#56: shareable synthetic test data sets: Epic clarity, i2b2, NAACCR, ...
--------------------------+--------------------------------
Reporter: dconnolly | Owner: bokov
Type: enhancement | Status: assigned
Priority: major | Milestone: data-quality-plan
Component: data-sharing | Resolution:
Keywords: | Blocked By:
Blocking: |
--------------------------+--------------------------------
Comment (by dconnolly):
I realized last night that the unit tests for the Data Builder (#134 #87)
include code to generate a mock i2b2 star schema. It's currently very
limited, but I had an idea:
Make a spreadsheet with diagnoses, the labs etc. used to diagnose them,
and the meds etc. used to treat them (insulin).
||= Condition =||= Indication =||= Treatment =||
|| Diabetes || A1C || Insulin ||
|| Diabetes || BMI || ||
|| MI || Pulse || Aspirin ||
|| Sepsis || || Antibiotic ||
And use those as the basis to generate test encounters.
Refinements:
- value distributions for labs etc. (mean, std dev)
- probabilities to go with the relationships
- sequences of related encounters: normal results, abnormal, diagnosis,
treatment, normal
--
Ticket URL:
<http://informatics.gpcnetwork.org/trac/Project/ticket/56#comment:15>
gpc-informatics <http://informatics.gpcnetwork.org/>
Greater Plains Network - Informatics
_______________________________________________
Gpc-dev mailing list
[email protected]
http://listserv.kumc.edu/mailman/listinfo/gpc-dev
