MEDLINE/PubMed will stop identifying partial retractions. Here’s why.

pubmedRetraction Watch readers may be familiar with partial retractions. They’re rare, and not always appreciated: The Committee on Publication Ethics (COPE) says that “they’re not helpful because they make it difficult for readers to determine the status of the article and which parts may be relied upon.”

Today, the U.S. National Library of Medicine (NLM), which runs MEDLINE/PubMed, announced that the vast database of scholarly literature abstracts is no longer going to identify partial retractions.

We spoke to NLM’s David Gillikin about the change:

Retraction Watch (RW): When did the NLM begin identifying “partial retractions” in MEDLINE/PubMed? What led to that move, and how many have you indexed since?

David Gillikin (DG): NLM began keeping separate statistics on partial retractions in December 2006 and reporting them on the MEDLINE Key Indicators page when a journal identified an article as partially retracted.  When a journal published a notice that an aspect of an article, not the full article, was retracted, we created a “partial retraction” link between the citation announcement and the article citation involved.  At that time, we expected that partial retractions would be a new designation that publishers would be using.  These were cases when some aspect of the article, like a table, chart, image, etc., was determined to be incorrect, but the overall methodology or conclusions of the paper were not affected – the overall paper itself was still considered to be worthwhile by the journal.

RW: The NLM has now decided to stop identifying partial retractions, opting instead to classify these as errata. What led to that decision?

Since 2006, only 42 articles have been designated as “partially retracted.”  In these ten years, it is obvious that the concept of partial retractions has not taken hold in publishing.  In addition, the COPE Retraction Guidelines state that partial retractions make it difficult for users to know what parts of an article are or are not reliable or what the status of the article really is.

Beginning in October 2016, we will be discontinuing this identification for partial retractions in MEDLINE/PubMed.  Instead, we will treat this type of notice as a published erratum.  A retraction label will not be added to the citation.  We will create erratum links between the notice citation and the actual article citation as we do for other erratum/correction notices.  These citations are found using the “haserratumfor” and “haserratumin”search phrases in PubMed.

RW: Have you discussed this upcoming change with publishers, and if so, what has been their reaction?

DG: We are basically reflecting what journals are publishing and what their publishing practice is.  Publishers are not routinely using partial retraction declarations.  In the past 3 years alone, 1,651 articles were retracted and identified in MEDLINE/PubMed, while only 7 articles were identified as partial retractions.  Meanwhile, in the past 3 years, 31,124 published errata were identified and linked in MEDLINE/PubMed citations.

We haven’t discussed this change with publishers for this a workflow issue for us. Partial retraction is not a declaration being used in publishing, so it’s not efficient to have a separate workflow for something that isn’t being used.

In addition, the COPE Retraction Guidelines concerning the partial retractions are reasonable. The importance of whether an article overall should be retracted or has been retracted should not be confusing for readers to determine.

RW: Do you anticipate that your decision will influence whether publishers use this correction mechanism?

DG: Publishers haven’t been using the partial retraction mechanism for the past 10 years even though it’s been identified in MEDLINE/PubMed.  It doesn’t seem likely that there will be any change in their publishing practices with our decision to handle a partial retraction as a published erratum.

Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.

2 thoughts on “MEDLINE/PubMed will stop identifying partial retractions. Here’s why.”

  1. Whet will MEDLINE/PubMed do with the 42 currently listed partial retractions? Reclassify them? Ask the publishers to reassess those cases?

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.