My Machine-Teacher, the Electronic Medical Record
Notes from a Family Meeting, Vol. 3, No. 8
Joyce Adams, an internist, just finished documenting her last progress note of the day at 11:00 PM. “56 yo M w/ cholecystitis…,” “98 yo F w/ weakness…,” “72 yo M w/ chest pain…” The one-liners rattled through her head as she lurched through the rest of what had become of her bedtime routine: checking email, checking messages within the electronic health record (EHR), and then slipping down the infinite waterfall of social media until she falls asleep.
The next day, she wakes up and, with coffee, does it over again, relying on patterns of illness to shuffle her patients into emergency or not, and then onward down their differential diagnosis. And then she documents it: “56 yo M w/ cholecystitis…” She doesn’t really know their names. She checks her list before entering their rooms during rounds, “Oh, Mr. Smith.” She has only moments to spend with them before retreating to the care of the EHR. Once there, she familiarizes herself with the case of Paul Smith, someone who had been admitted to the hospital for about a week now. She sifts through the notes of others, through lab values, through images. She investigates. Soon, she has a well-assembled and short list of possibilities driving his ailing health. She considers calling his wife, but then gets a page to admit another patient. She opens their chart and begins to sleuth again.
Jon Keller, though a nurse, has a similar routine. He rushes in and out of rooms, tending wounds, administering medications, and sometimes changing the television station for an irate, incapable patient. But much of his time is spent tapping away at a computer, ensuring everything he does is documented. “If it’s not documented,” they told him at orientation, “It didn’t happen.” He flips through Dr. Adams’ note. He doesn’t have time to read it all. Most of it appears to be copy-and-pasted, outdated information anyway. An email pops up: be sure to document all skin assessments. Highlighted and in red, for all the nursing staff. He knows he always documents his skin assessments. A call bell starts to ring, and then another. Notes will have to wait.
Enslaved to our own tools.
It’s such a powerful idea that it’s captured in many stories throughout history, from The Sorcerer’s Apprentice to Lord of the Rings to The Matrix to, yup, Calvin and Hobbes. That one-special-tool promises more and in less time. But then we find the promise slips just beyond the horizon. Sure, things get better in some ways, but in other ways, they change unexpectedly, sometimes for the worse. Tools end up setting the agenda for how we use our time, setting the frame through which we see the world, and even before we find answers, they give us the questions. Gerald McKenny made this observation about this unexpected power of our tools: “…the very technology that originates in the effort of the modern subject to bring the external world under his power ends with the power of technology to recoil back and destroy or radically refashion the very subject whose power it is.”
So we discover the atom and with it more power than we can responsibly handle. The atom enthralls our imagination and arms our fears. We invent the automobile; its emissions choke the planet and its roads shape our landscapes. The ventilator wasn’t just more medical technology; it changed how we think of medicine and of dying.
The electronic health record (EHR) makes promises too. Computers have done amazing things, so maybe algorithms could diagnose with a margin of error far less than the fallible human! Algorithms could offer the right treatments, all the time! Algorithms could prognosticate more accurately! All the information would be at our fingertips. And algorithms don’t complain.
But if you ask most physicians today, that’s not how it works. Instead, we lament how tethered we are to the EHR, how it keeps us from spending more time with our patients, how much of our documentation is neither for our own clinical purposes nor for the benefit of patients, but to satisfy others who have entered the clinical encounter. Billers/coders, insurance agencies, regulatory agencies, and researchers edge in through the technical apparatus with their own intentions. Everyone comes to the EHR with their hands held out, longing to receive what it promises. It gives them clicks, note bloat, and alarm fatigue. Presumably someone benefits, because it keeps expanding in complexity and prevalence. Or maybe not; maybe no one really benefits. Neil Postman warned that we can be tricked into believing “we are at our best when acting like machines, and that in significant ways machines may be trusted to act as our surrogates. Among the implications of these beliefs is a loss of confidence in human judgment and subjectivity.” With nowhere else to go, we keep returning to the devil we know.
