How to Write a Systematic Review Methods Section

Your systematic review methods section carries more evidentiary weight than any other part of the manuscript. It's the section a peer reviewer scrutinizes most closely, since it's what determines whether your findings are reproducible and trustworthy, regardless of how compelling your results turn out to be.

Structure your methods section around PRISMA 2020's method items

If you've registered a protocol and followed PRISMA-P, your methods section is largely a matter of converting planned, future-tense language into completed, past-tense reporting, organized around the same items PRISMA 2020 expects.

**Eligibility criteria.** State your finalized PICO(S) elements precisely. If anything changed from your registered protocol, note the deviation explicitly here rather than silently reporting the revised criteria as though they were the original plan.

**Information sources.** List every database searched, with the exact date range and the date of your final search. Include any supplementary sources: trial registries, reference list checks, citation searching, or contact with study authors for unpublished data.

**Search strategy.** Report the full search strategy for at least one database in the main text or a supplement, not a paraphrased summary. This is one of the most frequently under-reported items, and its absence is one of the fastest ways to draw a revision request.

**Selection process.** State how many reviewers screened titles and abstracts, and full texts, whether independently, and how disagreements were resolved. If you used screening software, name it.

**Data collection process.** Describe your extraction process: whether it was done independently by multiple reviewers, whether a pilot round preceded full extraction, and how discrepancies were resolved.

**Data items.** List every outcome and variable you extracted, including how you handled studies reporting multiple time points or multiple effect measures for the same outcome.

**Risk-of-bias assessment.** Name the specific tool used for each study design included (RoB 2, ROBINS-I, Newcastle-Ottawa Scale, and so on), and state how many reviewers applied it.

**Effect measures.** State which effect measures you used for each outcome type and briefly justify the choice if it isn't the obvious default for that outcome.

**Synthesis methods.** This is usually the longest methods subsection. Describe how studies were grouped for synthesis, your statistical model (fixed-effect or random-effects, with justification), how you assessed and planned to handle heterogeneity, and any planned subgroup, sensitivity, or meta-regression analyses.

**Reporting bias and certainty assessment.** State whether and how you assessed publication bias, and whether you applied GRADE or CERQual to rate certainty of evidence.

Writing style: precise, past tense, and verifiable

A methods section should read as a factual record of exactly what was done, written so specifically that another researcher could replicate your process. Avoid vague phrasing like "a comprehensive search was conducted" without the actual search string attached, or "studies were appraised for quality" without naming the tool. Every claim in the methods section should be checkable against a supplement, an appendix, or the search strategy itself.

The most common methods section weaknesses

Peer reviewers most frequently flag: search strategies described narratively instead of reproduced in full, risk-of-bias tools named without stating how many reviewers applied them, and synthesis methods that don't specify a model choice or justify it. Each of these is a completeness gap, not a design flaw in your review itself, which means they're genuinely fixable in editing without redoing any of the underlying work.

**[Get a quote](/systematic-reviews#review-editing)** for a full methods section review before you submit.