Evidence Synthesis

Keep a Search Log You Can Actually Reproduce

Record the search while it happens

A final sentence saying that several databases were searched is not enough to reconstruct the work. Keep a log from the first exploratory search through the final update. Distinguish informal scoping from the searches intended for the review.

Work with an information specialist when possible. The Cochrane search chapter addresses source selection, search construction, documentation, and study selection. A quick search in one database should not be presented as a comprehensive systematic search.

Save the details that are easy to lose

For each run, record the database and platform, the date, the exact strategy as executed, applied limits, returned count, exported count, and export filename. Database syntax differs, so preserve each translated strategy rather than assuming the same query works everywhere.

Document why limits were applied. Save updates as new runs with their own dates and files. If an export limit forces several batches, record the batch boundaries and reconcile the total.

Keep records and studies distinct

Several publications can describe the same study. Deduplication removes duplicate records; linking multiple reports of one study is a separate task. Keep a traceable identifier scheme so that a conference abstract, trial registration, and full report are not mistakenly treated as independent studies.

Record screening decisions and full-text exclusion reasons consistently. Reconcile counts before preparing the review flow diagram; do not adjust counts merely to make the diagram balance.

Prepare the reporting handoff

Download the search-run log worksheet. Maintain exact search strategies in an accompanying supplement and preserve the original exports according to your project’s data-management plan.

Use the official PRISMA 2020 resources when reporting the review. A complete log supports transparency, but it does not by itself establish that the search was sufficiently sensitive or that the review question is novel.

Related reading

Reading path: Review evidence. Guide 2 of 8. Choose any guide to suit your project.

Evidence Synthesis

Pilot a Data-Extraction Form Before Scaling the Review

Test an extraction form against real reporting problems and preserve a traceable path from source to analysis.

Evidence Synthesis

Pilot Eligibility Criteria Before Full-Text Screening

Turn protocol criteria into reproducible screening decisions with a small, documented pilot.

Browse this reading pathTools for this stageGuides and downloads for this stage