The EMR and the Case of the Disappearing Patient

The electronic medical record (EMR) is here to stay. Its adoption was initially slow, but over the past decade those hospitals that do not already have it are making plans for implementing it. On the whole this represents progress: the EMR has the ability to greatly improve patient care. Physicians, as well as all other caregivers, no longer have to puzzle over barely legible handwritten notes or flip through pages and pages of a patient’s paper chart to find important information.

With the EMR, it is easy to see what medications a patient is taking, when they were started, and when they were stopped. Physicians can easily find key vital signs – temperature, pulse, respirations, and blood pressure – plotted over any time frame they wish. All the past laboratory data are displayed succinctly. But it is not all gravy.

There is a Problem

I use the EMR every day, and I am old enough to have trained and practiced when everything was on paper. While overall, I am happy to have electronic records, there is a problem: The EMR is trying to serve too many masters. The needs of these various masters are different, and sometimes they are incompatible, even hostile to one another.  These masters include other caregivers, the agencies paying for the care, and those interested in medico-legal aspects of care.  What can happen, and I have seen it many times, is that the needs of the caregivers take a back seat to the needs of the payers and the lawyers. The EMR is supposed to improve patient care, but sometimes it makes it worse. Physician progress notes illustrate how this happens.

The Patient’s Story

Progress notes are the lifeblood of the medical record. They tell, from day to day, what physicians did to a patient and why. They are a narrative of the patient’s care. Three decades ago we sat down, pulled out a pen, and wrote out our daily progress notes. There were standard ways of doing this, but physicians were free to organize their notes however they liked. That was both a blessing and a curse. It was a blessing because not all patients fit the standard way of note writing, so you could modify how you recorded things; it was a curse because every physician was different, and some wrote very sketchy notes indeed, notes from which it was very difficult to figure out what happened.

I once did a research project for which I was reading physician notes from the nineteen twenties, thirties and forties. I recall one patient in particular who was clearly desperately ill. He had critically abnormal vital signs (which I could tell from the nurses’ graphic chart), needed several blood transfusions, and even stopped breathing once. His progress note for the day, written by a very famous and distinguished physician, was one line: “Mustard plaster didn’t work.”

Physician notes have evolved a great deal since 1930. Certainly in my medical career, which began in 1974, physicians were expected to make some reference to what they were thinking, why they did or did not do what they did. Sometimes the notes were cryptic jottings that made it very hard to follow what was happening. But most of the time you could understand what your colleagues were thinking.

Payers and Lawyers: Different Needs

But while this worked reasonably well for physicians, other users of the medical record complained loudly. Payers, such as insurance companies and Medicare, based their reimbursement upon those notes. They were unwilling to pay for anything that was not clearly documented. They also increasingly based their payment structure on the complexity of the medical decision making; if physicians wanted to be paid at a higher rate for managing a complex and difficult patient they needed to show in their note just why that patient was complicated. They needed to show what they were thinking, and what information, such as laboratory data and the physical examination, they used to make their decisions.

Finally, for the lawyers, the operative phrase was “if it’s not documented, it didn’t happen.” In theory, the goals of all three users – caregivers, payers, and lawyers – should be in alignment. But with the EMR the needs of the caregivers, which should be paramount, are losing ground.

The EMR, since it is on a computer, can be manipulated in all the ways a computer allows. Hospitals are laying out millions to implement the EMR, and to ensure maximum payment they want to make sure it is easy for the payers to find in the EMR all the things the payers want there. This is accomplished, among other things, through the use of templates and “smart text” for progress notes. For example, a physician writing a progress note in Epic, a popular EMR system, can open a template that has many components of the evaluation already filled in. The program can bring into the note all the previous laboratory values. It has all the categories of the physical examination sitting on the screen for the physician to fill in.

