FHIR Chat · tx.fhir.org - LOINC validation error · IG creation

Stream: IG creation

Topic: tx.fhir.org - LOINC validation error

Max Masnick (Nov 30 2020 at 00:36):

I'm running into an issue that I believe is an error with LOINC validation on tx.fhir.org.

Given this:

{
  "code": {
    "coding": [
      {
        "system": "http://loinc.org",
        "code": "48002-0" // Genomic source class [Type] <https://loinc.org/48002-0/>
      }
    ]
  },
  "valueCodeableConcept": {
    "coding": [
      {
        "system": "http://loinc.org",
        "code": "LA6683-2",
        "display": "Germline"
      }
    ]
  }
}

I get the following QA error:

Unable to resolve value set (from http://tx.fhir.org/r4)

LA6683-2 is from the preferred answer list for 48002-0. Strangely, if I choose another code from the preferred answer list like LA6684-0, there's no QA error.

To reproduce in situ:

• GitHub diff showing two branches of the mCODE IG where the only change is from LA6683-2 to LA6684-0. Names of these branches are below.
• QA report for branch genomics-error1-mfm (LA6683-2 with the QA error)
• QA report branch genomics-error1-somatic-mfm (LA6684-0, no QA error)

@Grahame Grieve is this something you can help with or is there someone else I can ask? Thanks!

Max Masnick (Nov 30 2020 at 18:47):

@Rob Hausam or @Robert McClure do you have any insight into this? Thanks!

Rob Hausam (Nov 30 2020 at 19:49):

@Max Masnick Hmm. I'm not entirely sure yet why the two different codes are giving different QA reports - with one showing "Unable to resolve value set ..." and the other not. But, in looking at the output of tx.fhir.org for all of the LOINC answer (LA) codes in the LL378-1 MG_2_Genetic variant source answer list, it seems that the descriptions being returned for all of them are for the wrong (completely unrelated) concepts. It appears to me that likely we have a problem with the loading code for LOINC on the fhirserver, at least for how it is handling the codes and value sets for the LOINC answer lists (likely not only these particular ones), and they are being loaded with entirely incorrect data. :frown: I expect that it's going to take a bit of investigation to figure out and fix (by me and/or @Grahame Grieve).

Max Masnick (Nov 30 2020 at 19:50):

Ok thanks for looking into it! Please let me know if there's anything I can do to help.

Grahame Grieve (Dec 01 2020 at 01:57):

it looks to me as if this is a cache error - it happened in the past and now it's embedded in the loinc cache file in tx-cache

Grahame Grieve (Dec 01 2020 at 01:57):

delete any entries in there with reference to it, and it should go away

Rob Hausam (Dec 01 2020 at 15:18):

The results I reported are from querying tx.fhir.org with Postman.
For example, GET http://tx.fhir.org/r4/CodeSystem/$lookup?system=http://loinc.org&code=LL378-1 returns:

{
  "resourceType" : "Parameters",
  "parameter" : [
    {
      "name" : "name",
      "valueString" : "LOINC"
    },
    {
      "name" : "version",
      "valueString" : "2.68"
    },
    {
      "name" : "display",
      "valueString" : "Pentadecanoate/Creatinine"
    }
  ]
}

but GET https://fhir.loinc.org/CodeSystem/$lookup?system=http://loinc.org&code=LL378-1&_format=json returns:

{
    "resourceType": "Parameters",
    "parameter": [
        {
            "name": "name",
            "valueString": "LOINC"
        },
        {
            "name": "display",
            "valueString": "MG_2_Genetic variant source"
        },
        {
            "name": "abstract",
            "valueBoolean": false
        },
        {
            "name": "property",
            "part": [
                {
                    "name": "code",
                    "valueCode": "answers-for"
                },
                {
                    "name": "value",
                    "valueString": "48002-0"
                }
            ]
        }
    ]
}

Rob Hausam (Dec 01 2020 at 15:47):

And GET http://tx.fhir.org/r4/CodeSystem/$lookup?system=http://loinc.org&code=LA6683-2 returns:

