Stream: implementers
Topic: model normaliszation and episodes of care
Erich Schulz (May 16 2016 at 01:05):
heya - this looks like where its happening... I just starting to poke at FHIR doing a little dinky project for my team. Very impressed by how coherent and workable it all looks. By way of background, I wrote an SO question (that I then answered myself) last night and @Lloyd McKenzie just helped me out with
Erich Schulz (May 16 2016 at 01:07):
so I was looking at the conceptual overlap of Encounter and EpisodedOfCare
Erich Schulz (May 16 2016 at 01:08):
it looks to me that EoC is just a pale shadow of Encounter
Erich Schulz (May 16 2016 at 01:15):
So I am curious about what I'm missing? and could/should they be merged? In my case it seems I could simply have 2 encounter.types ("APS episode" and "APS visit") using the partOf to link them together
Lloyd McKenzie (May 16 2016 at 02:33):
While there are similarities between them, they tend to be used quite differently in practice and manifest differently in systems. We tend to be guided by implementer practice when we evaluate when to merge and split concepts. Encounters are used for things like phone calls, office visits, hospital admissions and long-term stays. The focus is the contiguous interaction between practitioner or provider organization and the patient. Episodes of care tend to be condition-centric and may span multiple visits and multiple practitioners. Either may be used as a basis of billing and/or reporting.
Erich Schulz (May 16 2016 at 02:36):
I can imagine how EoC have grown up in admin systems and episode have come from clinical situation - but contained information is almost identical
Erich Schulz (May 16 2016 at 02:36):
the one key difference I can see is in the relationships
Erich Schulz (May 16 2016 at 02:40):
it seems that encounters support a tree structure (encounter can have many sub-encounters via partOf from the child) - but that there is an implied many-to-many between each encounters and episodesOfCare (from the Encouter.episodeOfCare field)
Erich Schulz (May 16 2016 at 02:43):
so it looks to me that if Encouter.episodeOfCare was changed from a reference to 0..* EpisodeOfCare to 0..* Encounterthen EpisodeOfCare would be redundant...
Erich Schulz (May 16 2016 at 02:44):
sorry this is a complet noob insight
Ewout Kramer (May 19 2016 at 09:41):
but contained information is almost identical
I don't know whether this is true (as I am not a domain expert here), but since Encounter is more about clinic/logistics and EoC is administrative, I'd expect them to be different (and in fact, the current definition of the resources is pretty different?)
Erich Schulz (May 21 2016 at 01:30):
I've done a little table to summarise the mappings: https://docs.google.com/document/d/1OlGcNBO3u3T6BrmErh28njK2ezyfVRZj2IlbbQGVHSY/edit?usp=sharing
Erich Schulz (May 21 2016 at 01:34):
in summary there is nothing that can be said with an EoC that cannot be said with an Encounter.
Chris Grenz (May 25 2016 at 16:07):
We've come to the same conclusion as we've looked at these, especially when the sequence of events places the EoC creation prior to the Encounter as has been suggested....
Chris Grenz (May 25 2016 at 16:08):
We do have a need for an EoC, but it's generally as more of a context for CarePlan and gathering the encounters and activities needed to evaluate care plan conformance.
Last updated: Apr 12 2022 at 19:14 UTC
