Systematic Review Protocol: What to Write Before You Screen

Your search is ready, the reference manager is open, and your supervisor asks one question before you go further: where is the protocol? You have a topic and a hunch about the databases, and not much else on paper.

Here is the short answer. A protocol for a systematic review is the plan for the review, written before anyone screens a record: the question, the eligibility criteria, the sources and search, how studies will be selected, what data you will collect, how you will judge the studies and combine what they found, and how you will record any change of plan. You make it public, usually by registering it, so that readers can later compare what you did with what you said you would do.

The template below follows PRISMA-P, the reporting checklist for protocols, grouped into the parts you write. The worked example after it uses an invented review: how does remote work affect employee productivity compared with office work?

Systematic review protocol template: the sections

Read it with the PRISMA-P explanation and elaboration, which gives an example for every item.

1.Writing a systematic review protocol

The sections are not written in their printed order. Start with the question, because every later section depends on it.

The methods sections are the easiest to leave vague. "Two reviewers will screen" is not a method until it says at which stage, independently or not, and who settles a disagreement.

2.Systematic review protocol example

Here are six sections of the invented review, written the way a protocol states them.

Each sentence answers a question a second reviewer would otherwise ask you halfway through the review.

3.Where to register a systematic review: PROSPERO or OSF

A registration is a short public record with a date and a number. The protocol is the full plan. They are different documents, and the PRISMA 2020 explanation (item 24) asks you to report both: where the review is registered, and where the protocol can be read.

Both are free, and PROSPERO publishes the record once every author has approved it. Finish the protocol, then register it before screening starts, so that readers can see the plan came first. PROSPERO's own cutoff is later, before data extraction begins, and it asks for a finished protocol rather than an early one.

If extraction has already started, PROSPERO will not take the review: make the protocol public by one of the routes below, and say in the paper that the review was not registered, which PRISMA 2020 item 24a allows.

For the full protocol, PRISMA 2020 lists the usual routes: attach it to the registration record, publish it in an open-access journal such as Systematic Reviews or BMJ Open, or deposit it in a repository such as OSF or Zenodo, which gives it a DOI you can cite.

4.Scoping review protocol

A scoping review maps what has been studied rather than answering one focused question, and its protocol changes in four places.

The detailed guidance is JBI's best practice guidance for scoping review protocols, which comes with a checklist.

5.When the plan changes

Some changes are unavoidable: the data may not support the meta-analysis you planned, or a criterion may turn out to be unworkable. Changes are allowed, as long as each one is written down and explained.

Those three elements are what PRISMA 2020 asks for under item 24c. Note each change in the registration record when it happens, then report them together in the review.

In Littview, the protocol's criteria, the number of reviewers per paper and the extraction form become the project's own settings, and the Methods draft reads them back from there.

FAQ

Three questions come up most.

Is a PROSPERO protocol the same as a PROSPERO registration? No. The registration is the form you fill in on the PROSPERO site, and there is no separate PROSPERO protocol template: every field must be completed even if you upload the full protocol. PRISMA 2020 suggests attaching the protocol as a PDF to the record, so the two stay linked.

Can a scoping review be registered in PROSPERO? No. PROSPERO does not accept scoping, literature or mapping reviews. Register a scoping review protocol on OSF Registries, or publish it in a journal that accepts scoping review protocols.

Does a literature review need a protocol? PRISMA-P was written for systematic reviews, and a narrative literature review has no equivalent requirement. A one-page plan still helps: the question, the sources you will search, what you will include and how you will organise what you find.

For this week: write the question in one sentence, turn it into criteria, run a pilot search, fill in the template above section by section, get every reviewer to agree to it, and register it before the first record is screened.