Thursday, September 8, 2011


In today's NY Times Well Blog, Dr. Danielle Ofri observes that healthcare is the only major industry not fully computerized…and she doesn’t wholly lament the fact.

To me, electronic medical records (”EMR”s) are like steroids: useful and at the same time potentially hazardous. Steroids can ease symptoms, even save lives, and they can also cause diabetes, insomnia, personality changes, and a host of other horrors. EMRs solve significant problems in medical practice, being portable, transmissible, and of course legible compared to physicians’ notorious scribblings. On the other hand, as Dr. Ofri points out, "…both physically and psychologically, [the computer] has placed a wedge in the doctor-patient relationship."

I’ve griped aplenty here about how crowded the medical examining room has become. Fifty years ago it was a sanctum of intimacy, occupied only by patient and doctor. Then an insurance agent stepped in to verify that the company wasn’t blowing its money. Soon an attorney dragged a chair in, followed by the other party’s attorney. Then came a smattering of bureaucrats, pharmaceutical reps, and whoever else felt they had a stake in the proceedings. And now, the ubiquitous computer. I can’t even count the number of people who’ve told me how they resent seeing only the back of their doc’s head as he/she types into the laptop.

Neither is Dr. Ofri wild about having to favor machine over patient. She writes that she compensates by devoting more attention to the physical exam. “Once the patient and I have broken free from confines of the desk, with its dictatorial PC, we have a more comfortable realm, that of touch. As soon as there is skin-to-skin connection, conversation flows more easily. In the absence of a machine lodged between us, the traditional doctor-patient relationship is restored.”

Hallelujah! But, alas, there is then the matter of arranging diagnostic testing, medications, and appointments, all mediated by computer. So the personal aspect of the medical visit—the touching, the contact, the deeper conversation—is now squeezed between the computer-dominated beginning and end.

“If you keep going where you’re going,” a nurse once advised me, “you’ll get where you’re headed.” Let’s look where medical practice is headed. Let’s suppose practitioners’ compensation continues to get pared down, further shortening the average visit time (now less than ten minutes). Designers of medical software—smart folks who wish to free up the doc’s time—will develop more sophisticated products such that the doc won’t have to key in a coherent narrative, only select a macro—or even speak a command—and a fuller report will appear in the record.

If I were a software developer, I’d go further by wondering why the doc needs to get involved even eliciting a history when the patient can check off a list while in the waiting room, which is converted instantly into a history on the EMR. And if, as Dr. Ofri suggests (and I agree), the exam is less useful than the history in securing the diagnosis, then as a secondary function it can be largely automated.

I won’t shilly-shally here: the ultimate destination is what in my darkest moments I’ve long predicted: healthcare from vending machines. Excuse me, I mean “healthcare” from vending machines.

Dr. Ofri’s to be commended on mourning the intimacy lost to computers. In that feeling, I suggest, she represents the bulk of practitioners, physicians and others who harbor implicit reservations about EMRs. I think that's what accounts for why healthcare is one of the last American endeavors to be computerized.

Other strategies are available. Here’s one alternative, which I admit might sound outrageous at first: smaller EMRs. Virtually all diagnostic tests are easily digitized, but as for doctors’ notes, what amount, after all, do we need? I remember the chief resident on a ward of the huge county hospital where I trained, a young guy who was promoted to that position because he flat-out knew everything. He’d see patients with the most complicated medical problems and then sum up their cases with a single sentence, a short, elegant statement which allowed any knowledgeable reader to instantly understand the situation.

I admit I don't have answers, only a few itchy questions. I'd love for more people to be in on the discussion. We’re drifting toward a vending machine future because those who are most involved in healthcare, the two original examining room occupants, are all but silent, having abdicated the chore to people who design, well, vending machines.


  1. I think I may disagree on this one - perhaps as a member of the first www generation my brain has been taken over by the microchips but here goes:

    First there are two benefits to EMR that I get excited about. #1 is portability - especially important for poorer patients (and migrant farm workers) who usually end up getting pretty slipshod treatment as they show up at various clinics around the country. And also it's important for bicoastal bobos like me. I want my new doctor to be able to see at a glance the name of that medication that I'd forgotten about. #2 is the possibility for improvement. Get everyone on the same electronic page and you create this massive supply to data that researchers will use to see what's working and what's not. Want to see if someone of your age and vitals really should be taking those expensive statins for the rest of their life? No problem. Even better, as we learn stuff we could link it to our charts. So the paper on the ineffectual statins comes out and when you and your doc look at your blood pressure history, there it is advising you to think twice before you buy. Rather than waiting a generation for outmoded practices to die with the doctors that use them, we could be teaching physicians in real time.

    Right now doctors are being taught to use decision trees in medical school. "If patient show X then look for y, if y then..." they are basically learning algorithms - being trained to be computers. Let's let the computers do the computer work and free up docs to do what humans do best: Caring, paying close attention, and seeing patterns in the depths of complexity.

