“Systematic review software” covers two different product families that a single search often conflates: full-workflow platforms built to carry a review from search import through risk-of-bias appraisal, and narrower tools focused on one stage (usually screening). Knowing which family a candidate tool belongs to matters more than any feature checklist, because the PRISMA 2020-reportable stages of a review – deduplicated import, dual independent screening, structured extraction, and quality appraisal – are what a review protocol actually has to document, and a tool that only covers one stage still leaves the rest to spreadsheets and manual tracking.
This guide surveys the landscape of tools used for that pipeline – commercial workbenches (Covidence, DistillerSR, EPPI-Reviewer), free and open-source options (Rayyan, ASReview, RevMan, Abstrackr) – and how to choose between them by review type and budget. Several of these tools also ship active-learning or machine-learning features; for a deep dive specifically on the AI-assisted screening, extraction, and risk-of-bias capabilities inside this pipeline, see AI Tools for Systematic Literature Reviews. If you’re choosing between a systematic review and a different evidence-synthesis format in the first place, start with Literature Review vs. Systematic Review vs. Scoping Review.
What a systematic review tool actually needs to support
PRISMA 2020 does not mandate any specific software, but its reporting checklist and flow diagram assume a documented, reproducible sequence of stages, and the Cochrane Handbook adds the methodological detail behind each one:
- Search and import. Running searches across multiple databases and grey-literature sources, then importing the combined result set and removing duplicate records before screening starts.
- Title/abstract screening. Two or more reviewers independently assess each record against pre-specified eligibility criteria, with conflicts resolved by discussion or a third reviewer – the dual-screening approach both PRISMA and Cochrane methodology expect.
- Full-text screening. The same independent, conflict-resolved process applied to the full-text reports of records that passed title/abstract screening.
- Data extraction. Pulling defined fields (population, intervention, outcomes, results, study design) from each included study into a structured, consistent form rather than free-text notes.
- Risk-of-bias / quality appraisal. Assessing each included study’s methodological quality using a recognized tool (e.g. Cochrane’s RoB 2, ROBINS-I, or a checklist appropriate to the study designs included).
- Synthesis and reporting. Narrative or statistical synthesis (meta-analysis) of the extracted data, and generation of the PRISMA flow diagram documenting how many records moved through each stage.
No single free tool covers every stage equally well, which is why most real review teams combine two or three tools rather than relying on one.
Full-workflow platforms
Covidence
Covidence is a web-based review workbench covering import/deduplication, title/abstract and full-text screening, data extraction forms, and risk-of-bias assessment in one connected workflow. It is the tool Cochrane recommends for authors preparing new Cochrane reviews, and it exports directly into RevMan/RevMan Web for the statistical synthesis stage. It’s a paid, per-review or institutional-subscription product – check current pricing directly, as this changes.
DistillerSR
DistillerSR (Evidence Partners) is positioned toward high-volume and regulated review work: literature monitoring, pharmacovigilance, and systematic reviews submitted as part of regulatory dossiers, where audit trails, validated workflows, and configurable multi-level screening forms matter as much as speed. Like Covidence, it builds data extraction around structured, configurable forms rather than free-text notes, and it supports machine-learning-assisted prioritization during screening. It is a commercial, enterprise-oriented product; pricing is quote-based.
EPPI-Reviewer
EPPI-Reviewer is developed by the EPPI-Centre at UCL and is a web-based platform covering the full pipeline from screening through data extraction, quality appraisal, and synthesis, including text-mining and machine-learning tools that assist screening prioritization and thematic coding. It is unusual among the tools here in its native support for mixed-methods reviews, qualitative evidence synthesis, and realist reviews, alongside conventional quantitative systematic reviews. EPPI-Reviewer runs on a subscription fee model administered by the EPPI-Centre rather than a per-seat commercial license.
Free and open-source options
Rayyan
Rayyan is a free-to-start screening platform: it imports records from reference managers or database exports, flags duplicates, and uses a machine-learning relevance ranking that improves as reviewers mark records included or excluded. It supports blinded, independent dual screening with conflict resolution and can generate PRISMA flow-diagram counts from the screening record, but it does not natively cover structured data extraction or risk-of-bias appraisal – teams typically pair it with a separate extraction tool or spreadsheet template.
