commander-beef
Very Active Member
can You give me a sample of excel database?
Bounty: Lil Red Notebook
- Thread starter Kumaki
- Start date
Kumaki
Well-Known Member
Scroll back... The other person who found this in the first place linked all that stuff.
I pulled this out by merging ICD10 and ICD9 with shortened and long text, it looks like this:
Really, 13 megs of text, could be shortened to 8 if we remove the shortened version of the text, which always looks similar.
To keep on the example, it pulls out this line:
S93.02XA 837.0 Subluxation of left ankle joint, initial encounter Subluxation of left ankle joint, initial encounter
And contains some of the crosslinks (first pass)...
Really, 13 megs of text, could be shortened to 8 if we remove the shortened version of the text, which always looks similar.
To keep on the example, it pulls out this line:
S93.02XA 837.0 Subluxation of left ankle joint, initial encounter Subluxation of left ankle joint, initial encounter
Last edited by a moderator:
Hmm... as for Prostate Cancer, searching on that only throws one of the raw files:
And here a good example of shortened text and the longer version:
With which you can search the code in the generated file:
The processing of the XML file is harder than I thought... for example, searching for "heart" && "attack" gives this (no code at all):
while myocardial infarction gives loads and loads of codes... one will need to add the nemod tag as tertiary text (or replace the secondary)?
And here a good example of shortened text and the longer version:
Last edited by a moderator:
Kumaki
Well-Known Member
Well, there you go. Obviously, y'all can do the best you can, and I, myself, may need to muck around with the data files myslef after all is said and done.
Malignasnt neoplasm of prostate, on ICD9 would be 185.0 - "personal history of" is a different diagnosis.
Similarly, on the subluxation of left ankle joint, initial encounter adds the "XA" but if it was subsequent, the "XA" would not have been added.
ICD10 is a real pain in the butt.
You see why we otherwise have to carry around 14-pound books?
And I'm not exaggerating, I weighed them.
Malignasnt neoplasm of prostate, on ICD9 would be 185.0 - "personal history of" is a different diagnosis.
Similarly, on the subluxation of left ankle joint, initial encounter adds the "XA" but if it was subsequent, the "XA" would not have been added.
ICD10 is a real pain in the butt.
You see why we otherwise have to carry around 14-pound books?
And I'm not exaggerating, I weighed them.
hm... yes, but why is Z85.46 not somehow linked to C48.0, which is the correct code, I assume:
C48.0 158.0 Malignant neoplasm of retroperitoneum Malignant neoplasm of retroperitoneum
however, that code is less specific to the neoplasmic area: http://en.wikipedia.org/wiki/Retroperitoneal_space
C48.0 158.0 Malignant neoplasm of retroperitoneum Malignant neoplasm of retroperitoneum
however, that code is less specific to the neoplasmic area: http://en.wikipedia.org/wiki/Retroperitoneal_space
Last edited by a moderator:
Kumaki
Well-Known Member
And this is JUST the problem with ICD-10 is the level of specificity.
This is designed by people who do not actually look at doctor's notes. They are often very NON-specific.
We have about a year to retrain our doctors.
@commander-beef Are you still supporting the software? There seems to be annual updates (for example, a new Covid19 code) https://www.cdc.gov/nchs/icd/icd10cm.htm
