What are the boxes in the PRISMA 2020 flow diagram?

PRISMA 2020 has three stages: Identification (records from databases and registers, plus records removed before screening such as duplicates), Screening (records screened, records excluded, reports sought for retrieval, reports not retrieved, reports assessed for eligibility, and reports excluded with reasons), and Included (studies included in the review, and the reports of those studies). A separate version adds a column for records found through other methods such as citation searching.

What is the difference between PRISMA 2009 and 2020?

The PRISMA 2020 update separates records identified from databases and registers, adds explicit handling of records removed before screening, distinguishes reports from studies, and includes a version that captures evidence found through other methods like citation chasing and contacting authors. It replaces the older four-phase 2009 diagram.

Why do my PRISMA numbers have to add up?

Reviewers and editors check that the numbers reconcile: records identified minus duplicates equals records screened; records screened minus those excluded equals reports sought; and so on down to the included studies. If the arithmetic does not balance, the diagram signals that records were lost or double-counted, which undermines the review's credibility.

Home › Guides › Systematic Review › PRISMA flow diagram

PRISMA flow diagram, explained

Q: What is a PRISMA flow diagram?

A PRISMA flow diagram is the standard figure that tracks every record through a systematic review — from the number identified in database and register searches, through duplicate removal and screening, to the studies finally included. It makes the selection process transparent and reproducible, and it is expected by most journals reporting systematic reviews.

By Dr. Rafiq Muhammad, MD, PhD · Updated June 2026

The PRISMA flow diagram is the figure that accounts for every record in your systematic review — from the raw database hits down to the handful of studies you include. Its job is transparency: a reader should be able to follow the numbers and see exactly how you got from thousands of records to your final set. The current standard is PRISMA 2020.

The three stages

IDENTIFICATION → records from databases/registers · duplicates & ineligible removed before screening ↓ SCREENING → records screened · records excluded · reports sought for retrieval · not retrieved · reports assessed for eligibility · reports excluded (with reasons) ↓ INCLUDED → studies included in the review (and their reports)

What goes in each box

Identification. Records identified from each database and register, then the records removed before screening — duplicates, automation-flagged ineligibles, and other reasons.
Screening. Records screened (title/abstract), records excluded, reports sought for retrieval, reports not retrieved, reports assessed for eligibility (full text), and reports excluded — each with a reason and count.
Included. The studies that met every eligibility criterion, plus the reports that describe them (one study can have several reports).

PRISMA 2020 also offers a version with a second column for records found by other methods — citation chasing, contacting authors, grey literature — so nothing you found outside the database search is invisible.

The numbers must reconcile

This is what reviewers check first. The arithmetic has to balance down the whole diagram: records identified − removed before screening = records screened; records screened − excluded = reports sought; reports assessed − excluded with reasons = studies included. If a number doesn’t add up, it tells an editor that records were lost or double-counted — and that quietly undermines the credibility of the entire review. Keep a screening log so every excluded report has a reason you can defend.

The free PRISMA Flow Diagram Generator takes your screening counts, checks that they reconcile, and renders a clean PRISMA 2020 diagram you can drop straight into your manuscript.

Get the free Systematic Review toolkit

A PRISMA checklist, a screening-log template, and a protocol starter from The Dissertation Literature Review Sprint. We’ll email you the download link.

Get the Dissertation Literature Review Sprint

Frequently asked questions

What is a PRISMA flow diagram?

The standard figure tracking every record from database hits through screening to the included studies — making selection transparent and reproducible.

What are the boxes?

Three stages: Identification, Screening, and Included — each with explicit counts (and reasons for exclusions at full text).

PRISMA 2009 vs 2020?

2020 separates databases from registers, handles records removed before screening, distinguishes reports from studies, and captures evidence from other methods.

Why must the numbers add up?

Reviewers check the diagram reconciles top to bottom; if it doesn’t, records were lost or double-counted.

Risk of bias & appraisal → Open the PRISMA generator →