Learn from other people’s mistakes, you cannot possibly live long enough to make them all your self.

Sam Levenson

I am still smarting from my respected friend LA s comment that Porphyria oh it is so boring, so I decided to take a non-clinical subject this month.  Why are mistakes done in medicine?  Are there bad doctors or are they impaired in a different way?  The basis for this essay are two outstanding articles entitled Five Pitfalls in Decisions About Diagnosis and Prescribing by Jill Klein in the 2 Apr 05 issue of the BMJ, and The Cognitive Psychology of Missed Diagnosis by Don Redelmeier in the 18 Jan 05 issue of the Annals of Internal Medicine.   It goes without saying that we all make mistakes that improvement not punishment is the right way of dealing with this, and that documentation is helpful, although not necessarily related to quality issues

1) Intro- remember the word heuristics. This is a fancy name for strategies and shortcuts that help us sort through the information that we are provided and arrive to the diagnosis.  These strategies help us sift through the irrelevant information, and streamline our thought processes.  In emergency medicine especially we must be very focused and to the point- we have little time with the patients, and must ask the right questions to get the right answers, as well as knowing when to quit the search.  Shortcuts are not a bad thing- indeed they are typically correct.

2) THE REPRESENTATIVENESS HEURISTIC  OR IRENE IS HERE YET AGAIN.

We all have seen the scenario- Irene has been to the emergency department 5 times in the last three days with the usual complaint- electrils going up and down her arm.  She has been under treatment with Risperdal for years, and collects empty plastic bags.    She has diabetes which she for which she does not take any pills. 

The problem: Psych patients complain a lot, but diabetics have MI s.  Which do you give more weight to?

The solution:  Be aware of the disease occurrences and risks before factoring in the additional confounding information.

Yea, I know, you were all brilliant and said that Irene did infarct.  Let us see how you do with your next drug abuser coming in with shortness of breath after a fight with his girlfriend!

3) THE AVAILABILITY HEURISITIC OR IT IS AN AORTIC DISSECTION NIGHT

We often most remember the most complicated or interesting case we saw recently.

That is, Jane saw a lady with neuro symptoms who had porphyria (sorry) and now everyone you see with parasesthias has the same thing in your mind.  The easiest retrieved is not necessarily the correct information.

4) OVERCONFIDENCEOR EVERYONE KNOWS I AM G-D S GIFT TO MEDICINE

Do not need to say much here- we are all overconfident according to the studies.  I say doubt yourself, try to punch holes in your reasoning, and do not ignore little added details that make the whole hypothesis collapse.  Then make your patients aware, and if possible part of the reasoning process.

Example:  I often tell patients- look you are very constipated now, and that is not to say that you do not have an underlying process, but with all this constipation, it confuses the picture.  Here is some laxative, and go to your doc tomorrow to be reevaluated.

5) CONFIRMATORY BIAS, OR YOU HAVE THE DISEASE I WANT YOU TO HAVE

Here we read the nurses note or the referring doctor s note and automatically put the person in a specific category.  This will affect what questions you ask, and how you interpret the information you are given. I would not say always to start again, because in the emergency department that is not always practical, but keep your ears peeled for pieces of information that do not fit.  This is also known as anchoring bias, and reflects the stuck in gear philosophy

Example: The ankle has no swelling therefore the X ray is normal.  I would suggest you look at the x ray first, then the ankle than the x ray again

6) ILLUSORY CORRELATIONS OR THE PATIENT GOT BETTER BEFORE MY EYES

This is the tendency to relate two events without considering if this is a coincidence or two non-related facts.  Examples include the improvement before the storm seen in epidurals and in iron poisoning.  Voltaren may have helped his pain in the back, but did nothing for his aortic aneurysm.  

7) FRAMING BIAS, OR YOU CAN NOT BE PREGNANT BECAUSE YOU SAY YOU ARE NOT

The way it is presented makes a difference, and is not that different than confirmatory bias.  For example if you present the chances of dying at 10% people will panic.  If you say it is a 90% survival, people breath easier.  Same with us- fever shortness of breath and cough makes PE leave our mind.  Be a devils advocate and thing that perhaps you are not hearing the full story.

8) BLIND OBEDIENCE OR IT IS MY WAY OR THE HIGHWAY

Here the attending told you what to do, or a consultant or a better doctor.    Think how ridiculous this would sound in a court of law, and what a star you will be for thinking of something else.  And guess what?  It makes us all better doctors

I would add just two other factors for EM- one is always, and I mean always think of the worse thing this could be and make sure you rule it out.  Also, remember the exhaustion factor- if it is the end of the day, do not take simple solutions- tough problems come in the end of the day too.

Many EM programs put in as the last line in the chart Medical Decision making and explain what they thought.  Maybe you should too!

It is OK to make mistakes, just do not become a legend at it!

Peter Viciellio, MD FACEP