What Covidence is
Covidence is a web platform that manages a systematic review from imported search results to a completed extraction table. It covers title-and-abstract screening, full-text assessment with recorded exclusion reasons, risk-of-bias appraisal, and dual-reviewer data extraction — and it maintains the PRISMA flow diagram automatically as records move between stages.
The distinguishing fact is institutional rather than technical: Covidence is Cochrane’s designated production platform for its own reviews. That matters because the tool’s shape follows Cochrane methodology rather than a general idea of “literature management”. Stages are sequential, dual review is the default rather than an option, and exclusion reasons are structured data. If your review is registered on PROSPERO and expects to satisfy a methodological reviewer, the platform’s rigidity is the product, not a flaw in it.
Why researchers use it
- One system for the whole review — no handover between a screening tool, a spreadsheet for extraction, and a hand-drawn flow diagram.
- The PRISMA diagram maintains itself — record counts at every stage are derived from your actual decisions rather than reconstructed from memory at write-up.
- Dual extraction with reconciliation — two reviewers extract independently into the same template and Covidence surfaces every field where they differ.
- Risk-of-bias tools built in — the standard appraisal instruments as structured forms, attached to the studies they assess.
- Audit trail by default — who decided what, when, and why, exportable when a journal or an auditor asks.
Where it fits in a research workflow
Covidence takes over at the point where a search ends. Upstream is your database searching and deduplication — done in EndNote or Zotero, and Covidence deliberately expects records to arrive already deduplicated. Downstream is analysis and writing: extraction tables export to CSV for meta-analysis in R or Stata, and the flow diagram exports for your manuscript.
The tools it replaces are the awkward middle of the review: the shared screening spreadsheet, the second spreadsheet for full-text decisions, the third for extraction, and the manual arithmetic connecting them. What it does not replace are search construction, deduplication, meta-analysis, or reference formatting, and it is not the right home for primary data collection — that is REDCap’s job.
Getting started
Half a day of setup, and one decision that must be right on day one.
- Check institutional access first. Many university libraries hold a Covidence subscription that covers unlimited reviews for their researchers, and paying per review when a site licence exists is the standard expensive mistake.
- Deduplicate before importing. Covidence flags some duplicates on import, but the clean set should be produced upstream so your PRISMA “records removed” number has a single, defensible source.
- Configure your extraction template before anyone extracts a single study. This is the decision that must be right on day one — adding a field after twenty studies are extracted means going back through all twenty.
- The step people skip: run two reviewers through three studies as a pilot and compare. You will find out that “sample size” means different things to different people while it still costs you twenty minutes.
Covidence vs the alternatives
| Alternative | Does it better | Pick it if |
|---|---|---|
| Rayyan | Free collaborative blind screening | You need screening only, with no budget |
| ASReview | Peer-reviewed active learning; runs entirely offline | The pile is enormous and prioritisation is the bottleneck |
| REDCap | Institution-hosted forms with full compliance controls | You are collecting primary data, not extracting published studies |
| Spreadsheets | Flexibility, and nothing else | Your review is small, single-screener, and unregistered |
Pick Covidence when the review is registered, funded, or Cochrane-bound and multiple people will touch it. Pick Rayyan when the honest answer is that you need screening and have no money.
Cost, licensing, and your data
Paid, with two routes: an institutional subscription that covers a university’s researchers, or per-review pricing for individuals — the latter runs to hundreds of dollars for a single review, which is the number that sends most solo researchers to Rayyan. Cochrane review groups have their own arrangements. Always check with your library before paying personally.
Your review lives on Covidence’s servers. For extraction of published studies this is routine, but note what you are uploading at the extraction stage: full-text PDFs, sometimes under publisher licences that restrict redistribution, and extracted outcome data that may be under embargo if the review supports an unpublished trial or a competitive grant. Export your data regularly rather than at the end — a review that exists only inside a subscription you may not renew is a fragile place to keep three years of work.
The honest review
Strengths. Covidence makes the methodologically correct path the path of least resistance. Dual screening, structured exclusion reasons, independent extraction with reconciliation, and self-maintaining PRISMA counts are all things a disciplined team could do in spreadsheets and almost never does consistently. Buying the platform is, in practice, buying the discipline — and for a multi-year review with rotating student reviewers, that is worth more than any single feature.
Limitations. The price is genuinely prohibitive without institutional access, and per-review pricing punishes exactly the early-career researchers most likely to be doing the screening. The rigidity that serves a Cochrane intervention review fights you in a scoping review, a rapid review, or anything with an iterative search. Extraction templates are inflexible once populated. The interface is functional rather than fast, and large reviews feel slow. And exporting mid-review gives you your data but not your workflow — the structure that made it valuable does not travel.
Verdict. Adopt it if your review is registered, has more than one reviewer, and your institution holds a licence. Skip it for scoping and rapid reviews, for solo work, and whenever you would be paying per review from a personal budget. The condition that flips the answer is entirely financial: at zero marginal cost it is the obvious choice, and at list price the calculation rarely survives contact with a student stipend.
Common questions
Is Covidence free?
No. It is available through institutional subscriptions that many university libraries hold, and through per-review pricing for individuals that runs into the hundreds of dollars. Cochrane review authors have separate arrangements. Check your library first.
Does Covidence produce a PRISMA flow diagram?
Yes, and this is one of its strongest features. Record counts at every stage derive from the decisions your team actually recorded, so the diagram stays consistent with your data instead of being reconstructed by hand at write-up.
Covidence or Rayyan?
Covidence if the review is funded or institutionally licensed and you want screening, full text, appraisal, extraction, and PRISMA in one audited system. Rayyan if you need free collaborative screening and will run extraction elsewhere.