Playing
with FHIR

November 20, 2014 -- One of the things I learned at AMIA's 2014 conference is that a thing named FHIR will change the way we do computing in healthcare.

Much of the communication that is done between the numerous and disparate software systems that run our hospitals, labs, clinics, payers, etc., etc. is via an antique and under-powered standard named Health Level 7 (HL7). (It was invented by Dr. Don Simborg, one of my UCSF grad school profs.) Standardized communication protocols were a breakthrough in their day but that day was decades ago. FHIR is HL7's effort to remain relevant. It just may work.

Today's massive, monolithic, and preposterously expensive EMRs and EHRs will, they claim, with time become relegated to commodity products serving as backend servers of medical and financial data, living in server farms in basements or even in the cloud. The front ends (the part users see) will be small, cheap, easily-deployed (and replaced) apps. Everybody is talking about apps.

Modularity in healthcare software has been predicted for many years yet no one has been able to make it work. The contrarian view -- and it is thus far proven by practice -- is that the data and the workflows are just too complex, and that's why our only choice is the fabulously complex and expensive software monstrosities of today.

Who is right? Only time will tell.

Go Don! Go FHIR!

Table of Contents