EHR Promise vs Reality

EHR System Technical Functionality vs. UsabilityDoes the EHR Have Value?

William A. Hyman
Professor Emeritus, Biomedical Engineering
Texas A&M University,
Read other articles by this author

On a recent visit to my dermatologist I found that he had abandoned his tablet entry EHR and was back to pen and paper. He said that he was on board with the two theoretical EHR values of generating good records (maybe easily) and sharing information such as receiving a comprehensive history on new patients. It was just that we weren’t there yet and he decided that as it stood the burden of using his system trumped the value.

During my first patient visit to him a few years ago he was then newly struggling with the EHR which involved a good deal of time looking at and poking around the tablet, with a little muttering, rather than being focused on my skin. He was apologetic about it then and said the system was new. I admit to procrastinating enough between visits so that I missed a presumed middle period where his facility with the system probably improved, but it still wasn’t giving him what he needed.

The things that made his system not have value to him, as opposed to not functioning at all or creating errors, have been widely recognized. One is that too many systems are too hard to use during both data entry and data review. They don’t match work flow and modifying the work flow to the product seems backwards, although it is conceivable that in some cases the EHR driven work flow could be an improvement over the previous non-EHR work flow. Clinical decision support (CDS) might be an example here, if the CDS was valuable and reliable. In this regard CDS is still for the most part a work-in-progress. Lack of real world usability has been attributed in part to programmers that didn’t understand, and didn’t bother to learn, the nuances of face-to-face healthcare delivery.

The second issue is the goal of the relatively seamless and useful exchange of patient data among providers. With the exception of a few highly integrated health systems this hasn’t happened, and many would argue it isn’t on its way to happening anytime soon. The key problem here is the lack of standardization of data structures and software such that the various EHRs cannot readily communicate information to each other. A related question is where and how the communication would take place if they could communicate, i.e. direct or via “the cloud” or with some mutual host. Even exchange itself has not been a clear concept. Does it mean for one EHR to send information to another HER on command, but for each to then continue to be independent thereafter? Or does it mean a constant (and automatic) exchange so that each EHR for the patient is fully populated with the same information (or some of the same information)? Or does it mean that there is one master EHR which gets populated and used by each provider? As I discussed previously, this was once the promise of the H in EHR, as opposed to EMR which was an individual provider’s record.

One might note here that highly integrated systems would be capable of the highly integrated perpetuation of errors. Also, we have recently read that for proprietary reasons some EHR vendors are less than enthusiastic about open exchange and perhaps intentionally make it difficult. One reason might be that if data is easily moved, changing vendors would be facilitated. A second is to in effect force local outside providers to adopt the same system as the dominant facility/hospital. This problem may be in part self correcting as the number of EHR vendors dwindles. Or at least it will leave fewer different systems that have to be integrated.

Sharing my dermatologist visit with another doctor I confer with, he concurred that the theoretical promise of EHRs was good but that in many ways the implementations were bad, or at least have not achieved a good deal of their promise. Perhaps in 5 or 10 years he thought there might be a system or a system of systems for which both direct user interactions and the sharing of data would be productive and useful. This of course means that today’s systems will not have staying power, unless we get stuck with what we have now without the promised value being fulfilled with what will have to be new products. In the meantime those who took the money and have the mandate will have to struggle on.