Use the registry history and publication together. A discrepancy is a finding to explain, not an automatic conclusion about intent, validity or misconduct.
Comparison sequence
- Confirm identity. Match registry identifier, sponsor, intervention, population, recruitment dates and locations.
- Check registration timing. Distinguish prospective registration from records first posted after enrolment began.
- Compare outcomes precisely. Review outcome name, measurement method, time point and hierarchy—not only similar wording.
- Inspect the history. Record dated amendments and whether the publication explains material changes.
- Preserve multiple reports. One trial may produce protocol, primary, secondary and follow-up papers.
Why timing matters
WHO trial-registration policy identifies a minimum data set and supports prospective registration. Registry platforms can also retain change histories. Interpretation should therefore separate what was registered before recruitment from later corrections or updates.
Worked use case
The publication identifies the same trial but reports a secondary outcome as the main paper outcome. Check the original registry version, protocol and statistical analysis plan. Record whether the hierarchy changed and whether the authors disclosed the reason.
Evidence boundary
Registry records may contain delayed updates, administrative errors or evolving terminology. A mismatch may signal selective reporting, but this conclusion requires protocol-level context and cannot be made from a single field comparison.
Primary sources
- World Health Organization. WHO Trial Registration Data Set. Retrieved 5 August 2026.
- ClinicalTrials.gov. Protocol registration data element definitions. Retrieved 5 August 2026.