Stream: implementers
Topic: geocode extension
David Hay (Jan 20 2020 at 20:17):
We'd like to use the Location.position element in national NZ project - but the national geocoding service uses a different datum than WGS84, which is specified in the spec. It's impractical to convert to WGS84 at this time, so we need an extension of some sort.
This could either be a complex extension at the root level, or a modifierExtension on Location.position.
What's the best approach?
Lloyd McKenzie (Jan 20 2020 at 22:12):
Should we push back against the WGS84 requirement and make the element more generic in R5? (Still would need an extension in R4)
Brian Postlethwaite (Jan 22 2020 at 23:22):
Or have a different extension for the other datum
Brian Postlethwaite (Jan 22 2020 at 23:23):
Which could also be used in the location type too.
Lloyd McKenzie (Jan 23 2020 at 00:06):
Which would be more intuitive implementers?
David Hay (Jan 28 2020 at 21:40):
Assuming that most of the world (especially google) uses WGS84, then leaving that seems most intuitive to other developers to me. In the absense of a strong opinion otherwise, I'd go for a separate extension at the root (rather than the modiferExtension on the element) - that way if we do eventually support WGS84 as well, we can populate Location.position as well...
Samantha Burchell (May 11 2021 at 01:48):
@David Hay at Healthpoint we use WGS84 (longitude and latitude) as specified in FHIR. Will that be fine and not to use the NZ extension? Cheers, Sam
David Hay (May 11 2021 at 02:28):
That would be best - if you can be compliant with the spec then so much the better!
Last updated: Apr 12 2022 at 19:14 UTC
