Home > Tutorials > Finding the Evidence: Evidence-Based Searching Methods

Finding the Evidence: Evidence-Based Searching Methods

Two of the most cited reasons for avoiding evidence-based practice are
the time it takes to do the literature searching
the difficulty of executing an effective search strategy
This tutorial is designed to help health practitioners acquire the skills they need to conduct successful evidence-based searches.
Formulating the Clinical Question
After the patient history has been taken and the clinical exam has been performed, the next step is formulating the clinical question. Make the question specific enough to be able to judge the appropriateness of the diagnosis or treatment for our particular patient, and for the target outcomes. For therapy or harm, we will want to know the outcome difference between different treatments, and/or no treatment (control).

Example 1: Diabetes & Hypertension

Example 2: Suspected Unstable Angina

Findings after obtaining a history and conducting physical exam

A 55-year-old white woman presents with type 2 diabetes mellitus and hypertension. Her glycemic control is excellent on metformin and she has no history of complications. To manage her hypertension, she also takes a small daily dose of therapeutic diuretic over a 6-month period. Her blood pressure hovers around a value of 155/88 mm Hg.

A 39-year-old man without previous chest discomfort comes to the emergency department after work. Early that day he had felt unwell and nauseated; he had had a vague sensation of chest discomfort and had begun to sweat profusely. The unpleasant experience lasted for about 2 hours, after which the patient felt tired but otherwise normal. At the end of his work day, feeling rather nervous about the episode, he came to the emergency department. The patient has no family history of coronary artery disease. He has had hypertension for 5 years that is controlled with thiazide, has a 15-pack-year smoking history, and has a normal lipid profile. His physical examination, EKG, creatine kinase level, and troponin I level are all normal.


Hypertensive type 2 diabetic patients without diabetic complications

Young men with atypical symptoms and normal EKG and cardiac enzymes presenting with possible unstable angina

Either beta blocker or ACE inhibitor Either admission to a monitored bed or discharge home, or time


Stroke, myocardial infarction, cardiovascular death, renal failure, total mortality

Severe angina, myocardial infarction, heart failure, or arrhythmia, all within the next 72 hours

Not all "evidence" is equal: a hierarchy of evidence
What type of evidence do we want? What type of studies will we accept? AMA builds a hierarchy for the different available information sources. While this hierarchy is not absolute, it helps us understand why certain evidence is more compelling.
Evidence for Treatment
What it is
N of 1 Randomized controlled trial A patient is given a target treatment during one period, and either an alternative treatment or placebo in the other period. The order of these periods is randomized. Both doctor and patient are blind to allocation.
Systematic reviews of randomized trials A review of the literature that critically assesses and evaluates the research on a particular clinical question. Conclusions about effectiveness are summarized to give a picture of all known research on the topic.
Single randomized trial Research in which individuals are randomly allocated to receive or not receive an intervention, and then followed to determine the effect.
Systematic review of observational study addressing patient-important outcomes Critical assessment of studies which measure the effects of certain conditions or treatments. Assessing harm, where randomization is not possible or practical, is often very important.
Single observational study addressing patient-important outcomes Patient or physician preference determines whether patient receives treatment or control. May be biased towards treatment, as physicians often select select patients with good prognosis to receive the treatment.
Physiologic studies Studies of blood pressure, cardiac output, exercise capacity, bone density and so forth.
Unsystematic clinical observations Allow only for weak inference for patient care decisions.

EMB Sources available from the MUSC Library
Best EBM Sources
Other Information Resources
MUSC ONLY Harrison's Online


ACP Journal Club MUSC ONLY
National Guideline Clearinghouse

Effective Evidence-Based Searching in Ovid MEDLINE
With Ovid, EMB limits can be set, which makes searching much easier and faster.
Enter one term at a time Ovid is not Google! Break your search question up into its important topics, and enter them separately
Use controlled vocabulary

Your search results will be more relevant if you track your topics to the appropriate MeSH (Medical Subject Heading). Be specific, selecting the narrowest term that fits the clinical question. Go into the subject tree, when necessary, by clicking on the term itself. Examples: Diabetes; Angina


"Explode" means to include more specific subjects related to your chosen topic. For example, exploding "diabetes mellitus" in MEDLINE will not only get articles on the general topic of diabetes mellitus (62,702 articles), but also articles on the more specific subjects of diabetes mellitus type 1, diabetes mellitus type 2, gestational diabetes, obesity in diabetes, etc. (173,722 articles).

