Five years of FHIRPublication date: Aug 11, 2016
The initial version of FHIR was published exactly 5 years ago. This post covers the time frame leading up to the publication of the initial strawman version.
Five years, on August 18th 2011 to be precise, Grahame Grieve published the initial version of FHIR (known as RFH at the time) on his website. The date of the initial version was August 11th - which is the reason for this post today. Congratulations to all involved for helping to create a success - FHIR has gained a lot of interest over the past few years, and a normative version will be published in the near future.
A bit of FHIR historyIn order to understand the origins of FHIR one has to understand the context of its origins: by 2011 we had a substantial number of HL7 version 3 implementations. However, the uptake had been stalling for years, and software implementers were reaching the conclusion that there were significant issues around its implementation. For example, see Grahame’s blogposts about the concept of drive by interoperability, HL7 v3 has failed, or my blogpost How to lower the hurdle for HL7 v3 implementers. The general sentiment amongst implementers about the future of HL7 was rather cynical at the time, as illustrated by these alternative lyrics (by Jean Duteau) to “the wreck of the Edmund Fitzgerald”.
HL7 as an organization, whilst committed to the further development HL7 version 3, was aware of these problems as well, and initiated the Fresh Look initiative, more or less a brainstorming exercise as to the future direction of its standards. At the time there was a certain level of frustration with the direction of this initiative, which is probably why Grahame did some thinking on his own as to how a future HL7 standard should look like.
Grahame said that “The seed that became FHIR was planted over late night drinks in the Orlando bar [the May 2011 HL7 Meeting] with Peter Hendler and the team from KP (though they didn’t know it at the time)”. That conversation was about the need to back away from HL7 version 3. I have no doubt that Peter also raised the concept of SMIRFs (a concept for the storage of HL7v3 model fragments akin to that of a FHIR resource) which we had been discussing in the RIMBAA WG (now: AID). Combine this with the core characteristics of the Highrise API, a successful non-healthcare oriented REST API, and the “best of” Hl7 version3” – these are the core of FHIR.
Grahame wrote the initial strawman version of FHIR between May and August 2011 – which was a bit of a challenge as this had to be done next to his day job. He asked four persons for their comments (his brother in law (a non-healthcare CIO), Lloyd McKenzie, me, and subsequently Ewout Kramer) both inside and outside HL7, and received strong encouragement to continue. I'm not a software architect, so in July 2011 I introduced Ewout to Grahame; Ewout immediately saw the power behind the FHIR concepts and he has been working with FHIR ever since. Still, at various times he thought of aborting this work, having doubts about whether or not the resulting spec would gain any traction whatsoever within the HL7 organization.
For example, introducing a REST framework for the exchange of healthcare data within HL7 was risky – hData (a REST interface for an EMR) was developed within the HL7 organization in 2009, it was however regarded to be too much of a transport mechanism for it to be acceptable within HL7, which is why the standard ended up at OMG. FHIR didn’t have this particular problem because it does focus on the structural definition of the various resources, and although it has a REST interface it isn’t actually tied to HTTP as a transport protocol.
In order to make FHIR more acceptable to the HL7 community we stressed its HL7v3 foundations (although we knew already that FHIR was only loosely related to HL7 v3 and the RIM). Truth be told – without the HL7 v3 models it would have been much harder to reach solid maturity levels of the various resources, and the design and requirements work done in HL7v3 related to data types and vocabularies could be re-used to a large degree.
The proposal was published in time for the Fall 2011 HL7 Working Group Meeting (WGM) which took place a couple of weeks later. The RFH proposal was well received. The HL7 Methods and Methodology (MnM) workgroup accepted his proposal as a project on Sept.11, 2011 – the formal start date of the HL7 work on the FHIR standard.
The MnM meeting on Sept.11, 2011, which accepted FHIR as a project
The MnM workgroup was aware there was a problem with RFH as a product name (“Resources for Health”; the name of the proposal at the time) and asked the Marketing committee to come up with a new name. We wanted to have “Resources”, “Health(care)” and “Interoperability” in the name, and after playing around with various acronyms we came up with FHIR. We didn't want to use "version 4" as that would be both presumptious as to its success, and it would suggest a level of backwards compatibility - whereas its introduction was a breaking change from HL7 version 3. I was the chair of the marketing committee at the time, and was subsequently rebuked by the HL7 board for coming up with a new name – instead of asking them. In terms of marketing FHIR (as a name) has turned out to be OK I guess ;-)
By January 2012 the FHIR core team was firmly in place (see below) and its development started to gain significant traction in the HL7 comunity.
FHIR Core Team - MnM meeting in January 2012
Looking aheadOver the past 5 years I’ve been involved with FHIR, one way or another, not as much as an author of the standard, but as a marketing person (positioning of FHIR as a product, its name), as a reviewer of the initial proposal, and as an educator (providing training courses and talks about FHIR). The past 5 years have been interesting (in a non chinese way) -- I certainly intent to be involved for a couple of years more..
PermaLink to this page: http://www.ringholm.com/column/FHIR_5_years.htm
Index of columns:
About Ringholm bvRingholm bv is a group of European experts in the field of messaging standards and systems integration in healthcare IT. We provide the industry's most advanced training courses and consulting on healthcare information exchange standards.
See http://www.ringholm.com or call +31 33 7 630 636 for additional information.
Rene is the Tutor-in-chief of Ringholm.