How to detect authorship problems?

Editors cannot police the lists of authors or contributors in each original manuscript that is submitted to journals. However, sometimes they may suspect that the list of authors is not complete, or that it includes some authors who do not deserve to be listed as such (guest authors or gift authors). The COPE flowchart "What to do if you suspect anonymous, guest or gift authors" suggests some actions to take in these situations. The points below are intended to increase editors' sensitivity to the possibility of inappropriate authorship, so that they can spot warning signs that may indicate problems.

Types of authorship problems

A ghost author is a person whose name has been excluded from the list of authors despite meeting the requirements for authorship. This is not necessarily the same as a ghost writer, since excluded authors often perform other tasks, namely data analysis. (Gotzsche et al. have shown that statisticians involved in study design are often omitted from manuscripts reporting the results of pharmaceutical industry-subsidized trials). If a professional writer has been involved in the article, whether or not he or she meets the criteria to be listed as an author depends on the authorship criteria applied. According to the CIERM (International Committee of Medical Journal Editors; ICMJE) criteria for research articles, medical writers do not normally meet the criteria for authorship, but their role and source of funding should be listed in the Acknowledgments.

Guest authors or gift authorship is a person whose name appears in the list of authors even though he or she does not meet the criteria for authorship. Guest authors are, in general, people whose name is included to lend luster to the list (even though their participation in the research or in the article has been little or non-existent). Gift authorship is often an indication of "mutual curriculum enhancement" (i.e., the inclusion of some contributors in articles in exchange for their inclusion in their own articles of the name of the contributor who included them in the first instance).

Indications that could point to the existence of problems with authorship:

• The author responsible for the article seems to be unable to respond to reviewers' comments
• Someone who is not listed as an author makes changes to the original (You can check with the properties option in Word files to see who made the changes, but note that there may be another legitimate explanation, e.g., use of a shared computer, or changes made by an administrator)
• The Properties option shows that the draft was created by someone who is not listed in the list of authors and who is not properly mentioned in the Acknowledgements (see, however, the previous point)
• An author so prolix - e.g., of review or opinion articles - that it is hard to believe (Check for the possibility of duplicate or overlapping publications) (This can be detected with a MedLine or Google search with the author's name)
• Several review, opinion, or editorial content articles similar to each other have been published by different authors (This can be detected with a MedLine or Google search with the article title or keywords)
• A specific role is missing from the list of contributors (For example, it appears that none of the named authors are responsible for data analysis or writing the first draft of the article)
• The list of authors is so long or so short that it seems implausible (For example, a case report with a dozen authors or a randomized clinical trial with only one author)
• A pharmaceutical industry-funded study with no authors from the sponsoring company (This could be legitimate, but it could also mean that some authors who deserve to be listed as such have been excluded. Review of the protocol may help to clarify the role of employees (See Gotzsche et al. and Wager's commentary)

References