Expand your search  If you found too few citations or didn't find the type of information you were seeking.

Limit your Search If you came up with too many references: 
 
 

Although by no means written in stone, we suggest the following sequence:

• expand your diseases (patient) search. For example, if you initially searched the disease as "myocardial infarction" you may want to  combine (Boolean OR) this term with others which are similar disease processes or descriptions, such as unstable angina, coronary artery disease, chest pain, angina, myocardial ischemia.  The Mesh Browser can be very helpful here.  If you are willing to put up with many irrelevant citations, try adding the disease as a text word, or as a key word (mp in OVID).  To do this check the key word box on the bottom of the ovid MESH list (which you will be sent to automatically when you enter anything in the OVID search box).  Or, if you are familiar with the proper notation,  you can type the notations directly.
• In a similar fashion, use the MeSH browser, text word, title, or abstract searches to build up the intervention (or test) search. 
• Try changing any focused search terms to exploded terms
• Make your search filter more sensitive and less specific.  For example, if you had performed a filter for therapy below
• Randomized controlled trial (pt)   (pt=publication type)

or random.tw
Then adding additional terms to your filter would achieve this goal. Be creative
• or  drug therapy

or therapeutic use

• Remove any other limits you had applied to your search
• Remove one or more components (disease, intervention, comparison, search filter) of the search.
• Re-run your search in a different database, such as EMBASE or HealthStar

To see all OVID  limits, click on the limit icon in the OVID search page.
The new version of PubMed has pull down limit options.

Try the following

• Apply limit  to Human Investigations
• Assuming you can't read or obtain them anyway, limit to English.
• Limit your search to the past few years. In OVID you can do this from the Limits options or by changing databases, which are broken into several year intervals
• Limit to AIM journals.  This restricts the search to a subset of major journals
• If not already included in your search filter, limit to controlled clinical trial or randomized controlled trial from the publication type limit box (OVID).
• Start removing terms from your search for disease and  intervention (test), try to stick with MeSH terms and avoid text word searches
• Remove terms from your search filter to make it less sensitive and more specific.