AND → Narrows search results
This connects two words by saying that BOTH of the terms are necessary to your search. Using AND will narrow down your search results. In many databases (such as PubMed), AND is implied by default.
NOT → Narrows search results by excluding unwanted terms
NOT excludes results that contain a specific term. Be careful using NOT -- you could potentially remove relevant results using this. In general, you should only use NOT in response a a problem with your search strategy, such as removing articles that are in a completely different domain. If you were getting a lot of results about studies on rats, you could also use (NOT rats) to ensure that nothing about rats show up in your results.
"Quotes" → Narrows search to an exact phrase
Quotations can be used to search for a specific phrase. For example, if you search "infant development," you will only retrieve articles that are phrased EXACTLY like that. So you would not find any articles that say, "the development of infants."
Proximity N# → Narrows search by making sure two or more words occur within a certain number of words from each other
OR → Broadens search results
OR broadens your search by telling the database that you are okay with multiple words that identify your concept. Anytime you are using synonyms, you will need to use OR to let the database know you are okay with either term. For instance:
brain development AND (noise OR sound)
In this search, your search results would include articles that discussed brain development, as well as either noise or sound.
Truncation* → Broadens search to include various endings and spellings of a word (NOTE: Using truncation in PubMed turns of automatic term mapping, so you may actually get fewer results! Truncation works best in other database.
Boolean operators are a way to construct a computer searchable question. Use Boolean operators (AND, OR, NOT) to combine or exclude search terms.
PubMed is unique in that it does something called "Automatic Term Mapping" (or ATM). When you do a keyword search, it will automatically add synonyms and MeSH terms to your search.
This greatly expands your search, and it means that you don't always need to know the MeSH term, as PubMed will try to add the MeSH term to your search for you.
BUT...
Using truncation, quotation marks and proximity searching will turn off PubMed's automatic term mapping!
So if you are using any of these methods in your keyword search in PubMed, you might want to try a search without these methods to compare the two strategies and see how PubMed's ATM might change your search results.
While PubMed automatically maps your search terms to subject headings to get you the MOST results, CINAHL works in the reverse: it is intentionally narrow, and if you want to search subject headings, you must intentionally add them to your search.
You may have noticed the use of parentheses in our previous searches. For databases like Pubmed where there is only a single search box, you may need to use parentheses in conjunction with your boolean operators (especially OR).
Take a look at this search:
Neonatal Intensive Care Unit AND Sound OR Noise
177,813 results
This search retrieved a TON of results!!! This is because it is searching for articles about Neonatal Intensive Care Unit and sound. OR... you might find articles that use the keyword sound, even if the article has nothing to do with the NICU.
To fix this, you can use parentheses to "nest" your synonyms. Just like a mathematical formula, this will tell the database to process the request inside the parentheses first.
Neonatal Intensive Care Unit AND (Sound OR Noise)
640 results
This new search will bring back every article in the database that contains the words Neonatal Intensive Care Unit and either the word sound or noise.
If you are searching a database like CINAHL, you can simply put each concept on a separate line, with synonyms all in the same box:
But if you are searching in PubMed, you'll notice there are not separate search lines -- the default search is a single search box. So you have a few options for doing complex searches in PubMed:
For example, let's say you are searching for information about spinal manipulation and exercise to treat lumbar pain.
If you want research that compares spinal manipulation and exercise, you want to connect your search terms with AND: lumbar pain AND spinal manipulation AND exercise. Using AND will retrieve all three search terms, which is important if you are looking for research that compares those treatment options.
If you want research on either spinal manipulation or exercise, you can use an OR: lumbar pain AND (spinal manipulation OR exercise). You might do this if you aren't comparing the two treatment options; rather, you want studies that include either of those types of treatment.
NOTE that I put parentheses around spinal manipulation and exercise. This is because databases read your search from left to right, like a mathematical formula. If you don't include the parenthesis, you might find information about lumbar pain and spinal manipulation... or you might just find information about exercise that has nothing to do with back pain!!
A good way to think about this is: What are your MUST-HAVE concepts? For this question, there are three: the problem (lumbar pain) and your two treatment options that you are comparing. You'll want to connect each MUST-HAVE concept with an AND. You can add alternative keywords to your search, but just make sure you are clumping together your concepts using parentheses. For example:
For this next example, you have only TWO concepts: the problem (lumbar pain) and the intervention (either exercise or spinal manipulation). You are not wanting to compare the two treatment options, but just see information about either of them -- in this case, you would nest those treatment options together with the boolean operator OR. Adding alternative keywords might look like this:
(lumbar pain OR low back pain) AND (exercise OR spinal manipulation OR "manual therapy" OR chiropractic)
You can also try combining keywords and subject headings as synonyms. For example, if you were searching for ED patients, you might use the term Emergency Department. But you could also search Emergency Room. If you are searching CINAHL, the CINAHL subject heading for emergency departments is Emergency Service; for Pubmed, the MESH term is Emergency Service, Hospital.
So if you were searching CINAHL, you could search:
MH "Emergency Service" OR "emergency room" OR "emergency department"
If you were searching Pubmed, your search could be:
"Emergency Service, Hospital"[Mesh] OR "emergency room" OR "emergency department"
You can do this for each PICOT concept you are using. Again, each PICOT concept should be connected using the Boolean Operator AND, while alternative search terms / synonyms are separated with an OR. If you are creating one long search string (rather than entering each PICOT component on a separate search line), make sure to nestle your concepts using parenthesis:
("Emergency Service, Hospital"[Mesh] OR "emergency room" OR "emergency department") AND ("Patient Compliance"[Mesh] OR adherence OR compliance)
Proximity searching allows you to search for words that appear within a certain proximity of one another. This is generally done using the letter n, followed by the maximum number of word distance between the two terms.
For example, in CINAHL, cloning N3 human would retrieve cloning of humans, human cloning, human reproductive cloning, etc.
The PubMed equivalent of this search would be "cloning human"[tiab:~3]