Late last month, the editors of The Lancet Oncology published an “expression of concern” regarding a paper published in 2007. This term, which appears to be unique to scientific publishing, was helpfully defined as “A statement issued by the editor of a peer-reviewed journal … [that] calls attention to a specific paper, especially to question the validity of that paper or portions of that paper” in an article in The Scientist.
A more recent and straightforward definition comes from Retraction Watch, a new blog that should be required reading for anyone interested in scientific journalism or the issue of accuracy. The blog’s authors—Adam Marcus and Ivan Oransky—describe an “expression of regret” as “a Britishism that might be better expressed as ‘Holy Shit!’”
Holy shit, indeed: it appears the offending Lancet article from 2007 is caught up in an unfolding scandal centered around Anil Potti, a Duke researcher who falsely claimed to be a Rhodes Scholar and may have “invented key statistical analyses in a study of how breast cancer responds to chemotherapy,” according to Retraction Watch.
Looks like they’re off to a good start.
Marcus and Oransky are well qualified to serve as retraction hunters. Oransky, a former editor at The Scientist (he assigned the article cited above), is the executive editor of Reuters Health, teaches medical journalism at New York University, and runs the Embargo Watch blog, which was the subject of a CJR article by Curtis Brainard.
Marcus is the managing editor of Anesthesiology News and contributes to other publications. He distinguished himself as a first-rate accuracy hound by breaking the story of Scott Reuben, a respected anesthesiologist and researcher who turned out to be a serial fabricator. Though the Reuben case drew far less attention than that of Jayson Blair, it “led to the retraction of more than a score of papers and deeply rattled an entire medical specialty,” according to Retraction Watch’s introductory post, “Why write a blog about retractions?” (Why the hell not, I say.)
“I’m an avid PubMed troller, and after covering the Reuben episode I began checking the literature regularly for retractions on the theory that you never know what you’re going to find,” Marcus said via e-mail. “And in this way, Anesthesiology News [where he serves as managing editor] became the first outlet to report on the retracted paper by Dr. Timothy Kuklo, a former Army surgeon who apparently fabricated a study using research money from Medtronic and got himself into extremely hot water with the military and Congress. But it’s hard to make those stories fit with my publication’s audience, so when Ivan suggested that we blog about it, I agreed immediately.”
Though the site just launched this week, the duo appear to have a lot of material to work with. “We’re not working with a deficit,” Oransky says by phone.
Along with recent retractions, they’ll dip into the deep well of historical recalls and highlight the notable and nasty. In a larger sense, though, they hope to draw attention to the issues related to retractions.
“One of messages we hope make clear is that a lot of what we’re looking for is inconsistency,” he says. “… Do you always do the same thing [with retractions] based on the last time you did it? That’s one of the major things I look for.”
The pair hope to bring awareness to the retractions, which, like their counterparts in magazines, newspapers, and online, often slip by unnoticed. That was one conclusion of a paper published two years ago in the Journal of Medical Ethics that examined retractions of scientific articles:
Although retractions are on average occurring sooner after publication than in the past, citation analysis shows that they are not being recognised by subsequent users of the work. Findings suggest that editors and institutional officials are taking more responsibility for correcting the scientific record but that reasons published in the retraction notice are not always reliable. More aggressive means of notification to the scientific community appear to be necessary.
The paper found the rate of retractions “remains low but is increasing.”
Oransky says there’s already a mechanism in PubMed whereby people can see if a paper has been cited by other papers—a kind of trackback—and he thinks there should be a way to notify people who have downloaded a particular article when it has been retracted.
“What I’m not sure they do is have a decent mechanism for alerting people who downloaded the original paper,” he says. “There is not as much linkage as there could be.”
Marcus has special insight into how journals handle retractions, thanks to his work on the Reuben story, which resulted in a remarkable flowering of article recalls.
“… I can say that I think Dr. Steve Shafer, editor of Anesthesia & Analgesia — which was hit disproportionately hard by the Reuben fraud — has the right approach to how to respond,” he wrote. “Shafer didn’t stick his head in the sand and hope that it would all blow over. He notified his readers of the retractions, revamped his journal’s author guidelines, and has been willing to talk about the case from the first moment. In my experience, other journal editors have been much less open. I suspect this reticence stems from a fear of the ‘gotcha,’ and as an editor I can understand that. But my feeling is that openness should be the default, not the exception.”
Hopefully, Retraction Watch will nudge scientific publishers towards increased openness. At the very least, though, they will provide an invaluable service by offering layman’s translations of retractions such as this:
The authors wish to note the following: “The locus AT1g35515 that was claimed to be responsible for the cold sensitive phenotype of the HOS10 mutant was misidentified. The likely cause of the error was an inaccurate tail PCR product coupled with the ability of HOS10 mutants to spontaneously revert to wild type, appearing as complemented phenotypes. The SALK alleles of AT1g35515 in ecotype Columbia could not be confirmed by the more reliable necrosis assay. Therefore, the locus responsible for the HOS10 phenotypes reported in ecotype C24 remains unknown. The other data reported were confirmed with the exception of altered expression of AT1g35515, which appears reduced but not to the extent shown in Zhu et al. The authors regrettably retract the article.”
Not exactly a misspelled name, huh?
Correction of the Week
“An article (9 October 2009), ‘Hunger striker’s £7m Big Mac’, reported claims that Mr Subramanyam was caught secretly eating burgers while on hunger strike during the Tamil protest in London, wasting significant police costs. We now accept that there was no truth in these allegations and we and other media have agreed to pay him damages and have apologised to Mr Subramanyam for the distress and embarrassment caused.” – Daily Mail (U.K.)
Craig Silverman is currently BuzzFeed's media editor, and formerly a fellow at the Tow Center for Digital Journalism.