The Review Protocol

PROSPERO Registration: Step-by-Step Guide

By Angel Reyes Last updated 4 min read

PROSPERO — the International Prospective Register of Systematic Reviews maintained by the University of York's Centre for Reviews and Dissemination — has become the de facto registration venue for systematic reviews with a health-related outcome. Most journals now require a PROSPERO ID at submission. Fortunately, registration is free and usually approved within a few weeks, provided the submission is clean. This guide walks through each field, what registrars are actually looking for, and the mistakes that cause rejection.

Who needs to register

You need PROSPERO registration if:

  • Your review has a health-related outcome (broadly defined — public health, clinical, social care, welfare)
  • Your review is a systematic review, rapid review, or umbrella review (not a scoping review — see our scoping review post)
  • You have not yet completed data extraction (PROSPERO accepts protocols up to the data extraction stage)

If your review is not health-related, register on the Open Science Framework (OSF) or INPLASY instead. If you missed the data extraction cutoff, your review is unregistered — note this as a limitation.

For the full systematic review process context, see our systematic review guide and reporting standards page.

When to register

Register before you start screening. PROSPERO explicitly rejects reviews where screening is complete. Ideally register after:

  • Your protocol is written and approved by your team
  • Your search strategy has been peer-reviewed (PRESS checklist)
  • Your inclusion/exclusion criteria are locked

Do not wait for the search to finish. Register, then search.

The submission form, field by field

PROSPERO's form has about 40 fields. The ones that actually matter:

Title

Format: "Effect of [intervention] on [outcome] in [population]: a systematic review [and meta-analysis]."

Registrars reject vague titles. "A systematic review of mindfulness" will come back for revision.

Review question

Write in PICO form, even if your review question is messier. Example:

Among undergraduate students (P), does mindfulness-based stress reduction (I) compared with usual care or no intervention (C) reduce self-reported anxiety (O)?

Searches

List every database you will search, with platform. "PubMed, Embase (Ovid), PsycINFO (APA PsycNet), CINAHL (EBSCO), Web of Science Core Collection, and grey literature via Google Scholar and OpenGrey."

Include planned date range and language limits.

Types of study to be included

Be specific. "Randomized controlled trials" is specific. "Experimental studies" is not.

Condition or domain being studied

Use MeSH or equivalent standard terminology. Registrars will correct ad-hoc terms.

Participants / population

Replicate the P of your PICO. Specify age, setting, and any exclusions.

Intervention(s), exposure(s)

Replicate the I. Specify dose, duration, and mode of delivery where relevant.

Comparator(s) / control

Replicate the C. "No intervention," "waitlist," "usual care," "active control."

Outcome(s)

List primary and secondary outcomes explicitly. Include the measurement tool and time point.

Data extraction and management

State: "Two reviewers will extract data independently using a piloted extraction form. Disagreements will be resolved by discussion, with a third reviewer if needed." If you are deviating from this, explain.

Risk of bias assessment

Name the tool: "Cochrane RoB 2 for RCTs," "ROBINS-I for non-randomized studies," "JBI Critical Appraisal Checklist for qualitative research."

Strategy for data synthesis

State whether you plan a meta-analysis and under what conditions. Typical phrasing: "If sufficient clinical and methodological homogeneity exists, a random-effects meta-analysis will be conducted. Otherwise, narrative synthesis following SWiM guidance."

Common reasons for rejection

  1. Screening already complete. Register earlier.
  2. Review question not in PICO form. Rewrite.
  3. No risk of bias tool specified. Add one.
  4. No second reviewer. PROSPERO strongly expects dual screening and extraction.
  5. Duplicate registration. Check the database first for existing registrations on your topic.

How long approval takes

Published PROSPERO turnaround varies from 1 to 6 weeks. Clean submissions often go through in under two weeks. Submissions with revisions take longer.

Once approved, you receive a PROSPERO ID (format: CRD420250XXXXXX). Cite it in your protocol, every manuscript, and every conference abstract.

Updating the registration

You are expected to update PROSPERO when:

  • Search is completed
  • Screening is completed
  • Extraction is completed
  • The review is published

You are also expected to log any protocol deviations. PRISMA 2020 requires that deviations be reported in the manuscript (item 24). If you changed the outcome set, added a database, or relaxed a criterion, log it and explain why.

A pre-submission checklist

  • [ ] PICO question drafted and checked
  • [ ] All databases listed with platforms
  • [ ] Search strategy peer-reviewed (PRESS)
  • [ ] Inclusion/exclusion criteria locked
  • [ ] Risk of bias tool selected
  • [ ] Dual reviewers confirmed
  • [ ] Extraction form drafted
  • [ ] Synthesis plan decided (meta-analysis conditions stated)
  • [ ] Funding and conflicts of interest prepared
  • [ ] PRISMA 2020 and PRISMA-P checklists mapped

Treat PROSPERO registration as a protocol-quality audit. If you cannot answer a field clearly, your protocol has a gap — fix it before registration, not after.

Related posts