ASReview
ASReview is a free, open-source active-learning screening tool built by Utrecht University’s AI-aided Knowledge Discovery lab. Unlike a fixed relevance ranking generated once, it retrains its model after every screening decision, aiming to surface the most-likely-relevant remaining records earliest in a large candidate set. Like Rayyan, it is a screening-stage tool rather than a full-workflow platform.
RevMan (Review Manager)
RevMan is Cochrane’s own free software for preparing and maintaining Cochrane reviews. Its core role is downstream of screening and extraction: entering extracted data, running meta-analyses, generating forest plots, and assembling the formatted review document to Cochrane’s editorial specification. Non-Cochrane reviews also use RevMan purely for its meta-analysis and forest-plot output, often after screening/extraction in Covidence or another tool.
Abstrackr
Abstrackr, developed at Brown University’s Center for Evidence Synthesis in Health, is a free, web-based screening tool that uses machine-learning prioritization similar in concept to Rayyan and ASReview. It has seen less active development in recent years than the other tools on this list, which is worth checking directly before committing a review team to it.
Comparison at a glance
| Tool | Covers | Cost model | Best fit |
|---|---|---|---|
| Covidence | Screening, extraction, risk of bias | Paid (per review/institutional) | Cochrane and most health-sciences systematic reviews |
| DistillerSR | Screening, extraction, risk of bias, audit trail | Paid (enterprise, quote-based) | Regulatory submissions, high-volume/pharma review teams |
| EPPI-Reviewer | Screening through synthesis, incl. qualitative/mixed-methods | Subscription (EPPI-Centre) | Complex reviews needing qualitative or mixed-methods synthesis |
| Rayyan | Screening (dedup, dual screen, ML ranking) | Free to start; paid tiers for teams/features | Budget-limited teams, students, screening-only needs |
| ASReview | Screening (active-learning ranking) | Free, open source | Large candidate sets where minimizing screening workload matters most |
| RevMan | Meta-analysis, forest plots, review authoring | Free (Cochrane) | Statistical synthesis stage, especially Cochrane reviews |
| Abstrackr | Screening (ML-assisted) | Free | Low-budget screening; verify current maintenance status first |
Choosing a tool for your review
- Submitting to Cochrane: Covidence for screening/extraction, exported into RevMan for synthesis, is the path Cochrane itself trains authors on.
- Regulatory or pharmacovigilance submission: DistillerSR’s validated workflows and audit trail are built for that context specifically.
- Mixed-methods, qualitative, or realist synthesis: EPPI-Reviewer natively supports methods the other platforms don’t.
- No budget, or a single-reviewer/student project: Rayyan or ASReview for screening, a spreadsheet or free extraction template for data collection, and RevMan if any meta-analysis is needed.
- Large search results (thousands of records) where screening time is the bottleneck: the active-learning approach in ASReview, or the AI-assisted ranking in Rayyan, Covidence, and DistillerSR – covered in more depth, including independent accuracy studies, in AI Tools for Systematic Literature Reviews.
Whichever combination you choose, register the review’s protocol (e.g. via PROSPERO) and confirm the eligibility criteria and planned systematic review methodology before screening begins – the tool supports the protocol, it doesn’t substitute for one.
Frequently asked questions
Is there a completely free way to run a systematic review?
Yes, though it means combining tools: Rayyan or ASReview for screening (both free), a structured extraction template or spreadsheet in place of a paid extraction module, and RevMan (free from Cochrane) for meta-analysis if the review includes one. It takes more manual assembly than a single paid workbench but is a genuinely workable path for a student or low-budget review.
What software does Cochrane require for a Cochrane review?
Cochrane doesn’t mandate a single tool for the whole pipeline, but it trains and recommends Covidence for screening and extraction, and RevMan (or RevMan Web) is required for the statistical analysis and final review document, since Cochrane’s editorial and publishing workflow is built around RevMan’s output format.
What’s the practical difference between Covidence and DistillerSR?
Both cover the same core stages (screening, extraction, risk of bias) with structured forms, but DistillerSR is built for higher-complexity, audit-trail-heavy use cases – regulatory submissions, pharmacovigilance, large enterprise review programs – while Covidence is the more commonly used, lighter-weight option for standard academic and Cochrane-style systematic reviews.
Do any of these tools replace the need for a registered protocol?
No. Screening, extraction, and synthesis tools support execution of a review; they don’t substitute for a pre-specified question, documented eligibility criteria, or (particularly for health-related topics) prospective registration of the protocol, e.g. through PROSPERO.
