I've got some PowerPoint presentations and Word documents pertaining to our favorite topic — computerizing records and healthcare processes — that I created in the early 1990s for lectures, Grand Round, and papers. I would desperately like to be able to get at some of that content today but I can't. Microsoft has regularly made "enhancements" to their file formats and the newer versions of MS Office (say > 2003) no longer contain the code necessary to render the older formats. Not only can't Office open these files, nothing else can either. Short of locating my Office 97 disk and reinstalling the apps (a procedure that should be performed in a virtual machine to avoid creating unexpected and damaging side-effects) I'm out of luck.

This situation hints at two basic principles that are important to medical information theory. First and foremost: content and process are conceptually distinct, and should be so in practice. The content is mine (or yours) and it should be stored in a way that it can be opened and rendered without requiring one vendor's proprietary software. I'm not insisting that it be easy, just that it be doable and straightforward. Second, even if the content is in a non-proprietary format it must not be locked inside a proprietary storage system the makes external access difficult or impossible. There should be complete independence of the way content is stored and the processes that might operate on it. The task of resurrecting an old file is but one example of a process that I might want to carry out using the file as the source of input data. The variety of useful processes that one might invent to make use of the content locked in those file is almost limitless.

In the past, efforts were made to define the various terms used to describe computerization of healthcare activities. A distinction was drawn between a CPR and a CPRS — a CPR being a Computerized Patient Record and a CPRS being a Computerized Patient Record System. The content of the charts was to be stored in the CPR and the CPRS would carry out various processes using the chart information as part of their input. You will undoubtedly recognize one glaring defect in this naming scheme — the confusion with Cardio-Pulmonary Resuscitation. In any case this critical distinction seems to have gone down Titanically along with the terminology.

When the term EHR emerged, the memory of the earlier, important distinction seems to have been forgotten. For most people, EHR equals CPRS minus the suggestion (requirement) that the content be thought about and managed separately from the processes. Today's EHRs focus on processes — mostly organizational processes. Clinical processes come in a distant second. The "details" of the medical record, such as how it is created, what it contains, and how it might be used to drive clinical processes, is little more than an afterthought. You can think of the chart as the rumble seat of an EHR. True, it provides seating capacity but it leaves its users out in the cold. Later when your vendor changes their software, your old records may not simply be out in the cold, they may be locked in a deep freeze.