Search and study selection
Identify and screen eligible studies
Keep records, reports and studies as separate counting units. The starter partition covers duplicate removal only: add explicit branches for automation or other removals if used. Supplementary searches may need separate identification routes. A final report-to-study mapping is not an exclusion count.
Flow diagram
Steps
Search specification
Define sources, complete reproducible search strategies, coverage and search dates. Specify any automation, other pre-screen removals, citation searching and search updates; extend the count branches when these routes apply.
Records retrieved
A record is a database or source entry describing a report. Combine retrieved records while retaining the search source and retrieval date; multiple records can describe the same report.
Duplicate records removed
Identify records describing the same report using documented rules and verification. Deduplication removes repeated records, not additional publications from the same underlying study.
Records screened
Pilot title and abstract criteria, use independent decisions and resolve disagreements.
Records excluded
Count records that do not meet the screening criteria at title/abstract screening. Retain reviewer decisions and resolution; do not mix these with full-text report exclusions.
Reports sought
Link screened records to reports; record report retrieval attempts.
Reports not retrieved
Record reports that were sought but could not be obtained after documented retrieval attempts. Not retrieved is different from assessed and excluded.
Reports assessed
Review the obtained full reports against prespecified eligibility criteria, resolving disagreements and linking related publications before the final study decision.
Reports excluded with reasons
For each excluded full-text report, record one primary reason for the count summary while retaining the full decision rationale in the log. Avoid counting the same report in several exclusion totals.
Included reports
Keep all eligible reports linked to their underlying included study. A study may contribute several publications, registry entries or follow-up reports.
Included studies
Count distinct included studies after linking their reports. This step changes the counting unit from reports to studies; it is not another exclusion partition.
Search and study selection
Define the search and selection scope, eligibility criteria and review procedures.
End flow
Summarize the included studies and their linked reports, ready for extraction and synthesis. Excluded records and reports remain terminal side branches.
Tables and supporting information
Search strategy register
Enter searches used, including supplementary routes and update searches.
Screening decision log
Keep record, report and study IDs distinct.
Selection definitions
Only reconcile partitions using the same count unit. Records-to-reports and reports-to-studies conversions need explicit mappings. The default records partition assumes duplicate removal is the only removal before screening. Add separate disjoint branches for automation or other removals rather than including them under duplicates.