Skip to content
SRN — Systematic Reviews Network, home

Beginner guides

How to write a review protocol

A plain-language guide for researchers building their first systematic review.

Beginner guides

A protocol is the plan for your systematic review, written before you begin. It states the question, the methods, and the decisions you will make — so those decisions are made on principle, in advance, rather than shaped by results you have already seen. A clear protocol is the difference between a review that can be trusted and one that merely looks thorough.

Why write one first

Every review involves judgement calls: which studies count, how to handle a borderline case, which outcomes matter. Made after you have seen the data, those calls are open to bias — conscious or not. Made and recorded beforehand, they are simply method. Registering the protocol publicly (PROSPERO is the usual home for health reviews) puts the plan on record and lets others see that it came first.

What a protocol should contain

  • Background and rationale — why the review is needed and what gap it addresses.

  • The question — stated precisely, usually via PICO (population, intervention, comparison, outcome).

  • Eligibility criteria — exactly what makes a study includable: designs, populations, dates, languages.

  • Search strategy — which databases, which terms, and any limits, in enough detail to be rerun.

  • Screening and selection — how many reviewers, working independently, and how disagreements are resolved.

  • Data extraction — what will be pulled from each study, and on what form.

  • Risk-of-bias assessment — which tool, applied by whom.

  • Synthesis plan — narrative or meta-analysis, and the conditions under which you would pool results.

A note on changes

A protocol is a commitment, not a cage. Reviews sometimes have to depart from the plan — a database is unavailable, a definition proves unworkable. That is allowed. What is required is that you say so: record what changed, when, and why, so the reader can judge whether the change was reasonable. An undisclosed deviation is what damages a review; a disclosed one rarely does.

Where to start

If this is your first review, you do not have to write the protocol alone. SRN's Beginner Academy walks through each section in turn, and the Mentorship Programme pairs you with someone who has written several. The templates in this library give you a structure to fill in.

All resources

Put it into practice.

Our courses and mentorship turn method you've read about into a review you've finished.

Explore programmes