Child pages
  • Foundations of Information
Skip to end of metadata
Go to start of metadata

Sometimes the foundations of openEHR can be difficult to understand especially in our "now" world. 

There is a lot of specific information on openEHR on this website but without the proper context it can make the specifications seem overly complex.  This isn't the reality and in fact these specifications are actually quite simple in the scope of health information.  There are many textbooks and papers that can help with the study of information theory and an understanding of openEHR foundations.  I believe that it may help newcomers to spend sometime taking advantage of YouTube and learning about information in general by viewing Gary Marchionini's lectures from UNC at Chapel Hill such as this one These are not healthcare specific but I believe that they can help those new to health information management and systems design understand more about why openEHR specifications are designed the way they are.

Certainly Claude Shannon is considered the father of information theory and this page is a good overview of the Shannon-Weaver model.  This quote is particularly compelling: "The Shannon-Weaver Model, in common with many others separates the message from other components of the process of communication. In reality, though, you can only reasonably examine the message within the context of all the other interlinked elements."  IMHO, the last sentence is the most important where it states that the context and the interlinked elements must be there to achieve communications.  Note that this is speaking about the "message".  However, without having the complete semantic context avaible (which is what the openEHR information model is about) then you cannot construct a complete and proper message.

  • No labels


  1. Anonymous

    Surely the semantic context includes the full context of the users of the systems at either end. In fact, surely the users are the most important components of the whole system. When seen in this light OpenEHR is nowhere near enough to provide a complete semantic context.

    One may try and argue that the electronic message itself is limited to this context, but I would disagree. A properly designed system supporting communication between people should not be limited by the ability of the system itself to interpret the semantic content of the messages.

    In other words, systems that support health care should be primarily aimed at supporting people in doing their jobs and not limited to the ability of the system to interpret the information content that may be thus generated.

    This does not mean that OpenEHR has no value, but rather that it should be considered as a part of a broader approach to providing information services in healthcare.

  2. First and foremost openEHR applications are considered to be patient-centric.  But the specifications do include components to capture the who,what, when, how as well as auditing of the movement of information components between systems.  So other than video recording of the user at the keyboard I fail to see how much more context you can capture.  As I think about it; you could do this video capture as well if you had a need. 

     I refer you to the EHR Information Model and supporting components.

    If you still find something missing then by all means let us know so we can put it into the pipeline for inclusion. 

  3. Anonymous

    I think you are missing the point. It is not that there are technical imperfections leading to some incompleteness in the modelling environment. It is that there are real world limitations that need to be taken into account in the overall design of systems.

    The point is that systems used to support people in providing health care must exist in the real world. The full context in which they are being used will ipso facto be imperfectly represented within the system. It is impossible to capture the entire semantic context in a model. Given this, a thoughtful software developer will think carefully about where to draw the line in modelling the information content handled by the system. Developers need to accept that systems may handle real and useful information that will not, for practical reasons, be "understood" by the system. I think openEHR is regularly overstepping the line and over representing the value of semantic interoperability in developing real systems to support real healthcare activities.

    I repeat, this is not to say that openEHR has no value, just that we should see it is part of the picture, not an end in itself.

  4. openEHR is certainly part of a broader context of solutions. However, if we want to get beyond using health ICT as a network of fax machines that simply put documents on screens to be seen by health professionals, and actually do some computing with the data - essential for decision support, preventative personalised medicine, and medical / population research - then we certainly have to make semantic intoperability a headline issue. This is where 'real healthcare' is going, and where it has to, if we are to have any hope of making the best use of limited health system budgets in a world with huge chronic disease problems.