With EBM searching, it is important to choose the most specific term relevant for your query. To view the narrow terms included when exploding, click on the subject term itself to explore all the narrower terms, and the related terms around it in the subject hierarchy. For Example 1 (above), it would be best to choose the specific subject "Diabetes Mellitus, type 2"


It is often helpful to focus your most important concept in EBM searching.

Checking the "Focus" box makes sure your subject is one of the major points of the article. MEDLINE articles are assigned 15-20 Medical Subject Headings (MeSH). 2 or 3 of those are chosen as the major points of the article, in other words, as the focus of the article.

Subheadings? Use subheadings to make your search more specific. For example, if you were looking for drug treatment, you would select the "/dt [drug therapy]" subheading after selecting the correct MeSH term for the disease.
Use Boolean Logic
Combine your separate sets using Boolean Logic
Use"AND" when both or all of the terms should appear
Use "OR" when either of the terms will be relevant
Apply EMB Limits

There are two ways in Ovid to apply evidence-based limits.

Clinical Queries
Choosing these limits applies a filter, a set of terms developed in a recent study by Haynes RB et al.
Choose the limit icon from the top menu
Scroll down to the "Clinical Queries" box.
Pick the appropriate type of search: therapy, diagnosis, etiology, or prognosis
Choose between sensitive(broad) or specific(narrow)

An alternative is to use the "Publication Type" limit, and choose "Randomized Controlled Trial" (RCT).

Double-Blinded? Although there is not a limit for double-blinded studies, there is a MeSH term. If it is important to look at double-blinded studies, you can add the subject term to your search. Another strategy is to make a set for "double-blind$.mp." which will catch this as a keyword in the abstract or title. Then "AND" that with your MeSH sets.
Other Limits?
Often, EBM searches return only a small number of citations. It can be more time efficient to simply look at the set than to apply more limits. However, these are some quick limits you might want to consider.
Limit by date? Only the latest, or could earlier articles be important?
Limit by age group. Is your topic about a specific age group; e.g., adolescents or the elderly? (use limit icon)
Others to consider
English only? Human only
Checking the Results Searching is not a numbers game, especially in EBM searching. The point is to find articles that answer your question. You can only know that by reading the abstract, the complete reference, and sometimes the entire article. Then you will be ready to go to the next step of evidence-based practice, evaluating the evidence.

The Cochrane Database of Systematic Reviews
Think of Cochrane's like the meta-EBM search. When using PubMed or OVID MEDLINE, you are searching for individual studies which meet the criteria for a well-constructed trial. A Cochrane systematic review, on the other hand, is a review of all the evidence for a particular issue. Cochrane surveys trials on very specific topics, including in their summaries only those that meet their quality standards. The Cochrane Library also includes several other databases, including a register of controlled trials and a database of abstracts of reviews of effects (DARE).
Access The Cochrane Library is accessible on the MUSC network. If you wish to access the Cochrane Library off campus, you must first register with Cochrane while on campus by using the "My Profile" feature. Look for the link to "My Profile" on the very top of the home page for Cochrane, on the right.
Using the Search box
The "Search" box is located towards the top, on the left. If choosing this search option, simpler may be better. Use quotes for phrases, since the default is to "and" terms. Examples:
"unstable angina"
Combining Terms
If you choose to search for more than one subject, you can use either Boolean operators (AND, OR, NOT), or adjacency qualifiers (NEXT, NEAR). Example:
"antihypertensive agents" AND diabetes
Browse feature
Another way to conduct a simple search is to use the "Browse" feature. This allows you to look for evidence-based reviews by topic from an expandable list. Look for "BROWSE ARTICLES BY" in the main section of the web page, to the left, and click on "Topics." Pull down the menu to select a general category. An expandable list of sub-categories will appear. Follow the subject category hierarchy by clicking on the triangles to expand the categories:
Treatment of essential Hypertension
Pharmacological Therapies
Beta-adrenergic blocking agents
Caution: Choosing to browse by "Cochrane Reviews" gives you a list by title, not by subject.
Advanced Search Under the search box, there is a link to the "Cochrane Advanced Search." This allows entrance of several terms, choice of Boolean operators, and field searching. Choosing "search all text" will result in more results than choosing "keyword."
Using MeSH to search
Look for the circle icon for "MeSH" at the top of the web page. (Note:This is only available on the first page; if necessary, click on The Cochrane Library 200_, Issue _ to return.)
* Type in the appropriate MeSH heading, then click on"Go To MeSH Trees." If your term is not a MeSH heading, type in one word, and click "Thesaurus" For example, "stroke" will track to several terms, including "Cerebrovascular Accident" in the Thesaurus.
*Your chosen MeSH heading will appear in the subject tree. The default mode it "Explode" which includes any more specific term indented underneath it in the subject tree.
* To select a subheading--for example, diagnosis or therapy--chose that term from the pull-down "Qualifiers" menu.
* Click on "Go"
Final hint There may be more than one review on your topic, and more than one way to look for them. Be sure you have the latest information available on your topic.