As Postman’s commentary suggests, it’s not just that the EHR is a machine imposed on an otherwise human environment. It also demands, in some sense, that clinicians become machine-like in their work. But this places us at odds with a powerful metaphor that has informed the identity of clinicians (well, physicians) for a long time. According to Lakshmi Krishnan and Michael Neuss, that metaphor is of the “virtuoso doctor detective. … The image of a ‘half detective, half virtuoso’ remains central to the teaching of clinical reasoning, medical licensing (which tests clues-based investigative skills), professional identity formation, and the linear problem-solving of daily practice.” As a trade-off, “the clues-based investigation decentres the patient in diagnosis, overlooking patient expertise and embodiment, illness phenomenology, and the social determinants of health.”
The EHR forces us to reconsider the metaphor of virtuoso doctor detective: “…many perceive it as a threat to clinician identity and well-being, arguing that it turns a physician from a diagnostician, interlocutor and healer into the ‘highest-paid clerical worker in the hospital.’” Krishnan and Neuss trace the history of this metaphor along with the development of the EHR to provide some greater insight into how we might move forward:
“As this litany of modest, diligent, but necessary tasks suggests, are we more aptly clerks or curators or archivists than consulting detectives? Are our most significant contributions through the steady, incremental work of caring for people on a daily basis, rather than marvelous feats of diagnostic code-breaking? Are there ways in which we can embrace these roles or make the EHR work for those whom it was intended to benefit?”
The hope of doing more and better of what we already do existed before the computer itself in the form of the consultant, “a ‘brain on demand’, to turn to - as [Sherlock] Holmes says - when more mundane minds are ‘out of their depths.’” When we consult with a specialist, we don’t want their heart; we want their brain. Yes, yes, bedside manner is nice, but what we really need is their competent (No! Magnificent!) expertise. But medical evidence and clinical data now exceed what any one person can possibly remember, let alone manipulate. Enter the computer.
Well, because computers still can’t do much of what people can, they rely on humans to collect data and feed it to them. So, clinicians spend a lot of time tending to the EHR. Whether in primary care or inpatient medicine, it’s usually about half the (long) day. This is a bit better (but not much) for something like cancer care. The evidence is sparse for other clinicians like nurses, but they do get to spend more of their time in direct patient care and less time documenting. It’s possible that the EHR has more of an impact on how physicians engage with patients than it does on nurses (though I’d be interested to hear anyone else’s perspective on this). The EHR has a ravenous appetite for data, and we’re the only ones who can feed it.
All that time spent feeding the EHR teaches a person some things.
Medical Education
You wouldn’t know by the abundance of lectures and tests, but Frederic Hafferty and Ronald Franks had it right: “medical education - at root - involves the internalization of new values, attitudes, and rationales, and much of what students learn involves matters of a moral nature.” Lurking in the corridors of such an education is the “hidden curriculum,” those forces that foster undesirable or negative professional traits in trainees, though the inculcation of “new values, attitudes, and rationales” happens for good or ill across medical training via socialization either way.
I remember standing in the corner of an exam room as a trainee. Our soft-spoken patient that day was doubly unfortunate. First, she was stricken with a life-threatening illness. Worse, the thing that most bothered her was a minor issue that wasn’t in the realm of my attending’s expertise. After each mention of it, the attending brushed it off: “You’ll need to see your primary care doctor about that.” Eventually, though, my attending had enough. Turning away from the computer, they shouted at the patient, “This is not going to kill you! The disease I’m trying to treat is going to kill you! Do you understand?” Sufficiently brought to heel, the patient didn’t mention it again. My attending did a great deal of teaching in that moment. That’s the hidden curriculum.
In less dramatic fashion, when you watch interns spend much of their day in front of a computer, figuring out “how to get stuff done,” you learn what it means to be a doctor. The problem-oriented medical record whispers the hope of becoming the “virtuoso doctor detective” we all admire, even if it takes it further from spaces of care and compassion.
This hidden curriculum is in contrast to the handful of lectures that taught me about the four principles of biomedical ethics. The hidden curriculum has in its favor the prevalence of exposure (both throughout training broadly and also the day-to-day work of a clinician), proximity to the clinical encounter (the valuable work of a clinician), and emotional force (which enhances learning). Formal ethics, professionalism, and medical humanities curricula, on the other hand, are neither prevalent (occurring primarily in a single course and then “refreshed,” perhaps, in ethical updates for credentialing purposes), proximate to the clinical encounter (occurring primarily in a classroom), nor emotionally forceful.
