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that keep track of their publications and when you start telling them their article (no matter if it has been superseded) is not able to be seen by searching their name, it isn’t going to go well.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.

Get Paid to Help Create a Next Generation Resource Tool

InContext is partnering with LWW/Ovid to redesign the next generation of electronic information resources for medical professionals.  Based on previous in-depth observational interviews, InContext has proposed some solutions and ideas to better support people who access journal articles online or conduct literature searches.  Because this is a “next generation” project, InContext is especially interested in working with people who own and smartphones and/or tablets  for work or personal use. 

They are testing their ideas for this next genration project by conducting in-person, paper prototype interviews.  The interviews normally would take 90-120 minutes.  Ovid/LWW is offering an honorarium for participation.  They will be doing interviews March 15-20 and March 29-April 4.

If the idea of being on the ground floor to help shape and create a better tool for accessing online journal articles and conducting literature searches is interesting to you AND you like getting paid for your ideas, then go to http://www.surveymonkey.com/s/LWW_Study to see if you qualify to participate.

New PubMed Tutorial

A new PubMed Tutorial is available. This tutorial was updated in December 2011, and reflects PubMed changes through December 12, 2011. To see a list of recent PubMed changes, go to PubMed’s New/Noteworthy.

Those of you who create your own PubMed tutorials unique to your institution’s settings may want to view this new one to see if you need to change or update anything.

MEDLINE End of Year Processing Information

NLM is involved in MEDLINE year-end processing activities which include changing MeSH, Substance Names and other global changes.  This always causes a temporary suspension in indexing citations. 

So here are some important dates to note:

  • November 16, 2011: NLM expects to temporarily suspend the addition of fully-indexed MEDLINE citations to PubMed. Publisher-supplied and in process citations will continue to be added.
  • Mid-December 2011: PubMed MEDLINE citations, translation tables, and the MeSH database will have been updated to reflect 2012 MeSH.

For details about the impact on searching from November 16 to mid-December, see: Annual MEDLINE/PubMed Year-End Processing (YEP): Impact on Searching During Fall 2011.

For background information on the general kinds of changes made annually, see: Annual MEDLINE/PubMed Year-End Processing (YEP): Background Information.

My NCBI Enhancements

The June 24, 2011 NLM Technical Bulletin reported on some enhancement to My NCBI’s My Bibliography.  My Bibliography will soon display links to free full text, related citations, and articles cited in PMC.  It will also have a “portlet” for related PubMed citations.  The citations in the portlet will be based on topics of the citations that are stored in My Bibliography. 

For more information including a picture of how it will look go to the NLM Technical Bulletin.

Back Door Method to Getting Articles in PubMed: Is Indexing so Important?

A very good friend of mine is a professor who researches and writes a lot on malaria.  He emailed me this morning to tell me that he had recently published an article in a journal that was not indexed in MEDLINE, but he was able to get the citation and abstract in the PubMed database anyway. 

His research is funded by the NIH and the article he published is open access, so he made it available for immediate release and submitted it to PubMed Central.  Voila, his article, although not indexed, is in PubMed. 

He ended the email saying, “You probably knew this but how come we are never privy to this trick.”  This is where I am embarrassed to say that I did not know you could get an article published in a journal not indexed in MEDLINE into PubMed by sumitting it to PMC.  I had no idea.  I knew there were non-indexed articles in PubMed, but I always understood those to fall into two categories, 1. new and waiting to be indexed 2. articles in indexed journals that aren’t medically related…for example Dynamics of magnetic domain walls under their own interia. Science. 2010 Dec 24;330(6012):1810-3 is in PubMed but isn’t indexed.

I had no idea that PMC articles were automatically added to PubMed.  I always thought PMC articles were in journals indexed in MEDLINE that were OA.  Now, my friend said in his email that he got his article indexed in PubMed.  He was wrong, the article is not indexed.  If you search for it in PubMed using only MeSH terms or if you are like me, an avid Ovid user, and you don’t often search the Ovid MEDLINE In-Process & Other Non-Indexed Citations file you are going to totally miss that article.

