Teaching diagnostic reasoning


Category : Medical Rants

Amidoc wrote this comment:

Thank you for sharing this with us.
How about focus on teaching how to avoid clinical errors during medical school and residency? I am sure someone smart can come up with a curriculum and the apply it in real life.

Yesterday I gave Grand Rounds at my alma mater – the Medical College of Virginia in Richmond (sometimes called VCU but I reject the relabeling). The title – Learning to think like a Clinician – is pithy, but may not convey the essence of the talk. In this talk I present patients whose diagnostic process helps us understand the source of diagnostic errors as well as the path to diagnostic excellence. The talk borrows heavily from cognitive psychology and particularly two books – Thinking Fast and Slow by Daniel Kahneman and Sources of Power by Gary Klein.

This talk and those books outline a curriculum for understanding the basis of diagnostic reasoning. As noted a physician as Jerome Kassirer, former editor of NEJM, has called for diagnostic reasoning to be included as a basic science throughout medical school. He and Rich Kopelman started the NEJM Clinical Problem Solving exercises (another great way to learn medicine and the diagnostic process).

But I would argue that writing a curriculum is not the answer. The answer must come from improved clinician educators. We assume that anyone who finishes a residency and/or fellowship can teach medical students and residents. But skilled medical education requires specific skills. One skill that some cannot master is the skill of making explicit ones thought processes. Our research on ward attending rounds, and my anecdotal experience in talking with many students and residents, teaches us that learners want to understand how the process works. So we need to trainer the educators on how to teach medicine. We should develop more rigorous training for medical educators so that they can help their learners grow into great diagnosticians.

Unfortunately, we who value the art of diagnosis are handicapped because diagnostic excellence is difficult to document with measures. We cannot measure diagnostic error rates, because diagnoses are often difficult and gold standards are difficult to determine.

But we do have a responsibility to try. We should value diagnostic reasoning more as our learners know that they need to learn these skills.

Comments (11)

From the skunk at the garden party:
Dr. C: in theory this is absolutely correct. You have identified metrics as an obstacle to instituting this type of teaching. And you’re right.It is a big obstacle.
But there is another, bigger problem – the whole evidence-based medicine concept. I was around when it was first proposed in 1992. It was made to sound great. the problem was it was just a gussied up way of saying read the literature critically, something good physicians had been doing for several centuries. The literature had to be studied, critiqued and integrated into the clinician’s experience.Fine.
Suddenly EBM was this whole new thing that was going to replace “clinical experience”. Some of us thought and still think that was ludicrous – learning medicine is a lifelong process that besides taking years involves judgment, making mistakes and acquiring wisdom.
All of a sudden that wasn’t important – and there is no sugarcoating it- many of the main proponents, and I could name them but I won’t, made no bones about that. All of a sudden any first-year resident armed with a study could ignore a clinician with decades of experience- with impunity. That was the most stunning part, residents thought nothing of tuning out experience, and were encouraged by silly superiors of our generation to do so.
There has always been a conflict between older and younger doctors in medicine, but smart doctors in both camps knew how to navigate it. It was instructive, and fun, to explore the old adage that the two biggest fallacies in medicine are “old is good and new is better”.EBM made it impossible for older doctors to do so and unnecessary for younger doctors to do so.
To read older literate or use the library (interestingly medicine is the one field where Google didn’t see fit to copy old journals), something great doctors did since the 17th Century, is not outmoded. I could kill 50 people and leave the bodies in the stacks of my medical library and no one would discover them for months.
And those things are as important to diagnostic reasoning as EBM. So I don’t know how you are going to accomplish it.
Osler- “He who studies medicine without books sails an uncharted sea, but he who studies medicine without patients does not go to sea at all.”

Here is the good news. The learners that I work with love these conversations and the diagnostic thought process. I am much less skeptical than you about the opportunity for success.

EBM can help us choose treatments much better than it can help us make diagnoses.

