What is screening?
You've now conducted all your searches and have a lot of papers to go through - some of them you can already see don't really answer your research question.
Screening is the process where you select the studies from the literature which will be included in your review. Each individual article must be assessed to see if it meets the inclusion criteria you've set out in your protocol.
PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) has been described before, but as it is fast becoming a standard for reporting of systematic reviews and meta-analyses (and other review types), it's worth mentioning again here. It includes a procedural checklist, and a range of flow diagrams to illustrate the screening process.
Consider also using one or more of the tools described on the Managing Your Review page of this guide:
After you have run your search in the chosen databases, you will need to remove duplicates from your complete list of results (this can be done in EndNote). Then begins the work of going through the results to find the most rigorous and appropriate to help answer your research question.
The reviewing team
To reduce the risk of selection errors or bias, you should have at least two reviewers (including yourself) to screen your search results. Your reviewers should:
Two stages of screening
1. Title/Abstract screen: This is when the titles and abstracts of each article are scanned to remove obviously irrelevant studies. At this stage you may not need to provide a justification for your exclusions, and some papers may remain undecided. Think about your review question. For example:
2. Full text screen: Here you need to locate the full text of the non-excluded studies and carefully examine each one for compliance with your eligibility criteria. Now the reviewers should document the reason for excluding any articles.
Your review paper should provide details of how many reviewers screened the articles and how any disagreements were resolved. This is usually in the methods section.