Writing a systematic review or meta-analysis for a journal? Most good journals now expect the review protocol to be registered before you start extracting data, and the most widely used register is PROSPERO, run by the Centre for Reviews and Dissemination (CRD), University of York, and funded by NIHR. Registration is free and gives your review a CRD number (for example CRD42026XXXXXXX) that you quote in your paper.
Which reviews can be registered?
| Review type | PROSPERO? |
|---|---|
| Systematic review of interventions (with or without meta-analysis) | Yes |
| Diagnostic accuracy, prognostic, risk factor, genetic association and epidemiological (prevalence) reviews | Yes |
| Review of reviews (umbrella review) | Yes, if it meets the usual criteria and gives full methods |
| Methodological review | Only if it has at least one outcome of direct patient or clinical relevance |
| Scoping review | No. Register it on OSF Registries instead. |
| Narrative or literature review | No |
The review must have a health-related outcome. Cochrane protocols are added to PROSPERO automatically and do not need separate registration.
When to register
- Best: before you start screening titles and abstracts.
- Last chance: PROSPERO accepts a review only if data extraction has not been completed. The form asks you to tick which stages (searching, screening, data extraction, risk of bias, synthesis) have started or finished, so be honest about where you are.
- Before you start, search PROSPERO for similar reviews. The form also checks this for you (see Part 13 below).
Step 1: Create your PROSPERO account and ORCID iD
You need a PROSPERO account and an ORCID iD (PROSPERO will not let you create a record without it). Both are free. Full guide: How to Create a PROSPERO Account: Every Field Explained.
Step 2: Fill in the registration form, part by part
These are the parts of a published PROSPERO record and what to write in each. Prepare them in a Word file first, then paste them into the form. You can save the form and come back later before you submit.
Part 1: Review title and basic details
| Field | What to write |
|---|---|
| Review title | Clear title with population, topic and design, e.g. “… in adult ICU survivors: a systematic review and meta-analysis” |
| Condition or domain being studied | Choose the condition tags from the list, then describe the condition in two or three lines |
| Rationale for the review | One short paragraph: why the topic matters and what gap the review fills |
| Review objectives | The review question, written as one or two clear questions |
| Keywords | Three to six keywords separated by semicolons |
| Country | Pakistan (the review team’s country) |
Part 2: Eligibility criteria
| Field | What to write |
|---|---|
| Population: included / excluded | Who the studies must be about (age, setting, condition), and who is excluded |
| Intervention(s) or exposure(s): included / excluded | The intervention or exposure of interest, with tags from the list, and what is excluded |
| Comparator(s) or control(s) | Placebo, standard care, another intervention, or “not applicable” for prevalence reviews |
| Study design | Which designs are included, e.g. RCTs only, or RCTs and non-randomised, cohort, case-control and cross-sectional studies |
| Context | Settings and countries included, e.g. all ICUs worldwide, hospital or community settings |
Part 3: Timeline of the review
Enter the planned start and end dates of the review. PROSPERO adds the date of first submission and the date of registration itself.
Part 4: Availability of full protocol
Give a link to the full protocol if you have published or uploaded one (e.g. on OSF). If not, select that a full protocol has not been written or is not available.
Part 5: Searching and screening
| Field | What to write |
|---|---|
| Search for unpublished studies | Whether you will search grey literature and trial registries, or published studies only |
| Main sources to be searched | Databases, e.g. PubMed/MEDLINE, Embase, Scopus, Cochrane Library, CINAHL, Web of Science |
| Search language restrictions | e.g. English only, or no restriction |
| Search date restrictions | From and to dates of the search |
| Other methods of identifying studies | e.g. reference lists of included studies, citation searching, contacting experts, or none |
| Link to search strategy | Upload or link the full search string for at least one database. A complete strategy reduces queries. |
| Selection process | How screening is done, e.g. two reviewers independently, disagreements resolved by a third |
Part 6: Data collection process
| Field | What to write |
|---|---|
| Data extraction | Who extracts data and how (e.g. two reviewers independently), and whether study authors will be contacted for missing data |
| Study risk of bias or quality assessment | Who assesses it and with which tool, e.g. RoB 2 (RCTs), ROBINS-I (non-randomised), Newcastle-Ottawa Scale (observational), QUADAS-2 (diagnostic), JBI checklist (prevalence) |
| Reporting bias assessment | Whether risk of bias due to missing results (e.g. funnel plots, Egger’s test) will be assessed |
| Certainty assessment | Whether certainty of evidence will be rated, e.g. with GRADE |
Part 7: Outcomes to be analysed
Main outcomes and additional outcomes, each with how it is measured and at what time point. Vague outcomes are the most common reason for queries.
