Separate reporting guidance from review methods
PRISMA 2020 guides transparent reporting of systematic reviews. Its checklist and flow diagram do not, by themselves, specify every method needed to conduct a review or certify its quality. Choose methods appropriate to your question and review type, and check relevant reporting extensions. The Cochrane Handbook guidance cited here concerns reviews of interventions.
Define the question and preserve the plan
Before selection begins, document the review question, eligibility criteria and planned groupings for synthesis. Define criteria clearly enough to support consistent decisions. Keep a dated protocol and change record: if a method changes, explain what changed and why. Do not quietly rewrite eligibility rules around the results you prefer.
Keep a reproducible search record
For each source, retain the database or register name, platform, search date, full strategy as run, limits and retrieved record count. Save the original exports and document deduplication. Record supplementary searching too, including citation searching and other discovery routes. Preserve the actual search syntax rather than reconstructing it from memory when writing.
Make screening decisions traceable
Describe who screened records and reports, how decisions were checked and how disagreements were resolved. Record reasons for excluding potentially eligible studies after full-text assessment. Distinguish search records, reports and underlying studies: several publications may describe one study. Reconcile these relationships before completing the flow diagram so that publications are not counted as independent studies.
Build an extraction record that can be checked
Pilot a structured extraction form before using it across the review. Capture study characteristics, relevant outcomes, results and information needed for risk-of-bias assessment. Link entries to their source locations. Define how conflicting reports, missing information and derived values will be handled; preserve calculation notes and correspondence rather than presenting inferred values as directly reported observations.
Connect the report to the evidence
Use the PRISMA checklist to locate each required reporting element in the manuscript or supporting material. Check that the narrative, study tables, flow counts and synthesis describe the same evidence set. Keep limitations visible, including missing reports or information, and explain where the review’s methods departed from its original plan.
Assemble a useful final record
Maintain an organised package containing the protocol, search history, screening decisions, study-to-report links, extraction forms and analysis records. Consult the official checklist, explanation and elaboration, and applicable extensions for the complete reporting requirements. This guide is a practical starting point, not a substitute for those documents or methodological review.
Sources & further reading
Consult the original guidance for its full scope and any updates. Institutional requirements and review methods vary.
- PRISMA — PRISMA 2020 statement and supporting resources
- Cochrane Handbook, Chapter 1 — Starting a review
- Cochrane Handbook, Chapter 3 — Defining the criteria for including studies and how they will be grouped for the synthesis
- Cochrane Handbook, Chapter 4 — Searching for and selecting studies
- Cochrane Handbook, Chapter 5 — Collecting data