Uploaded image for project: 'FHIR Specification Feedback'
  1. FHIR Specification Feedback
  2. FHIR-30098

QDM to QI-Core mapping for AdverseEvent error

    XMLWordPrintableJSON

Details

    • Icon: Change Request Change Request
    • Resolution: Persuasive with Modification
    • Icon: Medium Medium
    • US QI Core (FHIR)
    • 2.0 [deprecated]
    • Clinical Quality Information
    • QICore AdverseEvent
    • 8.2 Adverse Event Mapping
    • Hide

      The QDM AdverseEvent code should map to AdverseEvent.event as mapped. The description currently states that the code should map to suspectEntity. This is incorrect. The description needs to change as follows:
      "QDM includes an attribute code that represents the specific type of event that occurred, consistent with AdverseEvent.event. QDM does not include an attribute to address the additional elements available in QI-Core: AdverseEvent.suspectEntity (the suspected cause), or the AdverseEvent.resultingCondition. As an example to differentiate these elements:

      • AdverseEvent.event = fall
      • AdverseEvent.resultingCondition = fracture
      • AdverseEvent.suspectEntity = area rug

      QDM version 5.6 (and earlier versions) only address one of these elements, the event.
      Therefore, QDM AdverseEvent code maps to AdverseEvent.event. Measure developers seeking to retrieve data about the cause of an AdverseEvent may be able to relate the occurrence timing of a potential causative event and the AdverseEvent.event timing. Further detail about the AdverseEvent will require use of FHIR or potentially a subsequent version of QDM after QDM 5.6."

      Show
      The QDM AdverseEvent code should map to AdverseEvent.event as mapped. The description currently states that the code should map to suspectEntity. This is incorrect. The description needs to change as follows: "QDM includes an attribute code that represents the specific type of event that occurred, consistent with AdverseEvent.event. QDM does not include an attribute to address the additional elements available in QI-Core: AdverseEvent.suspectEntity (the suspected cause), or the AdverseEvent.resultingCondition. As an example to differentiate these elements: AdverseEvent.event = fall AdverseEvent.resultingCondition = fracture AdverseEvent.suspectEntity = area rug QDM version 5.6 (and earlier versions) only address one of these elements, the event. Therefore, QDM AdverseEvent code maps to AdverseEvent.event. Measure developers seeking to retrieve data about the cause of an AdverseEvent may be able to relate the occurrence timing of a potential causative event and the AdverseEvent.event timing. Further detail about the AdverseEvent will require use of FHIR or potentially a subsequent version of QDM after QDM 5.6."
    • Floyd Eisenberg/Peter Muir: 13-0-0
    • Correction
    • Non-substantive

    Description

      On this qdm-to-qicore mapping page, 8.2 Adverse Event, table shows that QDM.code maps to AdverseEvent.event, however, the descriptions above the table says “QDM AdverseEvent references only the suspectEntity.instance (as code) and category (as type).”, which suggests AdverseEvent.suspectEntity.instance would be the primary code, not the AdverseEvent.event. I see the fhir-modelinfo-4.0.1.xml shows the primaryCodePath=“event” that matches the table. It seems to me that definition of suspectEntity.instance makes it a better fit for primary code than the event element.

      The main issue is that the QDM definition of the AdverseEvent datatype is the suspectEntity but it should probably be AdverseEvent.resultingCondition (reference to Condition), I.e., what happened to the patient. And there is a missing attribute, suspectEntity ( AdverseEvent.suspectEntity.Instance ). As stated in QDM v5.5 and 5.6, AdverseEvent defines the suspectEntity, not the event itself, and it does not have the attribute to reference the resulting condition. Thus, the mapping from AdverseEvent code should be to  AdverseEvent.suspectEntity.Instance unless QDM were to update the definition.

      Attachments

        1. Adverse Event Modeling December2020.doc
          57 kB

        Activity

          People

            bryn.rhodes Bryn Rhodes
            feisenberg Floyd Eisenberg
            Watchers:
            2 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: