In-depth practical guides

Read a clinical trial register

Connect protocol, history, posted results and publications without counting one trial twice.

Updated :

Consultation of scientific publications and bibliographic references.

AI-generated illustration

Does the record describe a protocol, a result or a publication?

A register describes a study and updates. It may list objectives and recruitment without reporting results. ICTRP gathers registers: an aggregate and its original register may describe one trial.

  1. Identify the trial

    Record the original registry ID, secondary IDs and sponsor. A similar title is insufficient as a key; inspect links between registrations.

  2. Fix the version

    Separate registration, planned or actual start and completion, and last update. When history exists, compare primary outcomes and planned enrolment across versions.

  3. Distinguish documents

    Locate protocol, posted results and article separately. Link the article through the trial ID and check population, intervention and period before attributing its results.

Reading workshop

A changed endpoint is not a result

Fictional file DEMO-T18 compares protocol versions without a medical conclusion.

Fictional data and records for learning: none of these results comes from a live service query.

Eligibility age and primary endpoint timing

v1 is dated 1 March; v2, 15 July. Recorded start is 20 June. The history shows a change but does not by itself establish its reason.

If the table exceeds the screen, scroll horizontally; arrow keys work when the region is focused.

Eligibility age and primary endpoint timing
StepEvidence or valueReadingDecision
Study identityDEMO-T18 in both versionsOne tracked study file.Retain identifier and v1/v2 documents.
Eligibility criterionv1: ages 18–65; v2: 30–65Declared eligible population changes.Record the date and seek the protocol explaining the amendment.
Primary endpointScore S at 8 weeks, then 12 weeksMeasurement timing differs.Cite definition and timing with version; do not treat it as unchanged.
Available resultsNo score S value in the fileNo effect direction established.Seek posted results and linked publications without inventing conclusions.

Which sentence respects the available evidence?

Open each proposition to compare its diagnosis. More checks may still be needed.

The trial failed because results are absent.

No: absence in this file does not establish result direction.

The score is better at 12 weeks.

No: a scheduled time is not an observed value.

v2 changes declared population and timing; clinical result unestablished.

Appropriate. Read protocol and results without a treatment recommendation.

Retained protocol comparison

File: DEMO-T18; invented identifier
v1: 1 March; ages 18–65; score S at 8 weeks
Recorded start: 20 June
v2: 15 July; ages 30–65; score S at 12 weeks
Amendment reason and effect: to document
Score results: absent from supplied documents
Conclusion: protocol change, not a result

Attach version, population and measurement timing when citing the endpoint. Separate start date, amendment date and results consultation date.

Documentation of the rules, separate from fictional examples

Compare other situations and exercises

Two registers, one trial

Use secondary IDs to avoid counting two independent observations.

Completed study without results

Describe displayed status and available results; do not infer positive or negative outcomes from missing results.

Worked exercise

Planned enrolment, actual enrolment and results differ

Fictional example: the values and documents teach the method and do not describe a real organisation.

All information concerns the same fictional record. Determine what can be calculated and what remains unknown.

If the table exceeds the screen, scroll horizontally; arrow keys work when the region is focused.

Starting information
FieldObserved information
Planned participants80
Enrolled participants60
Recruitment statusFinished
Primary-outcome resultUnavailable in the records consulted

Your checks

  1. Identify planned and actual enrolment.
  2. Read the displayed status definition.
  3. Seek outcome results and their date.
Read the answer

Answer and decision

Enrolled participants represent 75% of the stated target. This measures recruitment, not intervention effectiveness. Recruitment status and unavailable results establish neither clinical benefit nor clinical failure.

If the case changes

For results published elsewhere, check trial ID, protocol, analysed population and version before linking them.

Record to retain

Keep IDs, qualified counts, status, outcomes, versions and actually consulted results.

Documentation reference : WHO — Trial Registration Data Set · Consulted on

Record to retain

  • Original and secondary IDs, sponsor and registry.
  • Version, planned/actual dates and study outcomes.
  • Separate links to protocol, results and article.

Official references

Related source profiles

Related protocols

Checks before reaching a conclusion

  • Do identifiers allow duplicates across registers to be detected?
  • Are the version and planned versus actual dates distinguished?
  • Are protocol, submitted results and article identified separately?

Frequently asked questions

Does ICTRP publish all results?

No: return to the original register and look for results documents and linked publications.

Is “completed” status enough?

It describes study progress, not conclusions or the presence of an article.

Why compare several versions of the recording?

To identify the changes of information useful in reading. Differences must be described and contextualized, without automatically concluding their cause or effect on the results.

TOOLS / REGISTERS

Information to record

Locate protocol, posted results and article separately. Link the article through the trial ID and check population, intervention and period before attributing its results.

  • Access date and time zone
  • Exact URL
  • Document identifier
  • NCT identifier
  • Producer / publisher
  • Version / revision
  • Event, publication and observation dates
  • Sample / population
  • Acceptance criterion
  • Observed status
  • Limit / uncertainty
  • Decision / next action
  • Review owner
  • Next review

DOCUMENTED REFERENCE

Clinical study: five fields with different dates

These ClinicalTrials.gov API fields belong to identificationModule and statusModule inside protocolSection. They distinguish identification, study conduct and publication; not all describe a medical event.

Clinical study: five fields with different dates
ReferenceValue or formWhat it identifies
nctIdNCT + 8NCT followed by eight digits: the registry identifier, distinct from the sponsor’s internal number.
studyFirstPostDateStructdateFirst record availability after NLM quality review; not the study start.
primaryCompletionDateStructdate + typeCompletion of data collection for primary outcomes, with estimated or actual type. Precision may be month-only.
resultsFirstPostDateStructdateFirst availability of summary results after quality review; some may appear earlier with QC comments.
lastUpdatePostDateStructdateLatest publication of record changes, different from their submission date.

Implication for the record

Keep field names, version and precision. Do not invent a day for a month-only date. Quality review and an NCT do not establish efficacy, safety or treatment approval.

Does a completion date establish when results became public?

No. Completion of collection, results submission and publication are separate events. Check the results posting field and any QC comments. The last update may concern another part of the record.

Primary documentation

Documentation consulted on .

DEEPER READING / PRACTICE

Read a registry record with its status and dates

A register describes a study and its protocol. Reading focuses on what is recorded, the available versions and the possible results published, without transforming the recording into a demonstration of effectiveness.

Updated on .

Priority checks

Release identifier, title, promoter, calendar, status and criteria indicated. Compare the version consulted with the documents or results to which it references.

Interpretation pitfalls

An announced protocol does not prove the completion of the study or its outcome. Dates and statutes can describe different steps and require careful reading.

Working record to retain

The reading sheet indicates version, dates, status, announced elements and links to the available results. The conclusions remain limited to the documentation actually consulted.