Canadian Medical Association Journal 1996; 155: 897-898
[résumé]
Paper reprints may be obtained from: Dr. Bruce P. Squires, 96 Frank St., Ottawa ON K2P 0X2
The full text may also be ordered from the Canada Institute for Scientific and Technical Information (CISTI) or Ovid Technologies.
© 1996 Canadian Medical Association (text and abstract/résumé)
In an attempt to clarify the definition of authorship, the International Committee of Medical Journal Editors (ICMJE) drew up guidelines in 1985.[1] These were subsequently modified in 1994, but they retained the principle that each listed author should be able "to take public responsibility for the content [of the article]."[2] The committee members could not agree on the best order in which to list authors and compromised with the recommendation that the order of authors be agreed upon by all coauthors.
As Davies and associates show, the ICMJE definition is not working -- presumably because authors do not understand it, do not know about it or do not adhere to it. The confusion surrounding the question of authorship was the subject of a meeting of medical authors and editors convened in Nottingham, England, on June 6, 1996, under the auspices of the British Medical Journal, The Lancet, the University of Nottingham and Locknet (the international peer-review research network). The keynote speaker, Drummond Rennie, deputy editor of the Journal of the American Medical Association, proposed that the current attribution of authorship be replaced with a far more complete listing of all contributors, including an explicit description of what each did. This, he argued convincingly, would help to eliminate gratuitous authorship (no department head would want to be identified as an author when he or she did not contribute to the study) and would encourage the inclusion of other substantial contributors, such as statisticians and graduate students, who are often overlooked. He also suggested that, in a very complex study, the contributors could name one or more of the group as "guarantors" to act for the group in approving editorial revisions.
Although participants at the Nottingham meeting seemed to agree with the concept of listing all contributors, not all were convinced that substituting "guarantors" and "contributors" for author lists and acknowledgements would fully resolve the problem. Wouldn't listing everyone under the catch-all of "contributors" serve only to obscure further the question of responsibility?
At a subsequent meeting of the ICMJE on July 7 and 8, also in Nottingham, committee members were clearly unwilling to alter substantially the ICMJE statement on authorship to comply fully with Rennie's proposal. Nevertheless, the committee did adjust the statement to say that authors may be required to specify what they contributed to the study (Table 1). The committee also encouraged journal editors to discuss Rennie's proposal and to discuss with researchers the concept of listing all contributors and their activities. At its next meeting, to be held in Boston, May 9 and 10, 1997, the ICMJE will reconsider the guidelines on authorship with a view to pursuing one of three options: retaining and enforcing more rigidly the current guidelines; revising the guidelines; or eliminating author lists and acknowledgements in favour of listing "contributors" and "guarantors."
In the meantime, readers and investigators are encouraged to discuss the issues surrounding authorship with their editors and colleagues. The ultimate goal is a more accurate and descriptive listing of all genuine participants in collaborative research.