Systematic Reviews: How-To

All articles included in your systematic review must undergo two rounds of screening:

• Title/abstract: typically screened by two individuals though single screening is permitted; during this phase it is common for 90% of articles retrieved by the search to be excluded
• Full-text: the full manuscripts are reviewed by two screeners and the result of their decisions determines whether the article is included in the review

In many ways, the search strategy created by the librarian operationalizes the inclusion criteria. The screening process does the same for the exclusion criteria. In the title/abstract phase, it is better to be overly permissive rather than overly conservative. Often, exclusion criteria are made more stringent during the full-text screening phase. 

Before you begin:

• The librarian will provide an overview of the screening software.
• The project manager/team lead should meet with all screeners to discuss how many articles everyone is expected to screen each week, and how and who will resolve conflicts for each phase. 

Project planning:

• While there are more items to screen in title/abstract, teams should allocate the same or more time for the full-text screening stage.
• This is due to the level of reading required and the amount of time it can take to retrieve all full-texts. It is expected that a portion, sometimes a significant amount, will need to be requested via interlibrary loan. 

Zero results after screening?

If you are left with zero qualifying results after your screening process, you should reflect on whether or not publishing then results as an "empty" review would be useful to highlight an evidence gap. A report by Cochrane suggests revisiting the importance of the review questions in consultation with key stakeholders, but recognizes that in some cases, "publishing these empty reviews may help to highlight gaps in the evidence base." (access summary of empty reviews and full report)