Early librarian me probably would have been extremely concerned because the article wasn’t indexed.  However, how important is indexing when you can get your article in PubMed anyway without indexing?  Let’s face it normal people don’t search PubMed correctly.  Almost every library user I see searching PubMed is doing their Google style searching in the database.  A simple Google search for malaria and my friend’s last name retrieved the article immediately (top result since it is a 2011 article). 

The article isn’t indexed in MEDLINE yet it is totally retrievable through PubMed and that is the DOC (database of choice) for biomedical researchers.  Researchers’ understanding of the differences of being in PubMed vs. in the MEDLINE database are already extremely blurry.  They interchange the two terms (and librarians do too) when in fact there is a technical difference.  PubMed and MEDLINE have become the Coke/Pepsi of medical databases.  Two different products but people use the terms interchangeably when ordering a cola soft drink. (Don’t even get me started on the Pop vs Soda debate.)  As I mentioned, you have an ever growing group of users who do keyword searching on a structured vocabulary database. 

So what is the value of being in MEDLINE when you are in PubMed and what is the value of having a journal article indexed when people don’t search that way anymore?  All scientists want is for their research to available to be read and cited.  Getting an article in PMC does that.  Perhaps it is time for us to let the indexing go.  Wow I can’t believe I am saying that as a librarian because I love using MeSH to search.  But, just because we love something doesn’t mean that its time hasn’t past.

—-Update—–

My friend gave me permission to repost his email to the blog, to better understand how he as a researcher feels about the whole thing.  (All identifying information has been removed or changed.)

From my end, the NIH really cares that you have a PCMID (and a link to the pubmed page) for all manuscripts on your Biosketch or the paper doesn’t count. At least they are heavily moving in this direction to keep people more honest.

 Also who cares if the MESH terms didn’t get indexed; the title, author names, and the entire abstract did.  My MESH terms would have been earth shattering terms like, malaria, antimalarial drug discovery, new drugs etc. all of which are in the abstract.

 I found it all these ways by searching pubmed.gov for: My name, Part of the title, Sentence from the abstract, and keywords

It is searchable from Google Scholar and is in Ohiolinks now too.

 All of which is nice because now people can find it and cite it (infact someone already has). And now that it is in PMC they can read it easily, more so than other articles which are not in PMC or open access.

Basically all he wants is the PCMID and his journal to be findable in PubMed (which it is). As he mentioned he doesn’t care about MeSH.  Hmmm something to think about librarians.

Confusion on Withdrawn Article from Cochrane

Yesterday I had a request to do some sleuthing on the article,”Androgens versus placebo or no treatment for idiopathic oligo/asthenospermia. Vandekerckhove P, Lilford R, Vail A, Hughes E. Cochrane Database Syst Rev. 2007 Jul 18;(4):CD000150.  In the PubMed citation there is a nice big WITHDRAWN in front of the title.  The doctor wanted to know why the article was withdrawn.

My first stop was The Cochrane Library on Wiley.  According to The Cochrane Library,”This review has been withdrawn from The Cochrane Library as it has not been updated since 1996.”  Ok makes sense, if the review article hasn’t been updated in that long then I can see withdrawing it.  However, I began to look a little more and of course got a little more confused.  Apparently another review article (same title, authors, and CD#) was published in 2000 and does not have giant WITHDRAWN printed in front of the title on the PubMed citation.

So my brain started to ask the questions…

  • If the 2007 wasn’t updated since 1996, was the 2000 article updated?
  • Why is it when I search for the 2000 article in PubMed there is no mention of it being withdrawn, but when I search The Cochrane Library for both the 2000 and 2007 review articles (both have the same CD#), the databse tells me it is withdrawn? 
  • Shouldn’t PubMed have a big ol’ withdrawn next to the 2000 citation too?

