On this page

What EPPI-Reviewer does

EPPI-Reviewer is a browser-based literature-review platform developed by the EPPI Centre at UCL. Its documented scope includes systematic, narrative and other reviews, with shared workspaces, coding, data extraction and synthesis tools. It also offers LLM-assisted screening and extraction, but the provider explicitly limits that feature to evaluation purposes and does not recommend it for real reviews without formal accuracy testing. This source-based profile distinguishes the broader review workspace from its experimental AI functions; we did not use a trial account or evaluate its outputs.

Where it fits, and where it falls short

Useful strengths

  • Supports several review stages in a shared project, including screening, coding and evidence synthesis.
  • The provider documents a modern browser workflow, a one-month trial and separate pricing for user access and shared reviews.
  • LLM outputs can be distinguished from human coding and reviewed in comparison reports, supporting a formal evaluation design.
  • The platform includes an OpenAlex-linked discovery route and Zotero integration according to its current product page.

Limits to consider

  • EPPI Centre says automated LLM screening and extraction are for evaluation only and should not be used in real reviews without formal accuracy testing.
  • LLM use requires purchased credits assigned to a specific review; available models and their costs can change.
  • Shared-review costs are separate from per-user fees, and all collaborators in a shared review need active accounts.
  • The security page contains legacy notes about earlier versions. Confirm the current Reviewer Web configuration, institutional terms and data agreement for sensitive work.

A suggested EPPI-Reviewer workflow

View the workflow diagram
EPPI-Reviewer: an illustrated workflow for checking sources and reviewing your work
An editorial illustration of the workflow.
  1. 1

    Write the protocol, search sources, eligibility criteria, screening roles and synthesis plan before setting up the review workspace.

  2. 2

    Import and deduplicate the records, then document the screening stage and disagreements using the project’s agreed rules.

  3. 3

    Code and extract data against a defined code set; verify important fields against the original study and preserve unresolved questions.

  4. 4

    If evaluating an LLM feature, create a human-coded comparison set, obtain the required review credits and compare outputs before using AI beyond evaluation.

Manage a review project across more than screening

EPPI-Reviewer is designed for several kinds of literature review, including systematic reviews, meta-analyses, narrative reviews and meta-ethnographies. The product overview describes a browser-based workspace. Its value is the broader project flow: organize references, coordinate review work, code studies and carry information forward into synthesis. The exact tools depend on the review type and configuration, so map your protocol to the features you actually need before inviting a team.

Keep the database search outside the assumption that a review platform makes a search comprehensive. Record which databases and supplementary sources you searched, the exact query and date, filters, deduplication rules and selection decisions. Importing a clean set of records does not tell a reader where they came from or whether the set meets the review protocol.

Plan reviewer roles before the project begins. For each decision, agree whether it is independent, how disagreements will be resolved and when the reconciled record is final. Preserve the rationale and source passage needed to explain eligibility or coding choices, rather than relying on a status label alone.

Supporting documentation: [1] [2]

Treat LLM screening and extraction as experimental evaluation

EPPI Centre’s own automated-coding guidance says the LLM feature should be used for evaluation purposes only. The provider does not currently recommend using it in real reviews without first testing accuracy formally. The page describes AI-generated screening and data-extraction coding as something to compare with human decisions, not as an already validated replacement for them.

If your institution permits an evaluation, create a representative human-coded comparison set, define how disagreements will be measured and keep the model name, version, prompt and input text. EPPI documents reports that can compare machine and human coding. A single successful example or a plausible-looking field is not a formal accuracy assessment.

The guidance says LLM-generated codes can be attributed to the model and marked incomplete or locked to distinguish them from human coding. That separation helps evaluation; it does not establish that the codes are correct. Keep a qualified reviewer responsible for the final extraction and follow your protocol, ethics approval and course or publication rules.

Current documentation describes purchased LLM credits assigned to a review, and notes that the available model list may change. Full-text coding requires PDF processing and sends substantially more text to the model, increasing both time and cost according to the provider. Confirm the present credit and model settings before any evaluation, and upload only material you are permitted to process.

Supporting documentation: [3]

Budget users, shared reviews and the trial separately

EPPI Centre’s current fee page separates the user fee from the shared-review fee. The page lists £10 per user each month and £35 per shareable review each month; all users in that shared review need their own active accounts. It also describes a one-month free trial, while specifying that a trial account can create unlimited non-shareable reviews. Contact support if you need to trial a multi-user review.

