Stream: IG creation
Topic: LOINC license agreement
Craig Newman (Jul 27 2020 at 13:42):
HTA sent out an email last week about updated wording from Regenstrief about the LOINC license agreement. I know the FHIR License page http://hl7.org/fhir/license.html contains similar wording, but I don't recall seeing this sort of boilerplate in individual IGs. Should IGs that use LOINC codes include this wording (and maybe similar wording for SNOMED, etc) somewhere in the IG or is it covered by the reference in the base standard?
Eric Haas (Jul 27 2020 at 15:57):
we have been using this in all of our ValueSets and CodeSystems in all of IGs we author.
This content from LOINC® is copyright © 1995 Regenstrief Institute, Inc. and the LOINC Committee, and available at no cost under the license at http://loinc.org/terms-of-use.
for example here: file:///Users/ehaas/Documents/FHIR/US-Core-R4/output/ValueSet-us-core-diagnosticreport-lab-codes.html
Craig Newman (Jul 27 2020 at 16:44):
What about where a profile fixes an element to a LOINC code like Observation.code for a BMI observation (eg http://www.hl7.org/fhir/us/core/StructureDefinition-pediatric-bmi-for-age.html)? Should the profile contain similar text or may just put once somewhere in the narrative of the IG?
Eric Haas (Jul 27 2020 at 18:29):
Yes I was wondering the same thing as I responded. I think there should be a blanket statement for when we fix codes not sure where the best place to expose this though. in the introduction or as an the end?
Grahame Grieve (Jul 27 2020 at 19:01):
I didn't see this email - I may have bulk deleted it. What exactly does it say?
Craig Newman (Jul 27 2020 at 19:23):
Dear Co-Chairs
The HL7 Terminology Authority (HTA) is notifying all co-chairs of HL7 WG that LOINC has, this June, updated its license agreement for use of the LOINC - Logical Observation Identifiers, Names and Codes - terminology. Specific wording must be used in any product or service utilizing LOINC; this wording is provided for you in the "Arrangement or agreements with HL7 for use of content" on the LOINC - Logical Observation Identifiers, Names and Codes page of the HTA External Terminologies pages.
Since many of our HL7 products reference the use of LOINC, this is of relevance to most of you; please communicate it within your WG and especially to all how are working on materials for ballot and/or publication.
The HTA is also taking this opportunity to update you about recent new content additions to the External Terminologies - Information pages on Confluence. These include the Dartmouth Atlas of Health Care, ICD-10-PCS, International Standard Classification of Occupations 2008 (ISCO-08)
With thanks and best wishes
Julie
on behalf of the HL7 Terminology Authority
Grahame Grieve (Jul 27 2020 at 20:03):
@Julie James The FHIR project negotiated with LOINC for a simpler statement, which is written into the normative version of LOINC (though not normative itself):
This content LOINC® is copyright © 1995 Regenstrief Institute, Inc. and the LOINC Committee, and available at no cost under the license at http://loinc.org/terms-of-use
The HTA version is
This material contains content from LOINC (http://loinc.org). LOINC is copyright © 1995-2020, Regenstrief Institute, Inc. and the Logical Observation Identifiers Names and Codes (LOINC) Committee and is available at no cost under the license at http://loinc.org/license. LOINC® is a registered United States trademark of Regenstrief Institute, Inc
Can we clarify with LOINC whether they are now asking FHIR implementation guides and resources to use the longer license?
And specifically: it's a nightmare that the the statement they have asked us to use says "© 1995-2020", because it's unstable. In a prior version I negotiated with LOINC, it would have said:
This material contains content from LOINC (http://loinc.org). LOINC is copyright © 1995+, Regenstrief Institute, Inc. and the Logical Observation Identifiers Names and Codes (LOINC) Committee and is available at no cost under the license at http://loinc.org/license. LOINC® is a registered United States trademark of Regenstrief Institute, Inc
which is much more stable for the tool smiths and editors. Can you please clarify that with LOINC too?
Julie James (Jul 28 2020 at 08:34):
Hello Grahame. Thank you for telling us about the FHIR project's "separate negotiations". It really makes the myself and the other members of the HTA feel valued and appreciated for all the voluntary work we do with external terminology organisations, including LOINC, on behalf of the whole HL7 community, even if 90% of our work currently is to try to sort out all the myriad terminology issues in FHIR.
Please not that this is not "the HTA version", this is the LOINC statement. :) We don't author statements for external code systems :)
Am I to assume (never a good thing, but the best I can do here) that your concern (I'll tone down the "nightmare) regarding "instability)" is because by saying "(C) 1995+" you never have to change the statement, whereas by saying "(C)1995-2020" there is a likelihood that the statement will need to be changed in 2021, 2022 etc?
Julie James (Jul 28 2020 at 08:39):
I'm not a copyright lawyer, but from the little I am aware of, one uses a time period in copyright to cover versions, hence the 1995-2020. I have no sense of the legality or otherwise of 1995+. I've never ever seen it on anything, I'm afraid and I suspect many a copyright lawyer would be somewhat unhappy, especially in the US.
