You should break down your question into its constituent concepts (like we did with the PICO method outlined in Tip 3). Consider any synonyms and related terms for these key concepts in order to develop a search strategy. More information on search strategies can be found on our "Searching for information" guide:

Medical research example

Most published systematic reviews include the full strategies used for each database, often in the appendices. You may be able to combine sections of strategies used in other reviews to form the basis of a good search strategy for your review.

Search free text using synonyms and truncation, and search in the title and abstract fields. All of these steps will increase the number of records retrieved.

Test your draft strategy using papers that you are already aware of and would expect to find. If your strategy does not identify them, try to find out the reason, and modify your search strategy accordingly.

Develop the strategy in one database and then translate it for other databases. If the databases use different controlled vocabularies you should adapt your descriptors but you should use the same free text terms.

You can limit your search strategy to different types of studies using search "filters". These are, effectively, pre-canned search strategies. The ISSG Search Filter Resource lists a variety of these, though it can be difficult to use and does not provide access to the actual filters.

Try to reach a balance with the number of records retrieved by your search strategy. Your systematic review must be exhaustive but you also have to be realistic about the time and number of records that you are able to deal with.

When setting your limits, make sure you refer back to your protocol. Ensure that all your filter limits are applied with good reason, as established in your protocol.