Introduction to Evidence-Based Practice



Literature Search

The Literature Search

the evidence

3. Select the appropriate resource(s) and conduct a search

In the previous module, we learned how to construct a well-built clinical question. Using that question, we will move on to the literature search.

For our patient, the clinical question is:

In elderly patients with heart failure and an ejection fraction of 40%, is irbesartan effective in reducing the need for rehospitalization?

It is a therapy question and the best evidence would be a randomized controlled trial (RCT). If we found numerous RCTs, then we might want to look for a systematic review.


Constructing a well-built clinical question can lead directly to a well-built search strategy. Note that you may not use all the pieces of the well-built clinical question in your MEDLINE strategy. In the following example we did not use the term "ejection fraction 40%" or "elderly." We also did not include the word therapy. Instead we used the publication type, randomized controlled trial, to get at the concept of treatment. However, you may consider the issue of ejection fraction or age later when you review the articles for applicability to your patient.


Clinical Question

Search Strategy

Patient / Problem

heart failure, elderly
ejection fraction 40%

heart failure
Limit to Aged


irbesartan or avapro

irbesartan OR avapro

Comparison (if any)

none, placebo, standard care



reduce need for hospitalization

reduce mortality



Type of Question


(see below)

Type of Study


Clinical Query - Therapy/narrow
Limit to randomized controlled trial as publication type

Select a resource

The practice of Evidence-Based Practice advocates that clinicians search the primary literature to find answers to their clinical questions. There are literally millions of published reports, journal articles, correspondence and studies available to clinicians. Choosing the best resource to search is an important decision. Large databases such as PubMed/MEDLINE will give you access to the primary literature. Secondary resources such as ACP Journal Club, Essential Evidence, FPIN Clinical Inquiries, and Clinical Evidence, will provide you with an assessment of the original study. These are often called "pre-appraised" or EBP resources. The Cochrane Library provides access to systematic reviews which help summarize the results from a number of studies.

Detailed list of resources for practicing EBP

To quickly find an answer, we might first look at an appraised resource, such as ACP Journal Club. ACP Journal Club's general purpose is to select from the biomedical literature articles that report original studies and systematic reviews that warrant immediate attention by physicians attempting to keep pace with important advances in internal medicine. These articles are summarized in value-added abstracts and commented on by clinical experts.  Studies included in this small database are relevant, newsworthy and critically appraised for study methodology.

A search of irbesartan and heart failure identified this citation: 2008 - Irbesartan did not reduce all-cause death or CV hospitalization in heart failure and preserved ejection fraction.  This is a critical appraisal of Massie BM, Carson PE, McMurray JJ, et al. Irbesartan in patients with heart failure and preserved ejection fraction. N Engl J Med. 2008;359:2456-67.  The conclusion is that Irbesartan did not reduce the composite endpoint of mortality or hospitalization for cardiovascular events in patients with heart failure and preserved ejection fraction.

If you do not have access to ACP Journal Club or other EBP resources you will need to do the search in PubMed or MEDLINE.  PubMed/MEDLINE is a very large database with over 17 million citations.  You will need your focused question and search strategy for this database.

Conduct the search in PubMed:

Step 1: Use PICO to formulate the search strategy; start with the Patient problem and Intervention

Enter the term for the patient problem and the intervention: heart failure AND irbesartan. PubMed attempts to map your terms to appropriate Medical Subject Headings (MeSH). MeSH is the standard terminology used by the indexer and helps find articles on the topic, regardless of the exact wording used by the authors. 


Step 2. Look at Details to verify MeSH terms

Click on Details (from the Advanced search page) to see the terms that PubMed actually used in its search. You want to be sure PubMed found the appropriate MeSH terms.  PubMed will automatically also search for your terms as words in the title and abstract. Heart failure is a MeSH term and irbesartan is a Substance Name.


Step 3. Limit to appropriate study design
This is a therapy question. We know from the previous discussion that the best evidence for a therapy question is a randomized controlled clinical trial (RCT).  Use the Limits function from the advanced page to limit to RCT as a publication type.


You can also use the Clinical Queries function to limit the results to study methodologies relevant to therapy questions. Copy your last search strategy heart failure AND irbesartan.  Click on Advanced search; near the end of the page under More Resources there will be a link to the Clinical Queries. Paste the search strategy in the Clinical Query search box and select the type of question (Therapy) and the default narrow, specific search. Click on GO. You may get more results using the Clinical Queries.


Step 4. Review the results

Both methods limit your results to RCTs.  The second citation is the Massie article that we found in ACP Journal Club. 


We have Videoa short video demonstrating this search.


If you are not familiar with searching PubMed, you may want to review the PubMed tutorials at

The next step is to read the study and determine if the methodology is sound so that we can consider the results.

Previous References Next

Revised July 2010