#155: correlating observations with encounters
----------------------------------+----------------------------
Reporter: lv | Owner: dconnolly
Type: design-issue | Status: assigned
Priority: major | Milestone: data-domains3
Component: data-stds | Resolution:
Keywords: data-quality cdm-etl | Blocked By:
Blocking: |
----------------------------------+----------------------------
Comment (by lv):
Replying to [comment:24 dconnolly]:
> Interesting approach from MCRF:
>
> > ... our approach from the beginning was to artificially create
encounters based on patient, date, provider, place of service - rather
than trying to pull events, which we've found on other projects/network
collaborations to generate noise.
> - [http://listserv.kumc.edu/pipermail/gpc-dev/2016q2/002935.html CDM
v3 procedures/encounter EPIC installations] discussion in gpc-dev
Just to clarify, "Place of Service" is Marshfield terminology. Local codes
are mapped to Visit Type Concept Id (OMOP data model) and Enc_Type
(PCORnet data model) of AV, ED, IP, OA, OT, etc. We talked about this at
Hackathon2.
It doesn't work for all data. For the tumor registry, we use Accession
Number and Sequence Number to define unique encounters.
--
Ticket URL:
<http://informatics.gpcnetwork.org/trac/Project/ticket/155#comment:25>
gpc-informatics <http://informatics.gpcnetwork.org/>
Greater Plains Network - Informatics
_______________________________________________
Gpc-dev mailing list
[email protected]
http://listserv.kumc.edu/mailman/listinfo/gpc-dev
