20151204 OO FHIR conCall
| HL7 OO on FHIR (for Orders and Observations) Call in details: |
Date: 2015-12-04 2015 - 02:00 PM (Eastern Time, GMT -04 DST) | |
| Chair: | Note taker(s): | |
| Quorum = chair + 4 | yes/no | ||||||||||||
| Co chairs | Riki Merrick | Rob Hausam | Lorraine Constable | Patrick Lloyd | Ken McKaslin | Hans Buitendijk | |||||||
| Fhir Editor/Facilitator | Eric Haas | ||||||||||||
| Attendees | ||||||||
| Riki Merrick | Rob Hausam | Paul Knapp | Scott M | |||||
| Eric Haas | Attendee6 | Attendee7 | ||||||
| . | ||||||||
Contents
Agenda
- Roll Call
- Agenda Check
- Minutes from
- Review Tracker Items - Device resource: [1]
*7254 2015May core #373 - UDI doesn't identify repository/jurisdiction (Paul Knapp) Persuasive with Mod In Person
- Connectathon
minutes:
- UDI 7254
Proposal:
keep separate Device.udi element since a) removing would cause confusion with implementers and easier for implementers if udi is a separate identifier element.
1 keep name Device.udi (earlier proposal for ‘udistring’ – again
2 change type from string to identifier
3 use this for the udi string data NO MATTER if is a type or identifier ( this is impure from HL7 but felt is too hard for implementers to choose between identifier and code )
4 Add following udi system to fhir identifier registry and map to existing OIDs : FDA UDI root - urn:oid:2.16.840.1.113883.3.3719 ( Paul to check if FDA wants to define URL like http://hl7.org/fhir/sid/us-udi)
5 provide clearer notes in element definition and notes section as well as inline example:
The udi string is stored in the udi attribute and the repository URL is the default system (for example for the FDA UDIs, the correct value to use in the Identifier.system is urn:oid:2.16.840.1.113883.3.3719 .”
keep separate Device.udi element since a) removing would cause confusion with implementers and easier for implementers if udi is a separate identifier element.
1 keep name Device.udi (earlier proposal for ‘udistring’ – again
2 change type from string to identifier
3 use this for the udi string data NO MATTER if is a type or identifier ( this is impure from HL7 but felt is too hard for implementers to choose between identifier and code )
4 Add following udi system to fhir identifier registry and map to existing OIDs :
FDA UDI root - urn:oid:2.16.840.1.113883.3.3719 ( Paul to check if FDA wants to define URL like http://hl7.org/fhir/sid/us-udi)
5 provide clearer notes in element definition and notes section as well as inline example:
The udi string is stored in the udi attribute and the repository URL is the default system (for example for the FDA UDIs, the correct value to use in the Identifier.system is urn:oid:2.16.840.1.113883.3.3719 .”
Example: <Device xmlns="http://hl7.org/fhir"> … <udistring> <system value =”urn:oid:2.16.840.1.113883.3.3719”/> <value="{01}0061414999996{17}910304{10}123ABC{21}1234567890"/><!—this is the HRF udi string--> </udistring> <identifier> <value value="1234567890"/><!—this is the serial number--> </identifier> ...
discussion
Will send the UDI proposal to the Ord list and vote on that next week
Updated the proposal to not change the name for the Device.udi element – leave as is
Suggest the OID and see, what FDA says
where there are already existing OIDs we should use what is there / URL vs OID format?
Proposed OID is FDA root OID assigned by HL7 – need to extend the OID? – assume it is good for now – Paul will confirm, as the OID looks weird – different OID for submitting body – should reach out to Ted?
Discussion about the difference between coding and identifier – a code can be used to identify a single instance, but usually is used to identify classes of things – but the name is identifier, which makes is confusing.
The first component of the UDI string identifies who issued it, which will give you information how to parse, but you cannot tell which country it is in (which repository do you look for this OID).
Add another element for the UDI system = part of the coding element
Discussion about leaving it as an identifier, because that is what FDA calls it, even if that does not jive with the definition of data types we have, as it can be used as one or the other – code is more broad – it is ALWAYS a code, that periodically is uniquely identifying a single instance of the “class”.
This document will be sent to the ord list – will discuss next week
FDA is now suggesting that the string is unnecessary as all the information is available in the separate elements already – this flies in the face of what several issuing organization are specifying for the string. Paul to chase that down.
Type of the string is not discussed here
Possibly create a US profile to give specific guidance and how that fits into the specification
Device example need to change ( to { - add more examples
- Connectathon Clarification:
Track lead duties:
Creating the test examples – classes to be tested, test scripts (can approach in different ways – for example observation type – several patient information) – Eric and Riki have create the test scripts for the NIST v2 tool testing
In session you are the coordinating resource working on the content – liaison to FHIR core members to resolve issues
Need to have to have a server to file the lab order. The first step is can two
Is someone brining the “lab” to allow results to be sent for the order I sent you?
Provide the content, but not create the resources to be exchanged.
The servers might have to have some logic in order to make this work
Rob H is working with HAPI server – create a dB on it and build a receptacle for a lab order and then generate a result for that order and then the orderer can post a get to the server
Eric is identifying the steps in the test case
Lead does not need to create all the resources etc, but needs to figure out who will be the client and the server, so you have all the pieces – first step might be just checking if systems can take in the resources
Minutes
Next Steps
| Actions (Include Owner, Action Item, and due date) | |||
| Next Meeting/Preliminary Agenda Items |
Back to OO_on_FHIR
© 2015 Health Level Seven® International
