Stream: ihe
Topic: ValueSet OID == CodeSystem OID
Simone Heckmann (Sep 01 2018 at 08:51):
Is that correct that the OID 1.3.6.1.4.1.19376.1.2.3 both identifies the ValueSet and the CodeSystem?
https://www.hl7.org/fhir/valueset-formatcodes.html
Simone Heckmann (Sep 01 2018 at 08:57):
Also: Why does the ValueSet get assigned a URL and the CodeSystem doesn't?
Simone Heckmann (Sep 01 2018 at 09:05):
Is there a logic behind that?
Grahame Grieve (Sep 01 2018 at 20:24):
it's wrong for the value set and code system to have the same oid.
Grahame Grieve (Sep 01 2018 at 20:25):
IHE assigned an OID for the code system
Simone Heckmann (Sep 02 2018 at 14:54):
...which doesn't mean they couldn't also assign a URL...
Simone Heckmann (Sep 02 2018 at 14:56):
Oh! I just realized, IHE did assign a CodeSystem URL in the current built: http://build.fhir.org/valueset-formatcodes.html
Simone Heckmann (Sep 02 2018 at 15:06):
Although the URL doesn't help my confusion with ValueSet vs. Codesystem: http://ihe.net/fhir/ValueSet/IHE.FormatCode.codesystem
When I follow the url, I get the following information:
The IHE FormatCode code set is defined to be the set of FormatCode(s) defined by IHE. This code set additionally includes format codes defined by the HL7 CDA. The set was listed in HITSP C80 Table 2-153 Format Code Value Set Definition. The use of the FormatCode specifies the technical format of the document. The FormatCode is a further specialization more detailed than the mime-type. The FormatCode provides sufficient information to allow any potential document consumer to know if it can process and/or display by identifying a document encoding, structure and template. The actual list of codes is intended to be extensible.
...which indicates that this is the definition of a ValueSet. However, the usage in FHIR says it's a CodeSystem:
<compose> <include> <system value="http://ihe.net/fhir/ValueSet/IHE.FormatCode.codesystem"/> </include> </compose>
The url itself apparently remains purposefully ambiguous...
Grahame Grieve (Sep 02 2018 at 19:14):
it's nastily confusing for IHE to assign a code system URL that looks like value set URL. But It does resolve to a CodeSystem thta identifies itself differently.
Grahame Grieve (Sep 02 2018 at 19:14):
make a task to fix the value set definition
Simone Heckmann (Sep 03 2018 at 07:52):
I'm not sure exactly what needs fixing here. You mean the link behind http://ihe.net/fhir/ValueSet/IHE.FormatCode.codesystem that points to ftp://ftp.ihe.net/TF_Implementation_Material/fhir/CodeSystem/IHE.formatcode.codesystem.xml, which describes a ValueSet instead of a CodeSystem?
What do you mean "But It does resolve to a CodeSystem thta identifies itself differently" For me it only resolves to the above mentioned ftp site or, if c&p'd into a browser, not at all.
Simone Heckmann (Sep 03 2018 at 07:53):
...suppose it should resolve to https://wiki.ihe.net/index.php/IHE_Format_Codes?
Simone Heckmann (Sep 03 2018 at 07:58):
Besides, since the codes in this systems are URNs (e.g. "urn:ihe:pcc:apr:lab:2008") Shouldn't the CodeSystem just say "this is a URN", like we do with OIDs?
Grahame Grieve (Sep 03 2018 at 09:09):
the url that identifies the code system looks like a value url, but resolve to a code system that has a different URL
John Moehrke (Sep 03 2018 at 17:14):
no nastiness intended. I have asked for guidance and got nothing... so what should I do?
John Moehrke (Sep 03 2018 at 17:15):
also, the current codesystem and valueset do have unique OIDs... I fixed that a while ago
Grahame Grieve (Sep 03 2018 at 19:11):
the problem is this :
<compose>
<include>
<system value="http://ihe.net/fhir/ValueSet/IHE.FormatCode.codesystem"/>
</include>
</compose>
Grahame Grieve (Sep 03 2018 at 19:11):
it should be
Grahame Grieve (Sep 03 2018 at 19:13):
<compose>
<include>
<system value="http://ihe.net/fhir/CodeSystem/IHE.FormatCode.codesystem"/>
</include>
</compose>
Simone Heckmann (Sep 03 2018 at 19:29):
I have to ask again: isn’t a CodeSystem for a urn style code redundant, since the code itself is globally unique, like an oid?
