Versioning in PubMed

I am really late with this news, my apologies.  A colleague of mine brought this to my attention.  According to the February 2012 Technology Bulletin, PubMed no supports versioned citations.  “Revisions, scientific updates, and updates of reviews are examples of content that could be versioned. Versions are not intended for correcting specific errors in an article, for which published errata notices should continue to be used (see Fact Sheet). ”

As each article is updated the next version will be placed in PubMed with a label next to the journal title indicating the version. (Version 2, Version 3, etc.) 

I guess the good news is that only the most recent version will be indexed and found in PubMed using normal methods (even Googly type searches that really normal people do).  So you won’t be seeing multiple versions in a results list. Thank God.

You can search for older versions by using three different searching techniques (See Tech Bull for images).

1. Search via the PMID number for the specific version. Example 20029611.1
   The .1 represents version 1.
2. Search via the PMID number for any version. Example 20029611.*
   The * is the wildcard.
3. Search via topic and the click on Other Versions link at the bottom of the abstract.

It is important to know that authors do not cross versions.  What I mean by that is if Michelle Kraft was an author in version 1 but not an author in version 2, you can’t do an author search and retrieve the version 1 citation.  My version is lost if you search via my name.  I predict this to be a bit of a problem.  There are too many doctors and institutions that keep track of their publications and when you start telling them their article (no matter if it has been superseded) is not able to be seen by searching their name, it isn’t going to go well.

So if you don’t have the PMID number it is going to get clunky to find an older version.  If you can’t find a citation you have to double check the citation (spelling, numbers, etc.) but now the possibility lies that you retrieved no results because it is an older version, NOT because the citation is incorrect.

There are also some funny things going on with date as well that people should be aware of.  “PubMed will set the DateCreated for the new version to the date the citation is added to PubMed. We will retain the Entrez Date, which defines the display order in PubMed, as the original date unless the publisher supplies a new PublicationDate. If the PublicationDate on the new version is different, we will modify the Entrez Date so the citation will display at the top of the search results.”

I understand and like the idea of having different versions within PubMed, but I am not sure this done in the best way.  I think you need to be able to find older versions other than just by knowing the stupid PMID number.  How many people know the stupid PMID number.  Usually we librarians are verifying the citation to GET the PMID number.  I think you should be able to do an author search and retrieve the older version.  Now there should be some serious marking on the citation indicating it is an older version and a new version exists (way more markings than NLM will slap on a citation).  By not including the older version of an article when doing an author search, not only are they saying to the authors that their article has pretty much disappeared but it makes finding articles from a citation list a real pain in the butt as well.  There are lots of researchers who look at the citations at the end of the article for their own research.

What do you think?  I am not as big of a PubMed searcher as other librarians. I am an Ovid gal (thinking how this will effect searching Ovid MEDLINE hurts my brian too much) so there are others with much better PubMed searching skills and I would love to hear there thoughts.