Compare each protocol criterion with the paper's reported sample and outcomes, record every deviation, then decide eligibility and state the rationale in your methods. The decision becomes defensible when the deviation is named, judged against the protocol's stated purpose, and reported alongside the included study.
Run the comparison in order
- Extract the protocol's eligibility criteria into a table with one row per criterion: population, sample characteristics, outcomes, design, and any date or setting limits. Enter the paper's reported value in the adjacent column so each match or mismatch is visible.
- Mark each row as met, deviated, or unclear. For a deviated row, quote the paper's reported sample or outcome and note where it appears in the report. For an unclear row, record what is missing and check the registration record or supplementary files before deciding.
- Judge each deviation against the purpose the criterion serves in your review question. A sample that differs in a characteristic the protocol treats as defining changes the answer; a difference in a characteristic the protocol does not restrict leaves the study eligible.
- Record the eligibility decision and a one- or two-sentence rationale for each study, including studies excluded at this stage. This gives your methods section and supplementary table a traceable basis for the inclusion choice.
- Report the decision and its rationale in the methods and supplementary table, and show how many records were excluded for sample or outcome deviations in the review flow diagram.
Where to verify reported details
ClinicalTrials.gov defines structured registration and results fields for study identification, status, design, eligibility, locations, outcomes, and reported results, so a registration record can supply the protocol-side values you compare against. PubMed supports field tags, filters, Advanced Search, and citation matching for locating the paper and related records. The Cochrane Handbook describes planning searches, selecting sources, and managing records for systematic reviews, which covers the selection stage where this decision sits. PRISMA 2020 provides a checklist and flow diagrams for transparent reporting of what methods were used and what studies were found and included.
Keep the record consistent
Apply the same deviation rule to every study so the supplementary table reads as one decision framework. When a paper reports an outcome the protocol did not list, note it as an additional outcome and state whether it enters any synthesis. When the paper omits a protocol outcome, mark the row unclear and check the registration record before finalizing the decision.