Effective Evidence-Based Searching in PubMed
Although PubMed and Ovid MEDLINE contain the same information, their search engines work very differently. PubMed runs a sophisticated algorithm on the terms entered, automatically tracking to subject headings, exploding subject terms and searching for keywords.

For EBM searching,
"Clinical Queries"

Before you begin entering your terms, look on the left margin for the term "Clinical Queries" and click there. Select your category, and choose the emphasis you wish. Then use the search box to enter your search, using the methods suggested below.
Separate your subject terms with "AND" or "OR".

PubMed is not Google. Break your search question into its important topics, and enter them separated by Boolean operators (and, or, not). You must use "AND" if you want all of your terms included in the articles. Use "OR" if both terms are acceptable.

Exploding is automatic in PubMed. As a reminder, exploding finds any more specific subject related to your chosen subject.

If you wish to know the narrower terms included with an exploded term, you can go to the MeSH Browser by changing the menu to the left of the search box to MeSH. There, one can read the scope of a subject (MeSH) term, see and select appropriate subheadings, and see the term's place in the subject hierarchy. Remember, it is important to choose the most specific term relevant for your query.
As with Ovid, use subheading only if your search is broad. If you know the appropriate subheading or its abbreviation, you can enter it at the time. If not, you can use the MeSH Browser to select the relevant subheadings for your query. Some examples:
/dt drug therapy
/th therapy (includes drug therapy, surgery, prevention and control, rehabilitation...)
/di diagnosis
/tu therapeutic use (includes admin and dosage, adverse effects, contraindications...)
Apply Limit?

Limits are seldom necessary in PubMed EBM searching. Should you wish, for example, to limit by language or date, click the "Limits" link under the search box.

You can also use limits to do an evidence-based search after a general one. Do your search, click on "Limits," use the "Publication Type" limit to choose "Randomized Controlled Trial."

Double-Blinded? Although there is not a limit for double-blinded studies, there is a MeSH term, "Double-Blind Method." If it is important to look at double-blind studies, you can add the subject to your search.
Fun Stuff Ever wonder what PubMed has done with your search statement when you do a search? Click the "Details" link under the search box to see.
Checking the Results Searching is not a numbers game, especially in EBM searching. The point is to find articles that answer your question. You can only know that by reading the abstract, the complete reference, and sometimes the entire article. Then you will be ready to go to the next step of evidence-based practice, evaluating the evidence.

Selected Bibliography
Bhandari M, Montori V, Devereaux PJ, Dosanjh S, Sprague S, Guyatt GH. Challenges to the practice of evidence-based medicine during residents' surgical training: a qualitative study using grounded theory. Acad Med. 2003 Nov;78(11):1183-90.
Bradt P, Moyer V. How to teach evidence-based medicine. Clin Perinatol. 2003 Jun;30(2):419-33.
Guyatt G and Drummond R, editors. Users' guide to the medical literature: a manual for evidence-based clinical practice. Chicago: AMA Press, 2002.
Haig A, Dozier M. BEME guide no. 3: systematic searching for evidence in medical education--part 2: constructing searches. Med Teach. 2003 Sep;25(5):463-84.
Haines SJ, Nicholas JS. Teaching evidence-based medicine to surgical subspecialty residents. J Am Coll Surg. 2003 Aug;197(2):285-9.
Wilczynski NL, Haynes RB; The Hedges Team. Developing optimal search strategies for detecting clinically sound prognostic studies in MEDLINE: an analytic survey. BMC Med. 2004 Jun 9;2(1):23.
Additional EBM Education Opportunities. The MUSC Library offers to students, faculty, residents and staff:
Course-related instruction
Presentations or collaborations at Departmental meetings or rounds
Individual appointments with a reference librarian
Class on evidence based searching, part of the monthly Information Management Series
To schedule a class or appointment, or for more information, contact Reference or call Bob Poyer at 792-2892.
Over Rider Under Rider Full Screen