Systematic Review Protocol: How to Write and Register It
A systematic review protocol is the plan you write and lock in before you begin screening studies. It states your review question, defines who and what you will include, describes how you will search, and explains how you will assess risk of bias and synthesise findings. Writing it first is what separates a systematic review from an informal literature summary, because it commits you to decisions before you can be swayed by which results turn up.
This guide is for postgraduate researchers, clinicians, librarians, and evidence synthesis teams who need to produce a defensible protocol and register it correctly. It covers what a protocol must contain, how to register with PROSPERO, a step-by-step drafting sequence, how to choose between registries, how to work from the Cochrane framework, and the mistakes that most often send drafts back for revision.
The value of a good systematic review protocol is practical rather than ceremonial. A registered plan lets peer reviewers see whether your reported outcomes match the ones you set out to measure, which is one of the clearest signals that selective reporting has not crept in. It also forces the whole team to agree on definitions early, so you are not renegotiating inclusion criteria halfway through screening several thousand records.
Before you start, gather a focused research question framed with a structure such as PICO, a preliminary scoping search that confirms the review has not already been done, agreement on the databases you will search, and a named team with defined roles for screening and extraction. If you plan to run a meta-analysis, decide in principle how you will handle heterogeneity now rather than later. With those pieces in hand, the sections below walk through building and registering a protocol you can stand behind.
Key takeaways
- A systematic review protocol sets out your research question, eligibility criteria, search strategy, and analysis plan before you start screening, which protects the review from bias introduced after seeing the results.
- Registering a protocol on PROSPERO or the Open Science Framework creates a public, time-stamped record that reviewers, editors, and readers can check against your final report.
- PRISMA-P and the Cochrane Handbook both provide detailed checklists, but they suit different review types, so match the framework to your question rather than defaulting to one.
- Most drafting problems trace back to vague inclusion criteria and an unregistered analysis plan, both of which are cheaper to fix at the protocol stage than after data extraction.
What Should a Systematic Review Protocol Include?
A complete systematic review protocol answers a predictable set of questions, and reporting standards such as PRISMA-P exist to make sure none are left out. Each element does specific work: the question and eligibility criteria define scope, the search strategy makes the review reproducible, and the analysis plan commits you to methods before you see the data.
| Protocol element | Decision it settles |
|---|---|
| Rationale | Why the review is needed and what gap in existing evidence it fills. |
| Review question and objectives | The precise population, intervention, comparator, and outcome under study. |
| Eligibility criteria | Which study designs, dates, and languages are included or excluded, and why. |
| Search strategy | Databases, keywords, and limits, recorded so others can rerun the search. |
| Risk of bias assessment | Which appraisal tool you will use and how disagreements are resolved. |
| Data synthesis plan | Whether results will be pooled statistically or summarised narratively. |
Specifying primary and secondary outcomes in advance matters most, because changing them after seeing results is the classic route to outcome switching. If you are unsure whether your question fits a full review, our systematic review versus meta-analysis explainer can help you decide before you commit.
How to Register Your Protocol With PROSPERO
PROSPERO is the international registry for systematic reviews in health, social care, welfare, and related fields. Registration is free and happens before you begin screening. You complete a structured online form covering your question, inclusion and exclusion criteria, search plan, planned outcomes, and expected start and completion dates. The register team screens the entry for eligibility and completeness, then issues a registration number you cite in your final paper.
Register early. PROSPERO will not accept a review once data extraction is well advanced, because the point of registration is to record intent before results can influence method. If your plans change legitimately, you update the record and the amendment is time-stamped, which preserves the audit trail rather than hiding the revision.
Two practical points the registry documentation understates are worth planning for. First, expect a review turnaround of some days rather than same-day publication, so do not treat registration as a last-minute step before submission. Second, a PROSPERO entry is deliberately shorter than a full systematic review protocol, so keep a longer internal version that holds your complete search strings and analysis detail. Registering commits you publicly to the plan; the full document is what your team actually works from day to day.
How to Write a Systematic Review Protocol Step by Step
Work through the draft in a fixed order so each decision constrains the next. First, frame the question using a structured format such as PICO, which makes your eligibility criteria almost write themselves. Second, run a scoping search to confirm the review is needed and to surface the vocabulary your full search will use. Third, translate that vocabulary into a documented search strategy for each database, ideally with a librarian's input; many institutions publish a systematic review libguide that lists supported databases and search templates.
Fourth, set your screening process, including how many reviewers assess each record and how conflicts are resolved. Fifth, choose your risk of bias tool and your data extraction fields before extraction begins, not after. Sixth, write the synthesis plan, stating clearly whether you will pool data or synthesise narratively and how you will handle heterogeneity.