Grahame Grieve (Sep 03 2018 at 19:43):
sort of in principle. But given an OID, you'd need to know where to look to find the definition.... you'd need something like a code system URL
John Moehrke (Sep 04 2018 at 12:47):
I can fix that. oops. thanks for reporting the issue
John Moehrke (Sep 04 2018 at 12:49):
This is what I have today
<compose>
<include>
<system value="urn:oid:1.3.6.1.4.1.19376.1.2.3"/>
</include>
</compose>
John Moehrke (Sep 04 2018 at 12:55):
@Simone Heckmann are you referring to the trailing ".codesystem" in the url? I think IHE will be dropping that in R4. I put it in because Forge was recommending that years ago.
http://ihe.net/fhir/CodeSystem/IHE.FormatCode.codesystem
vs
http://ihe.net/fhir/CodeSystem/IHE.FormatCode
We will also be removing the period and going to underbar
http://ihe.net/fhir/CodeSystem/IHE_FormatCode
or do you mean the CodeSystem prior? I am very happy to receive comments and improvement opportunities.,
Simone Heckmann (Sep 04 2018 at 13:23):
Hi,
no my comment about redundancy was about having a code system at all rather than treating it like an oid where the system just says "this is an oid" and the value is the oid itself, like we have
<masterIdentifier> <system value="urn:ietf:rfc:3986" /> <value value="1.2.276.0.76.3.1.191.0002.01.6097600100103.00000042.003.01.1452003006772782605" /> </masterIdentifier>
we could have:
<coding> <system value="urn:ietf:rfc:8141"/> <code value="urn:ihe:pcc:apr:lab:2008"/> </coding>
I mean, it's not like we need a system here to disambiguate the code.
But I understand Grahame's comment about the benefits of having a canonical url to resolve to the CodeSystem's definition.
Personally, I'd be super happy with http://ihe.net/fhir/CodeSystem/IHE_FormatCode (although the use of underscores seems to be very uncommon in canonical urls, but I can't find any rules against it) especially if this pattern is going to be used consistently throughout the IHE defined CodeSystems and ValueSets.
John Moehrke (Sep 04 2018 at 13:35):
Where are you seeing that? Can you point me to it? Are you talking about FormatCode, or are you talking about DocumentReference? If so, then what within DocumentReference? Where?
Grahame Grieve (Sep 04 2018 at 20:57):
This is what I have today
<compose> <include> <system value="urn:oid:1.3.6.1.4.1.19376.1.2.3"/> </include> </compose>
where? that's in the endpayload-types. In format codes - which is what I thought we were talking about, we have:
<compose>
<include>
<system value="http://ihe.net/fhir/ValueSet/IHE.FormatCode.codesystem"/>
John Moehrke (Sep 06 2018 at 18:14):
@Grahame Grieve I am confused on what is broken, and what it should be
Grahame Grieve (Sep 06 2018 at 21:58):
this is broken:
<compose>
<include>
<system value="http://ihe.net/fhir/ValueSet/IHE.FormatCode.codesystem"/>
Grahame Grieve (Sep 06 2018 at 21:58):
it should be
<compose>
<include>
<system value="http://ihe.net/fhir/CodeSystem/IHE.FormatCode.codesystem"/>
John Moehrke (Sep 07 2018 at 00:16):
that is NOT what I have on the IHE formal location ftp://ftp.ihe.net/TF_Implementation_Material/fhir/ValueSet/
John Moehrke (Sep 07 2018 at 00:16):
I thought you were now pulling from there?
John Moehrke (Sep 07 2018 at 00:18):
That is not what I have registered in Simplifier https://simplifier.net/ihepatientdemographi/iheformatcodevs
John Moehrke (Sep 07 2018 at 00:19):
as far as I know I have never had the IHE valueset that way
John Moehrke (Sep 07 2018 at 00:26):
You seem to be still pulling this from the FHIR Build. I have been fixing these and many other things in the IHE codesystem and valueset. You had indicated to me that you are pulling them from the IHE ftp...
John Moehrke (Sep 07 2018 at 00:28):
clearly a miscommunication. I should have checked... sorry. So, fix after ballot. Sorry world.
John Moehrke (Sep 07 2018 at 00:30):
IHE is getting really close to figuring out how to get our conformance resources published in http form rather than ftp.. but I suspect it is going to take alot longer. Struggling with what should be simple. Would love someone to sprinkle pixi dust and have these things properly published.
Grahame Grieve (Sep 07 2018 at 02:59):
I'll schedule some time for this after Baltimore
Jose Costa Teixeira (Sep 09 2018 at 09:01):
IHE is getting really close to figuring out how to get our conformance resources published in http form rather than ftp.. but I suspect it is going to take alot longer. Struggling with what should be simple. Would love someone to sprinkle pixi dust and have these things properly published.
@John Moehrke "Figuring out how" should not take a lot longer. - depends on what is the issue, it is very close. Just requested an ftp access to publish some stuff on ihe's site. Making sure infrastructure is in place, that could take a little longer., but I hope not . Depending on time, during this week IHE people might be able to publish conformance resources using the IG publisher.
If there is anything more urgent to decide (like a standard location for conformance resources), then we can accelerate that.
With that in mind, when we publish on ihe site, the "official" conformance resources will always be available. Here's my old example: http://zeora.net/blog/mma/ihe-mma-performer.xml.html
Last updated: Apr 12 2022 at 19:14 UTC
