Stream: implementers
Topic: Integrating wearable sensor observation codes
Robert Lyons (Nov 26 2019 at 18:17):
Dear community,
I'm implementing a data integration of wearable sensor observations from disparate sources and varied types of sensors (including actigraphy, cgm, ecg, spirometry, blood pressure monitoring, and others). I'm looking to standardize the names of the incoming observed properties. I'm guessing LOINC is a solid option and would cover many of these properties, but I am unsure the extent of coverage I can expect and what other controlled terminologies would help increase the coverage. Any help would be greatly appreciated! Thanks.
Lloyd McKenzie (Nov 26 2019 at 18:34):
@John Rhoads
John Rhoads (Nov 26 2019 at 20:37):
@Robert Lyons LOINC is indeed a solid option and covers many common (and less common) sensor observations. It gets wide use in the US, but is perhaps a bit less familiar internationally. LOINC is justly famous for laboratory observations; sensor observations are a little less its focus, although through the efforts of Dr. Swapna Abyankar it is getting strengthened in the device area. Another terminology resource you should be aware of is the ISO/IEEE 11073-10101 (and others in the -1010x series of standards) Medical Device Nomenclature (see https://www.hl7.org/fhir/mdc.html for a quick intro, and https://rtmms.nist.gov/rtmms/index.htm for an online reference database maintained by the US National Institute of Standards and Technology), which is almost solely device and sensor-oriented. It is strongest for devices and sensors used in acute-care settings but includes some personal health devices - coverage of wearables is incomplete but the database and supporting standards are open to suggestions for additions. Full disclosure: I'm chair of the IEEE 11073 Point-of-Care work group. Would be happy to discuss further.
Robert Lyons (Nov 26 2019 at 22:38):
@John Rhoads thanks for the response and particular ly the pointer to the IEEE 11073-10101 ref. I'd seen it in my googling and thought it looked interesting. Will dive in deeper and reach out again. Thanks!
Robert Lyons (Nov 26 2019 at 22:38):
@John Rhoads thanks for the response and particular ly the pointer to the IEEE 11073-10101 ref. I'd seen it in my googling and thought it looked interesting. Will dive in deeper and reach out again. Thanks!
Robert Lyons (Nov 30 2019 at 11:50):
@John Rhoads thanks for the response and particular ly the pointer to the IEEE 11073-10101 ref. I'd seen it in my googling and thought it looked interesting. Will dive in deeper and reach out again. Thanks!
John Moehrke (Nov 30 2019 at 16:17):
Try @Todd Cooper or @Keith Boone or @Paul Schluter
Daniel Vreeman (Dec 01 2019 at 12:28):
@Robert Lyons Yes, LOINC is good choice for this (and is free worldwide). LOINC is a community-driven effort (new content is added by end-user requests), so if there are variables you have that aren't in LOINC, a submission would be welcome (see https://loinc.org/submissions/). The ongoing LOINC/IEEE collaboration (https://loinc.org/collaboration/ieee/) has produced a mapping between IEEE variables and LOINC Codes. You can get that mapping via LOINC's FHIR terminology services too.
Umberto Cappellini (Dec 02 2019 at 10:43):
Hi @Robert Lyons , I think the OSS project from Microsoft for IoMT and FHIR may be interesting for you:
https://azure.microsoft.com/en-in/blog/accelerate-iomt-on-fhir-with-new-microsoft-oss-connector/
Robert Lyons (Dec 02 2019 at 17:33):
@Daniel Vreeman thanks for the pointers. @Umberto Cappellini great azure resource. Thanks!
Keith Boone (Dec 13 2019 at 19:45):
Depends on a couple of things, including your market. LOINC has the benefit of being interpretable by many EHRs used in the US and is consistent with US core specifications for many device measured. IEEE 11073 probably has more international acceptance especially for device based observations. There's a published cross-walk between 11073 and LOINC on the interwebs: https://loinc.org/collaboration/ieee/
Last updated: Apr 12 2022 at 19:14 UTC
