Related Article Searching

The case study in the January 2008 issue examines the search for information related to patient experiences and perceptions of electronic medical records. The search strings included in the case study represent the concept of the patient in the following manner:

(Patient Access to Records [mh] OR (Access to Information [mh] AND Patients [majr]) OR (Attitude to Computers [mh] AND Patients [majr]))

You'll notice that this part of the search string does not include any non-controlled vocabulary terms. As we were compiling the case study, this was an intentional decision, and of course there could be many ways to construct a comprehensive search strategy. It is quite likely that no one search strategy would accurately capture all the articles we'd consider. In this particular case, there were far too many additional articles that were added in the results when the term "patient" and its variations were brought in as a keyword search. In a few cases, there were articles that used the plural terms of a collective group of patients (i.e. patients') that were difficult to capture in the search results effectively. For example, this 2003 article by Flynn et al.discusses patient concerns to EMRs in the psychiatric setting. Due to the indexing, this particular article does not appear in our search results, but would appear if patients* [ti] were added to the portion of the search string shown above as it uses the plural for the collective group of patients in the title. However, this small change increases the search retrieval from just over 100 articles to more than 700 articles! Luckily, this Flynn article appears on the first page of Related Articles result from one of the articles included in the final packet, the Pyper C et al article from 2004. What are some ways you might address these type of nuances?

Using the Related Articles feature in this situation proved quite helpful for locating additional articles. As we were working on the case earlier this year, a post from the Shelved in the W's blog was quite timely. In this August 2007 post, the author details in depth some of the aspects of PubMed's Related Searching feature and highlights findings from a report out of the University of Maryland. It makes for an interesting read and serves as a good reminder on the benefits of using "Related Articles." Using this feature may help you locate relevant articles that may not be picked up by multiple search strings.