having a policy is one thing, applying it is another. In February 2000, the New England Journal of Medicine admitted publishing 19 drug therapy reviews in 1997 that were written by researchers with undisclosed financial stakes in firms marketing the drugs.

Given the importance of research findings, editors are ducking their responsibility as "ethical gatekeepers" by not stringently applying policies, says Ronald Collins, director of the science integrity project at the Center for Science in the Public Interest, a non-profit Washington DC public-health campaign group. Krimsky suggests set questions be used as the starting point for more rigorous conflict-of-interest regimes, including an opt-out-style signed form on which authors explicitly deny specific financial ties.

Such requirements may be unpopular, but Krimsky thinks drastic action is needed to restore "weakened public trust in science".

Rennie backs a radical tack to stop financial interests going undeclared in what is being passed off as definitive research. He says: "It is outrageous, wrong, unethical and not in the tradition of science, which should be transparent. Science's face should be burning with shame."