    But for EMRs to help this rather than hurt they have to satisfy (at least) two conditions:
    1. They must be fully accessible to the patient (us). These are our records after all. The only way they will be used in a truly patient-centered way is if we have control over them. I was recently involved in an innovative medical program where the patients picked up their charts at the beginning of the day, read through them, weighed themselves, took their blood pressure, and wrote it all down. So empowering! People actually started speaking up, asking questions that they might have been to cowed to ask otherwise. They could see for themselves what the doctors were talking about. There was a lot more trust for the medical system than I normally see. And getting this right depends on:

    2. They must be designed so that they enhance rather than detract from the human interaction. And I'm not sure how to do this (I bet Apple's designers could figure it out). But we certainly can't have the physicians staring at the screen, doing data entry while the patient shivers on the other side of the room.

  2. Nate,

    Thanks for your thoughtful and thorough comment. I couldn't agree more with everything you say.

    To be clear, I didn't write that EMRs themselves are harmful. In fact, they're a boon to medical intelligence, and we've barely begun to mine their possibilities.

    Being universally portable, it's easy to render them accessible to patients. Legally, the information is the patient's anyway. Some organizations, like Kaiser Permanente, makes patients' records available to them online, and I know several Kaiser members who've even caught errors in them and had them corrected.

    What I meant in writing about EMRs was exactly what you feel, that they "…must be designed so that they enhance rather than detract from the human interaction." In a perfect world, computers would, as you wrote, "…do the computer work and free up docs to do what humans do best: Caring, paying close attention, and seeing patterns in the depths of complexity." But one of this world's imperfections is that we docs get no training in therapeutic patient contact. It's great when computers crunch the algorithms, leaving us free to treat suffering, but we haven't been taught how to do that, and I suspect most of us, especially younger docs, don't even consider that their responsibility.

    My fear is that as EMRs get perfected, docs will indeed have time freed up, and that time will get filled with more and more patient visits. A strategy to alter that course is for both docs and patients to assert their humanity, loud and clear, calling for human contact as deep and skilled as our physical proficiencies.

  3. Absolutely. Of course as you've suggested elsewhere on this blog, if doctors don't learn to flex these empathetic muscles they stand a real chance of being replaced by machines. Already, I feel like my primary care guy is pretty redundant. The question is, how much worse does it have to get before it gets better.

  4. I was irritated that your comment didn't get more "Recommends" in the NYTimes. I get more confirmation every day that people above age 40 are truly dinosaurs, and the young ones view us as ~virtually~ extinct already (pun intended), because we harp so much on the difference between pre- and post- ubiquitous technology. I was startled enough by the emergence of the Walkman and cellphone (when they were the size of a cinderblock)--how little did I know it would come to this--
    This post links well with your SUNDS post, because the whole point is, the computer reinforces the conceit that all our ailments are knowable, measurable, conquerable, and health is just a simple matter of having the "latest" research at our fingertips--that is, at our computer fingertips, to the neglect of the therapy of touch, empathy, and total person interaction.

    One of my favorite doctors I work with is above age 70, and he constantly peppers what might be rather dry soliloquies on the latest medical research with dead-pan, off-the-wall dark humor--so that young residents may not know that he is joking, to the great merriment of older nurses. One of my favorites of his refrains is that the patient is suffering from testing hypertrophy--and that they would find something wrong with any of us if they did enough tests--so true!

    After all, the most unconscious assumption of all is the agreement between patient and doctor is there is a problem--the bias leans towards finding a problem, and away from declaring there is no problem.

    Yet for all our fancy research and bias towards the "latest" research, as opposed to the most time-tested methods (like those ageless classics such as friendship and strong communities with intrinsic meaning to work and life), I've never seen two patients with the exact same diagnosis live through it anywhere close to the same way. Each person experiences illness, as they do life, with their inimitable quirks and charm, or lack thereof. So I always want to say, "We'll throw this medicine at you and see how YOU respond. It's going to be an experiment, let's hope it doesn't turn into an adventure."

  5. DreamsAmerlia, I'm delighted you're still reading my humble blog.

    Your comment's points are well-taken, especially that "…the most unconscious assumption of all is the agreement between patient and doctor is there is a problem…"

    Many folks won't grasp the significance of that statement. After all, there certainly seems to be a problem. My body had been functioning just fine and now it's not, so I'd like my prior condition back, please.

    Let's consider a couple of examples. Say I work in an environment short of congenial. My supervisor routinely upbraids me in front of my colleagues. It's hard to take. When I sense him approaching, I turn my back and protectively raise my shoulders. Doing this several times daily for twenty years, my neck's develop a chronic tightness and pain, so I consult my doctor for relief.

    There's nothing wrong with my neck. It's behaved normally, hurting as a result of a chronic stress to which I've not responded adequately.

    Or, say, I sate my various hungers by ingesting a couple of Whoppers daily. The resulting sludge that amasses in my coronaries finally leads to an anginal episode, so I'm offered stents and an angioplasty. There's nothing wrong here. My body's behaving normally to chronic physiologic insult.

    When we lead unhealthy lives, our bodies respond accordingly even given their admirable margin for abuse. To define a normal response, then, as a "problem" instead of a natural, anticipatable consequence is to enable pathogenic behavior. Even though that seems an alarming way to describe a common style of medical practice, it's the sorry truth.