Stream: implementers
Topic: MessageHeader.event Coding expansion
jassy (Aug 12 2016 at 22:13):
I'd like to use an event Coding to represent EligibilityRequest/EligibilityResponse messaging, but I cannot find a suitable Coding from the current Preferred list "http://hl7.org/fhir/message-events". Could we expand the list? or could I define a custom Coding system?
Grahame Grieve (Aug 12 2016 at 22:44):
you can make up your own events, and/or you can review the connnectathon notes from previous connectathons to see what they are using
jassy (Aug 14 2016 at 05:12):
The previous connectathons do not show use cases with the "messaging" transaction paradigm I am looking for. For example, most use cases are using a REST get request and a separate pull for response. Thanks for the reference and suggestion.
Grahame Grieve (Aug 14 2016 at 10:30):
@Paul Knapp - I thought there was a message event for this somewhere?
Paul Knapp (Aug 14 2016 at 11:58):
No, the FHIR Message connectathon work pre-dated the current MessageHeader and $process-message endpoints. Most of you testing of late has been conventional RESTful Request/Response or WSI Web Services (or FHIR REST using POST and GET).
The Message Event codesystem is inconsistent, but inferring the pattern then something like 'EligibilityRequest-Process' for $process-request to 'Process an EligibiltyRequest and return an EligibilityResponse.' would be appropriate.
Lloyd McKenzie (Aug 15 2016 at 16:12):
@Paul Knapp The notion of event code hasn't changed since the inception of MessageHeader. In any case, you can't infer the existence of codes not defined in the current code system. So while you can follow the naming pattern, you'd still have to do it within your own code system with your own URI until such time as HL7 defines an event code. (It's not clear right now whether HL7 is going to take this on as not a lot of implementers have been seeking standardized event codes - almost everyone wants to go off and define their own anyhow.
Paul Knapp (Aug 17 2016 at 21:06):
I didn't think the notion of the event code has changed, I said the code name pattern was inconsistent. Yes, to get a code in the codeset we would have to add it to the codeset. Agreed, people going off and doing their own thing is a problem, probably the biggest barrier to interoperability and largest source of risk and waste - I think that we need to be active rather than passive in addressing this.
Grahame Grieve (Aug 17 2016 at 22:07):
generally, we are pretty passive in the messaging space because
- few implementers are using it, and those that are have not stepped forward to ask for changes
- no committee is standing forward and caring
Last updated: Apr 12 2022 at 19:14 UTC