Part 8: Planned data synthesis
How results will be combined: fixed or random-effects meta-analysis, effect measures (risk ratio, odds ratio, mean difference, standardised mean difference, pooled proportion), how heterogeneity is assessed, subgroup and sensitivity analyses, and narrative synthesis if pooling is not possible.
Part 9: Current review stage
Tick which stages have started and which are completed: pilot work, formal searching, screening against inclusion criteria, data extraction, risk of bias assessment and data synthesis. Then select the review status (e.g. planned or ongoing). Data extraction must not be complete.
Part 10: Publication of review results
The language in which results will be published, usually English.
Part 11: Review affiliation, funding and peer review
| Field | What to write |
|---|---|
| Review team members | Each member’s title, name, affiliation, ORCID iD and any conflict of interest |
| Named contact (guarantor) | The person responsible for the record, with email and ORCID iD |
| Review affiliation | The main institution of the review, e.g. your hospital or university |
| Funding source | Funder name, or that the review has no funding and is done in the authors’ own time |
| Peer review | Whether the protocol was peer reviewed, e.g. by a supervisor or committee |
Part 12: Additional information
Any additional conflicts of interest for the review as a whole. Medical Subject Headings (MeSH) are added to the record from the tags you choose.
Part 13: Similar reviews
PROSPERO automatically shows existing records that look similar to yours. For each one you must explain why your review is different (e.g. different population, outcome or design) or why it should still go ahead. These reasons appear on your published record, so write them carefully.
Step 3: Submit, affirm and get your CRD number
Before submitting, the guarantor affirms that the information is truthful. Deliberately giving inaccurate information may be treated as scientific misconduct. PROSPERO checks the record, and it is then published, returned for clarification, or rejected. In our experience, complete records are often published within a day or a few days. Your citation will read: Authors. Title. PROSPERO 2026 CRD42026XXXXXXX. Quote the CRD number in the abstract and methods of your paper.
The most common reasons a record is returned or rejected:
- The review is a scoping or literature review
- Data extraction is already complete
- Search sources or strategy are missing or too thin
- Outcomes or eligibility criteria are vague
- No clear reason given for going ahead when similar reviews exist
Let UPMED register your review
- Help in setting up your own PROSPERO account and ORCID iD (both stay in your name)
- Review question, search strategy, eligibility criteria and synthesis plan written to PROSPERO’s standard
- Complete entry of all parts of the registration form, including the similar-review justifications
- Replies to PROSPERO queries until the record is published
- Follow-up until your CRD number is issued, and later amendments when your review changes
- WhatsApp: 0304 2397393 (save the number as “UPMED”)
- Email: [email protected]
Frequently asked questions
Does PROSPERO charge a fee?
No. PROSPERO registration is free. UPMED’s fee is for preparing, entering and following up your record.
Can I change the record after registration?
Yes. Changes are made as amendments, and each version is kept in the record’s version history. Explain any important change from the protocol in your final paper.
Do all team members need an ORCID iD?
The person creating the record needs one. It is good practice for every team member to have one, as it links the review to their research profile.
I am doing a trial, not a review. Where do I register?
Trials and quasi-experimental studies are registered on ClinicalTrials.gov or a WHO primary registry. See How to Register Your Clinical Trial on ClinicalTrials.gov from Pakistan and WHO Primary Clinical Trial Registries.
Source: PROSPERO, Centre for Reviews and Dissemination, University of York.
Rate this article
How useful was it for you?
UPMED reads every review; some are shown on this page.
Log in to write a few lines too.