Stream: cql
Topic: CMS eCQMs and Unfiltered Context
Zach Smith (Feb 25 2020 at 21:32):
I've noticed that all of the CMS eCQMs are defined with context Patient, and never use the Unfiltered context, even though they are intended to be evaluated on a population of patients. Is this simply because none of the CQL implementations fully support unfiltered context evaluation yet? Or is there some other reason they're defined this way?
Chris Moesel (Feb 25 2020 at 22:17):
It's been a while since I thought much about this, but IIRC, the CQL-Based HQMF IG requires measures to be written in the Patient context only and indicates that they should be processed one patient at a time. This same IG indicates how the results of Patient-based CQL should be aggregated into population metrics; and the exact aggregate operations to use are actually defined in the HQMF (again, IIRC). As for why it is this way, it's hard to say, but I expect it may be partially for legacy reasons (i.e., this is how QDM worked). It may also have something to do with reporting requirements and patient-specific reporting via QRDA Cat I. But again, it's been a while since I've been in this space, so...
Bryn Rhodes (Feb 25 2020 at 22:55):
Yes, the CQL-Based HQMF IG (and the new Quality Measure IG) both require the use of Patient context. It's a simplification to allow the measure to be expressed from the perspective of a single patient. That doesn't mean that implementations are required to process them a patient at a time though, implementations can (and many do) run the measures as population-level queries. And yes, it does also help support patient-level reporting, it's easier to aggregate patient-level results than it is to disaggregate population level results. In FHIR, it's roughly analogous to the Patient compartment definition.
Mohammad Afaq Khan (Feb 27 2020 at 22:04):
(deleted)
Mohammad Afaq Khan (Feb 27 2020 at 23:16):
(deleted)
Last updated: Apr 12 2022 at 19:14 UTC
