Stream: questionnaire
Topic: Connectathon 27
Paul Lynch (Apr 22 2021 at 21:44):
A couple of items of note:
1) Registration for the connectathon this time ends early -- April 30th is the deadline. (See https://confluence.hl7.org/display/FHIR/2021+-+05+Connectathon+27).
2) The kick-off meeting for the Questionnaire track meeting (as with all kick-off meetings) is scheduled for the week before the connectathon. For the Questionnaire track, the kick-off meeting will be May 12th, 9-10 PM UTC. (All times for this connectathon are in UTC).
Paul Lynch (May 10 2021 at 23:00):
For the Questionnaire track agenda, the easiest way to see it is to go to https://confluence.hl7.org/display/FHIR/2021+-+05+Connectathon+27, and filter on "questionnaire". All times are in UTC time, which is currently 4 hours ahead of ET. Here is picture of the current schedule (which could change, though I don't expect it to):
image.png
Paul Lynch (May 12 2021 at 14:05):
This is just a reminder that Questionnaire track kick-off meeting is today, in Whova, at 9:00 PM UTC (5:00 PM ET).
Paul Lynch (May 12 2021 at 21:57):
At the kick off meeting today, we discussed what our goals are for the connectathon. With the aim of helping track participants "connect" to each other either around shared goals or useful tools, I created a child page on the Questionnaire track confluence page: Questionnaire Track Participants, Tools, and Goals. If you don't mind putting your name and goals on the public website, please go ahead and add a row for yourself to the table.
Ilya Beda (May 14 2021 at 15:02):
Hi @Lloyd McKenzie @Paul Lynch
Sorry for missing the call yesterday. I was quite busy with some unexpected tasks.
I agree that https://jira.hl7.org/browse/FHIR-31080 is an issue that should be resolved on the application level. It may not be a part of the specification. So closing this task is the right solution.
Can we continue the discussion about https://jira.hl7.org/browse/FHIR-31078 during connectathon 27?
Lloyd McKenzie (May 14 2021 at 16:10):
I won't be on the track, but discussion can absolutely happen there.
Paul Lynch (May 14 2021 at 18:31):
Yes, it can. However, I think @Lloyd McKenzie is a key person to have in that discussion, if possible. But, if Lloyd is not available, we could still see how far we can get with the issue.
Diane (May 18 2021 at 03:48):
See everyone soon at the Connectathon!
My project is at https://testingconnectathon.dev.cloud.duke.edu/testfiles/
I am trying to make a valid Questionnaire that uses a Questionnaire Response as the data source. I have it hard-coded right now to 1139, but eventually, I would like to pass a variable representing the ID of the Questionnaire Response that is the data source. The questionnaire that I am trying to populate is at:
https://testingconnectathon.dev.cloud.duke.edu/testfiles/populateProtocolQuestionnaire_test.json
A detailed explanation and my questions are at https://testingconnectathon.dev.cloud.duke.edu/testfiles/questions_test.html
Diane (May 18 2021 at 09:52):
In looking at my Questionnaire (created with NLM Form Builder)
https://testingconnectathon.dev.cloud.duke.edu/testfiles/populateProtocolQuestionnaire_test.json
I am doing expression-based population. This is my first try, so I am testing with a hard-coded questionnaire response # 1139.
Nothing is populating at https://testingconnectathon.dev.cloud.duke.edu/testfiles/SIRBForm_After_Intake_test.html?formType=2POP&initiateQR=1146
-
Do I have the launch context correct in the extension? Do I need a launch context? Is the purpose of this launch context to tell LHC-Forms which type of resource to look in?
-
For my first question Initiate_1, do I have the fhirpath expression entered correctly?. I want to get the data for the populate from questionnaire response 1139 linkId='Initiate_1'.
"QuestionnaireResponse.where(id=1139).item.where(linkId='Initiate_1').answer.valueString"
- Do I need to have the resource Operation-questionnaire-populate created on my HAPI FHIR server for the population to work?
Paul Lynch (May 18 2021 at 10:51):
I created a sign-up page for our scheduled demo slots, so we can have an idea of which slots more demos will be at. (I am expecting at least one, but typically we have three or four.) We will still accept last-minute demo requests, but it will help with planning if you can sign up in advance. https://confluence.hl7.org/display/FHIR/Demo+Schedule
Ilya Beda (May 18 2021 at 12:07):
Following our conversation about how QuestionnaireResponse may be used for populating another QuestionnaireResponse
I created an issue that describes how I operate launchContext.
https://jira.hl7.org/browse/FHIR-32644
Paul Lynch (May 18 2021 at 12:57):
I agree, but think we might need to go further. The discussion was around Diane's use case of wanting to prepopulate several QuestionnaireResponses with one QuestionnaireResponse. (The Questionnaires are collecting data about a research study, but contain information beyond what ResearchStudy holds.) So, using launchContext to pass in a "context" QuestionnaireResponse seems to be the way to go. However, if we just make the launch context list extensible as in FHIR-29664, I don't see how implementations will guess what it is they will be expected to support. I think there needs to be some standard way of saying "give me the QuestionnaireResponse that is the context and put in this variable name".
