The INFO Project Blog

Public health librarians couldn’t survive without Google. In a few seconds, we can paste in a title and immediately find the article, or match up an NGO with its projects, or find the address of almost any organization. By making fact retrieval so easy, Google has freed librarians to focus on their real jobs, retrieving and organizing information and helping to create knowledge.

Here Google is only part of the solution. Whether we need to research complex topics, to find the best guidance for practitioners, or to create our own knowledge resources, we need information that is at once highly comprehensive and highly relevant. Fortunately, the Internet connection that links us to powerful relevance-ranking search engines like Google also connects us to structured databases for more complex retrieval. But each database approaches  comprehensive-plus-relevant in different ways. And many searchers know only a few databases well enough to take full advantage of their unique performance features. INFORetriever–as we’ve come to call the project’s cataloging hub–would love to pool this diverse expertise and create a knowledge archive for structured database searching.

My own favorite discovery of an underused search resource happened when I tumbled onto the MeSH database. I’d searched POPLINE for years and felt confident that I could mine the database for the information that I needed. PubMed, though, was a different story. I’d searched it much less frequently and always feared that I might be missing something. Like many of my colleagues, I assumed that the main PubMed search box, plus limits, was the only search interface available. Once, though, out of desperation, I decided to try the MeSH database. I was amazed. I had expected a database for medical librarians with esoteric knowledge of the MeSH vocabulary and “MEDLINE” searching. What I discovered was a very usable and even user-friendly search interface, a tutor in medical vocabulary, and a tool for PubMed search strategy construction.

I clicked on two excellent and easy-to-follow tutorials. Then I decided to build a MeSH-based search strategy for the topic, “Contraception in Latin America”. When I searched “contraception” in MeSH, I found a list of six likely MeSH heading links with definitions. When I clicked on each link, I found the heading located within the MeSH classification, with a full display of links to broader and narrower terms. After a few clicks, I realized that a comprehensive search on contraception should include the broad terms “contraceptive agents”, “contraceptive devices”, and “contraception behavior” along with “contraception”.

When I expanded my “Latin America” search in the same way, added a few parentheses, and clicked on the “PubMed Search” button, I saw my usual PubMed retrieval display – with one big difference. Within my date limits, I retrieved twenty six relevant records (and eight other records that I wanted to know about) using the MeSH interface, and only two relevant records searching “contraception AND Latin America” in the PubMed search box. I became an instant MeSH convert.

Of course, expert free text searchers could design a much better PubMed strategy than “contraception AND Latin America” and could retrieve more relevant records. But that only suggests that adding free text phrases to the MeSH search could increase retrieval even more. MeSH even helps with that by suggesting additional text phrases  for each search.

Oh, and I almost forgot. When you use the MeSH database, you only need to click on each term to enter it. For lengthy strategies, the saving in keystrokes, time, and (if you’re like me) typos is incredible.

INFORetriever strongly encourages all PubMed searchers who haven’t used the MeSH database to find out more about it. Try the introductory tutorial. If you’re interested in searching this way, also check out the tutorial on combining MeSH terms.

Then investigate a subject that you want to know more about. Go to the MeSH database, enter your term, and follow any of the links that look interesting. Enter a practice search. When you do your next real PubMed search, you can hit the ground running.

Good luck, have fun (really), and tell me how it worked.