Systematic literature searching

• Use key papers which you have already identified – are they actually present in this database? And if so does your search strategy retrieve them?
• If your key papers have not been retrieved, look at them carefully to identify additional keywords or subject headings to incorporate into your search.
• Be aware: finding all of your papers doesn't necessarily mean you have a good search. It's just an indication of progress.  

Practical tip: Make a 'test set'

In each database, papers will have an ID number, e.g. a Pubmed ID (PMID).

1. Build up a list of the IDs of your key papers.
2. Search for these, combining with OR to get your 'test set'
3. Combine your full search with the test set using AND
4. If the number of results stays the same as your test set, your search retrieves all papers
5. If the number of results is lower than your test set, your search is missing key papers.

Asking other members of your research team to review the search strategy and results before you start screening is useful for picking up on obvious mistakes such as typing errors or misapplied Boolean logic. Colleagues might ask you questions about your decisions which will help you to reflect on your approach and any omissions or weaknesses in your search.

Useful questions to ask yourself:

• Does my search retrieve the key papers that I already know about and wish to include?
• Why have I applied the limits that I have, e.g. date?
• What impact could the limits have on my search? 
• What impact do my decisions about which database fields to search have on my search? e.g. All fields versus Title/Abstract
• Have I used keyword searching and subject heading searching where available? Why/why not?