20 templates
PRISMA 2020 flow templates
Native PRISMA 2020 systematic-review flow — and the only one that checks the arithmetic the standard is made of. A PRISMA figure exists so that every record entering a review is accounted for on its way out, which makes it an arithmetic claim: the engine balances all four stages (identification, screening, retrieval, eligibility), reports the exact count of records that vanished and where, draws the failing stage's arrow in red, and marks a stage it could not balance as a gap naming the missing term rather than passing it.
PRISMA — vaccine effectiveness review
Sample PRISMA 2020 flow for a vaccine effectiveness lit review.
Research & AcademicPRISMA — telehealth mental-health review
Lit-review flow for telehealth-delivered CBT.
Research & AcademicPRISMA — environmental review
Environmental science systematic review template.
Research & AcademicPRISMA — educational technology review
EdTech effectiveness review.
Research & AcademicPRISMA — blank template
Starter PRISMA template with placeholder numbers.
Research & AcademicSystematic Review — the Reference Flow
The one to copy. A review that searched databases, registers and citation lists, removed duplicates and automation-excluded records, screened and retrieved, excluded at eligibility with three named reasons, and distinguishes the reports it included from the studies those reports describe. All four stages balance to the digit, the funnel descends, and every one of the eight checks says what it covered. Reports clean.
Scientific PlotsCochrane Review — Awaiting Classification and Ongoing Studies
A Cochrane-style review, and the reason two extra terms exist in this notation. Cochrane requires two further destinations at full text — reports awaiting classification, and studies still ongoing — and both left the eligibility stage without being excluded. Without a way to state them, a correct Cochrane flow diagram cannot balance, and a checker that reported an error on every Cochrane review would be switched off within a week. Reports clean.
Scientific PlotsUpdated Review — Reports Carried Forward
An update of an earlier review, which is the one place a PRISMA funnel may legitimately widen. Twenty-two reports of studies already included in the previous version enter at the included box without ever being sought this cycle — so `reports_included` exceeds `fulltext_assessed`, and the funnel-descent check would fire on a perfectly correct diagram unless the document could say why. `previous_reports` is how it says so, and the engine then checks the widened arithmetic rather than waiving it. Reports clean.
Scientific PlotsTwo-Arm Review — Databases, Registers and Other Methods
The second of the four flow diagrams PRISMA 2020 ships, and the one this engine could not state at all until it was attacked. Citation searching, websites and organisation pages form a separate right-hand arm: those reports are never de-duplicated against the database yield and never title-screened, so they enter at "Reports sought for retrieval" rather than at the top of the funnel. `other_sought` is how the document says which topology it is using — `other_records` is the other one, for a review that screened both arms together. Stated wrongly, a faithful transcription of a published figure errors, and the only edit that silences the error is a count the review did not screen. Reports clean.
Scientific PlotsTeaching Case — Twelve Reports That Left With No Reason
The reference flow with one number changed, and the reason this engine exists. Twelve reports leave the review between assessment and inclusion with nothing recorded against them — in a figure whose only job is to record it. Every box is present, every count is plausible, every arrow is where it should be, and the figure is identical to the correct one except for two digits in one box. The engine names the twelve and draws the failing arrow in red.
Scientific PlotsTeaching Case — A Stage With Nothing to Balance Against
The half of the discipline that is easier to get wrong. This review never states how many reports it sought for retrieval, so two stages cannot be balanced at all — and the engine reports them as GAPS naming the missing keyword, rather than assuming a number and reporting clean. "430 = 430 if we assume nothing was lost in retrieval" is exactly the sentence a reporting standard exists to prevent. The box is drawn as "n = not stated", never blank and never zero.
Scientific PlotsTeaching Case — More Reports Assessed Than Records Screened
Records only ever leave a review funnel, so every count must be no larger than the one before it. Here 640 reports are assessed for eligibility out of the 380 that were sought — 260 reports appear from nowhere, which no sequence of events could produce. It is decidable with no remainder at all, so it is caught even where the ledger cannot run, and it reads as a perfectly ordinary diagram: the boxes are the same size whatever numbers are in them.
Scientific PlotsTeaching Case — Counts With No Account
The eligibility box exists so that every report excluded at full text has a reason written against it. This review lists one exclusion with no reason at all and repeats another reason on two separate lines — the figure draws four lines and a reader counts four reasons, when the document describes three. Both are reporting-practice findings rather than contradictions, so both are warnings: the arithmetic itself balances exactly.
Scientific PlotsMulti-Experiment Papers — Two Boxes, One of Them Larger
PRISMA 2020 draws two numbers at the bottom of the funnel because one study is often reported in several papers — and the usual direction is many reports per study. This review runs the other way: 92 studies from 74 reports, because an "Experiment 1 / Experiment 2" paper carries two independent studies, which in psychology and education is routine rather than remarkable. The engine says so as INFO and the document reports CLEAN. It used to warn, and a warning here punishes honesty: the only way to clear it was to omit `studies_included` and report less than the truth. The arithmetic is untouched either way — the eligibility stage is reckoned in reports.
Scientific PlotsTeaching Case — A Review That Included Nothing
An empty review is a publishable result and its flow diagram is a correct diagram: every one of the 96 reports assessed was excluded, with reasons, and the arithmetic balances exactly. The engine says so out loud rather than staying silent, because a figure ending in `n = 0` reads as an unfinished diagram — and a checker that returns nothing on it would leave the author unsure whether it had run. Reports clean.
Scientific PlotsTeaching Case — The Same Count Stated Twice
A number edited in one place and left in another. The first statement is the one drawn and the one checked, and the second is reported rather than silently preferred — the previous renderer let the LAST statement win, so a document like this was drawn from a number the reader had already scrolled past, with nothing anywhere saying which of the two it used. Which of the two is right is not something an engine can decide, so it does not try.
Scientific PlotsSystematic Review Screening (PRISMA)
PRISMA flow for a systematic review.
Research & AcademicML in Radiology Review (PRISMA)
PRISMA flow for a deep-learning literature review.
Research & AcademicLow-carb Diet Review (PRISMA)
PRISMA flow for a diet effectiveness review.
Research & AcademicFlipped Classroom Review (PRISMA)
PRISMA flow for an education review.
Research & Academic