Thursday, August 15, 2013

Communication Breakdown: Zebras, not Unicorns

In my last post I discussed the role of presentations in determining how a student is perceived by their evaluators.  Today I want to hone in on one particular part of the presentation:  the Differential Diagnosis.

The usual advice given to students on how to construct an "EM appropriate" differential is to go "worst first".  This means that the first diseases you discuss are those most likely to result in morbidity and mortality for the patient.  This is important in the ED as this is the place where all the sickest patients get funneled to.  In most out-patient settings acutely ill patients are a terrifying rarity.  In the ED they are the routine.  If you do not think of a life-threatening diagnosis, your chances of lucking into it are low.  "Worst first" is a safety net for us and our patients.

Every student has probably heard the truism, "When you hear hoof beats, think horses, not zebras."  The problem with this clever turn of phrase is that zebras are real.  If you have ever been to a zoo, or the Serengeti, you have seen them.  Positioned as we are, at the bottom of the medical pathology funnel, the ED is the Serengeti of medicine:  the place where zebras roam.  Or at least the zoo of medicine.

Not all uncommon diseases (zebras) are dangerous.  The uncommon and bad ones, like aortic dissection (~200 times less common than MI) are the ones we must be vigilant against and belong on our differentials when we "hear hoofbeats" (i.e. "chest pain").  Zebras that are not particularly dangerous do not need to appear on your differentials.

There are still diagnoses we should ignore, diseases that do not belong in an ED differential.  These are the ones that are neither common nor dangerous.  These are the ones we will only find in text books.  For EM, the truism should be changed to, "When you hear hoofbeats, do not think of Unicorns".

Once a student is able to reliably think of the dangerous zebras, the life threats, I encourage them to order their list by probability.  When first getting used to EM it is okay for your differential for a 25 year old man with pleuritic, reproducible chest pain and normal vitals to be "AMI, dissection, PE, pneumothorax, and espophageal rupture".  You have included the bad stuff.  But you have not given weight to prevalence of disease or your own clinical impression.  This is not what your supervisors are actually doing.

Instead, once you are reliably thinking of all the bad diagnoses, you can start presenting your differential the way the residents do.  Start with what you think the patient actually has, and why.  And that can be a common and less dangerous diagnosis.  You need to follow that "horse" with the less likely but more concerning differentials (the zebras), why you do not think the patient has them, and how far you want to go in tracking them down.   You can even tack a unicorn on at the end, so long as you do not want to actually pursue that mythical beast.

Happy hunting!

No comments:

Post a Comment