Stream: IG creation
Topic: Carin-BB: Terminology Errors
Saul Kravitz (Oct 07 2020 at 20:30):
The remaining errors in the build of Carin-BB are failures to validate ICD-10 codes. (ref https://build.fhir.org/ig/HL7/carin-bb/qa.html#_scratch_ig-build-temp-WL5SYY_repo_input_examples_ExplanationOfBenefit-EOBProfessional2UPMC ).
Similarly here is another terminology issue: https://build.fhir.org/ig/HL7/carin-bb/qa.html#_scratch_ig-build-temp-WL5SYY_repo_input_vocabulary_ValueSet-FDANDCOrCompound
Will these be resolved in the next several weeks, or can I get special dispensation to publish with these errors?
@Pat Taylor @Ryan Howells
Grahame Grieve (Oct 07 2020 at 20:58):
@Rob Hausam how is progress here?
Saul Kravitz (Oct 07 2020 at 22:17):
Similar problem in Plan-Net with NUCC Provider Taxonomy:
https://build.fhir.org/ig/HL7/davinci-pdex-plan-net/branches/v0.1.4/qa.html#_scratch_ig-build-temp-WNCKPC_repo_input_examples_HealthcareService-PharmChainCompService
Grahame Grieve (Oct 07 2020 at 23:33):
I think we have enough agreement with NUCC to be able to sort out our disaster with regard to this code system, and sometime in the next week or so these errors will disappear
Rob Hausam (Oct 08 2020 at 04:01):
I'm definitely behind where I intended to be with loading the ICD-10-CM codes. But I'm spending more time on it now and it should be resolved soon.
Saul Kravitz (Oct 08 2020 at 13:09):
@Robert Dieterle @Gail Kocher
Saul Kravitz (Oct 21 2020 at 13:57):
I still get errors:
- Plan-Net: NUCC Provider Taxonomy
- CarinBB: ICD-10
Shall I just list these as 'beyond my control' in my IG worksheet to Melva?
Lloyd McKenzie (Oct 21 2020 at 15:37):
@Rob Hausam
Lloyd McKenzie (Oct 21 2020 at 15:38):
@Saul Kravitz, irrespective of those issues, it looks like there are a bunch of display names you need to fix.
Saul Kravitz (Oct 21 2020 at 16:35):
The NUCC names in the IG are in line with the up to date NUCC content that will be incorporated into the terminology server. (got 'em straight from the NUCC chair, Gail Kocher).
Rob Hausam (Oct 21 2020 at 17:18):
@Saul Kravitz I'm juggling several things, but I'm working to get ICD-10-CM updated - it's pretty high priority, and I'll try to have it done within the next 24 hours. I'm not sure where we are at the moment with the NUCC content, but I will investigate that and try to help make sure that what we have is up-to-date, as well. Until then, I think they are beyond your control - so if you have to do something with it today you can indicate that (i.e. suppress warnings/info messages in the build).
Sarah Gaunt (Nov 24 2020 at 06:26):
Is this the same issue I'm having here: http://build.fhir.org/ig/HL7/fhir-bfdr/branches/master/qa.html#_scratch_ig-build-temp-UD4OD5_repo_input_resources_condition_condition-coded-initiating-fetal-death-cause-or-condition ?
I can't figure out if I'm doing something wrong or if it's something outside of the IG.
They are actual errors so I can't suppress them.
Sarah Gaunt (Nov 24 2020 at 06:28):
This is an example of what I have:
<coding>
<system value="http://hl7.org/fhir/sid/icd-10"/>
<code value="O00.0"/>
<display value="Abdominal pregnancy"/>
</coding>
And it's coming from this value set: https://phinvads.cdc.gov/vads/ViewValueSet.action?oid=2.16.840.1.114222.4.11.3593
Rob Hausam (Nov 24 2020 at 19:32):
@Sarah Gaunt Yes, I think this is primarily the same issue - the ICD-10 codes that you are using are not included in the very limited fragment of the code system that is currently represented and being validated on tx.fhir.org. However, I have created and pushed a complete copy of the ICD-10-CM code system in the tx support packages, but so far that isn't being picked up by the server (I'm working on that with @Grahame Grieve). However, that is for ICD-10-CM (the US variant), not for ICD-10 (WHO) which is what you are referencing (and the code system and value set in PHIN VADS appears to, as well). I don't yet have a full source for the complete set of codes for WHO ICD-10 (but @Jim Steel has a ClaML parser which I think we should be able to use to get that). But the last question is whether the PHIN VADS code system and value set should be using ICD-10, or ICD-10-CM, since as far as I know the PHIN VADS value sets are intended entirely for use in the US, and it is only ICD-10-CM that is used here. It seems to me that PHIN VADS probably has it wrong? And unless I'm wrong about that, you may want to go ahead and update it in the IG to ICD-10-CM (even if they haven't done that yet)?
