Liz Wager: Trouble with retractions

Liz WagerRetracting unreliable publications can cause headaches for journal editors and a recent case illustrates why they can be so tricky. According to reports in the BMJ and Nature, the drug company GlaxoSmithKline (GSK) has requested the retraction of an article published in Nature Medicine in 2010 describing cell biology experiments funded by the company and carried out by employees. The company has stated that data were misrepresented and has fired one of the co-authors and suspended others. But the sacked author, while admitting mistakes in the figures, claims that these were honest errors which don’t affect the study’s conclusions (in which case, they would normally be handled by a correction rather than a retraction).

According to Nature, although the GSK findings haven’t been exactly replicated by other groups,  independent evidence of similar studies supporting their conclusions has been published. So the journal is faced with the difficult task of deciding whether, as GSK claims, the findings are so flawed they should be retracted, or as at least one of the authors claim, they simply need a correction.

This case has some other interesting features. One is that the commercial sponsor is effectively acting as the publication’s guarantor and, it must be said, is taking a more robust and active role in ensuring the reliability of the literature than many academic institutions appear to. I have long thought that companies should be allowed (or even required) to act as guarantors for the research they control and sponsor. In such cases, quality control depends on employees following company procedures, so it makes sense for these procedures (rather than an individual’s expertise) to be held accountable. Another interesting aspect is how companies investigate misconduct, and how they liaise with journals. Sadly, academic institutions’ track records on this are less than perfect. Recognising the problems this can cause, COPE has developed guidelines on how journals and institutions should cooperate. I’m currently working on developing this guidance further and, to be honest, hadn’t thought, until now, that perhaps we should consider the role of commercial companies as well as academic institutions. I’ll be following this story (probably via Retraction Watch) to see if it can teach us any other lessons.

Additional conflict of interest declaration: I was an employee of GSK from 1999-2001 and, since then, have received fees from the company for running publication training courses (including two in Shanghai). I am an author of the COPE guidelines on retractions and on cooperation between journals and institutions. I received expenses from COPE to attend the World Conference on Research Integrity in Montreal this year.

Liz Wager PhD is a freelance medical writer, editor, and trainer. She was the chair of the Committee on Publication Ethics (COPE) (2009-2012).