Another question that is bouncing around in my head is in the wake of so many scandals regarding scholarly publishing, were the 2000 and 2007 articles ever updated from 1996?  The way The Cochrane Library has it listed it makes me think not.  Because The Cochrane Library says the article hasn’t been updated since 1996 makes me believe that the original review article was written and published in 1996 or before.  However when I search PubMed there are no articles by these authors on this topic before 2000.  Yet when you look at the 2000 citation it clearly points to a 1996 article:

Cochrane Database Syst Rev. 2000;(2):CD000150. Review. Update in: Cochrane Database Syst Rev. 1996;(4):CD000150.

Obviously the 2007 review article has been withdrawn and I would be hesitant to use the 2000 one as well.  But it is a little hard to figure out the story behind the withdrawal, other than it hadn’t been updated since 1996.  But there are still a lot of questions left hanging out there.  Now that things are going more digital it seems the breadcrumb path of the article is more nebulous which makes it difficult when those articles are no longer appropriate to use for treatment decisions.

Web Tools For Searching Biomedical Literature

Have you heard of iPubMed, PubGet, Bablemesh, HubMed?  They are some of the many alternative interfaces to PubMed, offering different search and retrieval methods currently not available within PubMed.  With some many of these aternative interfaces how do you keep track of them?  When would it be better to use the alternative interface over PubMed or vice versa?

John Dupuis alerted me to this article, “PubMed and beyond: a survey of web tools for searching biomedical literature” (free full text) from Database (2011) Vol. 2011, doi: 10.093/database/baq036

The article looks at and reviews 28 web tools for searching the biomedical literature and compares them to PubMed and each other and has a website dedicated to tracking existing tools and future advances in the area of biomedical literature search tools.

Abstract:

The past decade has witnessed the modern advances of high-throughput technology and rapid growth of research capacity in producing large-scale biological data, both of which were concomitant with an exponential growth of biomedical literature. This wealth of scholarly knowledge is of significant importance for researchers in making scientific discoveries and healthcare professionals in managing health-related matters. However, the acquisition of such information is becoming increasingly difficult due to its large volume and rapid growth. In response, the National Center for Biotechnology Information (NCBI) is continuously making changes to its PubMed Web service for improvement. Meanwhile, different entities have devoted themselves to developing Web tools for helping users quickly and efficiently search and retrieve relevant publications. These practices, together with maturity in the field of text mining, have led to an increase in the number and quality of various Web tools that provide comparable literature search service to PubMed. In this study, we review 28 such tools, highlight their respective innovations, compare them to the PubMed system and one another, and discuss directions for future development. Furthermore, we have built a website dedicated to tracking existing systems and future advances in the field of biomedical literature search. Taken together, our work serves information seekers in choosing tools for their needs and service providers and developers in keeping current in the field.

Not only does the article look at these 28 interfaces but it also looks at the recent changes to PubMed that were often influenced by these and other outside interfaces.

There is no way any library or librarian can teach or support every one of these interfaces, but this paper is free and is a nice resource to whip out when somebody asks about one of them.

The Mobile Web Is Not An Alternative

Wednesday’s post on medinfo alerted me to this interesting article in The Chronicle of Higher Education, “As the Web Goes Mobile, Colleges Fail to Keep Up.”  The article states that more and more college students access the web using the mobile devices.  From the graph in the article, in 2010 43% of college students use mobile devices daily to access the Internet compared to 10.2% in 2008.  That is a huge jump in mobile web usage.  Yet according to the article many colleges “treat their mobile web sites as low-stakes experiments.” 

