Show your interests

Scientists should disclose deals that may pose a conflict of interest, argues Sheldon Krimsky

Reports of conflicts of interest in the biomedical sciences are becoming routine in North America.

Just two recent examples: on December 1 last year, The Wall Street Journal disclosed that five out of eight members of a National Institutes of Health panel chosen to evaluate diagnostic tests for Alzheimer's disease had ties to the test's manufacturer - conflicts of interest not revealed in the NIH consensus report.

On December 10, the Toronto Star revealed that a Canadian hospital with financial connections to a pharmaceuticals company turned against a haematologist who had disclosed negative findings of a drug used in a clinical trial. The researcher faced legal action by the company for breaching confidentiality.

Since the early 1980s, US policies have encouraged academic researchers and their institutions to develop partnerships with for-profit corporations. These policies, in conjunction with the liberalisation of intellectual property rights, have led to increasing entrepreneurial activities among scientists at leading universities.

Studies published in Science, the New England Journal of Medicine and the Journal of the American Medical Association on the relationship between academic institutions and industry in the life sciences have revealed trade secrecy in universities, conflicts of interest, scientific misconduct, shifting emphasis to applied research and corporate gifts with strings attached. In 1988, the International Committee of Medical Journal Editors passed a resolution that authors should report any financial relationships that "may pose a conflict of interest".

But in a 1996 pilot study of 14 biomedical journals (Science & Engineering Ethics), we reported that 34 per cent of the 789 articles by Massachusetts-based scientists published in 1992 had at least one lead author with a personal financial interest in the content of the study. None of these interests was disclosed in the publications, only a few of which had policies on conflicts of interest.

That has changed somewhat as journal editors begin to respond to calls for more openness about researchers' hidden biases. Nevertheless, only a small percentage of science journals and fewer than half of all medical journals require authors to disclose financial interests related to their publications.

A recent Nature editorial said: "Until there is evidence that there are serious risks of such malpractice (fraud, deception or bias in presentation) this journal will persist in its stubborn belief that research as we publish it is indeed research, not business." But a year later, the New England Journal of Medicine printed a paper correlating funding sources of authors writing in 70 scientific articles about calcium channel blockers (hypertension drugs) with their evaluation of safety and efficacy. It found that 96 per cent of those who supported calcium channel blockers - as against 37 per cent who were critical of them - had financial relationships with the drug's makers. Only two of the 70 papers disclosed authors' financial interests.

Other studies have shown that financial benefits can explain why some physicians have a higher referral rate for tests, hospital admissions or operations than others and that corporate sponsorship of research in the drugs and tobacco industries is more likely to yield outcomes favourable to the companies.

Those opposed to disclosure policies argue that financial interest is only one of many interests held by scientists and should not be singled out. They argue that a sound scientific study speaks for itself and that the quality of the science can be evaluated independently of the author's interests. Those who favour reporting financial interests argue that full disclosure removes the suspicion that something relevant to objectivity is being hidden.

While financial interest in itself does not imply any bias in the results of a paper and should not disqualify it from publication, we believe disclosure allows readers and reviewers the opportunity to form their own opinions on whether a conflict of interest exists and what relevance it has to the study. For peer reviewers, a disclosure of financial interest may ratchet up a few notches the scrutiny they give the paper. Disclosures of an author's financial interests in promising new drugs and treatments can modulate wild swings in stock prices and contribute to more rational and informed investments.

Finally, complete transparency of financial interests of scientists in their publications and public-interest consultancies will help restore our society's trust and confidence in its scientific institutions.

Sheldon Krimsky is professor of urban and environmental policy, Tufts University, Massachusetts, United States.