How-to
How to Check a Health Story Against ClinicalTrials.gov
Compare a study announcement with its registered outcomes, record history and results. A reporting worksheet that keeps the clinical limits visible.
What can a trial registration establish?
ClinicalTrials.gov helps you compare a study's registered design with the way its findings are described. Start with the study identifier, the measured outcomes and the version of the record relevant to your question. The comparison can reveal a question worth reporting; it cannot by itself establish that a treatment works or that a sponsor acted improperly.
The registry's guide to reading a study record describes the Study Details, Results and Record History tabs. A study may have several primary outcomes. The record history matters because the current display alone does not show what the entry said at an earlier date.
How do you compare a press release with the record?
Match the NCT identifier first. Similar drug names, study phases or sponsors are not enough. Then save the announcement and the specific registry version you used.
| Announcement claim | What to inspect | Question to carry forward |
|---|---|---|
| A positive result | Primary and secondary outcome measures | Which outcome is the claim about? |
| Benefit in a subgroup | Protocol, analysis plan and available results | Was this analysis planned, and how was it handled? |
| Recruitment stopped | Recruitment status and explanation | What reason is recorded, and when? |
| A changed endpoint | Record History and relevant documents | What changed, when, and why? |
| Results are forthcoming | Completion dates and reporting context | What can the sponsor clarify about timing? |
The registry is a source to investigate, not a shortcut around reviewing the methods and obtaining qualified comment. Its research-use guidance also explains why registration data need careful interpretation.
What does a responsible example look like?
In a hypothetical trial, a company announces an improvement in a secondary measure. The record lists a different primary measure. You can accurately report that the release emphasizes the secondary result. You cannot infer from that omission alone that the primary outcome failed.
Ask for the primary result, the analysis population, uncertainty estimates and the relevant protocol or statistical analysis plan. Ask an independent specialist whether the comparison you intend to make is meaningful. If the evidence is unavailable, describe that limit instead of filling it with a clinical conclusion.
A status change deserves the same restraint. "Terminated" is a recruitment-status field to investigate, not a synonym for unsafe or ineffective. Read the explanation and seek clarification before interpreting the reason.
Which shortcuts should you avoid?
Do not claim every study registers before enrollment or that every record has the same results deadline. Coverage, exemptions, extensions and other circumstances matter; consult the registry's policy FAQ for the applicable context. A missing results posting is not enough to pronounce a legal violation.
Do not use a historical drug controversy as a stand-in for the study you are covering. Verify the actual endpoints and chronology. Keep treatment advice separate from your reporting about a public record.
What should the published piece let a reader check?
Give the NCT identifier, the record version or retrieval date, the exact claim under examination and links to the supporting material. Separate the sponsor's statement, the independent interpretation and your own analysis. Use the reported-angle worksheet before drafting: if the central claim depends on a clinical inference you cannot support, narrow the claim or keep reporting.
Keep reading