What is a PRISMA Flow Diagram?
A PRISMA Flow Diagram is a visual representation of the stages in an evidence synthesis. Using a PRISMA Flow diagram you can show how many records and reports were included and excluded in each step of your evidence synthesis. No matter what type of evidence synthesis you are doing, you should use the PRISMA flow diagram to report the number of studies identified, screened, and included in your final review.
Terms to Know
Record:
A bibliographic entry from a database. This would include things like the title and abstract of a potential article.
Report:
The published document associated with a study. This could take the form of a full-text article.
Study:
The research investigation itself. It is the original research that is described in a report. It is important to remember that one study can have multiple reports associated with it.
- PRISMA Flow DiagramThe flow diagram depicts the flow of information through the different phases of a systematic review. It maps out the number of records identified, included and excluded, and the reasons for exclusions. Different templates are available depending on the type of review (new or updated) and sources used to identify studies.
Filling Out a PRISMA Flow Diagram
- Record the number of citations that you pulled from your database search. This should be the total from all the databases you searched. For example, I pulled 25 articles from PubMed, 28 from Scopus, and 10 from Web of Science. I add it all together and report that I identified 63 records.
- After you've gathered all the records you've identified, you need to remove the duplicates. You can do this easily in both EndNote and Zotero. Make sure to record the number of duplicates that you removed. If I removed 8 duplicates I would put that here.
- After removing duplicates you can move on to your article screening. Here you should report the number of articles that are included in your screening. For my example I had 63 records, but I removed 8 duplicates. That means that I have 55 records that are included in my screening.
- Make sure that as you screen you are checking that the records match your inclusion/exclusion criteria. If you need to exclude an article be sure to keep track of how many you have excluded and the reason for exclusion. I am screening my 55 records and I have decided to exclude 28. I report this total as well as my reasoning i.e. Not peer-reviewed= 8, Nonhuman subjects= 14, and Non English language= 6.
- In this context "reports" means the full-text of a study or article. You need to report how many records you looked for the full-text of. Additionally you need to report how many you were unable to find. After excluding 28 records in the last step I sought the full text of 27 records and I was unable to retrieve 7 of them.
- Now that you have the full-text you can move on to the data extraction. This is where you should report the number of reports that you have included in your data extraction. In the last step I was unable to find 7 of my 27 records, so here I'll put that I assessed 20 reports for eligibility.
- Similar to the article screening you need to report how many reports are excluded during your data extraction. Don't forget you also need to report your reason for exclusion. Of my 20 reports I am excluding 8. I put this number down as my total and provide my reasons i.e Unrelated to topic=5 and Participants outside of age range= 3.
- Once you have completed your data extraction you should record the number of studies or reports that you are including in your review. Studies represent the individual research investigation and reports are the published documents describing those studies. For my example I have 12 reports. Of these reports three of them are describing the same study. As a result I would record that I am including 10 studies and 12 reports.