An umbrella review synthesises the findings of multiple existing systematic reviews on a topic, and it reports its selection process with a standard PRISMA flow diagram just like any other review. The structure is unchanged: records are identified, deduplicated, screened, assessed, and included across the same four phases. The one conceptual shift is that the unit flowing through the diagram is a systematic review, not a primary study, so the "included" box at the bottom counts reviews rather than trials. You can build the figure for an umbrella review in our free PRISMA 2020 flow diagram generator, and this guide explains how the familiar boxes are interpreted when your evidence base is itself made of reviews.
The reason an umbrella review needs its own explanation is the unit confusion. Researchers used to counting studies sometimes hesitate over what a "record" and an "included" item mean when the whole point is to sit one level above the primary literature.
What an Umbrella Review Actually Synthesises
A systematic review pools primary studies; an umbrella review, also called an overview of reviews, pools systematic reviews and meta-analyses. It is the right method when a field already has many reviews and the question is no longer "what do the trials show" but "what do the existing reviews collectively conclude, and where do they disagree". That places it at a higher tier of evidence synthesis than a single review, a distinction our comparison of meta-analysis versus systematic review helps situate. The method still demands a comprehensive search and transparent screening, which is why the PRISMA flow diagram remains essential.
How the Flow Diagram Units Shift
Every box behaves as usual, but you read the units one level up:
- Identification: records returned by your search, where each record is a citation to a systematic review or meta-analysis rather than a primary study.
- Records removed before screening: duplicate review citations removed.
- Screening: review records screened at title and abstract and excluded.
- Eligibility: full-text reviews assessed against your criteria, with exclusions grouped by reason.
- Inclusion: the systematic reviews included in the umbrella review.
The arithmetic still reconciles top to bottom exactly as in a primary review, and the studies-versus-reports nuance applies the same way: one included review can be reported across several documents. If the box-by-box logic is unfamiliar, our explainer on what goes in each PRISMA box sets it out, and you simply substitute "review" for "study" at the inclusion stage.
The Overlap Problem Umbrella Reviews Must Report
The distinctive challenge of an umbrella review is primary study overlap: two included reviews may both incorporate the same underlying trial, so the evidence is not as independent as the review count suggests. The flow diagram does not capture this, but your reporting must. Best practice is to add a citation matrix showing which primary studies appear in which included reviews, and to report a measure of overlap so readers can judge how much double-counting exists. The flow diagram proves the reviews were selected transparently; the overlap analysis proves you understood that those reviews share evidence.
Quality Appraisal at the Review Level
Because the included items are reviews, the quality appraisal targets review methodology rather than individual study risk of bias. An umbrella review typically appraises each included review with an instrument designed for that purpose, assessing whether the review's search, selection, and synthesis were sound. This sits beside the flow diagram in the methods, and it matters because an umbrella review inherits the weaknesses of any poorly conducted review it includes. Selecting an appropriate appraisal instrument, and reporting how it was applied, follows the same logic as in a primary review, and our guide to the instruments used to appraise each included review's methodology covers the tools and how to match them to what you are judging. The broader sequence of planning, searching, screening, and appraising is the same as in our walkthrough on how to conduct a systematic review, applied at the level of reviews.
Frequently Asked Questions
Does an umbrella review use a PRISMA flow diagram?
Yes. An umbrella review documents its selection process with a standard PRISMA flow diagram across the same four phases. The difference is that the records flowing through the diagram are systematic reviews and meta-analyses, so the included box counts reviews rather than primary studies.
What is the difference between an umbrella review and a systematic review?
A systematic review synthesises primary studies, while an umbrella review synthesises existing systematic reviews and meta-analyses. The umbrella review sits one tier higher and is used when a field already has many reviews whose conclusions need to be compared and summarised.
How do I count included items in an umbrella review diagram?
Count systematic reviews, not primary studies, at the inclusion stage. Each box reads one level up: records are review citations, and the included box reports the number of reviews that met your eligibility criteria, with any multiple reports of a single review noted separately.
What is primary study overlap and why does it matter?
Overlap occurs when two included reviews both contain the same underlying primary study, so the evidence is partly double-counted. The flow diagram cannot show this, so an umbrella review should add a citation matrix and an overlap measure to make the shared evidence transparent.
How is quality assessed in an umbrella review?
Appraisal targets the methodology of each included review rather than individual primary studies, using an instrument designed to judge whether a review's search, selection, and synthesis were sound. A poorly conducted included review weakens the umbrella review's conclusions.