All EHRs Aren't Created Equal
As some of you know, I teach about HIT and try to do it from a real world perspective so this is something I've though about quite a bit. (I'm also a "reformed" vendor.)
Frankly, I don't see the added value in surveys like this one. The result should be obvious pre-survey, given where we are. There are literally hundreds of EHRs and, from what I've been able to see (I've hardly seen them all), most aren't very good with respect to some key usabiliity issues.
Reader beware: Some of what follows may be a bit redundant with what David and others have already said:
1) They don't integrate well with workflow. Rather than becoming a seamless component of seeing patients and documenting their care and one that actually improves productivity they are essentially an off-to-the side-extra activity that physicians resent for taking up too much of their valuable time. This is partially do to poor EHR design and partially due to poor implementation. In my years as an HIT vendor it was rare for new clients to take the time to look at their manual processes and think through specifically how they wanted automation to improve them. I would argue that only once that's been done is a healthcare organization (or any organization for that matter) ready to look at systems of any kind. It is unfortunately at least equally rare that systems are designed around process. Far too many EHRs are designed to mimic the paper world (so that sales people can tout how easy they are to use and how little training is involved in learning to use them) rather than provide a tool to improve and optimize office processes. I won't name names but there ARE exceptions to this out there and physicians should be looking for them.
2) The vendors haven't thought creatively about data collection. In the end most physicians are complaining that its taking too much time and/or its too hard to enter data. Of course, in many cases, this is a false comparison between poor, inadequate documentation and forced, more complete documentation but that doesn't change the fact that there are creative EHR solutions out there with respect to data entry (again, I won't name names) and that, from what I can see, physicians really do like those systems. Caveat emptor.
3) This one doesn't get nearly enough attention. Once EHRs have the data they typically don't do much creatively to present it back to the physician in ways that both save time and help provide better patient care. Here I really haven't seen an exceptional EHR and it is something serious vendors should be spending more time on. The inception of EHR app platforms may offer real help to the extent that their APIs allow indendent developers explore creative approaches to visualization of the underlying clinical data.
BTW, many vendors will point to the high degree of user configurability as the "solution" to #2 and maybe even #3. Let's get real. Very few end users have the talent or the time to do system configuration well. The answer is better thought out system design and LESS user configurability -- just ask Apple!
Finally, wouldn't it nice if physicians -- who now are and forever will be essentially required to use EHRs -- were actually trained in health informatics? I guess its easier for me to be critical since Georgia Tech doesn't have a medical school but I think the time is long past that we keep producing physicians, nurses and other health professionals whose only experience in health IT is learning to use whatever system the teaching hospital they trained in happened to have. BTW, my hidden agenda here is smarter, more informed and more critical potential customers. That's the one sure way to get the vendors to improve!