Stream: implementers
Topic: Pulse Oximetry
John Moehrke (Sep 03 2018 at 17:43):
I have not found guidance on how to record Pulse Oximetry. In my case it is patient reported. Again a huge number of LOINC codes, not clear which to chose.
Bryn Rhodes (Sep 03 2018 at 21:14):
Coincidentally, I was working on a measure that needed exactly this today, so I found this in the US Core profiles:
Bryn Rhodes (Sep 03 2018 at 21:14):
https://www.hl7.org/fhir/us/core/Observation-oxygen-saturation.html
John Moehrke (Sep 04 2018 at 12:40):
I have more information than that profile describes. I have a comment, but also a symptom (shortness of breath, dizziness, etc). I have some details about the assistance device, but not enough for a Device resource. Do I just add these as more .component elements?
Lloyd McKenzie (Sep 04 2018 at 14:07):
You can send a comment with the existing profile using event-note extension. Shortness of breath, etc. would be a distinct Observation as it's interpretable independently from the pulseox reading.
John Moehrke (Sep 04 2018 at 14:15):
okay, one Observation would link to the other.. right? which direction?
Lloyd McKenzie (Sep 04 2018 at 14:23):
Unless one is derived from the other, there'd be no more an explicit link than there is between a blood pressure and a pulseox that happened to be made at the same time. You could create a "panel" Observation that groups them as members if that's important/relevant.
Eric Haas (Sep 04 2018 at 14:41):
I think there are better ways to do this like placing the symptoms etc in the procedure and using Observation to record the actual measurement and linking it to the procedure.
Eric Haas (Sep 04 2018 at 14:43):
you have extensions for reasonCode for symptoms or Observation.note extension for more notes ( we are discussing with PC Observation.note extension v Observation.comment at the WGM. )
John Moehrke (Sep 04 2018 at 14:44):
additional characteristic in my case is that these are patient reported... so there is no procedure
Eric Haas (Sep 04 2018 at 14:46):
sounds like Condition to me but that is fuzzy area
John Moehrke (Sep 04 2018 at 14:46):
on the UI for submitting Pulse Ox is a dropdown for selecting additional symptioms... so the patient is submitting a measured value, and then being asked to think about if they had one of those symptoms
John Moehrke (Sep 04 2018 at 14:47):
each time they enter...
Eric Haas (Sep 04 2018 at 14:48):
So .. . they may be reasons or just more indepenent ( other than the same encounter ) Observations or Conditions? W
Eric Haas (Sep 04 2018 at 14:51):
re not using device. you could hang the text blob Observation device reference display (although Lloyd might frown on that :-))
John Moehrke (Sep 04 2018 at 14:53):
that is exactly what I was thinking of doing... including Lloyd frown... Given that this is patient generated it is hard to justify creating (potentially) a new Device resource each time an Ox observation is made. I don't have an ability to relate each Ox to a specific device.
John Moehrke (Sep 04 2018 at 14:54):
but I mostly just have a 'type of device'... so not clear how that alone is recorded.
Eric Haas (Sep 04 2018 at 15:01):
I am surprised that an implementer has never opened a tracker for a deviceCode.
John Moehrke (Sep 04 2018 at 15:02):
I could do that
John Moehrke (Sep 04 2018 at 15:10):
Robert McClure (Sep 05 2018 at 15:27):
@John Moehrke Why "no procedure"
additional characteristic in my case is that these are patient reported... so there is no procedure
Patients can be the source of a procedure, no?
John Moehrke (Sep 05 2018 at 15:33):
technically, no problem... my point is that this is simply self entered Pulse Ox data. I have no way to relate it to a Procedure. There is no information that I can use to invent a Procedure. Lastly, no EHR that I know of allows a Patient to create a Procedure (most specifc to my case, Cerner doesn't).. but technically the current FHIR specification does seem to allow a Patient to be the recorder and asserter of a Procedure.
Michelle (Moseman) Miller (Sep 05 2018 at 19:29):
@Eric Haas Regarding GF#17576, do you know which quarter this discussion is planned?
we are discussing with PC Observation.note extension v Observation.comment at the WGM.
Michelle (Moseman) Miller (Sep 05 2018 at 19:33):
@John Moehrke regarding the comment
no EHR that I know of allows a Patient to create a Procedure (most specifc to my case, Cerner doesn't)
FYI, functionally speaking, Cerner does allow a patient to self-report a procedure (albeit via proprietary APIs at the moment). Once reconciled to the patient's chart, the practitioner gets "credit" for recording the procedure.
John Moehrke (Sep 05 2018 at 21:01):
what requires clarification @Michelle (Moseman) Miller ? You said it must be done by proprietary API. This is a FHIR discussion thread. full-stop.
Why would the practitioner get credit for a Procedure that a patient generated?
The overall use-case I have is purely patient generated pulse-Oximetry data; so I don't know what the use of discussing Procedure is. I am not saying that the system can't notice the patient generated Observation and update that Observation to point at an existing Procedure. I simply don't have that data, I can't record the Procedure. I can only record the Observation.
Michelle (Moseman) Miller (Sep 06 2018 at 11:06):
Sorry @John Moehrke for the confusion. I wasn't trying to influence the broader Pulse Ox discussion, but instead, I was trying to shed some clarification on the following statements you made:
Lastly, no EHR that I know of allows a Patient to create a Procedure (most specifc to my case, Cerner doesn't).. but technically the current FHIR specification does seem to allow a Patient to be the recorder and asserter of a Procedure.
To elaborate, the Patient Care workgroup added Procedure.asserter and recorder as part of GF#12633. The specific use case is around "procedure statements", which can be part of past medical/surgical history. The scope of Procedure indicates it can be used for current procedures (often asserted/recorded by a practitioner) as well as historical procedures (often asserted/recorded by a patient or related person). I used Cerner in my earlier comments as an example of an EHR that supports capture of patient-reported procedures (e.g. procedure statements). While our patient-recorded procedures are written using a proprietary API at the moment, we are moving towards supporting FHIR APIs for the writes.
John Moehrke (Sep 06 2018 at 11:43):
I would be interested in realistic (not theory) use-cases for a Patient authored Procedure. I think that is a different thread than Pulse Ox self-reported measurement. right? How would a Pulse Ox measurement be a use-case for creating a Procedure? I certainly see how it could be correlated to a CarePlan or prescription or other that initiated the use of Oxygen machine. Overall I would like to see a discussions of Patient authoring of Procedures on a different thread.
Last updated: Apr 12 2022 at 19:14 UTC
