Talk:Electronic health record

That may be a MeSH definition, but it's wrong.

Call it an EHR, EMR, PHR, or whatever, there is a massive difference between the machine-processable data that constitutes the record, and the human and automated medical equipment interfaces to it. In my practice of medical informatics, I often illuminate people who ask "how do I justify an EMR?" and I say "you don't."

If I dumped out an actual electronic medical record, it might be a collection of XML representation of HL7, which, on its own, is as comprehensible to most physicians as, alas, handwritten prescriptions. Now, if I present a history-taking tool, a set of workflow screens and schedules, a clinical decision support tool operating on the EMR and its knowledge base, I have something I can justify. When portions of the EMR need to be sent, in machine-readable form, to a third-party payor's computers, there is a justification.

We need to separate the presentation and use of the information in the record from the mechanisms of the record itself. Those mechanisms are nontrivial and important, since a major part of health cost is information transfer. Incompatible EHR formats require expensive manual intervention.

Howard C. Berkowitz 15:27, 9 October 2008 (CDT)

Security & privacy

Ross Anderson at Cambridge has done extensive work [1] on security & privacy of clinical information systems, some of it as a consultant to the British Medical Association. Some of this needs discussion here. Sandy Harris 13:43, 3 October 2010 (UTC)