FHIR Chat · Follow-up on #17513 · Orders and Observation WG

Stream: Orders and Observation WG

Topic: Follow-up on #17513


view this post on Zulip Christine D (Jul 27 2018 at 15:51):

Also checking in on #17513. Per specimen meeting on Tuesday 7/24, it was suggested to add these to specimen vs alter the dateTime. Have you heard any updates?

https://gforge.hl7.org/gf/project/fhir/tracker/?action=TrackerItemEdit&tracker_item_id=17513

view this post on Zulip Lloyd McKenzie (Jul 27 2018 at 16:05):

@Bryn Rhodes, where are we at with a generic "relative dateTime" extension?

view this post on Zulip Bryn Rhodes (Jul 27 2018 at 16:13):

Within PlanDefinition and RequestGroup we use the relatedAction element for that, which allows you to specify an offset and anchor relative to another action. A generic extension to support that would need to know what it was relative to, and how to interpret the "effective period" for that target.

view this post on Zulip Bryn Rhodes (Jul 27 2018 at 16:15):

So that would look like a target 1..1 Resource(Any) and a targetEffectivePeriodPath 1..1 String that would resolve to some temporal-valued element on the target resource.

view this post on Zulip Bryn Rhodes (Jul 27 2018 at 16:15):

Would that work?

view this post on Zulip Lloyd McKenzie (Jul 27 2018 at 16:19):

Plus a duration (positive or negative) to express the offset. Though in practice, the reference event is often expressed as a code. Using a reference is better though because you can have situations where there may be multiple reference points. E.g. "This observation happened 30 days post-transplant A and 10 days pre-transplant B" or "This procedure happened 30 days pre-start-of-treatmentfor trial A and 2 years post-start-of-treatement for trial B"

view this post on Zulip Lloyd McKenzie (Jul 27 2018 at 16:21):

I think all of the systems that use code should be able to resolve the StudyPubject or Procedure resource that's relevant and should know the path for the relevant time data element, so lets try that and see how it goes.

view this post on Zulip Bryn Rhodes (Jul 27 2018 at 16:24):

Sounds good, I'm working on extensions now.

view this post on Zulip Christine D (Jul 31 2018 at 20:03):

Hi @Bryn Rhodes , we were just talking about this tracker in the specimen call with @Riki Merrick . Could you please help us understand the extension and the status of the extension? Thank you!

view this post on Zulip Bryn Rhodes (Jul 31 2018 at 20:49):

Sure, the extension is defined here: http://build.fhir.org/extension-cqf-relativedatetime.html

view this post on Zulip Bryn Rhodes (Jul 31 2018 at 20:49):

What it allows you to do is specify the value of some date/time element as an extension that is calculated as an offset from some target.

view this post on Zulip Bryn Rhodes (Jul 31 2018 at 20:50):

So I could put it on an Observation.effective and point it at a Procedure to say this Observation took place two days after this procedure.

view this post on Zulip Riki Merrick (Jul 31 2018 at 20:55):

We need it on Specimen.collection to document that this collection occurred 30 minutes after administering Glucose or antibiotic X. Will that work?

view this post on Zulip Bryn Rhodes (Jul 31 2018 at 21:01):

After administering either of Glucose or antibiotic X?

view this post on Zulip Bryn Rhodes (Jul 31 2018 at 21:02):

As in, you're defining in the abstract when this collection should occur? Or you need to indicate in the concrete when it did occur, but relative to some other event?

view this post on Zulip Lloyd McKenzie (Jul 31 2018 at 21:40):

The latter.

view this post on Zulip Bryn Rhodes (Jul 31 2018 at 21:48):

So something like:

view this post on Zulip Bryn Rhodes (Jul 31 2018 at 21:48):

<Specimen>
  <collection>
    <collected>
      <extension url="http://hl7.org/fhir/StructureDefinition/cqf-relativeDateTime">
        <extension url="target">
          <valueReference>
            <reference value="MedicationAdministration/1"/>
          </valueReference>
        </extension>
        <extension url="targetPath">
          <valueString value="effective"/>
        </extension>
        <extension url="relationship">
          <valueCode value="after-end"/>
        </extension>
        <extension url="offset">
          <valueDuration>
            <value value="30"/>
            <unit value="min"/>
          </valueDuration>
        </extension>
      </extension>
    </collected>
  </collection>
</Specimen>

view this post on Zulip Lloyd McKenzie (Jul 31 2018 at 22:12):

Yes

view this post on Zulip Bryn Rhodes (Jul 31 2018 at 23:32):

(y)

view this post on Zulip Riki Merrick (Aug 01 2018 at 05:58):

Thank you!

view this post on Zulip Eric Haas (Sep 04 2018 at 02:37):

RE GF#17513 @Christine D where are we with this tracker? It sounds you may be able to withdraw it?

view this post on Zulip Eric Haas (Sep 04 2018 at 02:39):

and GF#17514 too

view this post on Zulip Lloyd McKenzie (Sep 04 2018 at 03:07):

