How to fix EMRs

I was talking to a colleague last week about his practice, and remarked that he was still keeping a paper medical record. Without hesitation, he made it clear that he not only liked the paper record, but he positively dreaded switching to an electronic record. He said sadly that he thought it was inevitable that he would be forced to switch, but hoped that the day would be far into the future.

Intellectually, I think most doctors (excluding the occasional luddite or those so set in their ways that nothing in their practices will ever change) understand the potential benefits of electronic record keeping: more complete information accessible to the clinician (and patient!) at any time, from any where; facilitated sharing of information among physicians caring for the same patient; the ability to provide clinical decision support (reminders about indicated services, drug-drug interactions, embedded care pathways, access to supporting clinical evidence); the ability to aggregate information for quality improvement purposes, and more.

And yet, reluctance to adopt an electronic record is prevalent. In general, the reasons – stated and unstated – include the common perception that an EMR slows clinicians down; the constraining nature of structured data entry; the tedium of typing (which often makes doctors feel like they are scribes); the barrier that the computer creates between the patient and the doctor; the frustration that the computer work-flow doesn’t match how doctors think or work; and the general reluctance to change what seems to be working (at least at the individual physician level). If it ain’t broke…

While this colleague and I put most of these issues on the table, he surprised me by saying that he also thought EMRs were bad because they promote fraud. He cited a computer-generated report that he had received from a surgical subspecialist that included a complete physical examination, including an assessment of the patient’s mental status. At the time, I conceded that it was unlikely (OK, it was absolutely impossible) that the surgeon had actually done all the things “documented” and had, instead, checked a bunch of boxes (or one “big box” that said everything was normal), but I insisted that it was unfair to blame the tool for its misuse. It was, I said, like condemning hammers because somebody smashed a windshield with one. After all, hammers are still pretty useful when you are faced with a nail.

I felt pretty good about the conversation, but kept thinking about the limitations of current EMRs, including their potential for abuse. Nearly all of the things that doctors dislike about them are “features” designed to capture information needed for billing purposes. That is, they are all about documenting what we did to or for the patient, not about how the patient was doing. How many elements of the physical exam were performed? How many systems reviewed? How much clinical reasoning demonstrated? Did I “do” enough to justify a level 3 office visit?

I recalled the utterly different EMR that I saw when I visited a primary care practice that was funded through a fully capitated contract with the union to which all of the patients in the practice belonged. The electronic record was basically a medication list and an annotated problem list, with narrative added to each problem as needed. That’s it.

Like so many other things that doctors hate about the current health care environment, the flaws of the current crop of commercially available EMRs are a consequence of how we pay for care. Since we are paid for “doing stuff,” we are constantly being challenged to prove that the stuff we are doing is justified, and that we actually did it. We are getting killed by the focus on process.

We ought to be focusing on outcomes. If we were compensated for caring for a population of people, and judged on their health outcomes (appropriately adjusted for the prevalence and severity of their illnesses), then we could be freed from the stifling limitations of so many contemporary EMRs, while still enjoying the benefits they can provide for us and our patients.

What do you think?

11 thoughts on “How to fix EMRs

  1. Bravo to that MD. There are a few benefits to an EHR/EMR, i.e. easy to read, easier coordination, save on fountain pen ink, easier to write prescriptions, but this comes a tremendous price. The highly skilled/trained professional now is doing secretarial stuff that takes much time. Even with dictation, there are numerous errors and it takes a good 15-30 minutes extra per patient to finish a note. Therefore, office volume is much less.

  2. The traditional physician record is meant to provide a narrative about an individual patient. The EMR is meant to collect data from large numbers of people. These different purposes are reflected in the alien and uncomfortable feel of many EMRs. In addition, many (most?) of the systems have been years in the making, and look as if 15 plus years of software engineering have passed them by (which they have).

    1. these are 2 excellent comments. The EMR’s are not even organized the way one has been trained to see apatient starting with a history, review of systems, physical exam, review of data, and then a thoughtful collection of thoughts for an inclusive differential diagnosis with a related plan of action. The EMR has all the elements, but does not lend itself to a flowing document and discussion. In addition, when you get an EMR consult letter or followup, they are typically many pages long because they are all vertical rather than a commentary format.

      1. I agree that the EMR feels foreign for the reasons I originally expressed and for the reasons others have added. Here’s another — it tends to drive the narrative out of our patient encounters. To me, the patient’s story should always be at the center of the encounter, and it is really hard to capture that with current systems, which are all about capturing “elements” but not “the story.” I think this is a real loss, not just because the story is the best way to make the diagnosis, but because it reinforces the fact (all too often lost) that we are dealing with PEOPLE with problems, not with problems.
        That said, I disagree that there are “few benefits” to an EMR. I do believe that they can facilitate communication among doctors and between patients and doctors, and that they can help us improve care in the ways I described. Having used paper for most of my career, and switching to an EMR, I would never choose to go back.

