Searching with PubMed Clinical Queries

Now that the basics of PubMed have been reviewed, rather than beginning a PubMed search on the homepage using a traditional or comprehensive approach, a valuable feature for answering clinical questions for busy professionals and students is PubMed Clinical Queries. Click on the link found on the Homepage under PubMed Tools (Figure 1). The Clinical Queries feature (Figure 13) provides specialized searches using evidence-based filters to retrieve articles. The built-in evidence-based algorithms streamline the process of searching for clinically relevant articles.

Figure 13. PubMed Clinical Queries.

Image: PubMed Clinical Queries

Of the 3 options on the Clinical Queries page, our focus is on using the first two options, the Clinical Study Categories and Systematic Reviews. Since our goal is to quickly find the highest level of evidence, the search should begin with reviewing the citations retrieved under the center column, Systematic Reviews. In addition to Systematic Reviews, citations retrieved under this heading include sources such as meta-analyses, reviews of clinical trials, evidence-based medicine, consensus development conferences, and guidelines. These secondary sources provide a higher level of evidence than the primary sources found under the Clinical Study Categories. Look for studies under the Clinical Study Categories column when there are no Systematic Reviews or none that answer your question, or those under Systematic Reviews are old.

Clinical Queries allows an individual with limited computer searching skills to find high levels of evidence by typing in a main topic or specific terms of interest. Using the same clinical scenario, "tooth erosion OR erosive tooth wear" is typed into the search box (Figure 14).

Figure 14. Clinical Queries Results of Searching for Tooth Erosion OR Erosive Tooth Wear.

Image: Results of Searching for Tooth Whitening

In this case, 62 citations were returned under the category of Systematic Reviews and 1552 under Clinical Study Categories. To narrow these, include Stannous fluoride (I) AND (Potassium nitrate sodium fluoride toothpaste (C), Figure 15a.

Figure 15a. Clinical Queries Results of Searching for the P, I and C components.

Image: Clinical Study Categories Results using a Narrow Scope for Tooth Whitening

In this case, no results were found under Systematic Reviews and only 2 under the Clinical Study Categories. These same 2 studies are the same found under the Advanced Search Builder, Figure 7, result #8. Therefore, it is possible in two steps using Clinical Queries (vs. 7 steps and the use of Boolean operators, Figure 7) to identify relevant studies to answer the question.

With no citations under Systematic Reviews, then review the abstracts found under the Clinical Study Categories. This search uses evidence-based search filters to find individual studies in the indicated category. The default settings are ‘Therapy’ under Category and ‘Broad’ under Scope, since the majority of questions asked fall under the category of therapy. The type of studies retrieved for this category and scope are randomized controlled trials (see filter table). If there were hundreds of studies, then change the Scope to Narrow (Figure 15b) and this will reduce the number of citations. A Narrow, specific search will return the most relevant citations, although it may miss some.

Figure 15b. Clinical Study Categories: Changing the Scope from Broad to Narrow.

Image: Narrow Search Under Clinical Study Category for Crest Whitestrips and Custom Tray Bleaching

In our case, with only 2 citations returned, there was no need to change the scope from Broad to Narrow.

Another important consideration is the type of question/category for which an answer is being sought, i.e., does the question pertain to etiology (or harm), diagnosis, therapy, prognosis or clinical prediction guides (Figure 16). Make sure the correct category is checked off.

Figure 16. Clinical Study Categories: Changing the Scope from Broad to Narrow.

Image: Type of Question/Category

Although there is no one correct way to conduct a search, how search terms are entered influences the results and the number of steps needed to refine the search.