No resolution is listed for GF#17654, which makes it hard to block-vote :)

view this post on Zulip Lloyd McKenzie (Sep 04 2018 at 03:07):

(nor for GF#17571)

view this post on Zulip Eric Haas (Sep 04 2018 at 03:14):

Thanks I updated - they are both preapplied

view this post on Zulip Christine D (Sep 06 2018 at 13:40):

I am not clear on the resolution for this one. Is the proposal to use the extension that Bryn listed above? Same question for 17514.

https://gforge.hl7.org/gf/project/fhir/tracker/?action=TrackerItemEdit&tracker_item_id=17513
https://gforge.hl7.org/gf/project/fhir/tracker/?action=TrackerItemEdit&tracker_item_id=17514

@Angie Romano what are your thoughts on this?

view this post on Zulip Angie Romano (Sep 06 2018 at 13:45):

for #17513, it looks like we just need confirmation from @Bryn Rhodes that the extension has been created to work at the specimen.collection level. If so, I'll give it a try on our test dataset.

view this post on Zulip Eric Haas (Sep 06 2018 at 15:42):

@Angie Romano The extension: http://build.fhir.org/extension-cqf-relativedatetime.html states "Context of Use: Use on Element ID Element" which means in laymans terms that is can be used anywhere in FHIR.

view this post on Zulip Eric Haas (Sep 06 2018 at 15:43):

... and what about GF#17514... I don't see how they are different.

view this post on Zulip Angie Romano (Sep 10 2018 at 15:47):

Hi again! @Eric Haas in using the extension http://build.fhir.org/extension-cqf-relativedatetime.html, I have a question about the offset. It's pointing to either valueDuration or valueRange. I've tried to use both, but they both reference currency and are looking for a code for money currency. I would expect to link to a code to pull the duration units (i.e. minutes, hours, days, etc...) Am I looking at it incorrectly? Thank you!

view this post on Zulip Angie Romano (Sep 10 2018 at 15:50):

@Christine D did you see Eric's question about if 17513 and GF#17514 were the same thing?

view this post on Zulip Lloyd McKenzie (Sep 10 2018 at 15:51):

They shouldn't be. Duration will be constrained to a UCUM code for time. Range should be open to your choice of code system. How are you seeing them be constrained?

view this post on Zulip Christine D (Sep 10 2018 at 16:07):

17513 was for the numeric representation (03-00 to represent 3 hours), while 17514 was for a textual description that included the event context ("3 hours post dose").
@Eric Haas @Angie Romano

view this post on Zulip Angie Romano (Sep 10 2018 at 17:04):

@Eric Haas and @Christine D. The extension: http://build.fhir.org/extension-cqf-relativedatetime.html includes a both of those. You get the reference to the resource with the event, the duration including units (minutes, hours, etc...), the relationship (i.e. before the referenced resource, after the referenced resource). I can show you an example if you would like to see one.

view this post on Zulip Christine D (Sep 10 2018 at 17:07):

@Eric Haas would we mention the extension in both 17513 and 17514 since the asks were slightly different in the two trackers? (but both requirements met by a single extension)

view this post on Zulip Angie Romano (Sep 10 2018 at 17:09):

@Lloyd McKenzie here's what I see and here's my json. It could very well be me :)

I look at the details for the extension at: http://build.fhir.org/extension-cqf-relativedatetime.html. I go to the json tab to view the structure of the json.
Then I click on duration and I'm taken to this link: http://build.fhir.org/datatypes.html#Duration which references the datatype money. I wonder if link is just going to the incorrect datatype.

Finally, here is that section of json:

"extension": [
{
"extension":
[
{
"extension":
[{
"url": "target",
"valueReference": {
"reference": "Encounter/TNCL789311"}
}]
},
{
"extension":
[{
"url": "targetPath",
"valueString": "Period"
}]
},
{
"extension":
[{
"url": "relationship",
"valueString": "after-end"
}]
},
{
"extension":
[{
"url": "offset",
"valueDuration":
{
"value": "60",
"currency":
{
"code":
{
"system": "http://hl7.org/fhir/ValueSet/duration-units",
"value": "min",
"display": "minutes"
}
}
}
}]
}
],
"url": "http://hl7.org/fhir/StructureDefinition/cqf-relativeDateTime"
}
],

view this post on Zulip Lloyd McKenzie (Sep 10 2018 at 17:18):

When you click on the link for Duration, look at the top of the screen. The bottom of the screen where it says "Money" is a different type.

view this post on Zulip Angie Romano (Sep 10 2018 at 17:48):

got it! Thanks @Lloyd McKenzie!

view this post on Zulip Eric Haas (Sep 10 2018 at 19:14):

@Christine D if this extension meets your needs for both trackers then with your approval, I can mark these trackers as "withdrawn".

view this post on Zulip Christine D (Sep 11 2018 at 18:32):

@Eric Haas please mark both trackers (17513 and 17514) as withdrawn. Thank you!

view this post on Zulip Eric Haas (Sep 11 2018 at 18:58):

done


Last updated: Apr 12 2022 at 19:14 UTC