A clear methodology for finding studies is an essential element of a systematic review. Your approach needs to be well documented (transparent) and as replicable as possible. This is the stage of a systematic review where consultation with a librarian is valuable. We have extensive experience with the databases most often used in reviews. We know what they cover (and miss) and how they vary. Through conversation and exploratory searching, librarians can help you find the right balance between search sensitivity and specificity, and they can assist in finalizing a valid search protocol.
Searching is an iterative process, and you will already have done some exploratory searching before you get to this point in the review process. During protocol development, you will have identified relevant databases, search terms, and studies. These will help you build your systematic search strategy that you will report out in your methods section (the more detailed and transparent you are about this process, the better, so it helps to keep track).
Don't forget, your librarians are here to help!
Compile identified terms and databases
Keep key concepts separate from one another rather than searching in phrases.
Connect synonyms for each concept with the Boolean operator OR
Population: (Latinx OR Hispanic OR Mexican OR Chicano)
Problem/Issue: (acculturative stress OR cultural assimilation OR marginalization OR discrimination)
Intervention: (community engagement OR community involvement OR civic engagement)
Test your search in your identified databases
Focus first on the number of results. Is this what you expected? Is this more search results than you expected or fewer?
Test your search strategy with the articles you identified when formulating your research question. Do those articles show up in your search results? If not, what is missing from your search strategy that might need to be added?
Tweak your search strategy based on what you find
Scan titles and abstracts in your search results to identify terms to add to or remove from your search strategy.
Try adding or removing quotation marks from phrases in your search strategy.
Test again, test a LOT
Don't get discouraged with this iterative process, especially if you are a beginner. Testing your search strategy can take many revisions, but there's no real right or wrong answer. What's important is that you feel relatively confident that you are capturing the relevant literature.
Adapt search strategies for different databases
Truncations and wildcards, and even Boolean operators can vary between databases, so you'll want to adapt/translate your search strategy accordingly. Use the Help link in the databases to identify the wildcards and truncations they use.
Conduct final search
Conduct searches separately in each database.
Keep track of the date, exact strategy used, and number of search results for each database search. You'll want to report this later.
Test your search methods with articles you know should match. Simply add the article title to the last bar of your search and run it again. If that article doesn't appear, find it in the database. Then look over the words in the abstract and subject terms to determine why you didn't find it. Adjust your search accordingly.
Document your search! Take advantage of database features to save searches. Keep a record of all your final searches. Keep a PRISMA flow diagram of your work.
For a systematic review, you want to find all possible permutations for your search terms to make sure you don't miss anything. Beyond scouring existing reviews and encyclopedia articles, you also want to make sure you find the appropriate controlled vocabulary (also called subject terms) for your concepts.
In PubMed, they are called MeSH Terms (Medical Subject Heading Terms):
In EBSCO, you can check to see if there is a Thesaurus or Subject Terms index for your chosen database(s).
In Embase, the subject terms index is called Emtree.
Use an organizational document or spreadsheet to list your keywords and subject terms as you brainstorm. Here is one or make your own:
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 Generic License.