Representation of life status

[buske]

#6 introduces an Individual.lifeStatus property of type Concept, but there is no guidance as to the terminology.

A. [Ideal] we would use an existing terminology with concepts that correspond to alive, deceased, and unborn
B. use a string enum with the stated 3 values
C. expand the pedigree relationship ontology to include a branch with life statuses and add these three concepts

[liberaliscomputing]

@buske, I am not sure if it will be informative for you, but the Kids First Data Resource Center (KFDRC) created a custom profile called VitalStatus off of Observation (based on that Observation should record events at a point in time). Thus, we considered an inndividual's vital status as an observation event where Observation.code is "Clinical status" from SNOMED CT ("263493007") and Observation.valueCodeableConcept accepts actual codes from SNOMED CT.

Please see the followings:

• Profiles

  • VitalStatus with PHI: https://github.com/kids-first/kf-model-fhir/blob/master/site_root/input/resources/profiles/StructureDefinition-kfdrc-vital-status.json
  • VitalStatus without PHI: https://github.com/kids-first/kf-model-fhir/blob/master/site_root/input/resources/profiles/StructureDefinition-kfdrc-vital-status-no-phi.json

• Extension

  • age-at-event (for de-identified data): https://github.com/kids-first/kf-model-fhir/blob/master/site_root/input/resources/extensions/StructureDefinition-age-at-event.json

• Examples

  • An example resource with PHI: https://github.com/kids-first/kf-model-fhir/blob/master/site_root/input/resources/examples/Observation-vs-001.json
  • An example resource without PHI: https://github.com/kids-first/kf-model-fhir/blob/master/site_root/input/resources/examples/Observation-vs-001-no-phi.json

[buske]

Thank you, @liberaliscomputing. That is very helpful to see, and fits well with the structure proposed for eventually supporting Events. In terms of coding the life/vital status, SNOMED-CT isn't an open terminology, so I don't think we should recommend it's use for international standards.

[liberaliscomputing]

Re naming, here are some examples from established ontologies:

• SNOMED CT: Clinical status
• NCIt: Vital Status