Please confirm to me if I have intuited your concerns correctly. If I have, please then confirm if you would like the HTA to speak to LOINC on your behalf, or whether you plan to do it directly as you did when you negotiate the statement for FHIR separately. Sorry to be blunt, but I really don't want to ask HTA members and LOINC to give of their time to work on something that we later find has been duplicated at best, or at worst, wasted. Thank you
Grahame Grieve (Jul 28 2020 at 09:40):
well, the license was negotiated outside HTA a long time ago when there wasn't any prospect that HTA would be interested in anything like that. And I'm going to assume that the HTA discussion with LOINC predated the last round of discussions regarding NUCC where the importance of looking at what is already happening in HL7 was raised.
Grahame Grieve (Jul 28 2020 at 09:40):
I can certainly raise this with LOINC directly but I thought it would be better to work with HTA, so let's keep working on it through the HTA
Grahame Grieve (Jul 28 2020 at 09:41):
And yes, you got my concern correct. And while I'm not sure about the ins and outs of the time period legally, the LOINC lawyers did agree to simply 1995+ already
Craig Newman (Aug 03 2020 at 12:22):
Did we ever decide if individual IGs need specific text included in them when value sets or profiles contain LOINC (and presumably SNOMED, ICD, etc) codes?
Robert McClure (Aug 03 2020 at 17:06):
@Craig Newman and the rest of you. Questions like this should be answered via HTA pages. For LOINC see: https://confluence.hl7.org/display/TA/LOINC+-+Logical+Observation+Identifiers%2C+Names+and+Codes
Julie James (Aug 13 2020 at 21:03):
As requested, the HTA have met with LOINC to discuss this. I have sent an email directly to yourself, Grahame, but I also put the pertinent content here: LOINC has confirmed that the wording circulated (by themselves and forwarded to the HL7 community by the HTA) for the copyright agreement must be used, with no exception.
All previously discussed copyright statements, including the one that you referred to ‘for FHIR’ and the (different) one currently displayed on the LOINC on FHIR page, should be replaced with the official LOINC statement.
Thank you
Grahame Grieve (Aug 13 2020 at 21:16):
this includes the rolling current year?
Lloyd McKenzie (Aug 23 2020 at 21:06):
@Grahame Grieve @Julie James was Grahame's question answered elsewhere?
Grahame Grieve (Aug 23 2020 at 21:13):
yes. the answer is that it does, and editors will need to change the LOINC notice on a yearly basis in their source
Lloyd McKenzie (Aug 23 2020 at 21:34):
Any chance the HTA could have a friendly discussion with them to change that onerous requirement - because it's a significant amount of busy-work for IG authors...
Grahame Grieve (Aug 23 2020 at 22:08):
I already did, but the price of this isn't paid by the people making the rule
Rob Hausam (Aug 23 2020 at 23:28):
Yes. I also have an open question to Swapna about the expectations regarding a yearly update to the copyright statement, after I made the change to their preferred text in the CI (and Connectathon) build.
Julie James (Aug 24 2020 at 09:01):
I think it all depends on what you call "busy work". It tends to be used as a derogatory term, but routine work has to be done to respect the processes of communities. Think of all the things co-chairs have to do to make their WG run, some of which, no doubt, could be considered by some to be "busy work".....Please remember that you are coming from one perspective and others have a different one and it is important to be respectful of it.
Vassil Peytchev (Aug 24 2020 at 14:09):
Isn't it possible to have active content that automatically calculates the current year?
LOINC is copyright © 1995-<span id="copyrightYear"></span>
var currYear = new Date.getFullYear();
document.getElementById("copyrightYear").innerHTML = currYear;
Lloyd McKenzie (Aug 24 2020 at 14:31):
Not in our artifacts. And not easily in our publications which generally use static HTML.
Eric Haas (Aug 24 2020 at 16:13):
(deleted)
Vassil Peytchev (Aug 24 2020 at 17:05):
assets/js/fhir.js seems to be available everywhere...
Kevin Power (Sep 03 2020 at 12:45):
Checking back on this - any common guidance for IG authors?
Swapna Abhyankar (Sep 04 2020 at 04:15):
I am checking with our legal counsel on whether we can change the date range to 1995+ in our copyright statement. Like Julie said, I am also not a copyright lawyer, and I don't know if there are legal ramifications of 1995-2020 versus 1995+. I am not trying to make anyone's life difficult or give them more work to do. I am simply following the recommendations of our legal counsel and trying to make things consistent so that we don't have different versions of the copyright statement in different places.
If he says we can say 1995+, that change will happen with the December LOINC release. I'll keep you posted.
Grahame Grieve (Sep 04 2020 at 04:50):
thanks
Swapna Abhyankar (Sep 05 2020 at 16:03):
Here is the message from legal counsel:
*Use of the copyright notice is optional under the US Copyright Act. However, if the notice is used, the statute prescribes the format:
© [Year] [Owner]
The "Year" is the year of first publication. However, for a compilation such as LOINC where new copyrightable material is added regularly, the "Year" field will be expanded to cover each year that new copyrightable material is added.