How will you plan to improve clinician educators? How do you define clinical educator? Won’t the medical students of today be the clinical educators of tomorrow? Or all the clinical educators come from space or ivory towers of medicine….

I agree that everyone needs to learn … The curriculum does not have to have a start date of medical school, residency, fellowship, clinical educator pathway. it should be something that will put the seed in every mind and some may die the first day but most will evolve from that and not get trapped in fancy words like EBM, patient satisfaction and patient centeredness. Eventually we will again have a critical mass or herd immunity of critical thinkers like you and Cory. I bet there were more thinkers in medicine 30 years ago than now.

I work in a reasonable sized community hospital and sometimes have the prevailed of working with nice younger individuals who teach me a lot of new journal stuff but it kills me when they are obsessed about potassium replacement, GI prophylaxis and dvt prophylaxis in the ICU and they tell me all the data behind it. I try to tell them that the patient did not come to ICU to get dvt prophylaxis. He came for a problem and if we do not focus on diagnosis and management of that, we will keep replacing potassium and giving GI and DVT px.

thank you for though provoked writing in clinical reasoning. I am an mediacl educator in small rual teaching medical school with limited high technology iinvestigation. we need this curriculum very much and urgently. I will follow you till success.


Somewhere around 1995-2000, ICU medicine took a turn for the weird. Because ICUs, with the exception of CCUs, were developed by non-internists, there wasn’t the classical medical concept of diagnosis, management treatment. It was post-op care, trauma care and ventilation.
Very few people appreciated how medical diseases presented in the ICU. About 1980, that started to change. The ICU was actually a place where you had to be able to think as well as act. The pulmonary physicians initially shied away from ICU management but it became financially expedient for them to change their profession to pulmonary and critical care (note the timing of the change of the name of the journal American Review of Respiratory Disease).
For about 10-15 years ICU medicine was a sophisticated subspecialty with difficult patients and problems. Then subspecializaiton set it. No single person is really managing the patient any more -its management by committee. And diffusion of responsibility in sick patients is not a good thing.
Then protocols came in and its management by committee and protocol. Then shift work. Patient gets lasix during the day and fluids at night. Every weaning protocol is different every other day. It’s not the ER where virtually all problems are solved or transferred in 12 hours, but they practice like it is.
And then the emphasis on terminal care came in, so when inexperienced doctors couldn’t make patients better in three days, that was it. A whole specialty of what should have been sophisticated medicine, with new learning curves and analysis of disease and management is pretty much a rarity. There are a few places to my knowledge that still practice it, but I’m betting not many.
Not good for critically ill patients, who don’t get better quickly.

Cory- Thank you for writing an enlightening post because I have wondered a lot about how we got here and this is the first time I have seen a logical reply to some of it. May be one day, you will write a more elaborate article (if you haven’t done already) on the evolution of ICU medicine.

Dr C- any way we can watch or read your talk on the topic of clinical reasoning?

As one of my mentors said, “pain is the biggest liar, syphyllis the biggest imitator”. This particularly applies to musculoskeletal medicine, where we can help patients with various injections and manual therapies, but if we don’t address the underlying biomechanics, these problems tend to return. This seems to me the “art” of medicine that doctors learn by experience, that is lost in the shuffle in the era of EBM. Most attendings in my training received no training in how to teach residents or fellows, so I really began to view my patients as the real teachers, not some academic doc on a power trip. I think we need to acknowledge that most of medical practice isn’t backed up by rigorous science, and embrace the “art” of medicine to problem solve for our patients.

Maybe I will write that article.
Also – by the way- next time, remember to check your “privilege”.

Great post. An important thing to consider is that many academic institutions still do not adequately reward clinicians for teaching excellence. In fact, our current system actively discourages teaching in favor of academic research. Of course many medical schools have teaching awards and some even use these awards to determine promotions. But teaching is still the ugly stepsister to academic research and until we find ways to truly promote and support clinical educators the status quo will remain unchanged.

You are spot on – and we need to advocate for the importance of education in educational institutions.

Post a comment