These two prevalent paradigms of educating physician trainees parallel a similar dialectic in the field of moral psychology between rationalist and intuitionist models for moral decision-making. The rationalist approach is skill-based. The intuitionist approach is motivation-based. Jonathan Haidt explains, “The first principle of moral psychology is intuitions come first, strategic reasoning second.” To capture this idea, he supplies a metaphor of a man riding an elephant. The elephant represents a person’s intuitions, while the rider represents strategic reasoning. Haidt and others have argued that social intuitionism has primacy over the rationalizations of higher-order thinking; snap judgments influence behavior more effectively than post hoc reasoning. Therefore, the rider becomes less of an advisor to the elephant and more of a “lawyer,” justifying where the elephant goes after the elephant has chosen the direction in which both are headed.
A rationalist approach to ethics education, speaking to the rider, would equip trainees with skills to manage ethical dilemmas, whereas an intuitionist approach, training the elephant, would cultivate character and virtue. The hidden curriculum and other socializing forces of medicine operate through the latter approach.
The EHR, a master teacher in the hidden curriculum, is an ever-present companion for most clinicians. It teaches us not only what to think but how to think. We really do come to see our patients as bundles of problems because that’s what we spend half the day writing about. I doubt the EHR is teaching anyone to be an evil clinician, but insofar as the aim of medicine is health, it might be teaching us to be bad clinicians, if caring for the EHR distracts us from caring for the patient. We become more skilled in clicking than comforting.
But maybe our professionalism, ethics, and medical humanities curricula will save us?
Warren Kinghorn argues that medical educators are mistaken to believe the moral formation of medical trainees is best accomplished through a “top-down teaching initiative” (e.g., lectures) as “it is not talking about the virtues, but only practicing the virtues, that leads to virtuous practice.” Modern medical education presumes that “professionalism” (that is, holistically good, virtuous clinical practice) is a product of that education. This is just like the rest of medicine: you teach about hyponatremia and then you diagnose and manage hyponatremia. Not so for moral formation: “Close, on-the-ground, concrete moral mentorship is therefore absolutely essential … if one is ever to become a person of moral excellence.”
Inefficient Mentorship
There’s one problem.
Human relationships, including mentorship, just aren’t that efficient. Sure, we can have efficient meetings, efficient rounds, efficient conversations. But if we’re going to have more than a merely transactional relationship, it’s going to be hard to efficiently manage. In our quest for efficiency, we break something essential about ourselves, as Jacques Ellul lamented:
“Technique has penetrated the deepest recesses of the human being. The machine tends not only to create a new human environment, but also to modify man’s very essence. The milieu in which he lives is no longer his. He must adapt himself, as though the world were new, to a universe for which he was not created. He was made to go six kilometers an hour, and he goes a thousand. He was made to eat when he was hungry and to sleep when he was sleepy; instead, he obeys a clock. He was made to have contact with living things, and he lives in a world of stone. He was created with a certain essential unity, and he is fragmented by all the forces of the modern world.”
Kinghorn echoed this same warning when he described the confusion in attempting to “produce” good physicians by way of technical educational interventions.
Our struggle with the EHR is both a symptom of and a contributor to our struggle with how to properly inculcate the “values, attitudes, and rationales” of medicine. It’s a symptom because without a firm grounding in such values, we become distracted by, in this case, anything that promises greater efficiency. I wrote about that confusion elsewhere. The EHR, rather than being a tool to bolster the health of patients, becomes a distraction from that very endeavor. We become enamored with the metrics that the EHR can capture, the ways it can be used to report workload, and its useless comprehensiveness.
It also contributes to the struggle. Our hope in technological salvation expands with medical knowledge and clinical data. We suspect that more knowledge, more efficiently collected and applied, will help us to solve our all our dilemmas. The more hope we invest in the EHR, the more frustrated we become when it fails to deliver. Yet we keep returning to it! But it’s not for lack of data that we stumble.
I don’t mean to throw records to the wind. We need our tools, the EHR among them. But the EHR is now also one of our most steadfast, effective teachers. It shapes our thinking by its standards of clinical documentation, giving us eyes to see our patients in certain ways (as chunks of time, RVUs, CPT codes, diagnoses). It’s a short step from “If it’s not documented, it didn’t happen” to “What matters is what’s documented.” In so ordering our thinking, we lose sight of that for which we’re using all this medical intervention. Some may forget, as they become beholden to the requirements of the machine. Others may never learn, for want of a good teacher.
