#385: Collect BC 'Patient Mapping File' from from breast cancer survey sites
--------------------------+------------------------
Reporter: dconnolly | Owner: vleonardo
Type: task | Status: closed
Priority: major | Milestone: bc-agg-1
Component: data-sharing | Resolution: fixed
Keywords: | Blocked By: 362, 420
Blocking: 356 |
--------------------------+------------------------
Comment (by vleonardo):
Received the following follow-up questions from Brad via email. After
looking into these, I had a follow-up conversation w/ him and both
questions were explained to his satisfaction. I've documented responses
below:
• It appears that two patients do not meet the criteria for
inclusion in the survey: 350339 (LCIS morphology, 8520/2 and seq = 02) and
450170 (seq = 02). Do you know how these may have snuck into our survey
dataset?
==> Pts w/ "only LCIS morphology" are excluded. If a pt has a tumor w/
any other morphology, then they are included. As we include all
encounters for every eligible patient, then some encounters (like those w/
LCIS morphology) will be included, as well.
• It appears that some patients have duplicate (or nearly
duplicate enough to be suspicious) records in the dataset, and this is not
an issue with just one site. One example: 350131. I’ve noticed that some
of these may have been duplicated when surgical treatment data were
updated. More troubling is that some of these nearly duplicate records are
linked with a different date of diagnosis (e.g., 300147). Any thoughts?
==> The duplicates reflect similar records as recorded in the site's tumor
registries. Encounter numbers are different for each of these, so the
datamart is reflecting correctly the source data, as it is submitted by
the site's tumor registries.
==> Regarding duplicate records having different 'Dx Dates': The value for
the field NAACCR 390, Dx Date, is a pass-thru from the source data. The
field gpc.dx.date then simply references the value stored in NAACCR-390.
--
Ticket URL:
<http://informatics.gpcnetwork.org/trac/Project/ticket/385#comment:23>
gpc-informatics <http://informatics.gpcnetwork.org/>
Greater Plains Network - Informatics
_______________________________________________
Gpc-dev mailing list
[email protected]
http://listserv.kumc.edu/mailman/listinfo/gpc-dev
