Stream: implementers
Topic: vitalsigns profile question
John Timm (Oct 03 2019 at 14:31):
The VitalSigns StructureDefinition.snapshot contains an ElementDefinition at this path:
Observation.component.value[x]
that includes the following ValueSet binding
"binding": {
"extension": [
{
"url": "http://hl7.org/fhir/StructureDefinition/elementdefinition-bindingName",
"valueString": "VitalSignsUnits"
}
],
"strength": "required",
"description": "Common UCUM units for recording Vital Signs.",
"valueSet": "http://hl7.org/fhir/ValueSet/ucum-vitals-common|4.0.0"
},
which seems like it would only be valid if the choice type was Quantity, yet the structure definition allows more choices:
"type": [
{
"code": "Quantity"
},
{
"code": "CodeableConcept"
},
{
"code": "string"
},
{
"code": "boolean"
},
{
"code": "integer"
},
{
"code": "Range"
},
{
"code": "Ratio"
},
{
"code": "SampledData"
},
{
"code": "time"
},
{
"code": "dateTime"
},
{
"code": "Period"
}
],
Should a separate ElementDefinition / path be used instead? Perhaps something like:
Observation.component.value[x].system or Observation.component.valueQuantity.system
To express this terminology constraint. The binding itself doesn't have anything that is computable that would indicate the context that under which it should be used. I suppose another option might be to create a FHIRPath expression for it:
Observation.component.value.is(Quantity) implies (Observation.component.value.as(Quantity).system = 'http://hl7.org/fhir/ValueSet/ucum-vitals-common|4.0.0')
Rob Hausam (Oct 03 2019 at 15:25):
I think it would make sense to declare the binding on the valueQuantity slice.
Chris Moesel (Oct 03 2019 at 16:13):
I agree that the binding at the value[x] element is ambiguous and validators would have to apply it to CodeableConcept and Quantity, which I'm sure is not the intent. I also agree that the solution is to apply the binding to a valueQuantity type slice instead.
That said, I'm looking at the VitalSigns profile now and not seeing that binding at all: http://hl7.org/fhir/R4/vitalsigns.profile.json.html. Are you looking at a version prior to R4?
John Timm (Oct 03 2019 at 18:37):
I'm using R4 and I'm seeing it here:
http://hl7.org/fhir/vitalsigns.profile.json.html
Chris Moesel (Oct 03 2019 at 18:54):
Oh sorry. Duh. I was looking at Observation.value[x], not Observation.component.value[x]. My bad. Sorry for the confusion.
John Timm (Oct 03 2019 at 19:01):
No worries.
Eric Haas (Oct 03 2019 at 19:19):
sounds like a tracker...
Last updated: Apr 12 2022 at 19:14 UTC