Making progress with competing interests

The BMJ and other journals are making progress with managing the problem of competing interests (or conflicts of interest, as most journals call them). Today we take two further steps forward by posting on our website the competing interests of editors, our editorial board, and our group executive (http://bmj.com/aboutsite/competing_interests.shtme) and by publishing a study we have conducted that shows that readers' reactions to research are strongly influenced by statements of competing interests.¹ We still, however, have some way to go to the fully transparent world that is desirable.

The history of medical journals and conflict of interest might be cruelly summarised as lots of rhetoric and not much action. The International Committee of Medical Journal Editors produced a policy on conflicts of interest as long ago as 1993,² but several studies have shown that such conflicts are rarely declared in most journals—despite good evidence that most authors have them.^3–7 The international committee strengthened its policy in 2001 by stating that journals should declare the exact role of sponsors (often pharmaceutical companies) in studies and decline to publish studies where the sponsors controlled the decision on publication.⁸,⁹ This policy too has yet to be widely implemented.⁷

At the BMJ all authors and reviewers of original articles, editorials, and most other material are asked to complete competing interests forms, and declarations of the competing interests of authors are made with every article.¹⁰ We have now instituted a system to ensure that the role of sponsors is made clear, and anybody submitting a rapid response is required by the electronic system to make a statement on whether they have competing interests. Our main means of managing competing interests is disclosure, but sometimes the conflict is so strong that it disbars somebody from being an author or a reviewer.

The slow progress of journals with managing conflicts of interest is bothersome because evidence is strengthening that they have a strong influence on the interpretation of evidence. A study of authors from many journals showed that people with financial relationships with manufacturers of calcium channel antagonists were much more likely to be supportive of the drugs.³ Yet they rarely declared those conflicts. The major determinant of whether reviews of passive smoking concluded it was harmful was whether the authors had financial ties with tobacco manufacturers.¹¹ In the disputed topic of whether third generation contraceptive pills cause an increase in thromboembolic disease, studies funded by the pharmaceutical industry find that they don't, and studies funded by public money find that they do.¹² A recent strange decision of the American Heart Association to declare alteplase a class I (definitely recommended) intervention for stroke despite controversy about its safety and efficacy seemed less strange when we learnt that the manufacturers of the drug had donated $11m (£7m; €11m) to the association and that six of the nine members on the panel making the decision had undeclared financial ties to the manufacturers.¹³

There isn't anything wrong in having competing interests. They may be near universal in medicine. Most doctors have had their lunch bought for them by a pharmaceutical company and write their prescriptions with pens provided by the industry. Many have been flown to very agreeable locations by the companies to listen to speakers, staying—at the companies' expense—at the best hotels. The problem with conflicts of interest is not declaring them.

We must recognise too that we are influenced by those conflicts. None of us likes to think that we are influenced by advertising. But we are. Double blind randomised controlled trials are necessary for testing new treatments not because those conducting trials are dishonest but because bias is pervasive and acts on us unconsciously. We deceive ourselves if we think that we understand exactly why we behave as we do. We are human beings, deeply flawed creatures.

Although as authors we may be slow to recognise the importance of conflicts of interest, it seems that as readers we do. In our study we sent randomly to BMJ readers the same paper with or without a statement of competing interests.¹ We asked readers to score the study for interest, importance, relevance, validity, and believability. We thought that the competing interest statement might lead readers to score the study lower on believability. In fact they scored it significantly lower on every measure. Our study had several weaknesses, and we are repeating it—but its results are very interesting. Those who worry that concern about conflict of interest is simply political correctness may have their worst fears confirmed: readers do seem to discount heavily studies with competing interest statements. On the other hand, the study perhaps shows that we all know that conflicts of interest can have a profound effect.

It is human to be concerned about the accountability of others and unconcerned about our own accountability. Thus editors have been keen to act on the sins of authors but unable to act on their own sins.¹⁴ Similarly we have been busy disclosing the competing interests of authors but have not disclosed our own. Today we do, although the process of getting everybody to disclose their competing interests has been protracted and readers will see some holes. We have declared the interests not only of editors and our editorial board but also of our executive team—because there are examples of publication committees getting into trouble over undeclared conflicts of interest.¹⁵ We haven't yet declared the competing interests of the board of BMJ Publishing Group Limited—because the company has only just come into existence and the board just been elected. Similarly we haven't declared the interests of our editorial advisers because we have recently appointed new ones. The competing interests of all these people will be posted soon.

These are, we believe, steps in the right direction, but important questions remain. Our policy requires people to declare financial conflicts of interest and encourage them to declare non-financial ones. This is partly because most of the evidence relates to financial conflicts of interest and because defining non-financial conflicts is hard. But should we go further? Another question is whether we should ask people to declare the scale of their financial conflicts of interest? Charles Warlow, president of the Association of British Neurologists, raised this question in the debate over the American Heart Association and alteplase.¹⁶ It seems likely that different degrees of conflict are probably raised by being bought a cheese sandwich by a company or by being flown on Concorde to New York to give a lecture and spend five days in the Ritz-Carlton Hotel. But at the moment we don't ask for amounts, partly because we British are even more embarrassed to talk about money than about our sex lives. But should we change?

Papers p 1391