Stream: netherlands
Topic: Metrics
Michael van der Zel (Apr 18 2017 at 07:29):
I did some model metrics on the FHIR resources and would like to see what people think. What stream would be best? I mainly see streams about implementation, which is good! But where to talk about these kind of things? See e.g. https://app.box.com/s/cak8oqk46mc4ow6tm8x8b0yi4pq3iz8p
Michael van der Zel (Apr 18 2017 at 07:30):
For me resources with a lot of elements, constraints, references are really complex to use. In the link above I filtered out the FHIR resources itself. The first column is the workgroup.
Grahame Grieve (Apr 18 2017 at 09:45):
interesting stuff.
Grahame Grieve (Apr 18 2017 at 09:51):
do any of these actually create a problem for you?
Michael van der Zel (Apr 18 2017 at 12:47):
Not persé. There is no minimum or maximum complexity. Resources are the way they are.
Michael van der Zel (Apr 18 2017 at 12:48):
There are resources that bring information together and those have more references, than the more self-contained atomic ones.
Michael van der Zel (Apr 18 2017 at 12:49):
The "coded ratio" is an indication of a resource of how useful it is for CDS, I think. I assume that resources that have a low coded ratio have more textual (human interpretation) elements.
Grahame Grieve (Apr 18 2017 at 21:10):
varies - some are more in the nature of aggregation/linking (e.g. encounter) and others are more atomic. It is logical that CDS resources will be more complex and abstract, for instance. (e.g. *Definition)
Last updated: Apr 12 2022 at 19:14 UTC
