Re-imagining the EHR
What would the EHR look like if medicine were invented after the computer?
Like most first generation computer systems, today’s EHRs took the way the world worked on paper and computerized it. Sure, there were some things that were added – like automatic drug-drug interactions – but for the most part the metaphor is the same as the paper chart. Results are in sections as they were in the paper chart – labs, vitals, medications, etc. The process of writing orders is separate from the process of writing a note. Lab displays look more or less like they did when they were hand written or printed out of a dedicated lab system.
And, just like paper, the system pretty much behaves the same way regardless of what type of physician you are, or what the patient’s condition and current state is. So a general surgeon caring for a patient in the PACU logs on and is greeted with the same set of screens, data and menu options as a hospitalist rounding at night on 40 patients, as is the cardiologist who receives a new patient consult from said hospitalist.
The result: Physicians have to look around and mentally assemble the data they need to make decisions and place orders. This is pretty much the way things have always been, albeit with clicks instead of page flips.
Today, we have computers far more powerful than was imaginable 20 years ago. In the rest of our lives, computers are learning about our habits, our likes, tracking everywhere we go, largely so they can make suggestions to us about what to buy, what song to play, or whom to vote for.
While this is more than a little bit scary (and also at times super convenient), imagine what we could do with this type of approach in an EHR?
After four years of development, last year we brought our first incarnation of this new approach – we call it PatientKeeper v9 – to a group of physicians in Roseburg, Oregon, and have been working to refine it ever since.
At the technology level, we had to completely re-write and re-architect the presentation layer of our software. We did so to enable three things:
1) Physician-specific dashboards – These allow physicians to see key elements of all their patients at a glance. This is somewhat similar to the notes that you see residents carrying on their patients or what some physicians used to use for paper-based sign out – show me at a glance, for all my patients, what is relevant now and for the care I am providing.
2) Patient dashboards – When you select a patient you get a one-screen view of what is important for this patient across all clinical elements. A cardiologist may want to highlight cardiac enzymes and the text of the most recent echo, whereas a pediatrician rounding on newborns may want to bring forward the weight change by hour, recently administered medications, and a bilirubin risk chart.
3) Plugable and configurable gadgets – As we learned more and more about the diversity of specialty-specific, patient-specific and physician-specific requirements, we quickly realized that a traditional software approach would not work. So we built a concept we call “gadgets”. They are the presentation layer; they are user-configurable and can be written by anyone; and they can be deployed at run time, which means that adding or changing them does not require a software release or even a server restart. Using gadgets, a system’s functionality can be instantly enhanced, consistent with the needs and priorities of any set of users.
While we were at it, we re-architected the reporting framework to be extremely fast with large data sets. The first place our user community will see this is in charge capture; we’ll move it to other reports in the system over time.
A change of this magnitude is not easy, nor is it possible to get it right on the first try. We’re taking a slow, deliberate approach and have kept the “classic” PK look and feel in v9, so users will be able to choose and we don’t have to do an all-or-nothing upgrade to get onto the v9 experience.
We’ve shown the software and the new approach now to hundreds of physicians, and the feedback has been universally positive. Which suggests that our efforts to re-imagine the EHR as a native tool for providers in the digital age holds great promise, and may ultimately make a significant difference in the delivery of patient care.