OK, a closer reading of the naaccr_facts_load.sql script verifies Angela's 
point, that Heron base just generates an encounter_num for the observation_fact 
record but no record is created in the visit dimension corresponding to that 
encounter_num, and no encounter facts are created either.  So that explains how 
it works now.

So an inconvenient question remains:  Should it?  Should NAACCR tumor records, 
each of which involves a tumor accession, be represented as "real" encounters?  
A use case might be a researcher who wants to see outpatient encounters for 
people who have a tumor accession listed in the registry.  Does that sound at 
all like a real world situation to you folks?

From: Bos, Angela [mailto:[email protected]]
Sent: Friday, January 30, 2015 9:37 AM
To: Lenon Patrick; [email protected]
Subject: RE: NAACCR Encounter type?

Using the HERON code base, we do not generate encounters (visit_dimension) for 
NAACCR tumor records.

-Angela | UTHSCSA

From: 
[email protected]<mailto:[email protected]> 
[mailto:[email protected]] On Behalf Of Lenon Patrick
Sent: Friday, January 30, 2015 9:13 AM
To: [email protected]<mailto:[email protected]>
Subject: NAACCR Encounter type?

Hey all,

I'm close to loading our NAACCR data into our I2B2 instance, which involves 
generating encounter records (per the tumor_reg_visits table referenced in 
KUMC's naaccr_facts_load.sql).  These encounters don't line up with any 
existing encounters, they're all unique.

My question is, should these NAACCR "encounters" be represented in the 
Encounters portion of the I2B2 tree?  I checked Babel and spot-checked a few 
setups and didn't see any reference to NAACCR under Encounter Type, which is 
where I'd expect to find it.

Has this been considered/discussed/rejected?  Your thoughts are welcome.



Patrick Lenon
HIMC Informatics Specialist
608 890 5671

_______________________________________________
Gpc-dev mailing list
[email protected]
http://listserv.kumc.edu/mailman/listinfo/gpc-dev

Reply via email to