Skip to Main Content

Systematic Reviews

This guide describes all steps involved in the conduct of a systematic review

Study selection

Study selection (also known as screening) is the process by which records identified in the search are assessed for inclusion. Objectively and systematically applying the pre-specified eligibility criteria to each record in order to determine whether the article should be included helps avoid selection bias.

Take note on the following points before screening search results:

  • Duplicate records of your search results should be removed before starting the screening process.
  • At least two team members should work independently on screening the papers to reduce the risk of missing papers or bias in the selection process.
  • Papers must be selected based on the research question and eligibility inclusion & exclusion criteria defined in the planning process.

Two stages of screening of papers

Title and abstract screening:

Each reviewer will need to read titles and abstracts of the papers to see if they match the inclusion & exclusion criteria. This is done separately by each reviewer. After both reviewers have evaluated the studies, votes are compared.

Full-text screening:

In the second round of screening, members evaluate the remaining studies by reading the full-text articles. Members must record the reasons for excluding studies based on the pre-determined inclusion & exclusion criteria.

Avoid selection bias:

To prevent bias, study selection is completed by two independent reviewers with a third person available for resolving conflicts. Alternatively, disagreement of study selection may be resolved by discussion between the reviewers.

Full text can be retrieved by EndNote. How to use EndNote to retrieve full texts is explained in detail in our Endnote LibGuide. If you fail to retrieve all full texts with EndNote, try How to get the pdf
When this is still not enough, you can see if your institution can borrow the article from other libraries via the Inter-library loans (ILL) program against a small fee. It should be reported if a full-text cannot be obtained.

For every article from which the full-text is screened, a reason should be listed for inclusion or exclusion.

Helpful tools within the Screening Process

Tool   Main characteristics
Rayyan Free / Subscription  Work remotely and collaborate with a distributed team
Covidence Subscription Primary screening and data extraction tool for Cochrane authors conducting standard intervention reviews
ASReview Free / Subscription Based on machine learning