Questioning academic integrityBMJ 1994; 309 doi: https://doi.org/10.1136/bmj.309.6969.1597 (Published 17 December 1994) Cite this as: BMJ 1994;309:1597
- Richard Smith
Be prepared to defend yourself on live television
Academic conflict of interest has suddenly been given a high profile in Britain. The media have discovered that the Portman Group, which lobbies for the drink trade, offered £2000 each to five academics to produce anonymous criticisms of a book on alcohol policy.1 The media, which have been awash with stories of “sleaze” in public life, are also concerned about the funding by the drink trade of the Alcohol Research Group in Edinburgh.2 3 These stories are forcing academics to reconsider the rules of academic life.
Alcohol Policy and the Public Good upset the Portman Group because it argues that increasing the taxation of alcohol reduces alcohol problems and that the current safe limits on alcohol consumption have a scientific basis.4 Published in association with the World Health Organisation, the book is the result of a two year review by an international group of 16 well known researchers chaired by Griffith Edwards, professor of addiction behaviour, University of London.
The Portman Group managed to obtain a prepublication copy from the WHO regional office in Copenhagen, and George Winstanley, director of strategy for the group, wrote to the academics, giving them a clear idea of what he wanted in the review:
“I do not think that this book can be allowed to go unchallenged. Apart from its excessive length, it is in my opinion, unsatisfactory in the way it selects the evidence and draws conclusions from it. Results are reported and research cited selectively. Evidence which does not support the views expressed is frequently ignored. But I am far from an expert ….”
Mr Winstanley told the academics that their 2000 word pieces would be collected into a “readable whole” and published by the group with no attribution except to say that “the group has consulted widely.” The only academic whom we know to have received such a letter was Nick Heather of the Centre for Alcohol and Drug Studies in Newcastle upon Tyne. He refused to cooperate, and I am quoting from the letter to him. Although the others are not known, Mr Winstanley presumably selected those likely to produce arguments that would please his masters.
The day after this story appeared the Independent ran a story about the Alcohol Research Group in Edinburgh receiving funds from the Portman Group and the Scottish Whisky Association.2 Professor Martin Plant and Dr John Duffy are the two main academics associated with the group, and Dr Duffy is known to be sceptical about the scientific validity of the safe limits that were originally recommended by three medical royal colleges in 1987.5 The group has never made any secret of the source of its funding, but the public is sceptical about the relation between alcohol researchers and the drink trade.
These two stories have prompted a vigorous debate over academic integrity and raise several complicated issues. The first issue is the propriety of lobby groups. Most people concerned about health are disgusted by the activities of organisations like the Portman Group or Forest (Freedom Organisation for the Right to Enjoy Smoking Tobacco), but these groups have a democratic right to campaign on behalf of their legal interests. They do, however, have access to much greater funds than most health groups, and they have disproportionate influence with government because of the economic muscle they represent. The Portman Group has also produced educational material on drinking,6 and many parents would probably prefer that such material came from a less biased source.
The second more difficult issue concerns academics accepting large sums of money to produce material that is used anonymously by a clearly biased group. Two thousand pounds for 2000 words is far more than the BMJ pays, but some columnists working for national newspapers would be paid more. Many surgeons would charge much more for operations that would need much less time than that needed to research and write 2000 critical words. In other words, pounds sterling 2000 may be the market rate.
What many will find more disturbing is the anonymity, but this issue is muddy as well. The BMJ, like most medical journals, uses anonymous peer review, and some journals—for instance, the Lancet—and most newspapers still publish anonymous editorials. Indeed, the highly respected Economist is almost entirely anonymous. Anonymity pretends a greater objectivity and authority than individuals can achieve. But in this case the audience for the Portman Group's book will inevitably doubt the authors' motives and the neutrality of their arguments. Anonymity is probably doomed to disappear from intellectual life but is currently alive and well.6
A third issue is the funding of research by organisations that have a clear interest in the results of that research. Nine years ago the BMJ published an anonymous editorial (which I'll now say that I wrote) arguing that it was wrong for researchers to take money from the tobacco or alcohol industries.7 The arguments were that the industry was given undeserved respectability and that the money might buy off a potential opponent, distort the research agenda, and possibly influence the results. Many people might still accept those arguments in relation to the tobacco industry, but they now sound precious in relation to the drink trade—mainly because of the accumulating evidence that small amounts of alcohol consumption are associated with reduced overall mortality.8
The fourth issue is conflict of interest. Substantial evidence has accumulated that doctors and others are influenced by financial and other conflicts of interest: bias is subtle and pervasive.9 The International Committee of Medical Journal Editors (the Vancouver group) has issued a statement on conflict of interest.10 The BMJ's policy is that such conflict cannot be eradicated but should always be disclosed.11 We now give the source of funding for all research papers published in the BMJ, and we send all authors of editorials and papers and referees a form asking if they have any conflict of interest. Such a form is not sent to all authors of letters—for logistic reasons —but we send a form to the authors of letters for publication whenever we think that there may be a potential conflict of interest. We have received many notifications of conflicts, and we have published several statements of conflicts. We will soon, we hope, be publishing a review of our experience. Broadly, the policy seems to be supported.
Academics: no place to hide
The final issue is the validity of giving public health advice based on fragile data. I was a member of the working party of the Royal College of Physicians that together with the royal colleges of psychiatry and general practice produced the safe limits, and perhaps we didn't acknowledge as adequately as we might have the paucity of the data we had on which to make a judgment. We chose to act as doctors must act everyday in giving advice on inadequate scientific data; and, as it turns out, the guess was probably a good one.
Perhaps the only confident conclusion from these complicated issues is that doctors and academics must consider very carefully the ethical implications of all their actions. A good guide when you are faced with a difficult decision is to consider whether you would be happy to be questioned about the decision on live television. Will those academics who have taken pounds sterling 2000 to write knocking copy on the WHO's book be happy to defend their actions on television (or in the letters pages of the BMJ)? If so, they should declare themselves.