Draft the systematic review protocol as a single working document and only then transfer the summary fields into your chosen registry. Writing the full version first stops you from discovering, mid-registration, that you have not decided something important. For the screening stage specifically, our practical guide to screening studies shows how the criteria you set here play out once records arrive.
Choosing the Right Registry for Your Review Protocol
The registry you pick depends on your field and how much of the protocol you want to publish. PROSPERO suits health and related reviews and is widely expected by clinical journals. The Open Science Framework, or OSF, is more flexible and hosts protocols across disciplines, including areas PROSPERO does not accept, and lets you attach full documents, search strings, and data. Some Cochrane reviews register through Cochrane's own editorial process instead.
The main trade-off is structure versus freedom. PROSPERO enforces a fixed schema that editors recognise instantly, which is reassuring but constraining. OSF gives you room to store the complete protocol and supporting files, at the cost of less standardised presentation. If your review sits outside health, OSF or a discipline-specific registry is often the only realistic option.
Whichever you choose, register in only one place to avoid duplicate public records, and record the registration identifier in your manuscript. Your library's systematic review libguide will usually name the registry your institution recommends and may explain any subscription-only tools it supports. Check that guidance early, because switching registries after registration is awkward and creates exactly the kind of version confusion the register is meant to prevent.
Writing a Protocol Using the Cochrane Framework
The Cochrane Handbook sets out the most detailed methodology in evidence synthesis, and even reviews that never reach Cochrane benefit from following it. Its protocol expectations go further than a minimum PRISMA-P checklist by requiring you to pre-specify how you will assess certainty of evidence, typically using GRADE, and how you will manage missing data and unit-of-analysis issues.
Working from the Cochrane framework means committing to specifics that lighter templates leave open. You state your risk of bias tool, name the summary-of-findings outcomes in advance, and describe your approach to statistical heterogeneity before any pooling. This front-loads effort, but it removes the ambiguity that peer reviewers most often query later.
If you intend to register systematic review plans through Cochrane, the editorial process reviews your title and protocol before you proceed, which adds oversight that self-registration does not. For non-Cochrane reviews, you can still adopt the Cochrane Handbook conventions and register systematic review details on PROSPERO or OSF instead. The point is to borrow the rigour, not necessarily the publication route. Our systematic review guide maps how these methodological choices connect across the whole project.
Common Protocol Drafting Mistakes and How to Avoid Them
The most common fault is vague eligibility criteria. If two team members can read the same criterion and disagree about whether a study qualifies, screening will stall and your decisions will look arbitrary. Write criteria specific enough that a reviewer outside the team could apply them consistently.
A second recurring problem is an underspecified analysis plan. Leaving synthesis decisions open invites later accusations of fitting the method to the result. Decide in advance whether you will pool data and how you will treat heterogeneity, even if you note that the final choice depends on what you find. The Cochrane Handbook is explicit that these decisions belong in the protocol, not the results section.
Third, teams often register systematic review plans too late, after screening is underway, which defeats the purpose of registration. Register before you begin, and log any genuine changes as dated amendments rather than silent edits.
Finally, avoid copying a search strategy from one database into another without adapting the syntax, and do not treat the registry summary as your only record. Keep the full working document current, because a protocol that quietly drifts from what you actually did is worse than no protocol at all.
Frequently asked questions
What should you know about systematic review protocol?
A protocol is the pre-specified plan that defines your question, eligibility criteria, search strategy, and analysis before screening begins. Its purpose is to lock in your methods so that the review's conclusions reflect the evidence rather than choices made after seeing the results. Writing and registering it is a standard expectation for reviews intended for publication.
Which factors matter most when evaluating systematic review protocol?
The factors that matter most are the precision of your eligibility criteria, the reproducibility of your search strategy, and the clarity of your pre-specified outcomes and analysis plan. Registries and editors scrutinise these first, because they are where selective reporting and scope creep usually enter. Getting them tight at the protocol stage saves far more time than correcting them later.
How can you compare the main options related to it?
To compare the main registration options, weigh discipline fit and how much detail you want public. PROSPERO is standard for health reviews and enforces a fixed schema; the Open Science Framework accepts a wider range of fields and lets you attach full documents and search strings; Cochrane offers editorial oversight for reviews going through its process. Register in only one place and cite the identifier in your manuscript.
What should you prepare before taking the next step with it?
Before your next step, prepare a focused question in a structure such as PICO, a completed scoping search, agreed databases and search terms, a named team with screening roles, and a chosen risk of bias tool. Check your institution's library guidance and decide your registry in advance. With those in place, you can draft the full document, transfer the summary into the registry, and move into screening with a plan that will hold up under review. Tools that keep your protocol, screening, and PRISMA reporting linked, such as those offered by Systematicly, can reduce the version drift that undermines otherwise sound protocols.