Paul Lynch (May 18 2021 at 17:34):
Continuing this idea at separate topic: https://chat.fhir.org/#narrow/stream/179255-questionnaire/topic/launchContext.20off-list.20.22names.22
Brian Postlethwaite (May 19 2021 at 05:17):
I'm going to have a go at the enablewhenexpression implementation, have others done that?
Do we have a sample?
I'd like to get it added to my EnableWhen tests questionnaire.
Jose Costa Teixeira (May 19 2021 at 08:31):
Should we create an overview of the features, as a table, and identify which implementations support what?
Paul Lynch (May 19 2021 at 10:11):
It seems to me @Brian Postlethwaite started a page like that somewhere.
Jing Tang (May 19 2021 at 11:46):
Brian Postlethwaite said:
I'm going to have a go at the enablewhenexpression implementation, have others done that?
Do we have a sample?
I'd like to get it added to my EnableWhen tests questionnaire.
Hi Brian - we've done some work on enablewhen but not enablewhenexpression in here: https://github.com/google/android-fhir/blob/master/datacapture/src/main/java/com/google/android/fhir/datacapture/enablement/EnablementEvaluator.kt. I really would love more real world examples to guide our implementation so if you can share any test questionnaires that'd be awesome. Thanks!
Brian Postlethwaite (May 19 2021 at 12:33):
This is the example EnableWhenExpression that we are going to try out.
<item>
<extension url="http://hl7.org/fhir/uv/sdc/StructureDefinition/sdc-questionnaire-enableWhenExpression">
<valueExpression>
<language value="text/fhirpath"/>
<expression value="(%resource.item.item.where(linkId='Q10').answer.value & %resource.item.item.where(linkId='Q10b').answer.value) = 'chicken'"/>
</valueExpression>
</extension>
<linkId value="Q10-ewe" />
<text value="Q10 enablewhen (Q10 + Q10b) = 'chicken'" />
<type value="string" />
<required value="true" />
</item>
Brian Postlethwaite (May 19 2021 at 12:44):
For those following along, this is the part of the spec where the SDC spec adds a few extra "goodies" for the expression evaluation
http://build.fhir.org/ig/HL7/sdc/expressions.html#fhirpath-and-questionnaire
Notably %context, %qitem and %questionnaire
Brian Postlethwaite (May 19 2021 at 12:45):
image.png
Should launch context variables (if present) continue to be available after pre-population is executed?
i.e during save - how? or during these enable when expression evaluations?
@Paul Lynch, @Lloyd McKenzie
Paul Lynch (May 19 2021 at 13:38):
I was assuming that launchContext variables should stick around to be used by calculatedExpression, etc. Do we need to clarify that in the specification?
Brian Postlethwaite (May 19 2021 at 13:43):
I'm not sure, but I don't see why they wouldn't, but glad I asked.
Paul Lynch (May 19 2021 at 14:02):
Second demo session starting in just a couple of minutes...
Jose Costa Teixeira (May 19 2021 at 14:32):
Paul Lynch said:
It seems to me Brian Postlethwaite started a page like that somewhere.
@Brian Postlethwaite where's that page?
Lloyd McKenzie (May 19 2021 at 16:18):
I think using launchContext in calculatedExpression is risky, but not prohibited. We should likely include a warning in the spec. Key thing is that launch context would vary each time the QR was opened.
Tilo Christ (May 19 2021 at 17:41):
Paul Lynch said:
It seems to me Brian Postlethwaite started a page like that somewhere.
You mean this one? https://confluence.hl7.org/display/FHIRI/SDC+Renderer+Capabilities+Matrix
Brian Postlethwaite (May 20 2021 at 00:59):
I also started experimenting with a rendering of a profile for a renderer that you could then use for verification also.
Paul Lynch (May 20 2021 at 12:57):
I think we were eventually going to use a CababiltiiesStatement (when the second version of CapabilitiesStatement is ready) to state the support available in the SDC implementations.
Paul Lynch (May 20 2021 at 21:01):
Lloyd McKenzie said:
I think using launchContext in calculatedExpression is risky, but not prohibited. We should likely include a warning in the spec. Key thing is that launch context would vary each time the QR was opened.
Brian Postlethwaite (May 20 2021 at 22:19):
I don't think a Capability statement has the level of granularity needed for this.
Lloyd McKenzie (May 21 2021 at 02:21):
CapabilityStatement2 would be needed
Brian Postlethwaite (May 22 2021 at 02:58):
For the renderers, I actually think a profile is the better answer.
Brian Postlethwaite (May 22 2021 at 03:00):
Maybe the tag to say SDC in the capability statement, and a ref to the profile for the level of detail.
Brian Postlethwaite (May 22 2021 at 03:01):
I'm creating the profile for ours, server and client.
Last updated: Apr 12 2022 at 19:14 UTC
