Uploaded image for project: 'FHIR Specification Feedback'
  1. FHIR Specification Feedback
  2. FHIR-31624

All-or-nothing scoring should use "implies"

    XMLWordPrintableJSON

Details

    • Icon: Change Request Change Request
    • Resolution: Persuasive
    • Icon: Medium Medium
    • US Quality Measures (FHIR)
    • 0.1 [deprecated]
    • Clinical Quality Information
    • Home [deprecated]
    • Hide

      Agreed, update all-or-none component numerator membership formulas to:

      define "ComponentMeasure1 Numerator Membership": // Repeat for each component
         ComponentMeasure1."Denominator Membership"
           implies ComponentMeasure1."Numerator Membership"

       

      And note that if "for which they are eligible" is interpreted as the initial population, then that would mean:

      define "ComponentMeasure1 Denominator Membership":
        ComponentMeasure1."Initial Population"
          implies ComponentMeasure1."Denominator Membership"

      and

      define "ComponentMeasure1 Numerator Membership":
        ComponentMeasure1."Initial Population"
          implies (ComponentMeasure1."Denominator Membership"
            and ComponentMeasure1."Numerator Membership")

       

       And add explanatory text:
      "Consider the example of a composite that includes a breast cancer screening measure and a colorectal cancer screening measure. For an individual that is male, they are only eligible for the colorectal cancer screening measure, so the fact that they do not appear in the denominator or numerator of the breast cancer screening measure should not remove them from the numerator of the composite measure."
      Provide a brief explanation of "implied": Note that implies is a logical operator within CQL. "X implies Y" roughly translates to narrative text as "if X is true, then Y must be as well"
      And a link to a definition of implies in CQL: https://cql.hl7.org/09-b-cqlreference.html#implies

      Show
      Agreed, update all-or-none component numerator membership formulas to: define "ComponentMeasure1 Numerator Membership": // Repeat for each component    ComponentMeasure1."Denominator Membership"      implies ComponentMeasure1."Numerator Membership"   And note that if "for which they are eligible" is interpreted as the initial population, then that would mean: define "ComponentMeasure1 Denominator Membership":   ComponentMeasure1."Initial Population"     implies ComponentMeasure1."Denominator Membership" and define "ComponentMeasure1 Numerator Membership":   ComponentMeasure1."Initial Population"     implies (ComponentMeasure1."Denominator Membership"       and ComponentMeasure1."Numerator Membership")    And add explanatory text: "Consider the example of a composite that includes a breast cancer screening measure and a colorectal cancer screening measure. For an individual that is male, they are only eligible for the colorectal cancer screening measure, so the fact that they do not appear in the denominator or numerator of the breast cancer screening measure should not remove them from the numerator of the composite measure." Provide a brief explanation of "implied": Note that implies is a logical operator within CQL. "X implies Y" roughly translates to narrative text as "if X is true, then Y must be as well" And a link to a definition of implies in CQL: https://cql.hl7.org/09-b-cqlreference.html#implies
    • Peter Muir/Bryn Rhodes:22-0-0
    • Correction
    • Compatible, substantive

    Description

      All-or-nothing scoring is "patients who received all preventive services for which they were eligible" This means that a patient should only be removed from the numerator for a component measure that they were also in the denominator for. So for patient-based logic, this means "Denominator Membership implies Numerator Membership"

      Attachments

        Activity

          People

            Unassigned Unassigned
            bryn.rhodes Bryn Rhodes
            Watchers:
            2 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: