Re: [Hardhats-members] EMRs gaining popularity among physicians

[Gregory Woodhouse]

On May 19, 2006, at 4:29 PM, Jim Self wrote: Gregory wrote: --- Jim Self <[EMAIL PROTECTED]> wrote: An EMR certainly could take time away from physicians if it is poorly implemented ... The quality of the user interface obviously plays a significant part in that, but even more important is the overall synergy of the system, the possibility that redundant data entry could be essentially eliminated, that all relevant data in its most timely and legible form can be instantly available anywhere it is needed at any time, that it is no longer necessary to interrupt people in the labs for updates on critical results. Isn't this all part of good interface design? No, it's the other way around. I guess I don't follow you here...Or maybe I do. I interpreted your comments to mean that the data should be instantly available to the user (i.e., not requiring that the user break his or her flow to go through a complex process to gain access to that new data). But I suppose there's another way of looking at it, too: The system should have the technical ability to retrieve data "instantly" as soon as it is available. If that's what you had in  mind, then I missed your point. User interfaces are essential parts of an EMR or HIS system and since everything that a user sees comes through a user interface, some users might get confused and think that the system is the user interface, just like some users think that the CRT or LCD disply is the computer, but developers (and I would think most users who have been around VistA very long) generally know better. Now this surprises me a bit, too. If you work in roll and scroll mode, then the changes you make are (logically, at least) committed to storage as soon as you make them. This is different from editing a form, where you can make changes that would not be reflected in the system, at least until you submit the form. So, it seems to me that VistA actually tends to foster the concept that what you see on the screen is what's in the system more than other products would. Alan Cooper argues strongly against what he calls the "two file" model, where the user modifies a local working copy and then saves those changes. I think his arguments have some merit, but I have a hard time with so radical a position and, to me, it's a bit a conundrum. Cooper uses the example of a book taken down from a shelf, and rightly argues that an old untouched copy isn't still there until such a time as the user returns the book (together with his or her notes) to the shelf. In our case, a paper chart might be a more apt example than a book, but the principle is the same: Once you pick up (open) a chart and start making notes (edits) those notes have indelibly been added to the one and only copy, even if they remain unsigned. Having worked with text editors since who knows when, I find this model uncomfortable (I like my :q!) but it is closer to the way documents really work. But does remaining faithful to the basic metaphor necessarily improve the interface? I'm not convinced that it does. What I am talking about are the deeper aspects of providing timely accurate reliable sustained communications between people working on different aspects of providing medical care. This has more to do with database design and functionality and is largely independent of user interface, which is why some people can still cling with some justification to a user interface (or more accurately, a lack of one) that seemed outmoded more than 20 years ago. So, perhaps we are saying essentially the same thing here. I'm still struggling to come to terms with Cooper's ideas here. He seems to advocate a multi-level undo process with checkpoints, but that raises all kinds of questions.  I appreciate your comments.  They are certainly thought provoking. Gregory Woodhouse [EMAIL PROTECTED] Metaphors be with you.