FHIR Chat · MedicationStatement - dosageAndRate not must-support? · IPS

Stream: IPS

Topic: MedicationStatement - dosageAndRate not must-support?


view this post on Zulip Morten Ernebjerg (Oct 01 2020 at 09:34):

Hi :wave: I've been looking at the IPS MedicationStatement profile and was surprised to see that dosage.doseAndRate (or sub-elements thereof) is not must-support, although dosage.timing and dosage.route are. I would have expected the "when" (timing) and the "how much" (doseAndRate) to have the same status. Indeed, the MedicationStatement example that comes with IPS has both filled. What was the reasoning behind this choice and what is the recommended way of specifying the intake amount (in a way that only uses "safe" must-support fields) if dosage.timing is filled?

view this post on Zulip Christof Gessner (Oct 01 2020 at 09:42):

Well, from a Patient Summary perspective this might make sense: The information that you were on a specific medication at a particular point in time is relevant. The precise dosage of that medication seems less relevant.

view this post on Zulip Rob Hausam (Oct 01 2020 at 13:10):

I agree with Christof that the basic rationale is that the IPS is truly a "summary" to provide the most essential data to support clinical care, not necessarily all of the specific details (but you are still free to include and send those if you desire).

view this post on Zulip Morten Ernebjerg (Oct 01 2020 at 15:24):

Thanks for the input! - that hierarchy of importance/focus definitely makes sense. What I still do not quite understand is the "relative question": why is dosage.doseAndRate on a lower level in this hierarchy than dosage.timing/dosage.route?

  1. If the specific use case supported is capturing the time period during which the patient was taking the medication, one could also (if I read it right) use MedicationStatement.effective[x](required in IPS anyway). Then the question becomes why dosage.timing needs to be must-support.
  2. Conversely, dosage.timing is also for recording the intervals at which you are taking individual portions of the medication (e.g. once in the morning). If such interval information is essential, why is the amount of meds the patient took (at this specific interval) not?

To concretely illustrate my confusion using the official IPS example of a MedicationStatement: Suppose we strip it down to only the must-support fields, as conformant systems handling the data would be allowed to. Then dosage.doseAndRate is lost and, as far as structured information goes, we are left with the statement that the patient took some unknown amount of the medication exactly once a day. That feels odd to me - is my "layman's clinical intuition" simply off?

view this post on Zulip Christof Gessner (Oct 01 2020 at 16:45):

I agree, in the end it is a matter of discussion. So during and before a ballot we asking as many experts and users as possible, and then come to a conclusion. BTW some of the decisions were derived from the functional requirements in the CEN IPS standard. Did you check there?

Relevance for timing could also be that in a summary you also want to know, if a medication was administered regularly, or on demand. Wouldn't that be expressed in .timing?

view this post on Zulip Christof Gessner (Oct 01 2020 at 16:52):

.route informs, if something was administered oral or i.v., for example. Apparently more relevant in a summary than the amount.

view this post on Zulip Morten Ernebjerg (Oct 02 2020 at 06:48):

@Christof Gessner Good point about CEN IPS - I will have a look there.


Last updated: Apr 12 2022 at 19:14 UTC