JAMA EHR Study: Misdiagnosis Poses Significant Potential for Harm

An important study in the Journal of the American Medical Association finds that misdiagnosis is more common than you might think.  According to the study, almost 40% of patients who unexpectedly returned after an initial primary care visit had been misdiagnosed.  Almost 80% of the misdiagnoses were tied to problems in doctor-patient communication, and more than half of those problems had to do with things that were missed in the patient’s medical history.

The results of this study shouldn’t be surprising if you’re a regular reader here – they are another example of a system that isn’t working as well as it could for patients, and doctors.  Doctors – and the medical professionals who help them in their work – are the best educated and best trained than they have ever been.  They have more access to medical information and technology than at any time in our history.  And yet, U.S. government data show that the typical doctor visit involves 15 minutes or less with your doctor.  Medical records are kept in fragmented, uncoordinated ways.

Never before have the stakes of getting the right diagnosis been so high, and yet our system is set up in a way that makes it increasingly difficult for doctors to do the jobs they were trained to do.  Seeing 40 patients a day, using uncoordinated medical records systems, and trying to keep up with continual advances in medicine is an enormous challenge under the best of circumstances.  And these aren’t the best of circumstances.  As the study authors point out, the greatest underlying cause is the failure to properly put together the pieces of a patient’s medical condition- exactly the type of thing you’d expect from people making high stakes decisions with not enough time or information.

The findings of this and other research underscore how important it is for patients to be active participants in their care, and to use every resource at their disposal to make sure they are not one of these many who are misdiagnosed.  Ask questions, know your family (and personal) history, and make sure you keep asking questions until you’re satisfied you are comfortable with what you are being told.

By involving yourself as an engaged, active part of your own care, you have a chance to help your clinical team avoid overlooking important facts, to help them avoid unwarranted assumptions, and to help them make sure they make the best decisions for you, with you.

Evan Falchuk is Vice Chairman of Best Doctors, Inc.. Prior to joining Best Doctors, Inc., in 1999, he was an attorney at the Washington, DC, office of Fried, Frank, Harris, Shriver and Jacobson, where he worked on SEC enforcement cases. This post originally appeared on Best Doctors, Inc.’s See First Blog.


Spread the love

21 replies »

  1. While electronic records will definitely increase the accuracy of our current healthcare system, misdiagnosis is still caused by communication errors the majority of the time.

    Interpersonal Education (IPE) is a program many healthcare organizations are resorting back to as a result of miscommunication between patients, their doctors, hospital staff, and outside healthcare personnel. IPE training can reinforce the communication skills that we sometimes forget as healthcare providers. IPE training in conjunction with more efficient electronic record systems will allow for a much more efficient healthcare experience for all involved.

  2. A great article, but I do think there is a bit of a caveat here when it comes to patients being ‘an active participant’ in their diagnosis. In this world of Google and WebMD, the potential for patients to become self diagnosing savants can pose some serious implications; especially when it comes to the extremes of over-reaction vs dismissiveness. Always ask your care provider questions, and never be shy about getting further information if something doesn’t make sense.

    • Good point about the growing popularity of patient’s self diagnosing themselves. Could you imagine if patients had their EHR available to them? People would be googling every elevated level in their blood work and come up with crazy diagnosis.

  3. Primary care nurses are a solution but they are the ones hobbled by design. There are 250,000 primary care RNs and a similar number of community health RNs. They are not coordinated as in other nations. They cannot impact care before, during, and after encounters as in other nations. The primary care RN is the largest primary care workforce, but often sits in back offices begging insurance companies for approvals for prescriptions to referrals to admission – or begs for ever more fragmented health information to be sent – each 5 years we add a new fragmented entity. Primary care nurses are much like MD, DO, NP, and PA – lack of specific primary care training.

    This article demonstrates false beliefs about what we have, fails to illustrate the multiple problems – and therefore will not help to develop solutions. The major problem in primary care and in health access and in public health – is that few of the designers actually understand most Americans as to how they live, seek health care, do not seek care, are damaged, and die.

  4. Bobby G,

    “Maybe if we enabled docs to focus on 8-10 pts/day instead of 24-30 …”

    Have you calculated what that will do to the supply of primary care docs in this country?

    Any idea of the cost of that idea?

    • Yes.

      Maybe it would have a salutary effect on the “supply.”

      You miss the obvious point (no surprise there). MOST of what routinely goes on in the outpatient setting (conservatively >50%) only requires the MD so the encounter can be billed. We are WASTING physician resources.


      • And would patients be happy to see a “mid level provider” rather than an MD?

        And would physicians be FORCED to sign off the charts of these “mid level providers” for malpractice reasons?

        What would happen if physicians REFUSED to sign charts of “mid level providers”?

        Have you missed these “obvious points”?

        • I did consulting work with the HAWC community health center last week. I spoke at length with a very bright, totally astute whitecoat in one of our clinic interviews.

          I thought she was a physician.

          She’s a NP.

          Spare me the all caps, OK? There’s a ton of shit that’s wrong in this system, as you rightfully allude to.

          Abilities of people for a lot of health care — irrespective of titles — is less important than other aspects.

          In case it (obviously) eludes you, I support MDs making that which their education, training, experience, and abilities warrant — WAY more than the chump change of the 99213

          • The caps were put in to emphasize certain aspects of our healthcare system that are not obvious to many people. Perhaps you are aware of these issues and the caps were wasted on you – but perhaps others are NOT aware of these issues.

