Developing the search

Developing a systematic review search is a case of trial and error, and it is rare to get it right first time.  You are attempting to retrieve as many 'on topic' papers as possible whilst also excluding as many irrelevant papers as possible.  

The most important step at this point is to correctly identify the correct search concepts.  It is also good practice to use a set of relevant papers for term selection and testing the search.

For a more detailed look at constructing a systematic search take a look at our introduction to systematic searching modules

Use this form to book an appointment with a SA Health Reference Librarian 

Establishing concepts


Considerations:

Concepts can be confusing to follow,  it is easy to focus too closely on individual terms.

Try using a venn diagram or logic grid to represent each concept.  Each paper that is retrieved needs to contain a term from each circle/column.

As a general rule the Outcome concept is not included in the search

Finding terms



Considerations:

Be very wary about including very general terms that could bring up incorrect results, such as patient or health.

Try testing the search terms in a database.  Restrict to title only and note the relevance to your topic.  

You can use the NOT operator to separate these results from the rest.

Constructing the search




Considerations:

Save the search in a database using a generic, shared user name and password.  This means you can work collaboratively with the your librarian and your other team members throughout the construction of the search.

Ovid databases contain a search history link function, which is even easier.

Testing the search



Considerations:

Testing the search is often neglected and yet super important.  It is especially important to be sure you have retrieved all known and similar papers.  

You can test the search using the test set of relevant papers. 

When removing terms always check what papers are going to be 'lost'