{
  "resourceType" : "Parameters",
  "parameter" : [
    {
      "name" : "name",
      "valueString" : "LOINC"
    },
    {
      "name" : "version",
      "valueString" : "2.68"
    },
    {
      "name" : "display",
      "valueString" : "Mycoplasma flocculare 16S rRNA gene"
    }
  ]
}

but GET https://fhir.loinc.org/CodeSystem/$lookup?system=http://loinc.org&code=LA6683-2&_format=json returns:

{
    "resourceType": "Parameters",
    "parameter": [
        {
            "name": "name",
            "valueString": "LOINC"
        },
        {
            "name": "display",
            "valueString": "Germline"
        },
        {
            "name": "abstract",
            "valueBoolean": false
        }
    ]
}

@Grahame Grieve

Rob Hausam (Dec 01 2020 at 15:51):

And, of course, the fhir.loinc.org display strings are the correct ones.

Grahame Grieve (Dec 02 2020 at 00:32):

oops. fixed next upgrade of the terminology server. hopefully today

Caleb Wan (Sep 30 2021 at 01:47):

The LONIC answer list LL4309-2 is used to bind the value field of the extension AssistanceRequired https://build.fhir.org/ig/HL7/fhir-pacio-cognitive-status/branches/master/StructureDefinition-assistance-required.html
In the QA report https://build.fhir.org/ig/HL7/fhir-pacio-cognitive-status/branches/master/qa.html I see two warnings as showed below. Please advise whether or not I should suppress these warnings. Thank you.

fsh-generated/resources/StructureDefinition-assistance-required.json Show Validation Information (1)
Path Severity Message
StructureDefinition.differential.element[3].binding warning The valueSet reference http://loinc.org/vs/LL4309-2 on element Extension.value[x] could not be resolved
StructureDefinition.snapshot.element[4].binding warning The valueSet reference http://loinc.org/vs/LL4309-2 on element Extension.value[x] could not be resolved

Lloyd McKenzie (Sep 30 2021 at 01:58):

@Rob Hausam @Grahame Grieve

Grahame Grieve (Sep 30 2021 at 02:04):

this is a bug in the validator that's fixed next IG publisher release

Caleb Wan (Oct 18 2021 at 15:31):

@Grahame Grieve I am still seeing a warning on the use of the LONIC answer list with the latest publisher release 1.1.82. Is this another problem with validator or a fix has not been put in yet?
StructureDefinition.differential.element[3].binding warning The valueSet reference http://loinc.org/vs/LL4309-2 on element Extension.value[x] could not be resolved

Grahame Grieve (Oct 18 2021 at 18:59):

have you cleared your txCache?

Caleb Wan (Oct 18 2021 at 20:23):

Yes. Here is the QA: https://build.fhir.org/ig/HL7/fhir-pacio-cognitive-status/branches/master/qa.html
Here is where the LOINC answer list is used: https://build.fhir.org/ig/HL7/fhir-pacio-cognitive-status/branches/master/StructureDefinition-assistance-required.html

Grahame Grieve (Oct 20 2021 at 17:54):

sigh. It was fixed, and then I did something on the terminology server to unfix it. It should be right now

Caleb Wan (Oct 20 2021 at 18:18):

@Grahame Grieve I just rebuilt our IGs. But but the warnings on the LL4309-2 answer list have not gone away:
https://build.fhir.org/ig/HL7/fhir-pacio-cognitive-status/branches/master/qa.html
https://build.fhir.org/ig/HL7/fhir-pacio-functional-status/branches/master/qa.html

Grahame Grieve (Oct 20 2021 at 18:52):

they will but you're about to get a new validation error:

Grahame Grieve (Oct 20 2021 at 18:52):

input\resources\CapabilityStatement-pacio-cs-cap.json Show Validation Information (1)
Path Severity Message
CapabilityStatement.url error Canonical URLs must be absolute URLs if they are not fragment references (/CapabilityStatement-pacio-cs-cap.html)

Last updated: Apr 12 2022 at 19:14 UTC