Innovation Pulse: Hardest part for FHIR lies ahead

'Even though the health IT crowd is crowing about joining the modern world, it's still a step behind'
By Tom Sullivan
10:44 AM
FHIR population health

Around the turn of this century a phrase manifested in certain IT circles: We're all going to agree on specification and compete on implementation.

And that sounded so good when Sun Microsystems multi-billionaire CEO Scott McNealy added it to his vibrant repertoire – other McNealy favorites of mine, to paraphrase, included "Budweiser, please" and "I just made my semiannual trip to Supercuts" – but the reality was much more complicated.

Those specifications included the likes of Java, XML, SOAP and the WS-x standards forming the Web services stack and they were engineered to enable integration, interfaces and, of course, interoperability.

Today, observing the emergence of FHIR, aka Health Level Seven International's Fast Healthcare Interoperability Resources, has taken me right back to that critical era of Web standards establishment.

"When vendors start adopting and implementing FHIR," Dave Shaver told me, "that's when things will start to get interesting." 

As president of integration vendor Corepoint Health and an HL7 Fellow who co-chairs the FHIR governance board, Shaver is correct, of course.

Predecessor: XML, et al.
At the risk of quelling the FHIR party it's time for a reality check on the fledgling interoperability standard, if only because what's happening now will, in hindsight, likely be the fun part.

One need look no further than those now-ubiquitous XML and associated standards, however, to understand some of the growing pains on the road ahead.  

At one point in time, the old guard of enterprise IT vendors -- IBM, Microsoft, Sun, Oracle to name the big guns -- came together with a number of specialty shops via standards bodies to craft those web services specs.

[Read more Innovation Pulse columns]

But then each vendor would take that work back to its headquarters and essentially bend the specification as far as they could toward their own will. Did the initial version of Microsoft's integration server seamlessly access information in an Oracle database or SAP's applications? Not so perfectly. What's more, Forrester, Gartner or IDC analysts were saying at the time that these companies, Microsoft specifically, weren't necessarily all that interested in interoperating with the so-called coopetition.

"I would suspect this will replay itself in healthcare," said Steven Waldren, MD, director of the American Academy of Family Physicians' Center for Health IT. "Layer on the confounding factors such as the multiple purpose of health records, ambiguity and uncertainty in the data, complexity of health information, misaligned business drivers, and it is going to be a rocky 2017-2020."

Technological obstacles
Despite the momentum and the promise, there are those who contend that FHIR has its share of challenges.

"Don't get me wrong, FHIR is a step in the right direction," workflow guru Charles Webster, MD, explained. "But the API is 15 years old. The technology FHIR is based on is about half that. So, even though the health IT crowd is crowing about joining the modern world, it's still a step behind."

FHIR is ideal for accessing data from a specialized software tool, such as a smartphone app or a module embedded in an EHR screen, Webster added, but that is not the same as true interoperability.

"Just as there is a big difference between a meaningful use certified EHR and a usable EHR," Webster said, "there may be a big difference between a FHIR enabled EHR and a usable EHR API."

Shaver's Corollary to Metcalfe's Law
It's not my intention to douse FHIR (forgive that one and only pun, please) in writing this column. Not at all.

HL7 Fellow Shaver, in fact, made another great point in our meeting. Intel co-founder Gordon Moore's Law, observed 50 years ago this spring, the short version of which states that processor speeds will double every two years while also coming down in price, may be the golden rule among all things technology but there is another not-as-widely-known theory of great import. That being Metcalfe's Law: the value of a network increases exponentially in direct relation to the number of attached nodes.

Bob Metcalfe, co-founder of networking equipment maker 3Com and a co-inventor of Ethernet, originally crafted his law with telecommunications in mind but it has since taken on a life of its own and others have deftly applied it to newer realms such as social networks. Shaver, for his part, claims a Corollary to Metcalfe's Law: The value of a heath data standard to a new adopter is equivalent to the number of previous adopters, squared.

Thus, not until EHR, integration, and application vendors start adopting FHIR en masse will a picture crystalize of just how valuable the standard really is -- or, ultimately, is not.

Future past
So how did all the vendors' differences around XML, SOAP and the Web services stack get ironed out?  

The companies kept hammering away, refining the specifications, bringing new and better products to market and, perhaps most significant, one by one Oracle acquired just about everybody, save IBM and Microsoft. 

What with IDC forecasting a wave of "3rd platform" EHR systems and applications that live in the cloud, and Black Book predicting massive EHR market consolidation for the last few years, we might just be watching history poised to repeat itself.

Previous installments of Innovation Pulse: 

Portals: The secret of population health success

The bold art of build your own IT

Hospitals opt to craft their own apps, rather than buy

Can platforms foster an ecosystem of IT advancement?