Sarah Gaunt (Nov 24 2020 at 19:51):
Hi @Rob Hausam I think they are the correct codes - this is for death reporting and the WHO codes are what the USA uses to report cause of death (both regular death and fetal death):
Beginning with fetal deaths occurring in 1999, ICD10 has been used for coding and classifying causes of fetal death. This revision of the Classification is published by the World Health Organization (WHO) and consists of three volumes.
(See: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&ved=2ahUKEwiQ6vjl95vtAhXCb30KHSIKAK4QFjABegQIAhAC&url=https%3A%2F%2Fwww.cdc.gov%2Fnchs%2Fdata%2Fdvs%2F2a_2017.pdf&usg=AOvVaw03TpIQl8xg7G-_-UjhaTLv)
What I can't understand is why it's throwing an error (not a warning). Other issues similar to this have just thrown a warning. Is it because the value set does exist and the validator can't find the code, rather than the value set not existing therefore it can't find the value set and has no way to even try to validate the code?
I did a little digging to find out why the regular death profile (which uses the same value set) isn't causing the same error: https://build.fhir.org/ig/HL7/vrdr/StructureDefinition-VRDR-Cause-Of-Death-Condition.html and it looks like the example just uses text rather than the code which prevent the error: https://build.fhir.org/ig/HL7/vrdr/Condition-f0cbc63d-ac4e-469b-8761-70b6ea8bf666.xml.html.
@Grahame Grieve can we get special dispensation to go to ballot with these errors (if we agree that there's nothing wrong with my xml)? Or is it an acceptable solution to just use text (obviously this wouldn't be acceptable to use as a real cause of fetal death, but maybe just for ballot)?
Grahame Grieve (Nov 24 2020 at 19:53):
don't use text just to avoid an error. If the code is correct, you'll get approval
Sarah Gaunt (Nov 24 2020 at 19:54):
Ok, was just checking! :)
Rob Hausam (Nov 24 2020 at 19:54):
OK. You probably are right that they've agreed to use the WHO code system specifically for death reporting.
Rob Hausam (Nov 24 2020 at 19:55):
It's confusing, though, as ICD-10-CM for the US will also have all of those same codes.
Sarah Gaunt (Nov 24 2020 at 19:56):
So should it be a totally different system value then maybe for ICD-10 (i.e. not the exact same system as ICD-10-CM if they really are different)?
Yeah, no idea how/why these decisions get made...
Sarah Gaunt (Nov 24 2020 at 19:57):
I'd say it's so they can compare cause of death worldwide, but if they are the exact same codes as in the other system, I guess that isn't the reason.
Sarah Gaunt (Nov 24 2020 at 20:01):
Looks like Australia uses it too. And here's an article about the history in the USA: http://library.ahima.org/doc?oid=91930#.X71mJFBS9hE (Ok, need to get out of this rabbit-hole!)
Sarah Gaunt (Dec 02 2020 at 02:22):
@Grahame Grieve further to these errors because of the missing ICD-10 (WHO) codes, I am getting one other error that I think is a direct result of the errors I'm getting in the Condition examples.
Composition.section:codedCauseOfFetalDeath.entry:codedInitiatingFetalDeathCauseOrCondition: minimum required = 1, but only found 0 (from http://hl7.org/fhir/us/bfdr/StructureDefinition/Composition-coded-cause-of-fetal-death)
One of the examples that is getting the ICD-10 error is the example that is supposed to fulfil the requirement for the minimum 1 InitiatingFetalDeathCauseOrCondition slice. And I'm pretty sure that because that example is erroring, it is causing the Composition to error also.
So I'm asking if I can add this error to the list of allowed errors for the BFDR ballot. Adds up to 5 errors all for the same reason (missing ICD-10 codesystem).
Grahame Grieve (Dec 02 2020 at 02:32):
iCD-10 should be sorted out by tonight
Sarah Gaunt (Dec 02 2020 at 02:34):
Oh! Exciting... Ok, will check it out tomorrow then.
Grahame Grieve (Dec 02 2020 at 02:34):
It's Tali's Birthday tonight, so I'm not entirely sure I'll have time
Rob Hausam (Dec 02 2020 at 02:35):
But, this will fix ICD-10-CM, not ICD-10 (unless Grahame has done some additional work specifically on that).
Sarah Gaunt (Dec 02 2020 at 02:35):
Happy Birthday Tali!
I think you should maybe leave it. As long as it works before ballot right?
Sarah Gaunt (Dec 02 2020 at 02:35):
Oh, yeah, this is definitely not CM.
Sarah Gaunt (Dec 02 2020 at 02:37):
@Melva Peters is already aware of the ICD-10 issue anyway and has put it on her ballot spreadsheet. So if it gets fixed it's gravy (vegetarian, obviously).
Rob Hausam (Dec 02 2020 at 02:39):
Once we have ICD-10-CM working correctly, we should be able to do something similar for ICD-10 - assuming that Jim Steel's ClaML parser (or something else) can get us the full source. Glad it isn't critical for you.
Grahame Grieve (Dec 02 2020 at 02:39):
oh. well.
Last updated: Apr 12 2022 at 19:14 UTC
