#158: usable view of LOINC lab terms
-------------------------+----------------------------
Reporter: rwaitman | Owner: nateapathy
Type: enhancement | Status: assigned
Priority: major | Milestone: data-domains3
Component: data-stds | Resolution:
Keywords: | Blocked By:
Blocking: 68 |
-------------------------+----------------------------
Comment (by nateapathy):
Jim,
This is good input and will help with the conversation later this week. I
have a couple of questions and comments.
1) We built the CMH hierarchy based on the LOINCPARTS ontology from UMLS,
and it is not nearly as flat as the one Nathan Wilson put together, if
that hierarchy is currently under the "GPC - Laboratory measurements"
hierarchy on Babel. I agree, the GPC - Laboratory measurements hierarchy
is far too flat, but there are less flat scriptable UMLS-based ontologies.
2) You are correct, we can extract the equivalent of flowsheet hierarchies
from Cerner, however those hierarchies will not align across sites since
different sites inevitably structure the views of their results in
different ways, so we have always aligned all of our clients with the
existing hierarchy that you see on Babel for CMH - Laboratory Results.
Since ontology paths must align for the GPC queries to be shared, I worry
about a hierarchy based on the UNMC lab results view that doesn't concur
with other sites' views of the same results. The flip side of that is that
we could use the UNMC lab results view as a base, and if we all agree that
that base makes most sense structurally, the codes would be the only
things to debate. Then, as long as all possible codes had a spot in the
hierarchy somewhere, the queries would align based on the paths (rooted in
the UNMC view). The downside there is that it is not based on a national
standard hierarchy of any type, so it may be harder to repeat at other
sites.
I think we're well on our way to figuring this out at HackathonTwo -
thanks for helping spell out the different options!
--
Ticket URL:
<http://informatics.gpcnetwork.org/trac/Project/ticket/158#comment:19>
gpc-informatics <http://informatics.gpcnetwork.org/>
Greater Plains Network - Informatics
_______________________________________________
Gpc-dev mailing list
[email protected]
http://listserv.kumc.edu/mailman/listinfo/gpc-dev
