Convert your FHIR JSON -> XML and back here. The CDA Book is sometimes listed for Kindle here and it is also SHIPPING from Amazon! See here for Errata.

Tuesday, April 22, 2014

Why don't we have interoperability?

I don't know how many times I've been asked that question, or attempted to answer it.  The short answer is fairly simple.  We have the technology to support interoperability today in our hands.  Not everyone agrees on what it is, but it does exist.  Even those of us who don't agree on which technology to use can still use it even if we don't like it.  The real problem is not the technology, but rather in its deployment.  It takes a lot longer to get the technology into the healthcare provider's offices than anyone really wants it to.

You can argue about whose fault that is, and no matter which argument you choose, there's probably some truth in it.  It takes time.  We will get there.  This is not a technology problem.

One thing that I've observed happen more than a dozen times since I got started in this field is that you can solve the technology problems in 1/5 the time it takes to develop and agree to the information sharing policies.  I've seen that happen over an over again.  It's not about bits and bytes.  That's the easy part.

The time is really about the words on paper that wind up having to be approved by the lawyers and signed by the big bosses.


  1. IMHO, the single biggest impediment to interoperability in Health Information Exchange is the poor quality of information entered into EHR systems - particularly in primary care. There are obviously many reasons for this, but technology is one of them - we have yet to produce the means whereby clinicians can enter high-quality, coded, clinical data within consultation time constraints. Furthermore, in their current iterations, the much-touted mobile technologies are slower than keyboard and mouse. Is the solution really at hand and, if so, why isn't anyone implementing it when a pot of gold lies waiting?

  2. I believe it's really simple. ONC and CMS do not have the stomach to really define or share a forward looking roadmap. Paving the road as we are driving on the highway is apt for errors.

    Have we really tested C-CDA? Nope, let's take it out for a national test drive. This looks like a national wreck to me. How many organizations have exchanged C-CDA documents? A really small number. Policy us really lacking here.

    We have too many choices to perform the basic interacts.

    Push - Direct, FHIR, Exchange (IHE XDR)
    Query - Response - FHIR, Exchange(IHE XCPD, XDS, XCA), Direct (push a query, push a response)
    Publish-Subscribe - FHIR (future), Exchange (HIEM), Direct (push a subscribe, push a publish)

    Too many ways to perform the same actions. This send conservatively and consume liberally is for the birds. It just means as a receiver if I want to maximize the end points that I want to interoperate with I have to support all options.

    Come on ONC and CMS, get a backbone and lock down standards. Too many choices and options leads to errors that we cannot afford as a nation. We really have one chance to get this right.