NLM Milestones

Congratulations to NLM on reaching a couple of milestones: the 20 millionth PubMed citation, the 2 millionth PubMed Central article: Two NLM Milestones

These two databases, along with Loansome Doc, Docline, and the rest of the free resources offered by NLM, are an amazing contribution to world health.


Mayo Clinic Library’s PubMed Tutorial

Mayo Clinic Library has posted an online tutorial for PubMed Advanced. It can be embedded in a web page or blog, like this:


Marketing ideas from Gale

Gale’s blog, The Sizzle, offers 10 ways to help you drive usage, a Powerpoint presentation (in PDF format) of marketing suggestions.  Many of the “rules” have application beyond promoting Gale databases.

For example, “Rule #1: Unleash your databases with widgets.”   A few widgets to consider adding to your library webpage:

Michigan libraries, including hospital libraries, can access Gale databases such as the Gale Virtual Reference Library, Health Reference Center Academic, and Health and Wellness Resource Center through Michigan eLibrary. Contact  Michigan Library Consortium for assistance in setting up your library’s access.

Semantic Medline

Cognition Technologies offers a free semantic search engine pulling results from Medline abstracts. They call it Semantic Medline, or Medline.Cognition; it has two URLs, and .  The help page delineates the proper use of capitalization, quote marks, Boolean,  proximity, pattern matching, wildcards, and required vs. optional search words for Cognition searching.

An interesting feature: a set of dropdown boxes appearing on the results page that allow the user to tweak the search by selecting a more appropriate meaning for a search term, where necessary.  For example, the search term “pain” has three meanings: unpleasant physical feelings (the default choice); vexatious person/hassle/annoyance; and unpleasant emotional experience. A “use all” option is also available.

My sample searches using Semantic Medline sometimes retrieved more than a matching PubMed Boolean search, and sometimes less.  It appears that foreign-language articles are not included in Semantic Medline.

Thanks to David Rothman for the pointer –

Upcoming PubMed Search Clinic

Attention PubMed users! Have you noticed the recent changes to PubMed, including the beta Advanced Search page and the changes to the automatic term mapping process? Would you like to know more?

On Thursday, July 17th at 2 p.m. EST, there will be a free online search clinic addressing these recent changes. This session, presented by the National Library of Medicine and the National Training Center and Clearinghouse, will discuss the changes to the automatic term mapping process, the citation sensor, and the beta Advanced Search page.

This online session will use Adobe® Connect™ and will last 30 minutes. There’s no registration required – you may enter as a guest. Due to technical limitations, the live session will only be able to support the first 300 participants. If you can’t attend this search clinic on the 17th or aren’t one of the first 300 participants, don’t worry! The session will be recorded and available on-demand after the 17th.

For more information and directions on how to access this online search clinic, visit this page from the NLM:

Catch up on PubMed Changes

NLM presents “PubMed Review,” at This 25 minute slideshow with audio & written transcript covers changes in PubMed since MLA 2007, including changes to the way automatic term mapping works, new MyNCBI features, and Advanced Search (beta).  Did you notice that new link to the right of the search box?


Unanswered Questions, part 2

The second research article is part of series published by Ely et al.:

Ely JW; Osheroff JA; Maviglia SM; Rosenbaum ME: Patient-care questions that physicians are unable to answer. J Am Med Inform Assoc 2007 Jul-Aug;14(4):407-14. Epub 2007 Apr 25 . PMID 17460122

This is a follow-up to a previous study in which the investigators had observed primary care physicians and recorded questions that arose during patient care. In the current study, the investigators worked with the questions for which physicians had pursued answers, but were unable to find them.

In the current study, the investigators analyze a portion of the unanswerable questions, categorize them, and try to find answers using a select list of resources that were “clinically oriented, generally available, and commonly used in practice.” I found the resource set interesting: UpToDate, Micromedex, Medline (actually PubMed, from the URL), National Guideline Clearinghouse, Harrison’s Principles of Internal Medicine, Nelson Textbook of Pediatrics, the Sanford Guide to Antimicrobial Therapy, the Red Book, eMedicine, and Google (!!). Three of the investigators report a current or past association with Thomson, which produces Micromedex.

The investigators were able to find answers to 70% of the questions studied in at least one of the 10 resources. They point out that they performed a “leisurely” search averaging 3.8 minutes per resource per question (up to 10 resources each), while the physicians who could not find the answers may at most have had a couple of minutes between patients. (Anyone who can get something useful from Google in 3.8 minutes is a more focussed searcher than I! I get distracted following interesting links.)

The authors reported their findings as an unanswered question taxonomy, recommendations for physicians, and recommendations for authors. I was most interested in their advice for physician/answer seekers.

The authors take issue with the PICO search strategy format, citing a poor phrasing of the question as one reason that physicians are unable to find an answer. While the PICO strategy is likely to find evidence-based answers, they contend, the answers don’t necessarily match the physician’s original question.

Their recommendations for physicians, to be more effective in finding answers, include selecting the most appropriate resource (handbooks and clinically oriented reviews rather than comprehensive textbooks); rephrasing questions to better match the type of information found in clinical resources (women with “fibrocystic disease” rather than “previous benign breast biopsies”); and using more effective search terms (treatment for “triglycerides” rather than for “very-low-density-lipoprotein cholesterol”).

I find PICO useful in reference interviews, myself. I may not structure my search strategy around the PICO format, but it helps me to ask the right questions, so that I know the patron asking for “pressure ulcers and surgery” wants to prevent skin ulcers in postop patients rather than to treat skin ulcers with surgery.

It seems to me that learning to break a question into component parts (which the article does touch on) and put it back together in a way that makes sense for the resource at hand would be a valuable skill for any searcher. The PICO format is one way, but not the only way, to do that. It also helps to know the resources well, know how to use the controlled vocabulary when it’s available, understand Boolean search logic … Searching 101.