The fee page says 20% VAT is charged on purchases within the UK, and that invoice purchases have a £100 minimum with a stated surcharge for invoices under £500. These are provider-published terms and can change. Recheck the page for current rates, tax, billing terms and whether your institution or collaboration is eligible for another route before estimating total cost.

Trial access expires after one month and then becomes read-only until the relevant account and review access is extended. Plan exports, project access and team membership with that transition in mind. This profile does not imply an individual or institutional entitlement and we did not create an account.

Supporting documentation: [1] [4] [5]

Check the current version and data terms before importing sensitive work

The product page identifies EPPI-Reviewer 6 as the current web version and notes that the older version still exists for a few unfinished features. Its security page says production data are held in Microsoft Azure’s West Europe region and describes encrypted backups in West and North Europe. These are the provider’s published statements, not an independent security audit.

The same security page contains notes about the previous version and its connection behavior, so do not generalize every statement there to the current web application. Verify the exact version, institutional agreement, privacy terms and connection settings that apply to your account before importing unpublished manuscripts, participant information or restricted documents.

We have not signed in, tried the product, inspected a live workspace, purchased credits or assessed data handling. Ask your data-protection contact whether the intended review material and any LLM processing are permitted under your organization’s rules.

Supporting documentation: [1] [6]

Choose by the review record your team needs to keep

EPPI-Reviewer is worth evaluating when the project needs a connected workspace for screening, coding, team coordination and synthesis. ASReview LAB is more focused on prioritizing the order of human screening for an already collected record set. Rayyan focuses on organizing screening stages, decisions and reconciliation. Elicit offers a different discovery and structured-extraction workflow. These are distinct task fits, not a benchmark or a ranking.

Before adopting any platform, map one realistic protocol from search export through screening, extraction and the final synthesis record. Check which parts are native, which need manual steps, what an export contains and what remains available when a trial or review license expires. A polished workspace does not itself establish methodological completeness or evidence quality.

Supporting documentation: [1] [2] [3]

Pricing, free access and plan checks

At the time of review, EPPI Centre lists a £10 monthly fee per user and £35 per month for each shareable review. Every collaborator on a shared review also needs a user account. Its fee page says 20% VAT is added to UK purchases; invoice purchases have a £100 minimum and may carry a surcharge under £500. A free one-month trial is available, but it allows unlimited single-user reviews; contact support to trial a shared review. LLM credits are purchased separately and assigned to a review. Recheck the provider’s current fees, tax, billing period and credit charges before budgeting.

Confirm billing period, renewal terms, limits and regional availability on the provider’s current plan page before paying.

Questions about EPPI-Reviewer

Does EPPI-Reviewer’s AI replace human screening?

No. EPPI Centre says its LLM-assisted screening and data extraction are for evaluation purposes and are not recommended in real reviews until their accuracy has been formally tested.

How much does EPPI-Reviewer cost?

The fee page currently lists £10 per user per month and £35 per month for each shareable review. Collaborators on a shared review need active user accounts. UK VAT and invoice terms may apply; check current pricing before budgeting.

Can I try EPPI-Reviewer for free?

The provider lists a free one-month trial with unlimited single-user reviews. A multi-user shared review requires contacting EPPI Support to enable it for the trial.

Are LLM features included in the monthly fee?

EPPI documentation says LLM features require purchased credits assigned to a review. Full-text coding can use more text and cost more than abstract-only coding. Confirm current credit rates and model availability.

Does EPPI-Reviewer handle the entire systematic review?

It provides a workspace for multiple review stages, but it does not automatically establish that your search, screening, appraisal or synthesis meets a protocol. Define and document the method and check the original evidence.

Did AI Essay test EPPI-Reviewer or its LLM accuracy?

No. This is a source-based profile. We did not use the trial, create a review, purchase credits or evaluate model outputs.

Sources and research notes

This guide synthesizes official product information with our editorial assessment of academic use. Suggested exercises are examples you can repeat; they are not results of tests we performed. We do not infer accuracy from a provider’s marketing claims.

  1. EPPI Centre: product overview and current web application
  2. EPPI Centre: EPPI-Reviewer user guide
  3. EPPI Centre: LLM coding, evaluation-only guidance and credits
  4. EPPI Centre: current fee structure
  5. EPPI Centre: trial limits and expiry behavior
  6. EPPI Centre: published security and data-location information

Found a changed feature or a factual error? Send the exact page and an official source to info@ai-essay.com. Read our editorial policy for how we handle corrections.