Literature Guide
How to Read a ClinicalTrials.gov Study Record
What each field of a ClinicalTrials.gov record means, how registration differs from results, and why an NCT number shows a study was registered, not that it worked.
[Research Use Only] This guide is for controlled laboratory research workflows only. It is not for human or veterinary use and does not provide applied-use guidance.
What a registry record is
ClinicalTrials.gov is a registry of studies, maintained by the US National Library of Medicine. A record is submitted by the study’s responsible party and describes what they intend to do, or did. It is a declaration, kept current by the sponsor, and it exists so that studies can be found and tracked whether or not they are ever published. That purpose is worth holding onto while reading one: the registry documents existence and design, not merit.
Identifying the study
The NCT number is the registry’s own identifier, assigned once the record has been released by the responsible party and passed review. Like a PMID it is opaque and stable, which makes it the right thing to cite. Titles come in two forms — a brief title written for a general reader and an official title carrying the formal design language — and the two can read very differently.
- NCT number: the stable identifier; the correct thing to cite
- Brief title: plain-language, written for findability
- Official title: the protocol title, usually stating design and population
- Other identifiers: sponsor protocol numbers and other registries, where cross-registered
Design fields, and what they constrain
The design fields describe what kind of evidence the study can produce. Study type separates interventional studies, where participants are assigned to an intervention, from observational ones, where they are not. Within interventional records, allocation, intervention model and masking describe how assignment and blinding were handled, and each is a constraint on what a result can support.
| Field | What it says | Why it matters when reading |
|---|---|---|
| Study type | Interventional or observational | Observational studies do not assign the intervention, so causal reading is weaker |
| Phase | Applies to interventional studies only | Describes the stage of investigation, not a level of approval |
| Allocation | Randomised or non-randomised | Randomisation is what makes groups comparable at baseline |
| Masking | Who was unaware of assignment | Open-label results are more exposed to expectation effects |
| Intervention model | Parallel, crossover, single group and so on | Determines what the comparison actually is |
| Enrollment | Number of participants, actual or estimated | A small enrolment bounds what can be detected |
Outcome measures
The primary outcome is what the study was designed to measure, specified in advance and at a stated time point. Secondary outcomes are additional measures. The order matters: a study is powered for its primary outcome, and a result on a secondary measure carries less weight than the same result on the primary one. Because the registry records what was specified before the study ran, comparing the registered outcomes with those reported in a later publication is one of the more useful things a reader can do with a registry record.
- Primary outcome: pre-specified, with its measurement time frame
- Secondary outcomes: additional, and not what the study was powered for
- Changes to outcomes are visible in the record’s history
- A publication emphasising a measure the registry lists as secondary is worth noticing
Status, dates and results
Recruitment status describes where the study is in its lifecycle — recruiting, active but not recruiting, completed, terminated, withdrawn. Status is maintained by the sponsor, and published analysis of the registry has found a meaningful share of records carrying an unknown status because they had not been verified within the expected window. Results information, where present, is submitted separately from the registration and is structured differently.
- RegistrationThe responsible party submits the planned study; an NCT number is assigned after review.
- ConductStatus and dates are updated as recruitment proceeds. Records can go stale here.
- Results postingSummary results may be submitted to the registry, separately and in a different structure.
- PublicationA peer-reviewed paper may follow, and may not agree with the registry entry.
What a record does not establish
Registration means a study was declared. It is not evidence that the intervention works, that it is safe, that any regulator has assessed it, or that the study was well conducted. A completed status means recruitment and follow-up ended, not that the result was positive — studies stop for many reasons, and a terminated record may be terminated for futility, for recruitment failure, or for reasons unrelated to the intervention.
- Not evidence of efficacy or safety
- Not an approval, and phase is not an approval stage
- Not a guarantee the record is current
- Not necessarily consistent with the eventual publication
- Not a complete picture: some studies are never registered at all
Research Checklist
- Confirm batch identity and records.
- Document all preparation inputs.
- Keep use within controlled laboratory workflows.
- Do not infer applied-use suitability from guide content.
Frequently Asked Questions
Does a phase 3 record mean a compound is nearly approved?
No. Phase describes the stage of investigation a study sits at, not a regulatory decision. Studies at every phase can and do fail, and registration says nothing about whether any authority has evaluated the intervention.
Why do the registry and the published paper disagree?
It happens, and it is documented. Registry entries and publications are prepared separately, at different times, under different requirements. Where they differ, the difference itself is informative — particularly if outcomes have moved between them.
Is a study with no results posted a bad sign?
Not necessarily. Results posting depends on the applicable requirements and on the sponsor, and absence may reflect timing rather than outcome. It does mean the registry cannot tell you what happened, and another source is needed.
Sources
The technical statements in this guide are drawn from the following. Where a definition is contested or a figure depends on method, the guide says so rather than picking one.
- How to avoid common problems when using ClinicalTrials.gov in research: 10 issues to considerPMC, National Library of MedicineUsed for the limits of quality-control review, record types and their proportions, the treatment of unknown recruitment status, the separation of registration from results information, and the documented discrepancies between posted results and other sources.
- When will the NCT Number for my study be assigned?National Library of Medicine Support CenterUsed for when an NCT number is assigned and the review that precedes it.