To re-order our attention, we need to ask, what is the EHR for? The answer is found with a deeper question: what is medicine for? Efficient mentors might seek an efficient answer. Others, though, may “live into” the question through their practice, modeling a more humane way of medicine.
Trajectories
Following a meandering reading-path, sharing some brief commentary along the way.
Nathan Gray picks out an interesting quirk of emotionally-fraught medical encounters: why do people desperately search for tissues once someone starts crying? He offers some thoughts. I agree with him: if they’re within an easy distance, I’ll pass them over, but I’m not going to run out of the room to grab some.
“Here’s why we’re not prepared for the next wave of biotech innovation”
Much of what this author describes a learning health system could do. The Institute of Medicine published a big report on learning health systems in 2007. If only they were more prevalent. Alas, we can’t run RCTs on everything. Our diagnostic nosologies are limited (e.g., for mental illness). The overlap between the biological and social-cultural skews the impact of purely biological interventions. And the author alludes to perhaps the biggest problem: not everyone is in this to improve health. Healthcare is a market with consumers. From research to development to marketing to documentation (EHR!), there’s a lot of money to be made. Perhaps the best evidence of this is that a lot of money is invested in developing expensive medications of questionable or no benefit (e.g., aducanumab) while other interventions of guaranteed benefit are overlooked (e.g., in-home care, nursing facilities). I’m not disparaging biomedicine, but without more clarity about what we’re pursuing with medical intervention, we risk falling over ourselves chasing after the ever-elusive promise that another technique will offer.
“Misapplying autonomy: why patient wishes cannot settle treatment decisions”
These authors helpfully delineate two contexts in medical decision-making that bedevil our debate about controversial interventions. The first is the pre-clinical administrative context, where society decides whether an intervention qualifies as a medical intervention. The second is the clinical context, where verified medical treatments are used to intervene on someone’s body for their health. The authors argue throughout the paper that the administrative context belongs primarily to clinicians alone, but later rightly recognize that it’s really a dialogue among clinicians and non-clinicians. I’d agree: everyone gets in the participate in the administrative conversation (e.g., regulators, politicians, ethicists, teachers, etc). As trust has eroded and we’ve lost sight of health as the purpose of medicine, clinicians have felt cramped as other voices have been elevated in that administrative context. We clamor about medical decisions being between a doctor and their patient, but fail to recognize these different contexts.
“Getting back to civil discourse will require Americans to be vulnerable and humble”
It should come as no surprise that this was the topic of a medical leadership conference. The years we’ve endured in the pandemic reveal that medicine is just as vulnerable to polarization and politicization as any other area of life. Disagreements remain. Unless we want our culture to devolve into ever more extreme polarization, we need to learn to talk and work with people with whom we disagree, even when our disagreements are over important issues. We need to figure out how to live together. Medicine has traditionally done this well because we’ve all shared a common goal: our patients’ health. However, we’re forgetting, losing sight, and diluting that goal, which means those with titles that might suggest they share this common goal (“physician,” “scientist,” “nurse,” “hospital president,” “health insurance agency”), may not. Recovering that shared goal is surely one of the early steps toward this more civil discourse.
In this study, the more frequently a patient saw their primary care clinician before a diagnosis of cancer, the less likely they were to have metastatic disease and the lower their cancer-specific mortality over the study period. If this were a medication, everyone would be raving about it and it would cost tens of thousands of dollars per treatment. But it’s “just” primary care, so things will continue business-as-usual.
“Normothermic regional perfusion”
There’s been increasing discussion about this practice which is used prior to organ procurement for transplantation. The American College of Physicians paper shares several ways the practice raises concerns. Ultimately, we can’t let the need for more organs drive to the determination of death (either at a policy level or an individual level). A good primer article to exploring this literature was written in 2009 by Alan Rubenstein entitled, “What and When Is Death?”
Closing Thoughts
“The reason that technological progress exacerbates our feelings of impatience is that each new advance seems to bring us closer to the point of transcending our limits; it seems to promise that this time, finally, we might be able to make things go fast enough for us to feel completely in control of our unfolding time. And so every reminder that in fact we can’t achieve such a level of control starts to feel more unpleasant as a result.”
Oliver Burkeman, Four Thousand Weeks