Retraction Watch readers may have noticed that we often cover retractions long before they appear in PubMed, the gold standard database for the life sciences literature. (In fact, we’ve taken to leaving comments on papers in PubMed Commons about retractions that haven’t been linked to their original abstracts yet.)
This can be an issue, because so many scientists use PubMed to find relevant literature. It may even contribute to the well-documented phenomenon of researchers citing retracted papers as if they hadn’t been retracted.
Until now, no one had quantified the time lag. In a new study, Evelynne Decullier, Laure Huot, and Hervé Maisonneuve — who have published on retractions before — looked at 237 retractions published in 2008. Their findings?
A search on notices of retraction issued in 2008 was repeated every 3 months during 15 months from February 2011. The first search resulted in 237 notices of retraction. Throughout the study period, 14 discrepancies with the initial search we’re observed (6%). We found that the number of retraction notices became stable 35 months after the retraction.
That means, of course, that researchers need to check publishers’ sites if they want to keep current. Of course, that won’t always work, either.
All the more reason to build the retraction database we’ve been planning. Here’s how you can help.