Intended for healthcare professionals

Editorials

Gift authorship: a poisoned chalice?

BMJ 1994; 309 doi: https://doi.org/10.1136/bmj.309.6967.1456 (Published 03 December 1994) Cite this as: BMJ 1994;309:1456
  1. Jane Smith

    Not usually, but it devalues the coinage of scientific publication

    The fruits of authorship are usually considered to be sweet. Authorship of a scientific paper leads to grants, jobs, and reputations. This explains why many people accept the “gift” of authorship on papers to which they have contributed nothing intellectually. And, as with all presents, the givers often derive something too. They may use authorship to repay kindnesses, in exchange for authorship of another paper, or— very commonly—to credit their head of department and in so doing gain a stamp of authority on their work. Last week's revelations questioning the scientific validity of papers in the British Journal of Obstetrics and Gynaecology (see p 0000)1 show how the gift can turn sour. Perhaps this scandal will finally undermine gift authorship. At the very least it should make researchers think hard about the responsibilities that come with putting their names on papers.

    The full details of the case, at St George's Hospital, London, have yet to emerge, but we know that an inquiry has found no evidence to support the findings of two papers written by Mr Malcolm Pearce and published in the August issue of the British Journal of Obstetrics and Gynaecology. Unfortunately the editor of the journal, Professor Geoffrey Chamberlain, is also a coauthor of one of the papers. We know nothing about Professor Chamberlain's role in the work, but he was quoted by a newspaper as saying, “The head of department's name is always put on reports out of politeness. I was not part of this work, but I have always trusted Mr Pearce.”2

    The fact that “everybody does it” does not make it right, but Professor Chamberlain is correct: heads of department often put their names on papers, irrespective of their input into the work. In this week's issue Goodman shows that in his study of 12 papers and their 84 authors six heads of department were included as authors without fulfilling any of the standard criteria for authorship (p 1482).3 Similarly, Shapiro et al found in the United States that on 184 papers with four or more authors 11 heads of department were included, although they had contributed nothing to the work.4

    Ironically, it was just such a predicament as Professor Chamberlain seems to find himself in that prompted the production of a standard set of criteria for authorship in 1985. In the early 1980s John Darsee falsified studies at Emory and Harvard Universities; many of the papers that were subsequently retracted included as coauthors prominent heads of department. These people had not fabricated data, but they had allowed their names to appear on work which they knew too little about.5 Partly as a result of this episode the International Committee of Journal Editors (the Vancouver group) drew up criteria for authorship, based on the concept that “each author should have participated sufficiently in the work to take public responsibility for the content.”6

    Editors, readers, funding agencies, and society generally expect someone to take responsibility for work that comes into the public domain: they want someone to defend it, to debate it, to admit mistakes if necessary. As Shapiro et al say, the signing of scientific papers by their authors “confers credit and denotes responsibility.”4 The Vancouver guidelines try to define the activities that allow an author to take credit and satisfy readers' demands for responsibility. They suggest that authorship should be based only on substantial contributions to “(a) conception and design or analysis and interpretation of data; (b) drafting the article or revising it critically for important intellectual content, and (c) final approval of the version to be published.”7 The guidelines explicitly reject fund raising, collecting data, and supervising the research group as justifications for authorship, though these and other contributions should be acknowledged. The guidelines do, however, make it clear that between them the authors must take responsibility for all aspects of the work: “any part of an article critical to its main conclusions must be the responsibility of at least one author.”

    The Vancouver guidelines say nothing about the opposite problem: that of the researcher who has contributed to the work but whose name is left off the paper. Arguments about who should be an author can often be acrimonious, and for this reason guidelines produced by the Swedish Medical Association and its journal recommend that researchers should decide who should be an author at the outset of the work and not when the paper is being written.

    Although many journals have incorporated the Vancouver criteria into their guidance to authors, many authors ignore them. In Goodman's study only 32 authors out of 84 definitely fulfilled the Vancouver criteria for authorship and 19 possibly did so.3 Shapiro et al found that 62 of their 1176 authors had made no substantial contributions to six major tasks (conception, design, analysis and interpretation, and writing and revision plus collecting data and providing resources), while a further 206 contributed only by providing resources or collecting data.4 The cavalier approach to authorship suggested by these findings is supported by our own experience at the BMJ. From the beginning of this year we have been asking authors of accepted papers to confirm that they meet the Vancouver criteria: this request has elicited remarkably few amendments to lists of authors. This silence contrasts with the large response to our conflict of interest statement8 -which has has clearly made people think about possible conflicts of interest in a new way.9

    One lesson might be that a set of guidelines drawn up by editors will not influence the behaviour of authors. Indeed, one of the groups in the recently formed peer review research network10 is studying what authors think authorship should mean.*RF ** Neverthless, it would be surprising and disturbing—if authors were to come up with criteria that did not acknowledge the importance of the activities at the heart of the Vancouver criteria: conception, design, analysis and intepretation of the work, and knowledge of how it is written up.

    No one should dispute that readers have a right to expect authors to be able to vouch for their work, and there are other ways of acknowledging important contributions (such as the fund raising role of a head of department). Perhaps most importantly, those who have really done the work have an interest in seeing that their role is not devalued by the inclusion of many who have done little. Finally, the events at St George's Hospital should remind “authors” that signing their names to something they can't defend is not in their best interests either.

    *This study is being coordinated by Professor Raj Bhopal, Division of Epidemiology and Public Health Medicine, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH, 0191 222 6000.

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.