Comment by perturbation | Hacker News Reader

Comment by perturbation | Hacker News Reader

Parent Full thread

See also (edit: just noticed that they include some of these links at the end of the article): http://docs.smartplatforms.org/

SMART on FHIR builds on top of FHIR for cross-EHR applications. In their words (http://smartplatforms.org/for-developers/):

"SMART on FHIR takes advantage of a new and powerful set of health data standards called Fast Healthcare Interoperability Resources, or FHIR. By using HL7’s FHIR for the clinical data layer, SMART offers a standards-based platform with broad reach, a rich set of detailed models, and the opportunity to create population-level as well as patient-level apps. SMART on FHIR also uses an updated set of authorization and authentication standards including OAuth2 and OpenID Connect."

and

http://jamesagnew.github.io/hapi-fhir/

(extension to the HL7 Java API "Hapi" which deals with FHIR and is (IMO) one of the more full-featured implementations of FHIR)

Disclaimer: I used to work for a major EHR company but no longer do so.

Just wanted to chime in here -- I'm the architect for SMART Platforms and happy to answer any questions. Over the past five years I've gotten much more deeply involved the healthcare standards development world than I ever would have imagined. I came at this space from the perspective of wanting to build applications that plug into real-world systems, but in some ways I've been digging lower and lower down the stack trying to lay firm groundwork. There's constant tension between laying groundwork and trying to get "real" work done.

Do you think that there's any chance of getting interop via FHIR (with or without SMART) as a condition of winning the DHMSM deal? My impression (rightly or wrongly) is that a lot of vendors are doing a lip service/wait-and-see approach to FHIR and truly opening up.

I'd like to hope so. I think we've seen this kind of interaction play out on a much smaller scale in the commercial world, and we're seeing more of it. For example, when Intermountain Healthcare selected a Cerner EHR, they did so as part of a multi-year partnership [1], under the condition that Cerner would provide open interfaces to their EHR data so that Intermountain could continue developing (and begin porting their existing) home-grown decision support tools. Part of the arrangement was an innovation fund to support development of the specifications needed. And I think that's actually a great pattern, since we certainly won't have all the answers right out of the gate. With respect to DHMSM, the timeline is long enough that it's very hard to name relevant technologies on the way in, without the concern that they'll be out of date (or at least out of fashion) by the time implementation hits.

1. http://www.cerner.com/Intermountain_and_Cerner_announce_stra...

Reply on news.ycombinator.com