If you are planning a systematic review or meta-analysis as a medical student or resident, the single highest-leverage 10 minutes of your project is checking PROSPERO before you commit. Most students have never heard of it. The ones who skip it discover, eight weeks in, that another team registered the exact same review six months ago and is about to publish.
What PROSPERO is
PROSPERO (the International Prospective Register of Systematic Reviews) is a free, public database hosted by the University of York's Centre for Reviews and Dissemination. It is where systematic-review and meta-analysis protocols are registered before the authors start screening papers. Registration is the equivalent of a clinical trial pre-registration on ClinicalTrials.gov — it locks in the question, inclusion criteria, search strategy, and analysis plan so the team cannot change them later to fit the results.
The database currently holds well over 250,000 registered reviews across every medical specialty. It is searchable by anyone. There is no paywall, no account requirement to search, and no embargo on the registration content.
Why it exists
Before PROSPERO existed, two unrelated teams could spend a year each conducting almost identical reviews and publish them within months of each other. The duplicated effort was enormous, and reviewers could quietly cherry-pick studies to fit a predetermined conclusion because nobody could compare the published review to a locked-in protocol. PROSPERO solved both problems. Most journals now require a PROSPERO registration number at submission for systematic reviews and meta-analyses. PRISMA 2020 reporting guidelines also expect it.
How to search PROSPERO
Go to crd.york.ac.uk/prospero. The simple search bar at the top accepts free text. For a more controlled search, use the Advanced Search link — it lets you filter by population, intervention, outcome, status (Ongoing, Completed, Withdrawn), and registration date range.
Search using your concept terms. Try 2 or 3 variations — PROSPERO records use free-text entries from registrants, not a controlled vocabulary, so synonyms matter. If you would search PubMed for "SGLT2 AND HFpEF," try PROSPERO with "SGLT2 heart failure preserved ejection," "dapagliflozin HFpEF," and "sodium glucose cotransporter heart failure" as three separate searches.
How to read a PROSPERO record
Open any matching record. You will see:
- CRD number (format
CRD42024XXXXXX) — the citation handle - Registration date and last edited date — tells you how long they have been working
- Status — Ongoing, Completed, Withdrawn
- Anticipated start and completion dates — how soon they will publish
- Review question and PICO — lets you compare to yours
- Search strategy — sometimes the full PubMed string
- Contact author — for collaboration outreach
Interpreting what you find
You find an Ongoing registration on your exact question
Check the registration date. If it is less than 3 months old, they are still in screening — you have 6 to 12 months before they could publish. If it is more than 12 months old, they are likely in synthesis or write-up and you should pivot or differentiate clearly.
You find a Completed registration
The review should be published or in press. Search PubMed by the author names from the PROSPERO record. If you find the published review and it answers your question, you do not have a project unless your angle is meaningfully different.
You find a Withdrawn registration
The team gave up. Sometimes they post a reason. This can be useful intelligence — if three teams have withdrawn from the same question, there may be a structural problem with the evidence base (no good RCTs, no consistent outcome measure, irreconcilable heterogeneity).
You find adjacent but not identical registrations
Read them carefully. A registration on "SGLT2 in HFpEF for mortality" does not eliminate your review on "SGLT2 in HFpEF for quality of life" or "SGLT2 in HFpEF in elderly subgroups." Look for a clear differentiation, write it down, and use it as the justification in your protocol.
How to register your own protocol
Once you have decided your review is novel and feasible:
- Create a free PROSPERO account
- Click "Register a new review"
- Fill in the structured form: review title, anticipated start and completion dates, PICO, search strategy, inclusion criteria, risk-of-bias tool, analysis plan, dissemination plan
- Submit. The CRD team typically reviews and approves within 1 to 3 weeks
- You receive a CRD number. Cite it in your protocol and final manuscript
Registration is editable later for minor changes (adding a reviewer, updating a contact email). Substantive changes (changing the PICO mid-screening) must be flagged transparently in the published manuscript.
Limitations to know
PROSPERO is not perfect. Some teams do not register, some register only after they have started screening, and some registrations are sparse. Searching PROSPERO will catch the majority of competing work but not all of it. Cross-check with the other four steps of the novelty check — PubMed for published reviews, Cochrane for upcoming reviews, ClinicalTrials.gov for ongoing trials that could obsolete your synthesis, and preprint servers for unpublished work.
If you are using ResearchChecker, the PROSPERO Registry section flags PubMed mentions of PROSPERO registrations and links you directly to the registry's search page so you can verify any specific result manually. The tool does not replace the registry — it just gets you to the right place faster.