How to Build a Better EHR

Did McCoy's tricorder have POE?
Did McCoy’s tricorder have POE?

A lot has been written about how awful Electronic Health Record (EHR) systems are. They are overwrought, overengineered, dreadfully dull Baroque systems with awkward user interfaces that look like they were designed in the early 1990s. They make it too easy to cut and paste data to meet billing level requirements, documenting patient care that never happened and creating multipage mega-notes, full of words signifying exactly nothing. They have multitudes of unnecessary, meaningless, oops, meaningful use buttons that must be clicked because the government says so. They have data formats that are incompatible with other EHR systems. Doctors fumble around trying to enter orders using electronic physician order entry (POE). There is terrible user support. And so on. At the end of the day there is decreased productivity, doctors are unhappy, and patients are unhappy. Big Brother in the form of the hospital and the state have more Big Data to look at, but certainly there doesn’t seem to be much benefit to patient care. The major benefit is to the companies that make these proprietary closed-source EHR systems. They get obscenely rich.

But surely there can be benefits to EHR systems? What about the ease of access to the patient’s chart? No more waiting for the chart to come up from Medical Records. In fact, no more Medical Records department at all! Aren’t we saving health care dollars by cutting out those jobs, as well as medical transcriptionist jobs and unit secretary jobs. Surely paper charts were worse?

Doctors should not turn away from information technology. After all, we use all sorts of sophisticated computer technology every day, from the internals of the ultrasound machine, to the software running an MRI scanner, to the recording system used in electrophysiology procedures. There is a role for technology in our record keeping as well.

The problems with current EHR systems are manifold. They are hack jobs, with nightmarish interfaces that obviously were never user tested. They are overly ambitious, trying to do all things and thus doing nothing well. They are ridiculous. I mean, having doctors enter orders directly into a computer — seriously? EHR companies have no incentive to improve their user interfaces, because government mandates requre that they be used no matter how awful they are. Those who don’t adopt these systems are penalized by loss of Medicare dollars.

I think it is an interesting thought experiment to consider how EHR systems would have been designed if they had been allowed to evolve naturally, without the frenzied poorly thought out incentives that exist in the real world. Imagine a world where physicians, the primary users of these systems, drove development and adoption of these systems. Imagine that there were no mandates or penalties from the government to adopt these systems. If a system was developed that improved physician workflow, it would be adopted. Nothing that slowed productivity, as the current EHR systems do, would ever be bought by a practice if the physicians made the call. Imagine EHR companies visiting practices, analyzing work flows, seeing areas that could be improved by computers, and recognizing areas that wouldn’t, at least with current technology. Imagine EHR companines testing their user interfaces using doctors from a spectrum of computer experience, as major software companies like Apple and Google do. Imagine them competing with each other not on how many modules they can provide, but on how few keystrokes or mouse clicks their system used to do the same work as another system. Imagine no government mandates for meaningful use, no dummy buttons that say “Click Me” but otherwise do nothing.

Think about how you would design a system. Certainly it is useful to have old records available online and we would want to keep that. The problem is how to get them there. Having physicians enter data is probably the least efficient way. Dictation and handwriting are still the fastest data entry methods. If Dragon is good enough (I’m not convinced it is) use it, or keep your transcriptionists around. They are very nice people who need jobs anyway. If handwriting recognition is good enough (I don’t think it is yet) use that, otherwise just store the written notes as pictures and be satisfied. In the ideal world, rather than force physicians to become typists and data entry specialists, we would wait until computer artifical intellegence was developed enough to allow the physicians to continue to do things the old way, with the computer processing the doctors’ notes transparently. If the technology isn’t there yet, develop it, but don’t push it on us prematurely.

Medical records primarily should exist to document important information about patients. It should not be primarily a means to ensure maximum billing of patients. If we eliminate that aspect, EHRs become much simpler. I would envision a small tablet that the MD carries everywhere with him or her. Keep the old workflow. Pull up patient records on the tablet. Write notes on the tablet in handwriting, or dictate into it. The tablet transcribes the input and files it appropriately. Need to give patient orders? Select from some templates or write them in. If the software is not good enough to transcribe written orders on a tablet, hire some unit secretaries to do this like they used to. Let them learn the intricacies of computerized order entry, and let the doctor deal with the intricacies of making diagnoses, doing procedures, and looking patients in the eye and grasping their hands when they are ailing — things that doctors do best! Minimize the interactions with the computer and maximize the interactions with the patients!

A good EHR system can simplify drug reconciliation, pull in drug data from patient pharmacies, and automatically identify patients who are being “overprescribed” pain meds. The system can look up recent relevant medical articles, can show appropriate medical guidelines, and can provide sophisticated medical calculators. There are so many good things computers can do for medicine. They’ve gotten an awfully bad rap from the current iteration of EHR systems. I think the technology exists or can exist to do all these good things, but there is no incentive if we remain satisified with the status quo. The current systems don’t do any of these things. They just get in the way.

If we lived in an ideal world it would be time to chuck the lot and start over.

By mannd

I am a retired cardiac electrophysiologist who has worked both in private practice in Louisville, Kentucky and as a Professor of Medicine at the University of Colorado in Denver. I am interested not only in medicine, but also in computer programming, music, science fiction, fantasy, 30s pulp literature, and a whole lot more.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.