From Michael Ryan (NCQA):

NCQA would like to suggest an agenda topic for the next QDM User Group meeting on 10/21. We are seeking guidance on the appropriate FHIR mapping for the QDM Datatype, Symptom.

The current QI Core mapping suggests using FHIR.Observation with category constrained to ‘symptom’, found here. The documentation does make it clear that this is a category code that would need to be added by the measure developer, as it is not included in ObservationCategoryCodes currently. We have concerns that, since this would be a user-defined category, we may run into mapping and data issues with end users.

Our use cases:

1. Symptom is used as one method of identifying a depression-related encounter. This is used in conjunction with three other unioned Encounter/Diagnosis combinations.

( ( ["Encounter, Performed": "Depression Case Management Encounter"] ) CaseManagementEncounter

   with ( ["Symptom": "Symptoms of depression (finding)"] ) DepressionSymptoms

        such that DepressionSymptoms.prevalencePeriod starts same day as start of CaseManagementEncounter.relevantPeriod

  )

 2. Symptom is used as a method of identifying Frailty, as part of a larger definition:

or exists ( ["Symptom": "Frailty Symptom"] FrailtySymptom

                                                where FrailtySymptom.prevalencePeriod overlaps "Measurement Period"

                )

Example 1 above is a Direct Reference Code using 394924000 (SNOMEDCT).

Example 2 references a value set consisting of mainly SNOMEDCT codes, along with a handful of ICD10CM codes.  

See attachment for a snapshot sample of the contents of the “Frailty Symptom” value set for context.

We are prepared to move forward with the suggested QDM-to-QICore solution, but would like clarification on the rationale behind this route, as well as discuss possible alternatives.