Stream: implementers
Topic: Condition Staging
Brian Kaney (Jun 23 2016 at 15:47):
Consider in oncology where there is diagnosis stage and pathological stage of a patient's problem. The Condition FHIR resource has one stage. So would this be best represented as two Condition resources for the patient, one with a category of 'diagnosis', and the other with a category of 'pathological' (adding that item to the condition-category valueset)?
Lloyd McKenzie (Jun 24 2016 at 02:44):
I think the existing "stage" element represents that diagnosis stage. Pathological stage would be an extension.
Erich Schulz (Jun 24 2016 at 08:05):
to me cancer staging feels almost like an observation...
Erich Schulz (Jun 24 2016 at 08:06):
like a GCS or Apgar score its made up of componenents then has a value "I, IIa, IIb, III, IV" etc
Erich Schulz (Jun 24 2016 at 08:07):
and for the same cancer there can be different competing (or evolving) stage classifications
Brian Kaney (Sep 08 2016 at 19:44):
@Erich Schulz According to analysts I spoke to, staging is semantically part of the patient's condition.
@Lloyd McKenzie what would the semantics behind category mean (considering one of the items in the valueset is 'diagnosis')?
Brian Kaney (Sep 08 2016 at 19:55):
One idea is to make a condition 0-* stage, and add to stage a category or kind.
Lloyd McKenzie (Sep 08 2016 at 20:35):
@Brian Kaney I think stage and category are very different things. Would it ever make sense to state a condition and convey both the diagnosis stage and the pathological stage? Could you give an example or two?
@Eric Haas Condition is a special sort of Observation - the stage, the severity, the verification status, etc. could all be considered observations, but it makes sense to group them all in Condition because in the end, it's really a single statement - "confirmed severe stage 3 whatever"
Eric Haas (Sep 08 2016 at 20:52):
forwading to @Erich Schulz
Brian Kaney (Sep 08 2016 at 20:55):
@Lloyd McKenzie The proposal would be to allow a Condition to have multiple Stagings, each with a kind. Our analyst listed a few, 'clinical', 'pathologic', 'recurrent', etc.
Lloyd McKenzie (Sep 08 2016 at 21:31):
That could work. Feel free to submit the change proposal
Thomas Wicker (Sep 15 2016 at 17:28):
Hey, all -- as one of the analysts @Brian Kaney spoke to, some thoughts:
(ah - straight return submits a comment; good to know ... )
1. "Staging" could be an observation; is currently in Condition, and could leave it there just fine.
2. For sending complete info in oncology, could easily have a Stage at Diagnosis, a Clinical Stage, and a Pathologic Stage:
Pt examined by Provider A; Provider A determines (clinically/pre-surgically) Stage IIA Breast Cancer.
Time passes (life happens); Pt examined again, new clinical stage = IIB.
Surgery happens.
Pathologist looks at surgical specimen, nodes, etc., determines Pathologic tumor & node classifications that result in Pathologic stage IIIA.
Three stages, one pt., all changes plausible.
Last updated: Apr 12 2022 at 19:14 UTC