Losing the Narrative

It is easy to “drag and drop” information from previous notes with simple keystrokes. There’s nothing intrinsically wrong with all this. It can make producing a complete progress note quick and easy. But it also can destroy the original purpose of the progress note – to give a narrative of the patient’s progress. It can stifle the conversation between physicians embodied in traditional progress notes

Recently I saw an example of the problems this can cause. A couple of weeks ago I heard I was getting a patient into the pediatric intensive care unit with multiple problems, most acutely a blood problem. One of these lesser issues was a heart problem that required surgery. Because of the other serious problems, though, the surgery had been postponed for the future. I read about all this in the patient’s EMR before she even arrived in the PICU, which is one of the great aspects of the EMR. We no longer have to wait for a clerk pushing a cart around the hospital to deliver the paper chart. The patient had been seen just that morning by her hematologist for the blood issue and the progress note in the EMR told me the plan for her heart problem was surgery sometime in the future when the child’s other problems had improved. It said so right there on the screen. In fact, all the notes had been saying that for over a year.

So imagine my surprise when I went in to see the child and saw an obvious and well-healed surgical scar on her chest, clearly from cardiac surgery. She had had her heart fixed two months before at another institution. I gave her hematologist the benefit of the doubt and assumed her doctor knew the surgery had been done, and that what had happened (I hope) was that the doctor had used the beguiling convenience of drag and drop on the progress note template to do the note. This particular incident was innocuous, but I think you can see the potential for mischief with this sort of thing.

This is not an isolated event. I have seen many examples– so many that I now cast a suspicious eye on all those uniformly formatted progress notes. The ease with which mounds and mounds of verbiage and laboratory data can be stuffed into a progress note may give the payers what they want, but it often does not give me what I want– and that is some evidence that all this information was processed through a physician’s brain and led to a carefully considered decision about what to do. I want a human voice, and that is getting harder and harder to find in the EMR’s stereotypic and bloodless documentation.

Medicine is about stories – patients’ stories. I was taught forty years ago that most of the time the history gives us the diagnosis. Osler reputedly said: “Listen to the patient. He is telling you the diagnosis.” (That attribution has been questioned, but the spirit is definitely Osler’s.)

Of course these days our wonderful scientific tools often give us the answer, and I certainly do not wish to toss all those things aside to go back to using only what Osler had. But medicine is not really a science. It is based on science, uses science, and is increasingly more scientific. But medicine also contains large measures of intuition, educated guessing, and blind luck. I do not think that aspect of medicine will ever completely disappear. When I read (or wade) through a patient’s record, I look for the story. When I cannot find a coherent story, I cannot give the best care.

For myself, even though I of course use the EMR, I refuse to use all those handy smart text templates. It takes me longer, but I type out my progress notes, organized as I did when I used a pen and chart paper. It takes me a little longer, but it makes me think things through. No billing coder has ever complained. More than a few colleagues have told me, that when we share patients, that they search through the EMR to find one of my notes to understand what is happening with the patient.

My advice to other doctors is this: don’t let the templates get in your way. Tell the story.


Christopher Johnson is a physician who has practiced pediatric critical care for more than three decades. For many years, Johnson served as the Director of the Pediatric Critical Care Service at the Mayo Clinic and Professor of Pediatrics at Mayo Medical School. Today, he devotes his time to practicing pediatric critical care as President of Pediatric Intensive Care Associates, P.C., i n St. Cloud, Minnesota, and as Medical Director of the PICU for CentraCare Health Systems. In addition, Johnson writes about medicine for general readers, both on his blog and in books such as HowYour Child Heals: An Inside Look at Common Childhood Ailments and How to Talk To Your Child’s Doctor: A Handbook for Parents.

Maggie Mahar is an author and financial journalist who has written extensively about the American health care system. Her book, Money-Driven Medicine: The Real Reason Health Care Costs So Much, was the inspiration for the documentary, Money Driven Medicine. She is a prolific blogger, and recently relaunched her HealthBeat Blog, where this post originally appeared.

9 replies »

  1. I, too, have noticed that people often swipe large sections of somebody else’s progress note to cut and paste into their own. Besides all the other issues associated with that, I have wondered what a copyright lawyer would think.