In short, the options are: (a) omit the copyright notice, which I do not recommend, or (b) continue to use the copyright notice in the format that you have been using.*
So bottom line, we have to include the date range. Sorry.
Vassil Peytchev (Sep 05 2020 at 18:45):
the "Year" field will be expanded to cover each year that new copyrightable material is added.
If © 1995+ does not satisfy the above (and I don't understand why it wouldn't), would © 1995 - current or © 1995 - present work?
Lloyd McKenzie (Sep 05 2020 at 19:49):
or 1995-2100 :)
Grahame Grieve (Sep 06 2020 at 00:39):
come on guys, stop trying to be sensible. this is the law at work here
Grahame Grieve (Sep 06 2020 at 08:37):
I can make the IG publisher fix this automatically in the resources. Fixing the HTML seems slipperier to me, but there authors can go for Vassil's solution, If it's legal
Lloyd McKenzie (Sep 06 2020 at 13:24):
You mean HTML for past releases? Is there anywhere the HTML for a new release would not drive from the resources?
Grahame Grieve (Sep 06 2020 at 20:58):
if an editor chose to do that
Grahame Grieve (Sep 06 2020 at 21:08):
ok. from the next release, if you put {{{year}}} in your copyright element in the source, the publisher will automatically replace it with the current year
Kevin Power (Sep 09 2020 at 21:38):
I assume we should include this agreement on profiles that use a fixed LOINC code? If not, ignore the rest :smile:
If the answer is yes, is the expectation that this license agreement shows face up? It looks like the HTML rendered for ValueSets (and I think CodeSystems) includes the copyright statement. However, for StructureDefinitions, you can only see the copyright statement when you look at the XML or JSON. Maybe that is OK, but wanted to ask.
Perhaps even easier, can we just include the agreement once in our IG index page?
Grahame Grieve (Sep 14 2020 at 06:25):
good point - will be addressed in the next release of the IG publisher.
Kevin Power (Sep 14 2020 at 13:30):
Thanks @Grahame Grieve
Kevin Power (Sep 16 2020 at 22:00):
@Grahame Grieve -- Did you add this yet? I tested by adding copyright to a StructureDefinition source and didn't see it in the HTML (other than the .json.html and .xml.html as I mentioned).
Not sure if it is related, but the latest version of the Publisher is throwing this exception during the build:
onGenerate.jira:
onGenerate.extend:
onGenerate:
java.lang.NullPointerException
at org.hl7.fhir.r5.renderers.ProfileDrivenRenderer.render(ProfileDrivenRenderer.java:109)
at org.hl7.fhir.r5.renderers.ProfileDrivenRenderer.render(ProfileDrivenRenderer.java:99)
at org.hl7.fhir.r5.renderers.ResourceRenderer.build(ResourceRenderer.java:55)
at org.hl7.fhir.r5.renderers.BundleRenderer.render(BundleRenderer.java:296)
at org.hl7.fhir.igtools.publisher.Publisher.getXhtml(Publisher.java:7767)
at org.hl7.fhir.igtools.publisher.Publisher.saveDirectResourceOutputs(Publisher.java:7112)
at org.hl7.fhir.igtools.publisher.Publisher.generateHtmlOutputs(Publisher.java:6562)
at org.hl7.fhir.igtools.publisher.Publisher.generate(Publisher.java:4849)
at org.hl7.fhir.igtools.publisher.Publisher.createIg(Publisher.java:871)
at org.hl7.fhir.igtools.publisher.Publisher.execute(Publisher.java:717)
at org.hl7.fhir.igtools.publisher.Publisher.main(Publisher.java:8276)
Generating Summary Outputs (01:46.0704)
Grahame Grieve (Sep 16 2020 at 22:59):
how do I reproduce that?
Kevin Power (Sep 16 2020 at 23:10):
You can recreate it with the latest of our IG: https://github.com/HL7/genomics-reporting/tree/master
Grahame Grieve (Sep 16 2020 at 23:57):
the IG publisher should handle this error better - and will in the future. but the underlying issue is that you are based on v 4.0 but using MedicationUsage, which doesn't exist in 4.0
Grahame Grieve (Sep 16 2020 at 23:59):
the current template doesn't display copyright for structure definitions - that's work in progress on a different thread
Kevin Power (Sep 17 2020 at 03:10):
Thanks for the help will fix that problem
Kevin Power (Sep 17 2020 at 13:41):
I took a quick look at our IG and do not see any references to MedicationUsage? We do have a profile of Task with a name in the IG of "Medication Usage Task", but it seems to render OK? http://build.fhir.org/ig/HL7/genomics-reporting/StructureDefinition-task-med-chg.html
I should also note that this error first occurred this week, and we haven't committed changes for about a month.
Also adding @Patrick Werner who has helped track down weird problems like this before in our IG.
Jose Costa Teixeira (Sep 17 2020 at 13:43):
medicationusage is the new name for medicationstatement
Kevin Power (Sep 17 2020 at 13:47):
Ahh, thanks for the clarification @Jose Costa Teixeira -- Our IG does have numerous references to MedicationStatement.
Last updated: Apr 12 2022 at 19:14 UTC
