Home > FAQ > How can I check whether a trial's registered outcomes match the outcomes reported in its paper?

How can I check whether a trial's registered outcomes match the outcomes reported in its paper?

October 26, 2026
trial outcome verificationregistered outcomes comparisonClinicalTrials.gov recordsWHO ICTRP searchevidence synthesis checks

Retrieve the trial's registry record, list its registered primary and secondary outcomes, and compare each one against the outcomes the paper reports before citing the paper's results. ClinicalTrials.gov defines structured registration fields for outcomes alongside study identification, design, eligibility, locations, and reported results, so the registry record gives you a defined place to start.

  1. Find the registration record. Search ClinicalTrials.gov by condition, intervention, location, status, study type, or dates, or use the WHO ICTRP Search Portal, which provides a single access point for records supplied by participating registries. Expected result: a registry entry with a registration number you can match to the paper's stated registration.
  2. Extract the registered outcomes. Copy the registered primary and secondary outcomes exactly as recorded, keeping their wording, time points, and any specified measures. Expected result: a written list you can compare line by line.
  3. Extract the paper's reported outcomes. List the primary and secondary outcomes the paper states in its methods and results, using the paper's own labels. Expected result: a parallel list in the same order as the registry list.
  4. Compare the two lists. Check whether each registered outcome appears in the paper with the same definition, and whether the paper reports outcomes that were not registered. Expected result: a set of matches, wording differences, and any unregistered or unreported outcomes.
  5. Record discrepancies before citing. Note each mismatch next to the result you plan to cite, so the discrepancy travels with the citation into your synthesis. Expected result: a documented comparison you can reference when interpreting the findings.

A wording difference alone does not establish that the outcome changed; treat it as unresolved until you check whether the registry record and the paper describe the same measure and time point. Where the registry record and the paper conflict, describe what each source states rather than assuming one is correct.

←
PreviousHow can I check whether a source in a due-diligence memo has been superseded by a newer version?
NextHow can I check whether a trial's registration record has been withdrawn or terminated?
→