  3. I began referring to EMRs as IHRs (Insurance Health Records) a couple of years ago, and got myself banned from a couple of health information technology-related interest groups on Linked-In. Fact is, though, that my observation was/is true. As the original poster relates above, it is possible to maintain a better EMR with much less hassle if you strip the insurance out of it. While his example seems excessively spartan, the previous comments here re: the patient narrative are on the money.

    Assuming one is using a “no insurance/no MU” EMR (and they do exist), two caveats remain, the most obvious being the difficulty of entering data. Many physicians are not good typists, but the keyboard remains the primary way for most of us to put data into the system. Voice recognition helps, and is getting better, but its use still requires patience and consistency, two traits which can be hard to come by. Related to the first caveat is the barrier an EMR potentially poses to the doctor-patient relationship, as noted above. This can usually be cured simply once the physician is aware of the problem.

    The bottom line is that it is not the EMR per se, but the regulatory/bureaucratic impositions of third party payers which have ‘broken’ EMRs, just as they have broken most of the rest of medicine. A surprising number of physicians have grown up with and accepted the notion that insurance = health care = health. Government and industry put a lot of effort in over the years to generate and maintain the misconception, and it becoming increasingly more important for the health of our profession that physicians recognize and reject the conflation.

    1. Many of the comments posted above represent valid concerns about EMR systems. At the North Shore-LIJ Health System, we refer to our EMR as the AEHR (Ambulatory Electronic Health Record). Our AEHR implementation team is currently in the process of addressing these concerns by improving the configuration and customization of the system, optimizing work flows to take full advantage of the potential benefits of the AEHR, and educating users about how to best use the system. Following are three examples of how we are enhancing the AEHR’s usability:
      1) Eliminating the delivery of paper reports. In the near future, physicians who use the AEHR will receive reports of radiology tests electronically in the AEHR. In addition, mammogram results will be entered as structured data. This electronic delivery method will eventually be used for many other types of documents, which will speed up their delivery, eliminate unnecessary paper, and enhance security and privacy that is put at risk when protected health information (PHI) is contained on paper documents.
      2) Removing meaningless tasks from the system. Several weeks ago there were nearly one million open tasks in the AEHR. Many of these were generated by the system because of work flows that are not currently utilized, such as notifying a physician after an order has been authorized. In addition to preventing further generation of these tasks, we are in the process of removing the ones previously generated. We hope to remove two thirds of these unnecessary tasks by the end of the year, and more thereafter. This will remove much of the noise that currently impairs physicians’ use of their task lists.
      3) Creating a POMR. Another approach we are taking to enhance the AEHR is neither technical nor related to work flow improvement. We hope to shift the paradigm of medical records to achieve what Lawrence Weed first advocated in the 1960s – the problem oriented medical record (POMR). Instead of relying on a note-centric approach to the chart, the AEHR could facilitate much more efficient data entry and retrieval. Almost all the information necessary to manage patients could be kept in the problem lists, where each problem has several annotation fields such as ‘description’ and ‘impression.’ These fields would allow items that should be in every note (the unchanging description of the problem) to automatically appear in the note, while items that vary (the impression of a problem during a specific encounter) would require entry at each visit. This would eliminate the need for a copy forward function, which in my opinion should probably be removed from the AEHR’s functionality.
      In the POMR, rather than searching through a variety of notes to see the history of a given problem, a user would simply view all the relevant information in the problem list. Medications and other orders can also be associated with problems, so all information relevant to a particular problem could be easily found. Most of this information is in the form of structured data, which is useful for queries and report generation. I agree, however, with the importance of narrative in telling the patient’s story. I believe that the history of present illness (HPI) section should remain the principal site for narrative and I would urge users to take the time to learn how to use voice recognition. In my experience, this technology not only allows for rapid entry of narrative history, but actually enhances the story beyond what can be accomplished with either handwritten or typed entries.
      There are many more initiatives underway to improve our use of the AEHR. Our health system’s goal is to continually increase the AEHR’s usability and usefulness, which will result in improved efficiency and effectiveness, more provider satisfaction, and ultimately better patient outcomes. This won’t happen overnight, but patience and diligence will pay off in the long run. I am confident that those physicians who are willing to invest the effort will reap the benefits.

  4. Mitch — thanks for your comments and your leadership in making our AEHR better for our physicians and better for our patients. I agree that focusing the record around the problem list would be big improvement. I also like the idea of making voice recognition easy enough to use that we can keep — or enhance — the narrative elements of the chart.

Join the Discussion! Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s