Stream: genomics
Topic: task recommendations
Mullai Murugan (Nov 20 2019 at 00:06):
Hi all - see attached example - we have multiple components (i.e. Gene Panel, PRS and LPA) in our report and we have a finding and recommendation for each component. We plan on using grouper to house the Observations related to each component and would like to use Task/Recommendation for the recommendation. However, task/recommendation is currently only available as a DR extension. What is the best way for us to correlate the Observation and associated recommendation for each component?
Report Sample
Appreciate any guidance for this. Thanks! Mullai
Kevin Power (Nov 20 2019 at 23:50):
Sorry I missed this Mullai. It looks like the Tasks can refer back to the appropriate Observation using this:
Kevin Power (Nov 20 2019 at 23:52):
Or this one:
http://build.fhir.org/ig/HL7/genomics-reporting/task-med-chg-definitions.html#Task.reasonReference
Patrick Werner (Nov 21 2019 at 11:09):
:+1:
Mullai Murugan (Nov 21 2019 at 14:13):
This fits our requirement perfectly, thanks!
Mullai Murugan (Nov 26 2019 at 13:45):
We noticed that the cardinality for reasonReference is 0..1, can this be changed to 0..*? This would accommodate cases where multiple Observations might result in a recommendation.
Jamie Jones (Nov 26 2019 at 15:32):
Makes sense to me
Kevin Power (Nov 26 2019 at 15:35):
Our profile has to stay within the limits of the base resource. In this case, it is defined as 0..1 on Task, so we can't make it 0..* I'm afraid. You could make the proposal to extend it on Task, but that would go to O&O.
Patrick Werner (Nov 26 2019 at 15:35):
We noticed that the cardinality for reasonReference is 0..1, can this be changed to 0..*? This would accommodate cases where multiple Observations might result in a recommendation.
you could group these observations with the grouper and then point to the group in Task.
Jamie Jones (Nov 26 2019 at 15:37):
we haven't talked much about Task at all in recent discussions
Jamie Jones (Nov 26 2019 at 15:39):
Grouper isn't a "Finding" so I'm not sure we can recommend pointing to it there without tidying up our profile (http://build.fhir.org/ig/HL7/genomics-reporting/task-rec-followup.html)
Patrick Werner (Nov 26 2019 at 15:42):
hmm good point, should add both to the Task: grouper and Finding
Lloyd McKenzie (Nov 26 2019 at 17:41):
I wouldn't want to have to introduce a grouper just to get around a cardinality limit. It makes more sense to fix Task
Patrick Werner (Nov 26 2019 at 17:42):
@Mullai Murugan do you want to log an jira issue?
Jamie Jones (Nov 26 2019 at 17:59):
For convenience, trackers are now logged via the 'create' button on Jira (eg https://jira.hl7.org/browse/FHIR-25250), and make sure you select "Genomics Reporting (FHIR)" under "Specification"
Kevin Power (Nov 26 2019 at 18:32):
Actually, in this case, you would want to select FHIR Core since this is a change to Task.
Jamie Jones (Nov 26 2019 at 18:33):
Oh, is it restricted on the base Task resource? It's true we cannot increase the cardinality in our profile beyond the base. Thanks for checking. (Would go to O&O)
Mullai Murugan (Nov 26 2019 at 19:08):
Thanks for feedback, all. I have opened a JIRA item for FHIR Core.
Bret H (Nov 29 2019 at 16:16):
we have reason reference restricted to 'Medication Implication' for TaskMedChg. just a reminder that in our IG the variant observation driven task for medication change is driven from a medication Implication observation. not directly from a variant.
Last updated: Apr 12 2022 at 19:14 UTC
