Uploaded image for project: 'FHIR Specification Feedback'
  1. FHIR Specification Feedback
  2. FHIR-44746

Change in AuditEvent type/subtype to category/code issues with DICOM mapping

    XMLWordPrintableJSON

Details

    • Icon: Change Request Change Request
    • Resolution: Persuasive with Modification
    • Icon: Highest Highest
    • FHIR Core (FHIR)
    • R5
    • Security
    • AuditEvent
    • Hide

      Revert to previous model using type and subtype to support better the broad application of auditEvent that benefits from a small distribution of types of audit events, and a further refinement of subtypes. This also better supports compatibility with ASTM, IETF, DICOM, and IHE. The AuditEvent is a cooperative resource with these other standards. 

      Show
      Revert to previous model using type and subtype to support better the broad application of auditEvent that benefits from a small distribution of types of audit events, and a further refinement of subtypes. This also better supports compatibility with ASTM, IETF, DICOM, and IHE. The AuditEvent is a cooperative resource with these other standards. 
    • Kathleen Connor / Johnathan Coleman : 10-0-0
    • Correction
    • Non-compatible

    Description

      In R5, AuditEvent.type was changed from cardinality 1..1 to AuditEvent.category with cardinality 0...  AuditEvent.subtype was changed from cardinality 0.. to AuditEvent.code with cardinality 1..1.

      This causes an inconsistency in the mapping between the DICOM AuditMessage and the FHIR AuditEvent Resource.

      The mapping in R4 is:

        EventIdentification.EventID 1..1  --> type 1..1
        EventIdentification.EventTypeCode 0..* --> subtype  0..*

       changes with R5 AuditEvent, as reported in https://www.hl7.org/fhir/R5/auditevent-mappings.html#dicom

        EventIdentification.EventID 1..1 --> category 0..*
        EventIdentification.EventTypeCode 0..* --> code 1..1

      The cardinality in DICOM for the elements, and the EventTypeCode is 0..* based on the schema, and also is U = This element or attribute is user optional. The creator may include it or omit it.

      So, the question is: In which way do I have to evaluate the "code" element if the EventTypeCode is omitted in the DICOM Audit or if in the DICOM audit there are more than 1 EventTypeCode ? 

      Perhaps this change can be reverted for R6 or type and subtype added back in as additional elements for this mapping?

      Attachments

        Activity

          People

            john_moehrke John Moehrke
            lduncan Luke Duncan
            Watchers:
            3 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: