Why health IT must consider physicians in the design stage.
Physician burnout is a growing problem — and health tech must fight it.
As a practicing physician and co-founder of a health-tech company that supplies data to doctors at the point of care, I find myself having to constantly weigh the value of that data in relation to its impact on the physician-patient interaction. I understand firsthand the real value in an engaged physician who keeps the patient the main focus during an appointment. I also understand the power of using data to make an informed decision or treatment recommendation in real time. We are struggling as an industry to balance this dynamic, and we seem to be moving away from this personal interaction, one small step at a time.
One of the most vexing tasks in a clinician’s role is taking an unstructured patient presentation and giving it structure. Artificial intelligence (AI) has not even come close to solving this problem in medicine.
>> READ: It’s Time to Help Our Healthcare Providers
For example, I recently saw a patient who has Parkinson’s, weight loss, a history of kidney stones, pain on the left side and constipation. Her chief complaint was, “I just don’t feel good, doc.” That’s a very unstructured scenario with disparate symptoms and issues to work through. As the first doctor to see her, I had to determine that she’s constipated because of a new medicine, but the significant weight loss is due to the fact that she’s depressed.
But how do you structure the presentation in getting to that analysis? What database or decision support would a computer program access to diagnose five different symptoms that can be interrelated in any number of ways? A series of electronic medical records (EMR) alerts certainly can’t do it. Not even the much-hyped promise of AI can do that at present, and it will likely fall short of that goal even in a future promised by popular healthcare visionaries. The art of taking a proper patient history and then structuring it into a narrative cannot be accomplished by AI-based technologies.
I believe that the art of medicine today, and for the foreseeable future, still lies in a clinician’s ability to take all of that unstructured data and sift through the noise to bring it down to something clean and meaningful that benefits the patient. But that requires focus and time, both of which are eroding thanks to current trends in healthcare. In reality, I see as many as 18 patients in a clinic block. We need to find a better way to document the facts and findings of the physician-patient interaction without disrupting it.
I agree with many of the physicians I meet: The EMR has taken us away from what we natively want to do, which is ask patients questions in an interview-style format and then work off those questions and answers to generate other investigative questions. It’s this kind of simple bidirectional interaction that leads physicians to the next thought of what could be going on and how to treat it. It’s like solving a crime. The symptom a patient presents with is the crime, the patient is the scene of the crime and the physician is the criminal investigator. The history, physical exam and diagnostic testing are tools that may lead to important clues. Checking boxes in the EMR, figuring out ICD-10 codes, spending time on buzzword initiatives are really just “noise.”
What’s happening in the workflow now is that you’re opening the EMR screen, and an alert pops up in your face before you’re even allowed to start the thought process as to why the patient is there by listening to their chief complaint. You constantly must stop some alert or close some box on the EMR and then start to ask questions. Instead of working off of listening to the patient’s answer, you’re typing or dictating, seeing where your cursor is located or considering whether you’re on the right screen.
All of this leads to a lower-quality interaction with your patient and a derailed physician-patient interaction replaced with a physician-machine interaction model that has far less value to both the physician and the patient.
>> READ: Leaders Weigh In on Solutions for Physician Burnout
The EMR currently requires physicians to work around how an EMR is built rather than the EMR being built around a physician’s thinking and process. The best clinicians I met in training homed in on amazing diagnoses by never interacting with the EMR. They heard a presentation from the resident about the patient, reviewed lab results the resident had gathered, and then asked a few targeted questions of the patient along with a physical exam. Then — bam! — the doctor would say something like, “This could be amyloidosis. Let’s biopsy the most affected organ, in this case the liver.” The EMR hasn’t come close to replacing the way a good resident could present information to an attending.
Many physicians went into the profession because of the prospect of real human interactions, which requires engendering interpersonal skills that are now being taken away by having to instead interact with the EMR. As a result, we’re seeing a high rate of burnout among my fellow physicians due in large part to all the disruptive, sometimes even worthless, data and alerts taking up precious real estate on the EMR screen. It demotivates clinicians because the process of documenting the interaction with a patient supersedes, and in many cases overshadows, the interaction itself.
Physicians now feel like data entry specialists, constantly filling in boxes, checking boxes and stopping best practice alerts. Research shows that the EMR adds an hour and half to two hours a day to the physician’s workload. Add to that the seemingly endless list of health IT companies integrating with the system to provide insights, not realizing where a physician is in the thought process of that patient’s treatment. I see this as one of the biggest challenges in the health IT today: the deluge of additional data added to the EMR and how to present it to clinicians without annoying them or, more important, disrupting the act of practicing medicine in a clinical setting.
We are clearly not going to reverse course with regard to the use of technology or electronic health record systems in healthcare settings. Health systems have invested billions of dollars in these systems. So, the real challenge is to present valuable data, essential data, on top of the EMR screen, in addition to meeting all of the existing government mandates.
Health IT developers need to spend more time trying to figure out how not to annoy or distract physicians than thinking about what other data points we can put into a product. I know from experience that this is a technological feat. By focusing on the interaction with the technology, you can accomplish presenting the most actionable data, integrate user context and consider providing a choice to engage the information or not. This design approach empowers physicians to make more informed decisions.
But it requires a shift in mindset for the developer that seems counterintuitive: Err on the side of a physician not using the technology rather than risk interrupting an interaction with a patient. Data alerts should not be so arrogant that they require adjudication just because a physician opened the EMR.
I’ve learned many things over the last few years while balancing a dual career as a gastroenterologist and health-tech entrepreneur, but none as important as the idea that there are core missions that physicians need to focus on, so leave them alone when they’re doing it. If there is ancillary yet valuable data to present, don’t present it in disruptive ways or at times when it’s not requested. These are lessons that I think are worth remembering and putting into practice for every health IT solution. No patient or physician benefits from the dehumanization of medicine. We are losing great clinicians, which is very alarming to me — but we have the chance to change this.
Mukul Mehra, M.D., is the CMO and co-founder of IllumiCare and a practicing gastroenterologist in Birmingham. As managing partner of his practice, Southeast Gastroenterology, and the director of endoscopy at St. Vincent’s Health System, Dr. Mehra understands the limited access physicians have to cost and risk data. He designed the Smart Ribbon, now used in 120+ hospitals, to assist in, not impede, smarter fiscal and medical decision making where physicians can control cost. Dr. Mehra received his M.D. from the University of Alabama and completed his residency and fellowship at Washington University School of Medicine in St. Louis.
Get the best insights in healthcare analytics directly to your inbox.
Medical Scribes Can Cut Physician EHR Time and Boost Productivity, Satisfaction
Helping Healthcare Workers Reach Their Full Potential
Doctors Aren't Luddites. But Their EHRs Are Broken