FHIR Chat · CodeSystem for adverse-event-outcome · terminology

Stream: terminology

Topic: CodeSystem for adverse-event-outcome

view this post on Zulip Diane (Nov 14 2021 at 02:18):

I am trying to find a codeSystem URL to put into the answerOption.valueCoding.system property for a questionnaire item about adverse event outcome. I would like to use a URL that conforms to FHIR standards and that will be accepted by the IG publishing tools :persevere: .

The values that we want the user to select from in the questionnaire item are at http://hl7.org/fhir/R4/codesystem-adverse-event-outcome.html

Questions:

  1. The defining URL http://terminology.hl7.org/CodeSystem/adverse-event-outcome leads to a 404 error. Does that 404 error mean that this URL is not to be used in the answerOption.valueCoding.system property?

  2. Should I use http://hl7.org/fhir/R4/codesystem-adverse-event-outcome.html instead because it is a valid URL that doesn't give a 404 error?

  3. Why doesn't https://terminology.hl7.org/codesystems.html have a link for the adverse-event-outcome codesystem? Does this mean that http://terminology.hl7.org/CodeSystem/adverse-event-outcome is not approved?

view this post on Zulip Lloyd McKenzie (Nov 14 2021 at 03:10):

  1. There is no requirement that canonical URLs resolve - though the fact this one doesn't is odd. @Ted Klein - any idea what's happening here
  2. Definitely don't use http://hl7.org/fhir/R4/codesystem-adverse-event-outcome.html. The web page that describes the code system is NOT typically the canonical URL.
  3. Not clear what's happening here...

view this post on Zulip Diane (Nov 16 2021 at 01:45):

@Lloyd McKenzie So the IG will accept any url? I could just put something like www.google.com? One of the validators we tried over the summer didn't like external urls. Presumably, www.google.com would not pass the IG voting requirements, though?

@Ted Klein Ted...When you get a chance, can you comment on my questions #1 and #3? Also, the concepts (but not the codes themselves) in http://terminology.hl7.org/CodeSystem/adverse-event-outcome are very similar to those in https://terminology.hl7.org/CodeSystem-v2-0241.html for Patient Outcomes. Is there any overall attempt to be sure that multiple parallel code systems are not published?

view this post on Zulip Lloyd McKenzie (Nov 16 2021 at 02:36):

The requirement is that the URL be the official URL for the code system. If you're making up your own, it should be in the IG's canonical space. If it's external and you expect other implementations will also be using the code system, you should be submitting a request to the HTA to register an official URL (which they'll do after discussion with the terminology source if they can) that everyone using FHIR will be required to use. One of the criteria for the URLs is "it's nice if they resolve to something useful to an end user". But that's a nice-to-have. And there are lots of reasons why sometimes they don't.

In terms of avoiding having HL7 create multiple code systems that say things that are similar, in theory the UTG process can help with that. In practice, not so much. Quite often the v2 terminologies don't necessarily follow 'good terminology practices' so there's a deliberate decision to create something new. Other times, it happens by accident and because review is done by volunteers who happen to have the time to bother, in practice stuff gets missed. Good quality review requires time by knowledgeable people whose time tends to be scarce. It's a problem we haven't figured out a good solution to :(

view this post on Zulip Ted Klein (Nov 16 2021 at 02:48):

@Diane upon examination, it would appear that when the subset of FHIR R4 CodeSystems had their canonical URL changed to be anchored in THO (http://terminology.hl7.org) and their source moved to THO, this one was changed but never moved (the move was a manual operation, and it must have just been missed is my surmise). Sorry this happened, but happy you found it (no one else has reported this in all this time, ugh). I will get this object migrated into THO asap. It is absolutely supposed to be there and this is simply a technical error in the content which must be addressed. @Grahame Grieve any issues with moving it into THO? This is pro-forma, does not need any kind of consensus review...what other element value mapping might need to occur besides the <version> element? And for everyone reading this thread, I want to call attention to the last paragraph of Lloyd's comment above; we are suffering many pains due to this particular problem.

view this post on Zulip Diane (Nov 17 2021 at 11:59):

@Lloyd McKenzie Thank you for the explanation of the URL and its usage in the event of an external code system.

view this post on Zulip Diane (Nov 17 2021 at 12:03):

@Ted Klein Thank you for finding the answers about this code system and working to get it migrated into THO. There is so much for volunteers to manage that I am sure that minor glitches can occur.

view this post on Zulip Grahame Grieve (Nov 18 2021 at 01:36):

@Ted Klein in principle I'd generally agree but in fact, this code system is already dropped in favor of SCT. @Diana_Ovelgoenne see here: https://build.fhir.org/valueset-adverse-event-outcome.html - I recommend you follow this approach.

view this post on Zulip Diane (Nov 18 2021 at 04:38):

Will discuss with the BR&R team. Meanwhile, I have two questions, @Grahame Grieve and @Ted Klein

  1. Our questionnaires are R4 but this code system that you are suggesting is R5. Will this cause problems with our IG publishing? We have already balloted and the code system for the Adverse Event questionnaire is what is holding up final publishing.

  2. The defining URL on the page https://build.fhir.org/valueset-adverse-event-outcome.html is pointing to http://hl7.org/fhir/ValueSet/adverse-event-outcome. If you look at the codes at http://hl7.org/fhir/ValueSet/adverse-event-outcome, you will see that they are the old (ICSR) codes and not the SCT codes. What is the defining URL for the new SCT codes supposed to be?

view this post on Zulip Grahame Grieve (Nov 18 2021 at 04:42):

given the situation - that arises because AdverseReaction is still early in the development cycle - I'd clone the R5 value set into your IG

view this post on Zulip Ted Klein (Nov 18 2021 at 17:08):

So since the R5 value set is anchored in hl7.org/fhir and the code system is SCT, THO is not involved. That having been said, since the R4 value set has the old code system in the logical definition anchored in THO, should we add the R4 code system to the THO ci build with a <status> of 'retired' so it gets into the THO 3.0.0 release so folks can see it, and also see the retirement comment that it has been retired because the value sets no longer use it? If not (as this seems like just 'busy work') how do folks that may run into this problem not again waste time getting all this disconnection sorted? Personally, IMHO, all the stuff that was published in R4 that is anchored in terminology.hl7.org should be visible there; the <status> and commtents about deprecation should clarify if not longer straightforwardly active,

view this post on Zulip Ted Klein (Nov 18 2021 at 17:10):

or is this sufficiently handled with your URL redirect that references the R4 code system page when the system is clicked on from the logical definition of the R4 value set?

view this post on Zulip Ted Klein (Nov 18 2021 at 17:10):

Note there is no server redirect so the URL direct gives a 404 error...

view this post on Zulip Grahame Grieve (Nov 18 2021 at 18:37):

I'm not sure it's worth it, but yes, we could. We just have many other problems to deal with

Last updated: Apr 12 2022 at 19:14 UTC


An HL7 Website -