Stream: social
Topic: 21st Century Cures Act has dropped
David Pyke (Mar 09 2020 at 13:15):
"e. Application Programming Interfaces (APIs)
We have adopted a new API certification criterion in § 170.315(g)(10) to replace the “application access—data category request” certification criterion (§ 170.315(g)(8)), and added it to the updated 2015 Edition Base EHR definition. This new “standardized API for patient and population services” certification criterion focuses on supporting two types of API-enabled services: (1) services for which a single patient's data is the focus and (2) services for which multiple patients' data are the focus. The API certification criterion requires the use of the Health Level 7 (HL7®) Fast Healthcare Interoperability Resources (FHIR®) standard Release 4 and references several standards and implementation specifications adopted in § 170.213 and § 170.215 to support standardization and interoperability. This certification criterion will align industry efforts around FHIR Release 4 and advance interoperability of API-enabled “read” services for single and multiple patients."
David Pyke (Mar 09 2020 at 13:42):
Dave deBronkart (Mar 09 2020 at 14:31):
Thanks @David Pyke ... Here's a tiny post I did with links to summary sheets and full text of both rules. No analysis in this post - just the links and a comment that woohoo, FHIR is now Federal policy!
Re this thread's title - technically the rules have dropped (from ONC and CMS). The 21st Century Cures Act itself was passed in the Obama administration; today we got the regulations to implement the law. To my limited knowledge only the ONC rule is explicitly related to that law.
Please someone correct me if I'm wrong on that.
In any case, the news is great!
David Pyke (Mar 09 2020 at 14:33):
Correct, the law was passed in 2016. These are the regulations (and changes to regulations) that implement the law. Also, the CMS Rule isn't specifically tied to this Rule but refers and adheres to it.
Lloyd McKenzie (Mar 09 2020 at 16:06):
Does the rule allow adoption of anything newer than R4 or is the U.S. stuck on R4 until there's a new rule? (Because that would suck...)
Chris Moesel (Mar 09 2020 at 16:22):
I don't know the answer to that question, but it seems to me like agreement on the version of FHIR would be important to maintain interoperability across the broadest set of systems. It wouldn't be helpful if 40% of systems were on R4, 40% of systems on R5, and 20% of systems on R6. That said, if R4 is required as a minimum, that wouldn't prevent systems from also supporting more recent versions. (Again, haven't read the rule -- so this is not a comment on the rule itself, but rather the implications of being flexible on versioning).
David Pyke (Mar 09 2020 at 16:56):
It requires 4.0.1 unless the ONC specifies a later version in a later rulemaking
David Pyke (Mar 09 2020 at 16:57):
Both the ONC and CMS rules state it's R4.0.1 so the US us stuck with it for a while
David Pyke (Mar 09 2020 at 16:58):
It also specifies U.S. Core 3.1.0
Josh Mandel (Mar 09 2020 at 17:13):
Cross-post from my Twitter:
Reading through #CuresRuleONC today? Me too! Let's make sense of things together via live discussion today at 2p ET. (I'll have PDFs open and screenshare on.)
{ "where": "http://aka.ms/jmandel.chat", "when": "2020-03-09T14:00:00-04:00", "who": ["@joshcmandel", "$you"] }
Abbie Watson (Mar 09 2020 at 18:08):
That's awesome to get the official word on R4.
Dave deBronkart (Mar 09 2020 at 22:08):
Lloyd McKenzie said:
Does the rule allow adoption of anything newer than R4 or is the U.S. stuck on R4 until there's a new rule? (Because that would suck...)
Time to get slicey-dicey about this. What is "allow adoption"?
I'm sure I have the terms wrong, but I'll guess there's such a thing as a FHIR Request Sender and a FHIR Request Receiver, roughly also known as FHIR Data Wanter and FHIR Data Haver. And I'll guess that FHIR 4.0.1 provides a certain level of clever and granularity and later versions are probably a superset (though some 4.0.1 might get deprecated in later versions).
So good so far? Please correct.
Now, if I'm a slicker shinier newer Data Haver, so I can Receive Requests at a higher level, I'd say I've adopted something higher. Are you asking if I'm ALLOWED to do that? (Why wouldn't I be?)
Or maybe more exotic: Let's say I'm Mankind's Greatest Hospital and I have super-awesome biz partners and we BOTH can talk FHIR beyond 4.0.1. Are you asking if that's allowed?
Or are you asking whether I, a studly FHIR Data Wanter, am allowed to DEMAND something beyond 4.0.1?
Dave deBronkart (Mar 09 2020 at 22:09):
Josh Mandel said:
Reading through #CuresRuleONC today? Me too! Let's make sense of things together via live discussion today at 2p ET. (I'll have PDFs open and screenshare on.)
Oh bother, I missed THAT window ... is there an archive?
Josh Mandel (Mar 09 2020 at 22:13):
There are discussion notes but not a video recording . We'll probably schedule another one for Thursday or Friday.
Josh Mandel (Mar 09 2020 at 22:14):
Re: upgrades, a few key points:
- Future FHIR versions won't be a strict superset of functionality in a technical sense -- there will be breaking changes, for all but the "normative" level resource
- ONC defines a "standards advancement process", which gives ONC the option to allow vendors to certify to newer versions of FHIR even before a new rule is published; but only if ONC specifically allows a version and the vendor choose to certify to it. Otherwise, things fall back to R4 (as named in rule).
Josh Mandel (Mar 09 2020 at 22:15):
This was one of the first topics that cam up on today's call; raw notes here.
Vassil Peytchev (Mar 10 2020 at 03:38):
That's good news, since there is a lot of interest in the subscription/notification framework in R5
Last updated: Apr 12 2022 at 19:14 UTC
