Chris Courville (Epic) - presiding co-chair

Brian Reinhold (LNI/Continua)
Michael Faughn (Prometheus Computing / NIST)
Paul Schluter (Center for Medical Interoperability
John Rhoads (Philips) - notetaker, circulated to attendees, corrections incorporated
Ana Kostadinovska (Philips)
Koichiro Matsumoto (Nihon Kohden)
Myron Finseth (Medtronic)
Uwe Meyer (MT2IT)
John Dyer (Draeger)
Stefan Karl (Philips)


  • Discussion of current state of FHIR Device resource proposal spreadsheet from joint meeting with Orders and Observations 2018-07-24 see minutes and spreadsheet in OO Conf Call Minutes - Device(Definition) - 20180724
  • Resolution of comments on FHIR PoCD Implementation Guide - January 2018 Ballot

Discussion Notes


After a point-by-point Q & A session on the proceedings of the 24 July joint meeting of Health Care Devices and Orders and Observations for the benefit of those who were not present, there was a line-by-line discussion of items in the right side columns concerning Device (instance).

A group statement was proposed by Brian Reinhold, seconded by Michael Faughn, and voted 10-0-0 as follows:
All present said they were satisfied with the contents of the spreadsheet as it stood after the 24 July joint meeting with O & O (and hoped very much that it would stay as it is through the rest of the process – we do not think there is excessive duplication of fields between DeviceDefinition and Device and that the presence of these fields was necessary for the intended use in PHD and PoCD implementations). We also looked at the column Q descriptions and found them satisfactory.

<end of statement>


The group then went down the left (DeviceDefinition) section. Several present thought that O&O members who had worked through particular use cases for collections of devices ought to be “in control” of the DeviceDefinition material – a number of the items there would no doubt be useful in HCD use cases to “look up” pieces of “kind” information, but such HCD use would not be the primary use of DeviceDefinition. Therefore the sentiment was to mostly defer to others in the DeviceDefinition decisions. A couple of specific comments, though, were to observe that the mentions of “settings” in the description of “property” in DeviceDefinition would probably not be used in that way in HCD use cases: “property” would be used in PHD for a narrower set of data items than is probably contemplated for “property” in DataDescription. It is usual in PHD to report device settings as observations, and in Point-of-Care devices, the notion of “settings” covered such dynamically manipulated things as ventilator settings (e.g. Peak End Expiratory Pressure) that may be changed by clinicians in accordance with clinical judgment and are likewise treated as observations. So they wouldn’t be the kinds of things that would go in a DeviceDefinition as there might not be a meaningful factory or recommended setting.

But it seemed to the group quite acceptable for exact usage and sense of like-named fields to vary between DeviceDefinition and Devices, appropriate to the use cases, so we raise no objection.

Comment resolution (PoCD IG):

Tracker Items (see links to trackers for details)

15352 Motion to accept proposed resolution - S. Karl / Brian:10-0-0

15354 Motion to accept proposed resolution - S. Karl/Rhoads:10-0-0

15378 Motion to accept proposed resolution - Reinhold/S. Karl:10-0-0

Action items

  • John Rhoads Email Hans Buitendijk, OO co-chair of referenced meeting, with results of this meeting