FHIR Chat · More SNOMEDCT woes -- intensional value sets

Stream: IG creation

Topic: More SNOMEDCT woes -- intensional value sets

Saul Kravitz (Oct 18 2021 at 22:21):

@Rob Hausam Take a look at http://build.fhir.org/ig/HL7/fhir-mCODE-ig/branches/master/ValueSet-mcode-cancer-staging-system-vs.html

It is built from one intensional statement:
Include codes from http://snomed.info/sct where concept is-a 254292007 (Tumor staging)

SCT#254292007 has a child:
897275008 | American Joint Commission on Cancer, Cancer Staging Manual, 8th edition neoplasm staging system (tumor staging) |

It is not included in the above value set, leading to other QA errors in mCODE.

Rob Hausam (Oct 20 2021 at 05:46):

@Saul Kravitz Maybe something changed since you posted? I see the 897275008 |American Joint Commission on Cancer, Cancer Staging Manual, 8th edition neoplasm staging system (tumor staging)| code in the expansion of the mcode-cancer-staging-system-vs value set in the IG page that you linked to.

Saul Kravitz (Oct 20 2021 at 12:01):

@Rob Hausam hmmmm…still getting Qa errors relating to that code.

Grahame Grieve (Oct 21 2021 at 06:36):

I see it there now if I build it locally. Wait for the new version coming out tomorrow and try again

Saul Kravitz (Oct 21 2021 at 21:01):

Still getting this on the master branch:
Error parsing response message: This does not appear to be a FHIR resource (wrong namespace '') (@ /) for 'http://snomed.info/sct#897275008'

Saul Kravitz (Oct 22 2021 at 18:26):

Hi @Grahame Grieve @Rob Hausam Indeed al of the errors of the form 'Error parsing response message' are gone.

However, now I'm back to having errors relating to codes that are in the July 2021 release of SNOMED considered missing because they are missing from the March 2021 release of SNOMED. If we could get past that, I think my bundle gets a clear bill of health.

Unable to find code 1156506007 in http://snomed.info/sct (version http://snomed.info/sct/731000124108/version/20210301); The code "1156506007" is not valid in the system http://snomed.info/sct; The code provided (http://snomed.info/sct#1156506007) is not valid in the value set 'All codes known to the system' (from http://tx.fhir.org/r4) for 'http://snomed.info/sct#1156506007'

Saul Kravitz (Oct 22 2021 at 18:27):

Reference this build/qa.html

Saul Kravitz (Oct 22 2021 at 18:28):

Here is the link to one of the missing concepts in the July SNOMED build.

Saul Kravitz (Oct 25 2021 at 17:33):

@Grahame Grieve

Good news: Major progress. All of my QA issues are wrapped up, except those relating to 3 SNOMEDCT codes
Bad news: SNOMEDCT: 1156506007 and 1156530009. 1149099005-- The code "<code>" is not valid in the system http://snomed.info/sct

Reference: http://build.fhir.org/ig/HL7/fhir-mCODE-ig/branches/pendingsctcodes/qa.html

Rob Hausam (Oct 25 2021 at 17:35):

I'll look at these, @Saul Kravitz.

Rob Hausam (Oct 25 2021 at 18:04):

@Saul Kravitz It's looking for these codes in the 20210301 version of the US Edition, which no longer exists on tx.fhir.org (it's now the latest which is 20210901). So the question is, where is the 20210301 version string coming from? You may want to clear your cache, if you haven't, to be sure there is no issue there. And see if you are specifying 20210301 anywhere in the IG. If not, then I think I know where to look and likely how to fix it.

Saul Kravitz (Oct 25 2021 at 19:19):

@Rob Hausam
These errors are from the CI build. They have nothing to do with my local configuration.
Why is it looking at the US edition? These codes are all in the international edition from July 2021.
The only reference to SNOMEDCT in the IG is us of the officially sanctioned codesystem -- http://snomed.info/sct -- without a version.
This refers to the most recent published international snomed edition, right?

