Evidence Based Medicine Guide: From PICO to PubMed Search Terms
NLM's PubMed Tutorial
Features and tools especially useful in searching for clinical evidence
Identify Core PICO Elements
Successful literature searches tend to have a simple search statements that include only two to three core concepts (1). Beginning EBM searchers often make the mistake of trying to incorporate every concept from their PICO analysis in their search query — which often results in few or no search results.
- In general, start your search with no more than three search concepts.
- These concepts should reflect your question's most important PICO elements. To help set priorities, some suggest underlining the the key terms in your search statement and others recommend listing them in rank order of importance.
(1) Doig GS, Simpson F. Efficient literature searching: a core skill for the practice of evidence-based medicine. Intensive Care Med. 2003 Dec:29(12):2119-27. PMID:129551882
Choose Search Terms Carefully
Once you have identified the main concepts of your question, you'll need to choose good search terms to represent those concepts. Unsuccessful or incomplete searches are often caused by by using overly specific terms or subject headings, or by the failure to specify alternate terms when the initial search fails (1).
Search Tip: Start with broader concepts
For example, begin searching with cardiac arrhythmia AND anticoagulation. Then, if you retrieve too many results, use more specific concepts (atrial fibrillation AND warfarin).
Search Tip: Identify alternative ways to express your main concept
Connect synonyms or related concepts in PubMed with OR (e.g., warfarin OR anticoagulants OR coumarins).
1) Doig GS, Simpson F. Efficient literature searching: a core skill for the practice of evidence-based medicine. Intensive Care Med. 2003 Dec:29(12):2119-27. PMID:129551882
Use Search Terms That Map to MeSH Terms
PubMed broadens your search when it can translate your search vocabulary into MeSH terms. PubMed employs Automatic Term Mapping in the background of every search. Whenever possible, Automatic Term Mapping automatically matches, or maps, thousands of search words and phrases to MeSH terms. PubMed then adds the synonyms from MeSH to your search terms and searches for both.
When your search terms map, your search is more comprehensive.
Mapping usually increases the number of search results and makes it less likely that you will miss important articles. For example, a search of deep vein thrombosis maps to venous thrombosis. PubMed searches for both: deep vein thrombosis OR venous thrombosis, and retrieves over 60,000 results.
If your search terms do not map, you may miss many relevant articles.
For example, a PubMed search for DVT, the common abbreviation for deep vein thrombosis, retrieves only about 6,000 results. A search for deep vein thrombosis retrieves over 10 times that number.
To facilitate automatic term mapping:
- Avoid abbreviations. PubMed does not translate many common abbreviations, including UTI, STD, or DVT, into MeSH. It's best to spell out the words in your search.
- Check your spelling. PubMed has a spell checker, but it doesn't catch everything.
Check the Search Details Box
In this screen shot of a Search Details box, we see that PubMed correctly recognized "deep vein thrombosis" as an entry term for "venous thrombosis" and mapped the search appropriately.