First of all, apologies for not reporting at the end of July - this was right in the middle of preparing for the publishing the FHIR R4 Ballot #2, which was our biggest editorial effort yet. The ballot is now open - along with 14 different implementation guides.
This is the last ballot for R4. The way the normative balloting works under HL7/ANSI rules is that we can't make substantive changes to the specification - that is, anything that would impact on specifications - without reballoting, and we won't be reballoting before we publish R4 on schedule at the end of this year. Balloting this round is limited to the changes that were made between this ballot and the last ballot. Each substantive change that was made is listed on the ballot introduction page:
If we don't get consensus from the HL7 balloters to the changes we made, we'll drop the relevant sections of R4 back to Trial Use and publish anyway.
Many individuals and committees worked really hard and/or effectively to produce the ballot ready version of FHIR. It's hard, from outside the process, to appreciate just how much work goes into the specification, but we processed over a 1000 ballot comments and editorial changes covering things such as typos, and html style issues all the way through to some major changes to identifying URLs. We'll be working to update the contributer list when we publish the R4 final specification but the list is so long now that it's hard to know what do with it: FHIR is a big community.
Ballot sign up is now closed - if you have issues that you feel needs to have the weight of a ballot comment, but you aren't signed up for the ballot you may be able to find a balloter who will take you issue on board. Note that this means that they'll represent you issue in an ongoing fashion. If you'd like help finding someone, contact me privately (at email@example.com).
The FHIR foundation publishes a few implementation guides directly - see:
Several more are on the way. The FHIR foundation will also publish other IGs, using the following rules:
FHIR Foundation members can request this by emailing me at firstname.lastname@example.org.
We are working on creating the first formal FHIR foundation project around the area of FHIR data analytics support. This is appropriately a FHIR foundation project not an HL7 project because there seems to be no intent to produce specifications, but rather white papers around implementation techniques. The group is meeting regularly and you can track/join progress here:
Note: there's events that cover or mention FHIR all the time now - I can barely keep with the mentions in social media etc. It's really the case that any conference anywhere in the world about health IT will cover FHIR to some degree or other now. Here I mainly focus on events focused on FHIR itself, or that were attended by one of the editors of the core FHIR Modules.
OpenHIE held their first "HackConnectathon" in Arusha, Tanzania and Bryn Rhodes attended on behalf of the FHIR team (thanks for squeezing family holiday time to do this), and provided the following report:
FHIR was a big part of the event. The event was organized similar to a FHIR Connectathon with multiple tracks focused on different areas of the OpenHIE ecosystem. Several of the tracks ran scenarios that involved FHIR directly:
The OpenHIM track, among other scenarios, build a "mediator" (integration connector) that transforms ADX (an IHE profile for aggregate data reporting) messages to a FHIR MeasureReport. At the end of the track, the mediator was functional but needs polish, and could also serve as a template for mediators to converts other types of messages to/from FHIR.
The OpenLMIS track added the HEARTH (open source NodeJS FHIR Server) server as a microservice within the OpenLMIS microservice architecture, and built a script to transform API endpoints that allows them to have the representation of facilities on a FHIR server.
The Intro to OpenHIE track began as an introduction to FHIR and Docker, and ended up as a massive educational learning and sharing academy.
The ADX on FHIR track looked specifically at representing a particular ADX IHE profile for HIV public health indicators in the FHIR Measure and MeasureReport resources. The track worked with the authors of the IHE profile to express the HIV contents using terminologies developed with Apelon as well as a CQL library and a FHIR Measure to represent the HIV indicators.
All in all it was a very successful event, and the OpenHIE community is enthusiastically looking forward to another one!
This was in Washington, and Josh Mandel attended and reported on this one:
On August 13, the CMS Blue Button 2.0 team convened the first-ever Blue Button 2.0 Developer Conference at the Eisenhower Executive Office Building on the White House grounds. Blue Button 2.0 is a CMS initiative to put health care claims data in the hands of over 50M Medicare beneficiaries, building on FHIR and OAuth 2.0 to allow for seamless sharing with apps of the consumer's choice. CMS Administrator Seema Verma opened up the meeting with a forceful and personal reminder about the importance of consumer access to enable data portability and better individual opportunities to seek high-value care. We heard from members of the CMS team who presented on the motivations, the business processes, and the technology behind the Blue Button 2.0 API, and we heard from app developers who are beginning to use this API in diverse use cases including personal health apps, research participation, and insurance coverage shopping advisors.
There was strong energy and excitement at the conference about the government's ability to push forward on the state of the art, launching a production API service in 2018 for the entire Medicare fee-for-service population, and unlocking any number of downstream use cases. I was delighted for the chance to attend and to participate in two ways. First, I had the opportunity to present an overview of the Sync for Science project, which helps patients share health data with researchers. And second, to participate in Microsoft's joint announcement Amazon, Google, IBM, Oracle, and Salesforce committing to removing barriers for the adoption of technologies for healthcare interoperability.
On that subject... half a dozen of the biggest names in technology – Amazon, Google, Microsoft, Salesforce, IBM, and Oracle – joined together to pledge speedy progress towards true health data interoperability. This was associated with the CMS Developer Conference (see above). See write up.
For further commentary, see Josh's blog:
Quoting from this:
Open standards, open specifications, and open source tools are essential to facilitate frictionless data exchange. This requires a variety of technical strategies and ongoing collaboration for the industry to converge and embrace emerging standards for healthcare data interoperability, such as HL7 FHIR and the Argonaut Project.
We understand that achieving frictionless health data exchange is an ongoing process, and we commit to actively engaging among open source and open standards communities for the development of healthcare standards, and conformity assessment to foster agility to account for the accelerated pace of innovation.
Together, we believe that a robust industry dialogue about healthcare interoperability needs will advance this cause, and hence are pleased to issue this joint statement.
As you'll notice, this announcement has no specific deliverables - I think that's good. There is a call for dialogue - that's a great opportunity for our community - if you have an opinion about how this group could match the accomplishments and impact of the Argonaut community - feel free to join discussion here:
For a while, it was clear that the answer was: we're on the first part the curve, and hype is flying into the stratosphere (well, compared with our initial expectations). But in the last year or so that hasn't really been the case. I routinely survey FHIR implementers informally when I'm speaking about this, and we're starting to get non-trivial numbers of people voting that we're on the 'slope of enlightment'.
If we really are (if), then that's great because it means that the trough has been relatively shallow, which was always our priority: to ensure that the quality of the specification and the community meant that FHIR was ready for real use.
The "(if)" part is because the interesting thing about FHIR is that it has such a broad impact, and different parts of the specification & community are in different places in the cycle. There's some very clear regards in which FHIR is still just hype (where's the FHIR bus in healthcare provider institutions - only a very few have one right now). Our challenge is to grow the community and infrastructure base with solid fundamentals so that each individual area of work has as shallow a trough as possible.
The HL7 board has asked the FHIR Foundation board to review the relationship between HL7 and the FHIR foundation. I'll prepare a separate brief for FHIR Foundation members about this, seeking your opinions.