Saul Kravitz (Oct 25 2021 at 19:54):

@Rob Hausam @Grahame Grieve Here is the missing code response from tx.fhir.org: https://tx.fhir.org/r4/CodeSystem/$lookup?system=http://snomed.info/sct&code=1149099005. It claims that the code is not present in the 2210731 version.

image.png

Here is the code in the 0731 build of SNOMED: https://browser.ihtsdotools.org/?perspective=full&conceptId1=1149099005&edition=MAIN/2021-07-31&release=&languages=en

Rob Hausam (Oct 25 2021 at 21:50):

@Saul Kravitz Yes, this is obviously still wrong. I think the SNOMED CT International Edition 20210731 version import will need to be re-done. I can do that later this evening.

Michael Lawley (Oct 26 2021 at 07:53):

@Saul Kravitz If you don't specify a version for SNOMED CT then you have not specified any particular Edition. If you want to tie things to a specific Edition, then you need to specify a version. In the case of the International Edition the version would be http://snomed.info/sct/900000000000207008, which should be interpreted as the latest known version of that Edition.

Rob Hausam (Oct 26 2021 at 11:56):

Agree, @Michael Lawley. But I don't think that's Saul's issue? The code 1149099005 isn't being found on tx.fhir.org for SNOMED CT version http://snomed.info/sct/900000000000207008/version/20210731, even though it's clearly an active concept in that version of the edition. The lookup message is:

Unable to find code 1149099005 in http://snomed.info/sct version http://snomed.info/sct/900000000000207008/version/20210731

The same query works fine on Ontoserver. I'm going to work on that now.

Grahame Grieve (Oct 26 2021 at 21:27):

so the first thing is that the IG is a US IG, so it is published based on the US version of snomed. That's a policy decision made by snomed, not us. Second, that's enforced in the IG configuration, and reflected in the call made by the IG publisher:

<Parameters xmlns="http://hl7.org/fhir">
...
  <parameter>
    <name value="profile"/>
    <resource>
      <Parameters>
        <id value="exp-params"/>
        <parameter>
          <name value="system-version"/>
          <valueUri value="http://snomed.info/sct|http://snomed.info/sct/731000124108"/>
        </parameter>
      </Parameters>
    </resource>
  </parameter>
</Parameters>

Grahame Grieve (Oct 26 2021 at 21:28):

so the question that matters is, is the code in the US distribution, and I don't see that it is.

Rob Hausam (Oct 26 2021 at 21:30):

Yes, it's in the US Edition. I'm working on getting this fixed.

Grahame Grieve (Oct 26 2021 at 21:47):

where? i don't see it in the RF2 distribution

Saul Kravitz (Oct 26 2021 at 21:50):

We have a fair number of new codes that haven't made their way into the US distribution yet (sadly, there is a lag).

Rob Hausam (Oct 26 2021 at 21:50):

I'm getting it from the US Edition RF2 distribution in sct2_Concept_Snapshot_US1000124_20210901.txt. There's a single active row for the 1149099005 concept.

Rob Hausam (Oct 26 2021 at 22:35):

@Saul Kravitz @Grahame Grieve I've re-imported and updated the US Edition on tx.fhir.org, and this now works correctly when the US edition is specified. It is not, however, working (for 1149099005, at least) in the International Edition, where it also should be working. I'll do the same fix there.

Rob Hausam (Oct 27 2021 at 03:51):

@Saul Kravitz The "missing codes" issue with the International Edition is also now fixed.

Rob Hausam (Oct 27 2021 at 03:52):

Let us know if that does fix the issue for you - or if you're seeing anything else (terminology server related).

Saul Kravitz (Oct 28 2021 at 22:09):

All is well. Thanks @Rob Hausam and @Grahame Grieve .

Last updated: Apr 12 2022 at 19:14 UTC

Main menu

FHIR Chat · More SNOMEDCT woes -- intensional value sets · IG creation