Monday, July 21, 2008

PubMed Search Clinic Guide & Why the Kitchen Isn't Falling to the Sea

PubMed Search Clinic

Other than a shoutout over at for maximum reminder exposure about the July 17th Search Clinic: PubMed® Update on Automatic Term Mapping, Citation Sensor, and Advanced Search (Krafty Librarian has posted a good review of the objectives) things have been quiet around here because I was with my family on a European vacation. Things didn't quite go according to plan since we were quarantined for part of it, but more on that in another post this week.

I am a fan of guides so I can jump straight to the point when streaming video lasts for more than 3 minutes and includes more than one subject. Therefore, as with the NLM Theater presentation breakdown, here is one for the recent Search Clinic:

  • 2:35 - Citation Sensor
  • 4:37 - Old vs. New ATM (compare/contrast discussion and illustration)
  • 6:20 - Example: Burns laryngoscope 2005
  • 7:10 - July 2nd enhancements to ATM were based on user input (we do have a voice that results in additional change), details at the tech bulletin
  • 7:40 - Examples: ferrous gluconate & complement factor B
  • 8:20 - What new ATM means in general: larger retrieval
  • 8:30 - Why was it done?

    Short answer: in FAQ.

    FAQ navigation: (more direct route than the clinic but not by much) at the top of the PubMed page is the U.S. National Library of Medicine page, from there under 'The World's Largest Medical Library' banner is a link to the FAQs, select MEDLINE/PubMed and other databases, then Why was PubMed's Automatic Term Mapping (ATM) changed?

    For future sanity: Bookmark NLM FAQs at

  • 10:20 - Result of change: MyNCBI search results may be larger, check and refine as needed.
  • 10:30 - Advanced search (to the right of the search box) in beta, please provide feedback. In my opinion the features here are mostly intuitive, with the exception of the ability to set limits as ‘sticky’ or ‘unsticky’ (15:50), and may be subject to change as a beta.
  • 18:30 - Questions & Answers, with additional answers to be posted at a later date.
The bottom line, according to the ATM FAQ, is that the previous ATM “caused thousands of citation-type searches to fail every day.” Revised ATM testing found that “no significant problems were discovered, nor complaints received.” Since the ATM was changed, “the Citation Sensor is now used in several thousand searches per day. From all our available indicators, the enhanced ATM has considerably improved searching PubMed.”

I am quite happy that the NLM distance education resources page has been updated to highlight the visibility of PubMed information which must be in high demand now; I didn't know about BabelMeSH and the multi-lingual guides to PubMed!

Kitchens falling to the sea? What?

Once upon a time there was a beautiful castle on the Antrim coast of Northern Ireland (click to enlarge, thumbnail doesn't do it justice). It was everything money could buy perhaps as far back as the 13th century and definitely by the 16th century.

However, deciding to build a castle on an outcropping of unstable basalt 100 feet above the ocean wasn't exactly a smart engineering move. One day in 1639 the kitchen suddenly fell into the ocean and took the majority of the servants with it. The lady of the castle decided she no longer wanted to live there after that unfortunate event and who could blame her? Dunluce Castle was abandoned 3 years later & gradually fell into these picturesque ruins that inspired a Led Zepplin inner album cover and the setting for many local wedding photo sessions.

What does this kitchen in Ireland have to do with PubMed?

It's all about the core foundation and awareness of the environment. Building a massive building on shifting ground? Not the best of decisions that led to dire structural consequences. Having a strong foundation of biomedical information and not responding to the fact that thousands of access attempts a day aren't working for users? Also not a smart move. We are no longer in medieval times where defense is a way of life, and PubMed is not the exclusive domain of health information professionals. We have & know the tools to refine PubMed queries beyond the basic search box that we can teach to users as needed to show them there's much more than citations that scratch the surface. Lower the drawbridge instead of shouting from the tower that users are doing it wrong, then think Nordstrom's vs. majordomo to show them the value of resources inside the impressive outer walls to halt their march to Château de Google as a primary resource.