Stream: implementers
Topic: HL7 V2 messaging schemas
Lucas (Jul 20 2021 at 23:12):
Hey all! This chat was linked to from the hl7.org site. Can I ask general HL7 questions here? If not, where can I?
Specifically I am looking for HL7 v2 messaging schemas. I found these XML schema downloads on hl7.org: https://www.hl7.org/implement/standards/product_brief.cfm?product_id=213 ... however, I think there may be a bug in the 2.7.1, 2.8 and 2.8.2 downloads. In 2.5.1, in datatypes.xsd, the CX.9 field has a type of "CWE":
<xsd:attributeGroup name="CX.9.ATTRIBUTES">
<xsd:attribute name="Type" type="xsd:string" fixed="CWE"/>
<xsd:attribute name="LongName" type="xsd:string" fixed="Assigning Jurisdiction"/>
</xsd:attributeGroup>
I believe this is correct. I am cross-referencing it with this site: https://v2plus.hl7.org/2021Jan/data-type/CX.html
However, in the 2.7.1, 2.8 and 2.8.2 downloads, CX.9 has a type of "CX":
<xsd:attributeGroup name="CX.9.ATTRIBUTES">
<xsd:attribute name="Type" type="xsd:string" fixed="CX"/>
<xsd:attribute name="LongName" type="xsd:string" fixed="Assigning Jurisdiction"/>
</xsd:attributeGroup>
This appears incorrect to me. I don't think a recursive field makes sense.
So I am wondering where I can report issues about HL7 org downloads? Is there a public JIRA or something like that? And then where can I get correct/complete schemas for HL7 v2 messages?
It looks like the HAPI library has correct version 28 schemas: https://hapifhir.github.io/hapi-hl7v2/v28/apidocs/ca/uhn/hl7v2/model/v28/datatype/CX.html
And this website is generating HTML version of the schemas: https://v2plus.hl7.org/2021Jan/data-type/CX.html
So where is HAPI and hl7.org getting their information from? Is there an official/canonical source for HL7 v2 schema (in a machine readable format e.g. XML or DB export)?
Thanks so much for any help!
Lloyd McKenzie (Jul 21 2021 at 01:56):
Most of the folks who hang out on the particular stream are FHIR-focused. (And the v2 folks find most of the traffic irrelevant as a result.) You might want to post your question on #V2. Also @Frank Oemig
Frank Oemig (Jul 21 2021 at 14:43):
You are right, this is incorrect. Generating new schemas does not show this again. We should update them.
Frank Oemig (Jul 21 2021 at 14:44):
The schemas are generated from a database that uses the Word docs as input. The db is also the source for generating some tools like HAPI. Therefore we are trying to get that complete and correct. Unfortunately the docs are inconsistent and needs fixing
Frank Oemig (Jul 21 2021 at 14:45):
PS: please place V2 questions in the V2 stream, thx
Lucas (Jul 21 2021 at 16:56):
Thanks for the help! @Frank Oemig this background information is very helpful, thank you. I am not sure how to move this post but I can create a new one in the V2 stream. I am wondering what is the best way to access a corrected schema data set? Is the DB accessible or is my best option to use a tool like HAPI? I could also try to work with the Word docs, but am not sure which docs you are referring to. I didn't find any docs after a quick look through the v2 downloads on hl7.org.
Frank Oemig (Jul 21 2021 at 17:50):
I have asked Lynn to update the download.
Frank Oemig (Jul 21 2021 at 17:53):
The problem is to extract the data from the original documents so that they are correct, complete and consistent. That requires manual tweaking because of Word. Everything else is then generated or extracted from the db. Of course, this is not normative anymore and may contain errors. We are doing our best to identify and correct them.
Frank Oemig (Jul 21 2021 at 17:55):
With V2+ we intend to provide the whole spec in FHIR that allows for using FHIR tools to work with it. But again, that also requires some fixing... :-(
Lucas (Jul 21 2021 at 19:18):
Ok, great! Thank you so much. V2+ sounds like an interesting project. Is there a way for me to access the V2+ FHIR spec yet?
Frank Oemig (Jul 22 2021 at 19:21):
Not yet, this is tooling. We are currently discussing the new form.
Last updated: Apr 12 2022 at 19:14 UTC
