How do I resolve sample-size or outcome discrepancies between a trial paper and its registration?

Extract the sample size and primary outcome from both the registration and the paper, then check the paper's methods and results sections for explanations such as protocol amendments before deciding whether the discrepancy is explained.

1. Pull the same fields from both records. From the registration, record the target enrollment, the primary outcome as written, and the registration date. ClinicalTrials.gov defines structured registration and results fields covering study design, eligibility, outcomes, and reported results, so use those field labels to locate each value. If the trial is registered elsewhere, the WHO International Clinical Trials Registry Platform search portal provides a single access point to records supplied by participating registries, which helps you find the correct record and its registry of origin.

2. Compare like with like. Match the paper's stated sample size to the registration's target enrollment, and the paper's primary outcome to the registered primary outcome. Note whether the paper reports enrolled, randomized, or analyzed participants, and compare each against the corresponding registered figure. Record the exact wording of both outcome statements so a rewording is visible alongside a substantive change.

3. Search the paper for an explanation. Read the methods and results sections for statements about protocol amendments, changes to the primary outcome, or revised sample-size calculations. A registered target and a final analyzed sample can differ for reasons the paper states; the explanation determines whether the discrepancy is documented or unresolved.

4. Check the registration history where available. Look for version history or amendment records in the registry entry. These show when a field changed relative to the registration date, which helps you place the change in sequence with the reported analysis.

5. Document the resolution and cite reconciled values. Write down the registered value, the published value, the explanation found, and the section where it appears. Cite the reconciled values with both the registration identifier and the paper, so readers can trace each figure to its source. Where no explanation appears, describe the discrepancy as unresolved and state which record you checked.

When the discrepancy remains unexplained

Report the specific fields that differ and the sources consulted. COPE retraction guidelines describe purposes, grounds, and responsibilities for correcting the research literature, which gives context for how journals handle documented corrections; use that context to decide whether to raise the issue with the authors or editor.