            You and I seem to be in agreement on how NP and PA should be used. And I am not opposed to them, like some physicians. I think a well integrated system in which NPs and PAs see the simpler cases and MDs see the more complex cases is a good idea.

            HOWEVER, in most situations that I am aware of, someone wants a DEEP POCKET. Spend a little time thinking about that and don’t let the implications ELUDE you. (more caps!)

          • ” Perhaps you are aware of these issues”

            Not “perhaps,” bro. Acutely so. My (fully linked) record could not be much more clear.

  5. Maybe a nurse wouldn’t be so hobbled by the MD’s traditional Iron Man know-it-all heuristic fallacy blinders.

    Maybe if we enabled docs to focus on 8-10 pts/day instead of 24-30, they could better use their extensive training and knowledge.

    Maybe if we dispensed with the juvie bump sticker straw man negativism we could turn our attention to better care.

  6. Primary care workforce is not the best trained ever and has the least primary care experience of all time. From start (training) to end (lower revenue support) the design is grossly insufficient.

    Only a small portion have significant clinical skills training. The first 2 years are spent in course work that has minimal resemblance to clinician training. There are different models where medical students are trained in hands on evaluation from week one and the learning is arranged around the ways that patients present to physicians (clinical presentation model). NP and PA situations are often worse.

    Also tracking the scheme from clinical presentation to diagnosis may require multiple encounters, tests, or consultations. One would expect “errors” when studies examine only a single early encounter.

    Communication can always improve – the major point of the article – but also patients can improve too. Also as anyone knows from home visits, a tiny part of a patient’s life is revealed in the typical encounter.

    Also primary care except for family medicine, is least experienced by design. Our US primary care training is mostly not focused upon primary care (hospital) and the primary care workforce is 70% non-primary care in result. The experience gained in primary care is often lost in IM, NP, PA, PD, and MPD as graduates move steadily away from primary care in the years after graduation. Massive expansions of NP and PA result in even less experience as the workforce is newer out of graduation.

    FM encountered by patients averages 40,000 visits on the way to 80,000 to 90,000 over a career. Others average 20 – 35% of this number of visits without corrections for rapid expansion or loss of the most experienced. But all employed family practice (MD, DO, NP, and PA) is least supported with funding directly or indirectly. Health spending in the US is multiple times less per person in the 30,000 zip codes where 65% of people are found and over 50% of family practice – where less than 25% of other MD, DO, NP, and PA are found.

    Like the major advances in CT and MRI – there are more toys and more information, but researchers often have no clue about what they are seeing.

  7. I have been doing this for 30 years. I am very good at it. If everyone who was not good at it would stop telling me what I must do, that has nothing to do with the patient, things would be much better. I must document a tome of useless crap just to get paid. I must enter my own orders instead of talking to the patient. I must now dictate and edit my note from a voice recognition system that is not ready for prime time. I must document into a template that has little capacity for the myriad of variations and subtleties between and among patients.

    The patient matters less and less.

    It used to be doctors did not worry about finances because they got paid by almost everyone Now they have to worry about finances because they get paid by almost no one.

    The EHR is prepared AFTER the care is provided. If you don’t know the answer before you do the record you certainly do not know it after.

    No amount of documentation changes bad care into good.

  8. The time devoted to patient – 15 minutes on the average should make it rather clear that the patient’s care is not what the doctor has in mind. It is rather how many patients he/she can see on a daily basis to meet his/her financial goal that matters. I don’t necessarily think that the doctors’ training is best in years. How do we measure that? Medical school in Europe lasts 5 years and produces doctors as good as here without forcing them to work 80-hours plus during their residency. That alone should raise the red flag – I’d never want, let alone trust a judgement of a doctor who’s been doing rotations for the last five days and he/she sees me today on the fifth day of continues work and after 17 hour day….

  9. It seems that whenever these problems are identified, the solution is always for the patient to change, not the system. A patient can only do so much to improve communication with their doc. 12 mins is 12 mins. If a pt is on track for a misdiagnosis, bec the doc isn’t listening to the pt’s “theory” or the pt’s recollections of other info pertinent to the presenting problem, or is following their own uninformed hunch – what, exactly, can a patient do? Uterine cancer misdisagnosed as fibroids and similar fiascos happen all the time. Again, it all comes down to the health ins industry turning up the speed on the assembly line. Gotta process those widgets.

  10. Think Zebra.

    My misdiagnosis adventure began with shortness of breath, and swollen ankles.

    “Exercise!” they said. “Asthma!” they said. “Lose Weight!” they said.

    So I exercised, and inhaled my inhalers and lost 40 pounds. I still got black spots in my vision when I climbed the stairs.

    Finally I scheduled an echo on my own, hoping something would show up. It did. A 775 ml pericardial effusion. Scleroderma. Pulmonary Hypertension.

    Twelve years of symptoms and nobody looked for a zebra.

  11. i wish people would stop throwing around figures that have no meaning. while i agree totally with the article it uses the term 40 patients a day like that is the number that causes primary care docs to stumble. Simple math would suggest at that a doc that works 5 days a week, takes 4 weeks vacation, 1 week of cme and has about 5 holidays off would be seeing 9200 patients a year. The 100th%tile of mgma for internal medicine is probably less than 5000 visits a years. That being said the rest of the article is right one. Fragmented care and cumbersome ehr (epic) still dont allow the physician to communicate with the patient.

Leave a Reply

Your email address will not be published. Required fields are marked *