EHRs should automate the business of medicine and eventually the science of medicine, while protecting the art of medicine. Margalit Gur-Arie shares her insights.
By the time the next decade rolls in there will be no paper charts. There will probably still be paper floating around in various capacities, but there will be no one charting on paper. The term “charting” itself may become obsolete, like yonder or popinjay. The term EHR, which is what replaces the paper chart, won’t last either because it doesn’t roll easily off the tongue like say, email instead of letter or missive. EHRs don’t do anything else easily, so chances are EHRs themselves won’t last much longer, relatively speaking. Sooner or later, the national spotlight will shift to something other than health care, and other electronic critters will emerge from the shadows. What will they be? What should they be? Those could be two very different answers.
What Should Be
Have you noticed how people advocating for EHRs use the word quickly in practically every sentence? Mega EHR allows you to quickly document XYZ, and Super EHR can quickly gather all historical data and display it in a summary dashboard which allows you to quickly assess the status of the patient. You can quickly send prescriptions to pharmacies, quickly order a bunch of tests, and quickly print out (the horror!) education materials. Your staff can quickly schedule patients, quickly answer questions, quickly verify eligibility and quickly drop claims. Your patients can quickly get answers, quickly schedule visits and quickly have their concerns resolved.
Of course, most of these things are aspirational at this point, but the language is indicative of the thought process behind EHR design: hurry up and get it over with. Time is money and quickly disposing of each patient is absolutely imperative in a high functioning health system. You could argue that there is nothing wrong with speeding up, or even eliminating, administrative tasks, and you would be right. But is reviewing historical information on a patient and documenting today’s encounter an administrative task? When reviewing histories is compressed into briefly glancing at a bunch of sparklines like they have for the stock market, it may seem like you are engaging in an administrative task. Buried somewhere deep in the dark chart though, there may be a note you made about Mary’s difficulties to navigate the front porch stairs last year.
Why did you make that old note? It’s not required for billing. It’s not demanded by insurers. What made you type that in? Do you usually make notes about irrelevant things that you fully expect to never see again? When you used paper charts, you had to flip through many pages to gather the information you can now quickly glean from your infographic patient dashboard. And while you were paging back and forth through that chart, chances are you would have stumbled upon that little note that seemed important enough for you to scribble down last year. And maybe you would discover other little notes too.
Is reading a book the same as reading the SparkNotes? Oh, you will pass the test either way, and may even get better scores with the SparkNotes version, but is it the same? Is reading this: “Ultimately, he is unable to bear the psychic consequences of his atrocities”, the same as reading this: “Tomorrow, and tomorrow, and tomorrow, /Creeps in this petty pace from day to day, /To the last syllable of recorded time; /And all our yesterdays have lighted fools /The way to dusty death. Out, out, brief candle! /Life’s but a walking shadow, a poor player, /That struts and frets his hour upon the stage, /And then is heard no more. It is a tale /Told by an idiot, full of sound and fury, /Signifying nothing.”? Which version leaves you better equipped to address the story at hand, and the story-teller himself as a whole person? And which version makes you, the reader, feel more like a whole person?
Whatever else the EHR of the future might do, it should quit trying to quantify, summarize, highlight and decontextualize the soul out of the medical record. Disposing of patients quickly should not be our goal. Other than being demeaning and dehumanizing for all involved, seeing patients quickly is the root of all our health care woes. In an environment increasingly hostile to human interaction, EHRs should ferociously fight to create more time for patients to spend with their doctors. EHRs should automate the business of medicine and eventually the science of medicine, while protecting the art of medicine. And by art, I don’t mean compassion and advocacy. I mean the application of professional judgment, without which the science is incomplete and will always be incomplete, regardless of how many genes we can map or how many bots we can implant in ourselves.
Assuming that the business of health care will be marching along the glorious path currently laid out by our betters, the EHR of the future should endeavor to become a silent background processing machine. Natural language processing should be the first and foremost feature to be implemented to perfection. The EHR should parse and extract useful information from doctor-patient interactions to maximize physician reimbursement by maximizing claim values (yep, I just said that), and by scouring all opportunities to obtain incentives and bonuses from the overlords, and automatically applying for all, without user intervention. Not an easy task, but IBM Watson could drop out of medical school and take this on.
The flip side of maximizing revenue is to cut overhead. No practice needs a scheduler. Let patients schedule themselves online or on the phone. Let them check in and let them room themselves (think about those little restaurant gizmos that beep when your table is ready). Let them take their own vitals and answer all those preliminary questions on their own. For the outliers, the frail and the elderly, one medical assistant can cover these tasks for bigger practices. It doesn’t sound much like an EHR because it isn’t.
What else costs money, but has practically nothing to do with medicine? All the school notes, the back to work notes, the disability forms, trade forms and all other forms, authorizations, pre-authorizations, eligibility checks, statements and everything in between, can be delegated to computerized self-service. Get the medical records online and let patients have at it. Make it user friendly like all other trivial consumer facing apps that have absolutely no bearing on enterprise technologies. Make them colorful and fun. Show pictures, animations and ads. Put them on the iPhone. Monetize the heck out of everything, and remove this purely administrative burden from the practice.
And then comes a moment when restraint needs to be religiously exercised. Stay the heck away from the exam room. Let people say whatever they want to say. Let doctors ask whatever they want to ask. Forget about boxes, no matter how useful they look. Forget about structure too, because Watson will be taking care of that. Don’t make documentation easy, because taking notes is not just about documentation. It is mostly about the background thought process that makes one decide what to note and what to discard. Remember, the way you would use different styles of handwriting and text of different sizes or boldness, and how you added critical notes in the margins, or big pointy arrows? Actively taking notes helps you synthesize information, internalize, memorize and understand the narrative (programmers, think back to your college days).
If EHRs want to be helpful, let them be secretaries. Arrange the notes in a way conducive to better information retrieval. Don’t summarize and don’t impose your (or your machine’s) notions of what is or is not important. Collate and bind everything into a patient book. Remember that this is a reference book and the user can read at a 30th grade level. It is not a cookbook and it is certainly not a picture book. Add a table of contents. Use the computer to make it dynamic. Make it easy to flip pages in an electronic context. Make the fonts nice and large, and do learn from beautifully maintained old paper charts. Go out and look at some before they are extinct. You can’t improve that which you don’t know.
What Will Be
Of course none of the above is going to materialize in a meaningful way, except maybe the infantilized consumer facing EHR. Health care is going retail and health care will care about its customers precisely as much as Walmart and Amazon do. And health care will care for its employees as much as Walmart and Amazon do. When providers are all transitioned to having their employer paid for value, EHRs will become very easy to use. A few big buttons, at the bottom of pages prefilled with stereotypes and sprinkled with colored sparklines, should quickly do the job (i.e. Order, Refer, or Deny). There will be nothing for you to document and nothing for the patient to convey because the system already knows everything there is to know. Eventually, it will dispense with the buttons altogether, and your job will be to quickly explain to your customer why the system did what it did. And you will never have to use a computer ever again. Mission quickly accomplished.
Margalit writes regularly about intersection of healthcare & technology on her site: On Health Care Technology