Best Practices


Accepting data from partner information systems


  • Primary or ICMJE-approved registries should be prepared to develop mechanisms for accepting data from Partner Registries or other appropriate data providers.
  • Primary or ICMJE-approved registries that accept data from Partner Registries or other appropriate data providers must ensure that these data meet the WHO criteria for data content, quality and validity before being incorporated into the Primary Registry database.

Details and discussion: DataFromPartners

Audit trails


  • Significant changes that must be included in audit trails are changes that alter the meaning of text in a data item.
  • Registries may need, for legal reasons such as data privacy laws, to make significant changes to some fields and to suppress the original information. In such cases the audit trail should indicate that the change was made and when, but should state that the original information is no longer being displayed for legal reasons and that the original information is held on file by the registry.

Details and discussion: AuditTrails?

Authenticity of the registrant


  • Registries must contact and receive replies from the person submitting the registration request using details cited in trial registration request to verify that this will be an appropriate contact.
  • When possible, registries should obtain institutional contact details

Details and discussion: RegistrantAuthenticity?

Authenticity of the trial


  • Registries should obtain evidence of third-party confirmation that a trial exists.

Details and discussion: TrialAuthenticity?

Data standard: Countries of recruitment


  • Primary Registries adopt country names listed in ISO-3166-1
  • Allow for local customizations if necessary by making use of the ISO-3166 user-assigned code elements feature

Note: this feature is already implemented in the Rebrac Clinical Trials application

Data standard: Secondary Identifiers


  • Ensure that the registry software allows responsible registrants to record multiple Secondary Identifiers for a trial and that an issuing authority can be recorded for each Secondary Identifier.

Details and discussion: DataStdSecondaryIds

Prospective translation of the TRDS


  • For registries requiring the responsible registration to submit "free text" data items in all languages used by the registry:
    • The registry must not accept the registration or an update unless data items are available at least in English, but preferably in all languages served by the registry.

Details and discussion: TrdsTranslation?

Provisional registration


  • Provisional registration is not recommended and should not be used by registries.

Retrospective registration of trials


  • Registries can choose to accept retrospective registrations if they wish, but there is no compulsion to do so.
  • If registries accept retrospective registration then they should:
    • make sure that the 'Date of Registration' and the 'Date of First Enrolment' fields show that a trial has been registered retrospectively,
    • alert readers by highlighting or flagging retrospectively-registered trial records, that is when a trial is registered after the expiry of a grace period of 30 days after first enrolment,
    • promote prospective registration as the norm.

Details and discussion: RetrospectiveRegistration?

Unique trial registration


  • Registries ensure that their guidance and electronic forms emphasize the importance of the Secondary Identifiers field and that it can be used to enter trial identifiers.
  • Registries have processes to ensure that a trial is not already registered by searching their own databases. They should not allow the same trial to be registered more than once on their own database.
  • Registries either check themselves and/or ask the registrant to confirm that they have checked the ICTRP search portal to see if similar trial registration records exist. This could be a search for key words in title, intervention, sponsor, source of funding or contact names. If matches are found, registries should insist that registrants provide the relevant Secondary Identifiers.

Details and discussion: UniqueTrialRegistration?

Validity: Is the information registered correct and complete?


  • Registries should contact registrants, their organizations or regulatory agencies as part of the editorial process whenever there is a need to clarify unclear or dubious items of information.
  • Registries should have a clear quality control process to ensure meaningful data in trial registration records
  • Registries should encourage registrants via the registration process and via subsequent communication to ensure all 20 Trial Registration Data Set items have been provided.

Details and discussion: RecordValidity?


Data standard: Health condition(s) or problem(s) studiedDataStdHealthCondition?
Data standard: Intervention(s)DataStdIntervention?
Data standard: Recruitment statusDataStdRecruitmentStatus?
Data standard: Study type (including study design and phase)DataStdStudyType?
Registering observational studiesObservationalStudies?