Skip to Main Content

The Dartmouth Institute for Health Policy and Clinical Practice probably isn’t the most collegial of places these days. That’s because Dartmouth officials agreed with Samir Soneji, an associate professor at the New Hampshire university, that his colleague Dr. H. Gilbert Welch, a leading expert on cancer screening and overtreatment, used data that belonged to Soneji and a collaborator in California in a subsequent publication without even a hat tip.

The misappropriation, which was disclosed by STAT and Retraction Watch this week, occurred in a 2016 article in the prestigious New England Journal of Medicine. But rather than retract — or even correct — the paper after being notified of the “research misconduct,” as Dartmouth described it, the journal punted. It acknowledged the findings of the inquiry, but declared the matter an “authorship dispute” and opted not to take any immediate action. (Welch, for his part, denies any misappropriation, saying the paper was a “natural progression” of his work.)

In other words, NEJM did pretty much what Soneji has accused Welch of doing — showing disdain for the powerless — and in so doing, the journal underscored how often science abuses junior researchers with impunity. In a world where publications in prestigious journals are the coin of the realm, that matters.


Arguments about authorship are hardly rare. The Committee on Publication Ethics (COPE), which advises journals and editors about proper practice, has more than 100 such cases in its files: researchers claiming they were unfairly left off manuscripts; researchers claiming they were inappropriately included on manuscripts; theft of the work of a postdoctoral scientist; and the tantalizing “author impersonating corresponding author without knowledge of coauthors.”

Those 100-plus cases are a no doubt a dramatic undercount, given that only a fraction of authors would take the time to contact COPE. Even 20 years ago, a paper in the Journal of the American Medical Association found that such disputes were on the rise. Despite the frequency with which these episodes occur, however, institutions tend to find ways not to deal with them.


The U.S. Office of Research Integrity, which oversees investigations into allegations of misconduct involving federal funding, won’t take up authorship disputes. Universities — Dartmouth notwithstanding, to its credit — would rather look the other way, and journal editors, as exemplified by NEJM’s response, often turn into publishing Solomons. Granted, whether an author’s purported contribution was too minor to be included on a paper generally is a subjective call. But in this case, Dartmouth’s finding of misconduct — it called what Welch did “plagiarism” — undercuts the rationale for baby-splitting.

Even if Welch didn’t think Soneji and his colleague’s work merited a co-authorship, he still could have acknowledged them in the paper. But he didn’t. And Soneji’s claim for authorship rather than a mere nod is made stronger by the fact that when — at Welch’s request — he provided Welch the disputed figure several years ago, he’d asked to be included on any manuscript that incorporated the data. In a 2015 email to Welch asking about co-authorship, Soneji explained that he’d “had a few negative experiences this year when sharing results.” To which Welch replied: “No worries about this appearing in a paper – this is for class … Sorry to hear you have been burned in sharing results.” (Underneath that sizzling you hear is the sound of Soneji typing up an agreement on data sharing for future use.)

To be fair, contributing a single figure to a paper typically isn’t sufficient to earn someone a place on the roster of authors. Generally, a researcher has to have significantly more involvement in the work. So say bodies like the International Committee of Medical Journal Editors and the Council of Science Editors, to name two. However, both groups, and others, say that people who contribute data and other assistance to a manuscript should at least be acknowledged somewhere in the article.

The fact that a junior researcher didn’t get due credit is the mirror image of a related phenomenon: prominent scientists who get more credit than they deserve. Critics of academic publishing have a whole class of sin to describe the unmerited appearance on papers of scientists — almost always senior faculty — who did little, if any, work on the article. These so-called “honorary authors” might have a famous name, or a hefty haul of grant funding, or the academic muscle to force their way onto the manuscript.

According to a 2011 article in the British Medical Journal, a quarter of studies published in six medical journals in 2008 included at least one honorary author. A 2012 editorial in Science called for the end to the “fraudulent and unethical practice,” which some institutions consider to be a form of research misconduct. Evidently, few listened. A 2014 study in the Journal of Medical Ethics found that roughly a third of researchers surveyed admitted to having added authors who didn’t deserve credit.

Miguel Roig, an expert in research ethics who teaches at St. John’s University in Queens, N.Y., called the NEJM’s handling of the Welch case “less than satisfying.” In comments on Retraction Watch, Roig wrote that “even if the editors of NEJM believe the case to be an instance of an authorship dispute, a finding of research misconduct by the institution should have automatically led the editors to, at the very least, issue an editorial note or a formal expression of concern.”

Failure of the journal to take a more muscular stance, he added, “will undoubtedly lead some … to wonder about NEJM’s commitment to uphold the highest levels of scholarly and research integrity.” (NEJM declined to comment when asked for a response.)

As always, the problem in these cases is not so much that there was a dispute, as how everyone involved handled it.

Ivan Oransky and Adam Marcus are co-founders of Retraction Watch.

  • Who’s stupid enough to send a figure representing your data before you have published, if that is your intent. I assume it was, since these guys eventually submitted an MS for publication. Nevertheless, once you release it, you lose control, period. The fault with Welch is: why was he arrogant enough to use it to prove his point? He’s relying on data from a slide someone sent? How the hell much ‘data’ can be in a slide? Bizarre.

  • It’s odd that the authors of this piece claim that contribution of a single figure typically isn’t sufficient to justify authorship, when the sources they use to support that claim say no such thing. The “substantial contribution” qualifier is extremely subjective, and obviously how “substantial” the contribution is would depend on the number of figures in the paper and the significance of the figure and it’s conclusion within the context of all the others. Further, if a contributor had his/her figure misappropriated, it would be impossible for that contributor to ever satisfy the other criteria for authorship while being wrongly excluded from the drafting, review and approval process. The ICMJE is very clear on this point in their guidance; that authors who make substantial contributions should not be denied the opportunity to satisfy the other criteria. This matter should be weighed based on it’s specific circumstances, not misapplied generalities.

Comments are closed.