  2. Rbaer, Bobby G Cascadia, Bird & everyone –Thanks for your comments.

    I think rbaer hits it on the head: “As Dr. Johnson and MM point out, it’s still possible to generate thoughtful notes, by typing or dictation; and if you proofread what you cut and paste, there is no harm in cutting and pasting . . . ”

    I would add that Iits up to physicians to adapt EMR to serve their (and their patients’) needs.
    There is no way that Health IT can tell the story for them. Only they know what happened to the patient. They need to intergrate that old-fashioned narrative into the EMR, as Johnson does.

    As Rbaer adds: “Training (not computer training, but physicians training med students and residents) must emphasize note proofreading, feedback on minor and consequences for meaningful c+p errors.”.

    Bobby G– Thanks . What you descrbe sounds ridiculous. IT over-kill. There are many things IT can’t do and one of them is telling stories.

    Like Chris J., I think that truly excellent EMRs can be extraordinarily useful.
    Unfortuinately, too many have been created by software people who have no understanding of what it’s like to be treating patients in an ER, in a smal
    primary care practice, or on a busy hopsital ward… . . .

    Some medical centers have excellent EMRS. (I think of the VA, Mayo.) But I suspect we need very different EMRs for different settings (though of course
    those EMRs need to be able to talk to each other).

    Cascadia– Interesting question. My guess is that the VA’s EMRs are less concerned about billing becuase they don’t have to persuade an external payor
    that what they did was needed.
    It’s also worth noting that VA doctors practice evidence-based medicine, following VA guidelines. So there’s less need to justify what they’re doing. . .

    Bird– Sounds like you’re doing what Chris J. does. My guess is that you’re
    finding so many errors on EMR records because docs are just learning how to use them properly See Rbaer’s comment.

  3. Pretty sure that copy/paste of one providers note into your own as ‘documentation’ of your visit isn’t going to pass muster if your payers / CMS knew about it.

    Using CPOE to generate your chart note is one thing
    passing off someone elses documentation as your own is another.

    Maybe im wrong.. whos going to be the first to ask CMS?

  4. “consequences for meaningful cut and paste errors?”

    hmmm think about where that one’s going to lead us …

    penalties for distracted autocompletes?

    sanctions for spellcheck failure?

  5. At the Health 2.0 SF Conference I saw a demo of an app that purportedly uses “NLP” (Natural Language Processing) to data mine progress note narratives for billable items that may not have made it to the claims, for subsequent re-billing.

    So, we now have EHRs that take in structured data and convert some of them to fake progress note open-ended text “narratives,” but we’ll subsequently add an app that converts the stuff back to structured dx and px codes?


  6. But that’s not Epic’s fault, or the fault of any other EMR (as a disclaimer, I do not like Epic in particular). As Dr. Johnson and MM point out, it’s still possible to generate thoughtful notes, by typing or dictation; and if you proofread what you cut and paste, there is no harm in cutting and pasting. Otherwise, it is like blaming an increase in speeding violoations to the increase of horsepower of the average car.

    It is our – the physicians – fault. Training must emphasize note proofreading, feedback on minor and consequences for meaningful c+p errors.

  7. Wonderful article!
    I have been on epic for over 10 years, I am convinced that the error rate on the EMR is significantly higher then the error rate on the hand written charts. I find medical errors on a daily, actually hourly basis. They generally are not greivous errors that effect outcomes but they are errors never the less. I have had to go back to dictation and have been much happier. The fact that anyone on the “care team” can abstract to the chart has caused serious problems. The phrase garbage in garbage out has never been more true.

  8. Paradox – interesting how we use the narrative to describe this problem instead of a series of numbers and other providers thoughts cut and pasted into this article?

    I also wonder what the EHR note looks like in those situations where billing is less of an issue – direct pay primary care, the VA etc.. Is there still the focus on trying to capture everything and as a result just creating noise?

    Do people feel that the voluminous notes are an artifact of our billing system, the EHR itself, physician culture or something else entirely?