Of course right away my mind is thinking, “If colleges are treating the mobile web as a low stake experiment, what are the libraries doing?”  Depending on the library’s relationship with the college, it may beholden to the college IT department or it may have its own IT department.  That relationship will help drive a lot of the mobile web direction.  However, what is also driving the libraries’ mobile web direction are the library resource vendors.  How many ILS systems have GOOD mobile web platforms?  In the days of shrinking budgets (state and institutional) how affordable is it to add these ILS companies’ mobile platform to the library’s system?  How can a library justify that extra cost when it is faced with a flat or shrinking budget and may have to cut journals, books, hours, staff, etc? 

How many databases and online books are available/optimized for mobile devices?  Let’s ignore the Nook and Kindle like devices, students ARE NOT using them as mobile devices.  They aren’t carrying them around all the time like they are their smart phones.  They are going to use their smart phones to order Chipotle, text a friend about meeting up or an upcoming test, then they are using it to do research (usually on Google) to find a title/resource and read it.  So how many online medical text books are smart phone optimized?  Not many.

Libraries are beholden to not only their institution’s response to the mobile web but also to their own profession’s resource vendors’ response.  I remember talking to one rather high ranking sales rep for a major medical database/journal/online book provider.  I asked him if his company had created an mobile optimized version of their search database and whether there were plans to gradually optimize their many online books and journals.  He said that quite frankly that he couldn’t see why anybody would want to search that way or read an article or book chapter that way.  He didn’t see as important.  That was about a year ago.  I was gracious and said that I don’t think that way of searching and reading is for everyone but I see it as a large growth area and I know we would eventually get people asking about it.  

Well guess what Mr. Sales rep, the college students of today are my residents and staff physicians of tomorrow.  They are also the current users of your products in college libraries NOW.  Their mobile web usage has jumped tremendously and you along with the libraries are missing out.  If my users don’t usage statistics on your resources drop below a certain line, guess what we drop your resources.  If people aren’t accessing your resources that I subscribe to because they aren’t mobile friendly and they are using the mobile devices, your usage statistics will drop.  How far?  Is it below that magic dropping line?  I don’t know but usage won’t grow, and you and I both want usage to grow.

Just to be fair, NLM’s PubMed smart phone app isn’t burning up the 3G networks either.  Just today, Wouter Stomp MD and Nick Genes MD, PhD who reviewed the 6 of best PubMed apps for iPhone and iPad for iMedicalApps.com said, “Although Pubmed has a mobile version of its website, it looks outdated and is not the easiest to use.”  So just because a library or vendor creates an app or mobile interface doesn’t mean that rest easy.  They need to find out how users use it and what other competitors or libraries are doing to improve their product. 

Are we starting to feel that we are missing the users?  I don’t know, I would guess it depends on your users and your library technology.  But I don’t think this mobile web access is a passing fad.  I think librarians, libraries, and library resource providers are behind the curve on this.

Good Example of Phrase Searching in PubMed

Just like some people like their cars to have a manual transmission while others prefer automatics, librarians tend to fall into one of two MEDLINE camps, those who prefer Ovid and those who prefer PubMed. 

I am an Ovid kind of gal.  Don’t get me wrong I can do a PubMed search and have done them and still do them frequently when I need to, but my MOC (MEDLINE Of Choice) is Ovid.  Since I am in Ovid often enough I tend to see and remember better certain things in my daily searches that might be good teaching methods or examples . 

Because I am not in PubMed as often as Ovid, I don’t have the experience of running across good search examples that I can pass on or use while teaching.  That is why I try and pay particular attention to good PubMed teaching examples as the come up.  I either try and blog about them, tag them, or print them off and save them for later. 

The NLM Technical Bulletin has a nice example of how to do effective phrase searching in PubMed.  This is nice because certain things like “text messaging” (their example) are best searched as phrases.  As the Tech Bull notes it is important to look at the Search Details to know whether your term is being applied in the MEDLINE database as you want/think it to be. 

Really I tell everyone when I teach PubMed to look at the Search Details.  Sometimes I wonder how much they really do that or whether the nodding of their head is not in agreement with my point but instead to the beat of some song they have stuck in their head.