Practicalities of Searching

Session: International Clinical Librarian Conference 3: Practicalities of Searching for Clinical Librarians, Informationists, and Embedded Librarian

Prior to the conference, members of the expert searching listserv responded to a call for searching questions that they would like to have help with or discussed in this session. Each of the four panelists in this session took two questions and presented solutions for each query. This was a good practical session for the beginner and/or intermediate searcher.

Several of the panelists prefaced their presentation with the acknowledgement that they use PubMed for the majority of their searching so some of the techniques may not apply to other databases.

Below are some of the questions asked and answered:

What are the guidelines used to support systematic reviews?
The presenter reviewed the
Cochrane Handbook of standards for systematic reviews as well as other resources. Duke University has an exceptional guide to the systematic review process. Other issues to keep in mind when conducting systematic review searches are: managing the patron’s expectations, the amount of time involved in the process and the searching expertise of the librarian.

How to find statistical/numerical information for a given country?
Start with Google where searches can be limited to geographic region. This approach can retrieve some grey literature and many of the sources found were open access. Check the website of an authoritative body—ask yourself, who would publish statistics on your topic of interest?

What is the best practice for reporting and sending out search results?
It is important to format the search results and include your library’s branding (logo). Some librarians include a disclaimer and a request for the client to contact the librarian to let him/her know if the results saved them time, money or helped make a decision—sort of like a ROI. Everyone in the room agreed that it is difficult to get feedback from clients. However, always report any feedback to your manager so that they can share it up the management chain.

Why do I get different results when performing the same search?
There are a few variables to consider when searching between two or more systems. The coverage of the database in each system is not the same. For example, NLM Pubmed is updated daily Tuesday through Saturday. This may not be true for Medline on other platforms such as Ovid.  Also determine the structure and indexing principles of the database system. The type of  search terms used is another issue for consideration. Terms are either controlled vocabulary or free text terms—how are you searching these? Does the system automatically explode terms and what fields will be searched? Another example, Ovid Medline’s default searches all fields as keywords but there is no automatic explosion of terms. When a term is truncated in NLM PubMed, mapping is turned off. For these reasons and others, get to know the database and the platform you are searching.

How to use subheadings in searches?
Using MeSH terms with subheadings may limit your results to records that are indexed in PubMed. Sometimes two separate searches are warranted in order to retrieve records that are not indexed or in process. A keyword search can be performed separately to retrieve new records. Floating subheadings have been used in searches but this can lead to false drops and expand the number of search results. It can be helpful to have a peer critique your search strategy in order to ensure relevant results.