Stream: Covid-19 Response
Topic: MeasureReport for multiple Measures
Michael Donnelly (Mar 31 2020 at 22:10):
The canonical measure in MeasureReport is 1..1. If it can actually fulfill multiple Measures, there's no way to convey it. Epic and Cerner both have measures we've defined internally. When I generate a MeasureReport, internally it's linked to the corresponding Epic Measure.
What if the definitions of the Epic Measure and the CDC Measure are the same? For interoperability purposes, we're going to want MeasureReport.measure to be the CDC Measure. That means we lose the Epic Measure. This isn't a big deal in this case - nobody cares that it's our Measure. But what if CDC and FEMA each define Measures for the same metric? Does a reporter need to create two copies of the same MeasureReport identical in every way but the measure?
Keith Boone (Mar 31 2020 at 23:02):
Right now CDC and FEMA approaches are dealing with different data sets. Measure.relatedArtifact can be used to reference related things (e.g., derived-from), though the relationships aren't exact. The CDC measure isn't derived from the Epic measure and visa-versa, but we could punt that just to get the links made.
I also think this is a valid space to look for extensions that might help reference source data.
Keith Boone (Mar 31 2020 at 23:05):
An icky alternative would be to use evalutatedResource(Measure) to point to the Epic measure if indeed the CDC Measure was actually computed based on Epic's measure rather than directly.
Might be valid, not really the intent.
Michael Donnelly (Mar 31 2020 at 23:06):
I agree - ick.
Keith Boone (Mar 31 2020 at 23:16):
Rethinking ... if you are computing the CDC MeasureReport from one or more Epic MeasureReport instances, then derived-from is exactly the right code for relatedArtifact, but MeasureReport doesn't have relatedArtifact ... but there is a relatedArtifact extension already defined that is exactly that.
Keith Boone (Mar 31 2020 at 23:17):
See https://www.hl7.org/fhir/extension-workflow-relatedartifact.html
Grahame Grieve (Mar 31 2020 at 23:36):
it would be the measure that is derived not the report
Last updated: Apr 12 2022 at 19:14 UTC
