Covidence and Rayyan are screening platforms that record every decision you make during title, abstract, and full-text review, which means the numbers your PRISMA flow diagram needs are already sitting in your project. The task is knowing which figure in each tool maps to which PRISMA box, then transferring those counts into a clean diagram that reconciles top to bottom. Covidence can generate a PRISMA-style summary directly, while Rayyan reports the underlying counts you assemble into the figure yourself. You can take the numbers from either tool and build the final diagram in our free PRISMA 2020 flow diagram generator, and this guide shows exactly where to find each count.
The reason researchers get stuck here is a mapping problem, not a data problem. The platforms track everything, but their labels do not always match PRISMA's wording, so it is easy to put a number in the wrong box.
Why the Screening Tool Holds Your Diagram
A PRISMA flow diagram is a record of screening decisions: how many records you imported, how many duplicates you removed, how many you excluded at each stage, and why. Covidence and Rayyan exist precisely to log those decisions, so by the time screening is complete the counts are determined. Your job is transcription and reconciliation, not recounting. The discipline of making those decisions reliably in the first place, through dual independent screening, is covered in our guide to systematic review screening best practices, and clean screening is what makes the exported counts trustworthy.
Mapping Covidence to PRISMA Boxes
Covidence is built around the PRISMA workflow, so its stages line up closely. As you work through a review, the relevant counts appear at each stage:
- Import: the records you brought in from each database, which feed the identification box. Record the per-database counts before Covidence deduplicates so identification is defensible.
- Duplicates removed: Covidence flags duplicates on import; this number is your records-removed-before-screening count. Capture it as a standalone figure rather than letting the tool silently fold it into a post-deduplication total, because the records-removed box must show both the identified count and the number removed. Our guide to documenting the deduplication count defensibly explains why reviewers expect that removal step to be reported explicitly and reproducibly.
- Title and abstract screening: the counts screened and excluded map directly to the screening phase.
- Full-text review: studies assessed and the exclusion reasons you assigned map to the eligibility phase, and Covidence lets you tag each exclusion with a reason.
- Included: the studies that pass full-text review.
Covidence can produce a PRISMA diagram summary itself, but it is still worth verifying the chain reconciles and rebuilding the figure in a generator if you want full control over layout and export.
Mapping Rayyan to PRISMA Boxes
Rayyan is more of a screening workspace than an end-to-end PRISMA tool, so you assemble the counts from its filters and labels:
- Imported references give your identification total; capture the per-database counts at import time.
- Rayyan's duplicate detection provides the records removed before screening.
- The included / excluded / maybe decision labels at title and abstract give your screening counts once you resolve the maybes.
- For full-text exclusions, use Rayyan's exclusion reason labels, which you define yourself, so each excluded full text carries a reason that sums to the total.
Because Rayyan does not draw the figure, the natural next step is to enter these counts into a generator. Our step-by-step guide to creating a PRISMA flow diagram walks through each field so the Rayyan counts land in the right boxes.
Covidence vs Rayyan at a Glance
The two tools solve the same problem from different ends. Covidence is a full review-management platform that follows the PRISMA stages and can output a diagram summary; Rayyan is a fast screening workspace that holds the counts but leaves the figure to you. The table contrasts the points that matter when your goal is a clean flow diagram.
| Dimension | Covidence | Rayyan |
|---|---|---|
| PRISMA diagram output | Generates a PRISMA-style summary | No built-in diagram; you assemble the counts |
| Duplicate detection | Automatic on import | Automatic, with manual confirmation |
| Screening model | Structured dual-reviewer stages | Include / Exclude / Maybe labels, blind mode |
| Full-text exclusion reasons | Tagged per study at full-text review | Custom exclusion labels you define |
| Identification counts | Per-import record totals | Per-import reference totals |
| Workflow scope | End-to-end review management | Screening-focused workspace |
| Typical access | Paid, free for Cochrane authors | Free tier plus paid plans |
| Best for the diagram | Counts ready to verify and redraw | Counts you compile, then draw in a generator |
Either way, the counts feed the same four boxes; the difference is only how much assembly you do before drawing.
Reconciling Before You Draw
Whichever tool you use, run the reconciliation check before committing to the figure. Records identified minus duplicates equals records screened; records screened minus excluded equals reports sought; and the full-text exclusion reasons must sum to the total excluded at eligibility. A mismatch usually means a count was read from the wrong filter, an unresolved "maybe" was left out, or duplicates were counted twice. Our explainer on what goes in each PRISMA box defines each unit so you can trace exactly where a stray number came from. Catching the discrepancy in the tool is far easier than explaining it to a reviewer later.
Frequently Asked Questions
Does Covidence make a PRISMA flow diagram?
Covidence can generate a PRISMA-style summary of your screening counts because its workflow follows the PRISMA stages. It is still worth verifying the chain reconciles and, if you want control over layout and export, rebuilding the figure in a dedicated generator from the same counts.
How do I get PRISMA numbers out of Rayyan?
Rayyan does not draw the diagram, but it holds every count you need: imported references, detected duplicates, the included and excluded decision labels at title and abstract, and your full-text exclusion reason labels. Assemble these counts and enter them into a flow diagram generator.
Which Covidence number is the duplicates box?
The duplicates Covidence flags on import are your records-removed-before-screening count. Record your per-database import totals first, because those feed the identification box, and the duplicate count is then subtracted from that total.
Why do my screening-tool counts not add up?
Common causes are reading a number from the wrong filter, leaving an unresolved "maybe" decision out of the totals, or counting duplicates twice. Reconcile the chain in the tool before drawing the figure, checking that each stage subtracts cleanly into the next.
Can I move counts from Covidence or Rayyan into a generator?
Yes. Both tools give you the counts each PRISMA box needs; a generator turns those counts into a publication-ready figure that redraws if a number changes. This is the cleanest way to get a polished, exportable diagram from your screening data.