Wilcox, Allen J.
In a recent issue, the Editors provided a draft policy on conflict of interest and invited comments from readers.1 In this issue, we present 5 commentaries in response.2–6 We thank our commentators, who have raised thoughtful criticisms. Our revised policy is provided below.
We do not expect every author to be free of conflicts of interest. Such conflicts, like other imperfections in the scientific process, are not incompatible with good science and like other imperfections, a conflict of interest is worst when it is hidden. Yes, there is a risk of “profiling” authors by making their associations explicit. Disclosure at publication has been criticized as a distraction from the scientific work or, worse, something that erodes trust in the whole scientific enterprise. However, the troubles seem even larger when potential conflicts are revealed after the fact.7 Disclosure at the outset can protect the work (and its authors) from worse damage later.
If potential conflicts are to be disclosed, how much should be disclosed? If accepting a million dollars is a conflict of interest, what about a hundred dollars? Does it matter if it happened last year or 20 years ago? There are no generally accepted guidelines. In the meantime, detailed author disclosure forms are aggravating. To those who fear the slippery slope, we say amen.
We will be minimalists. We ask our authors to disclose whatever potential conflict of interest is relevant to their work and interpretation of their findings. Aside from the standard questions on employment and study funding, we will not define relevance. We put that responsibility on our authors, where we think it belongs.
Authors may protest that such information can hurt their paper's chances at the journal. It may be that an author's conflict of interest will lead us to conduct an especially careful review. (We already do so for other reasons, for example, with papers that may have direct public health impact.) We will strive, however, not to allow information on potential conflicts to affect our editorial decisions.
The following statement is being posted on our web site as part of our instructions for authors.
We ask authors to report their place of employment and the study's sources of funding. We also ask authors to disclose potential conflicts of interest that could raise questions about a paper's credibility if discovered later.
1. Our policy on conflict of interest: open for discussion [Editorial]. Epidemiology. 2005;16:273–274.
2. Acquavella JF. Why focus only on financial interests? [Commentary]. Epidemiology. 2006;17:248–249.
3. Brunekreef B. He who pays the piper, calls the tune.… [Commentary]. Epidemiology. 2006;17:246–247.
4. Gilman P. On a conflict-of-interest policy for EPIDEMIOLOGY [Commentary]. Epidemiology. 2006;17:250–251.
5. Neutra RR. What to declare and why? [Commentary]. Epidemiology. 2006;17:244–245.
6. Olsen J. Kafka's truth-seeking dogs [Commentary]. Epidemiology. 2006;17:242–243.
7. MacReady N. Criticism of US fetal pain study escalates. BMJ. 2005;331:532.
© 2006 Lippincott Williams & Wilkins, Inc.