Right to be wrong: being comfortable with uncertainty



There are occasions when having the confidence to be wrong, and to be seen to be wrong, is advantageous to the patient.

Two cases:

I once saw a patient with signs of liver failure, but there was something about her that didn’t make sense. No risk factors, no alcohol. She had too much fluid in the body (‘overloaded’), and the veins in her neck were distended, a classic sign of heart failure. So, I thought, the liver signs could be secondary to the heart, and what she really needs is a cardiac assessment. I requested that assessment, and gave my reasons…but the cardiologist wasn’t convinced it was her heart either. The story was all wrong, she had had a healthy heart scan within the last 6 months, and there were no risk factors! We talked, we went to and fro, and in order to ensure that an urgent echocardiogram was performed, I said,

‘Well I’m pretty certain it’s her heart, there is no history of liver disease, and at the moment heart failure is the most likely diagnosis…’

Thus I nailed my colours to the mast, and backed it up by writing unequivocally in the notes something along the lines of, ‘…presentation most consistent with heart failure, needs urgent investigation…’ It was a form of brinkmanship, an unsubtle distribution of responsibility, such that my colleague felt compelled to attend urgently in order to check whether this patient’s problem did indeed stem from heart failure.

The cardiologist attended an hour later, while I was still on the ward. He called me into the patient’s bed space, behind the curtains, and showed me the images. The heart was beating perfectly normally. With equal emphasis, he wrote something along the lines of, ‘…valves and chambers normal, ejection fraction 60-65%, unchanged from __ /20__.’ We discussed the details, and agreed that in fact what we were probably looking at was a case of fluid overload due to kidney failure. I went back to the notes, looked at my big, bold words on the page, and thought…’You were completely, absolutely, wrong!’ But…but…by being bold, by taking a position, I achieved what was required – an urgent heart scan and a narrowing down of the differential diagnosis.

Another example. A patient with unexplained chest pain is admitted through the ED. It isn’t a heart attack, it’s not an ulcer, it’s not a pulmonary embolism. The thought develops that it might be a thoracic aneurysm; potentially lethal, very hard to rule out. Do I really think it is? No, I’m not convinced, the pain is already settling, perhaps it is a bit of acid reflux, but I feel uncomfortable. She said it felt like a tearing sensation. She needs a scan. It’s late. I’m going to have to make a good case to the radiologist.

‘Hi, it’s Dr______. I’ve got a patient with chest pain, severe, and we’ve ruled most things out, but I’d like to exclude a dissecting thoracic aneurysm…no, the chest x-ray looks normal but I can’t rely on that…the blood pressures in both arms are equal…’

I’m not doing well. How badly do I actually want this scan? If I’m not convinced (just concerned), so why push it? But if she is dissecting, and we miss it, she’s as good as dead.

‘…I just can’t sit on her all night without ruling it out. We don’t scan every chest pain, but the way she described it…it’s the only remaining diagnosis I can think of.’ The scan is agreed to. To back up my verbal conviction I write in the notes, ‘…plan: exclude dissection.’ and I wait for the result. It is negative. Of course it is negative. Perhaps I used up a bit of credit with the radiologist, perhaps next time they won’t take my request quite so seriously, but now, this evening, I achieved what I felt I needed to achieve.

Being wrong is nothing to be proud of, obviously. But what I describe here is a willingness to make a diagnosis and push hard for the investigations that are required to prove or disprove them. The process of forming a list of differential diagnoses and eliminating all but one (the actual diagnosis) will by its very nature involve barking up several wrong trees. Being wrong is therefore a necessary corollary to discovering the right answer. However, being wrong is not something that medical students or medical trainees are very comfortable with. Our training is focussed, entirely, on retrieving facts with accuracy and providing the ‘right answer’ immediately. Then, once we enter the real world, we encounter uncertainty, a whole series of possibilities stemming from each and every clinical encounter. It is physically impossible to pick out the right explanation for each collection of symptoms and signs. To make progress we must make a stab at the problem, have a go, test a series of theories, and, one by one, eliminate those that are wrong. If the discomfort that trainees feel in the face of uncertainty results in a form of paralysis, and the ‘differential diagnosis’ section at the end of the clerking is left blank, progress cannot be made. Progress requires an acceptance that medicine is uncertainty, a willingness to bark up those trees, and the maturity to absorb any sense of embarrassment that arises when the someone shouts down from the branches, ‘Wrong one!’


Some free resources on uncertainty in medicine:
1) Clinical uncertainty- Helping our learners by Dale Guenter, Nancy Fowler and Linda Lee (Canadian Family Physician)


2) Tolerance of Uncertainty and Fears of Making Mistakes Among Fifth-year Medical Students by Maarit Nevalainen, Liisa Kuikka, Lena Sjöberg, Johan Eriksson and Kaisu Pitkälä (Family Medicine)


3) The value of medical uncertainty? By Caroline Welbery (Lancet) – On the role of art


4) Uncertainty Is Hard for Doctors by Danielle Offri (NEJM)



cover to tweetSPOKENcreatspaceCOVER


Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s