We don’t need no Semantic Interoperability

What we need in healthcare is “Semantic Interoperability”. That’ll solve all our problems. That’s what they all say (here’s the most recent case).

Really? I beg to differ.

We’ve had semantic interoperability in healthcare since we started having healthcare. Since the beginning of healthcare (by whatever definition you can use), healthcare practitioners have exchanged data using spoken and written words, and the semantic meaning has been clear (well, as clear as it can be given that human knowledge is limited).

So whatever it is that we are doing, it’s not introducing semantic interoperability. In fact, what we are doing is introducing a new player into the mix: computers. And not, in actual fact, computers, but the notion that there is something to be gained by processing healthcare information by persons or devices who don’t properly understand it. So, in fact, what we are actually doing is seeking for unsemantic interoperability.

Unsemantic Interoperability

Yes, that’s right: what we’re trying to achieve is un-semantic interoperability. That’s what people actually mean when they say Semantic Interoperability.

Of course, from a computer perspective, we are trying to get “more semantic”, and so that’s the focus. But from a people perspective, we are actually trying to get less semantic – to constrain the semantics. And that’s why clinicians (who are truly semantically enabled) have problems with this.

And since interoperability is all about the people, it follows that it is important to remember that all this stuff we are trying to achieve is about unsemantic interoperability.

Aside: I know another case like this – it’s the word “Clinical”. From the point of view the general population, “Clinical” means cold, unemotional, not people focused. From the point of view of laboratory medicine, “Clinical” means people focused, messy, emotional.

I don’t know that looking at the interoperability this way makes much difference to what happens technically. But I think it makes a difference to how we assess our overall goals and progress, and how we socialize what we do in interoperability.

So the word “semantics” is a bit lacking in precise semantics of it’s own. No wonder we can never get anywhere.


  1. Thomas Beale says:

    Well I did say “what modern healthcare wants in terms of IT is the following” when I said the industry is looking for Semantic Interoperability. I don’t think clinical people want less semantics. They just want IT systems to do some of their routine work, and in order to be able to trust the IT to do that, it is recognised that it has to have some grasp on the semantics of shared information (much better than today). Semantic interoperability is about solving that problem.

  2. Grahame says:

    I wasn’t really picking on you. I agree that clinical people don’t want less semantics, and they do want IT system to do some of their routine work. But you can’t have your cake and eat it too.

  3. Rene Spronk says:

    Healthcare, being a ‘people issue’ concerns a lot of messy information. Some of it is easily captured in a data model (what room is this patient in? what procedure has been done?) – other stuff is not.

    What if we were to say that somewhen in the not so distant future we’d have near zero cost cloud computing, and we where to apply ‘Google style’ DSS interpretation/correlation/aggregation of (mostly unstructured text) clinical data?

    What is it why most people don’t like such a scenario, and go for an attempt to exhaustively structure and annotate their current clinical data? Do we really only want to trust a human to look at 1 GByte of data and create knowledge based on tons of data?

    If your answer is no, then why should we go for full semantic interoperability?

  4. Grahame says:

    http://www.meaningfulhitnews.com/2011/05/01/park-defends-plain-text-format-of-blue-button/ I think I’m going to be commenting on that. But there are many aspects to the answer – it’s more a book than a blog post

  5. Thomas Beale says:

    The question is not about trying for ‘full semantic interoperability’, indeed, a key part of the issue is to work out how much is needed and to work on that.

Leave a Reply

Your email address will not be published. Required fields are marked *

question razz sad evil exclaim smile redface biggrin surprised eek confused cool lol mad twisted rolleyes wink idea arrow neutral cry mrgreen


%d bloggers like this: