Monthly Archives: March 2013

The art of prognostication

One night in the ICU, I told two loving sisters to a 43 year old sickly cancer patient that their little sister was doing well. As the patient’s admitting doctor, the sisters were banking their hopes on me. They were trying to ask me probing questions, but fumbling. Then one of the sisters managed to spew out her fear. “Okay doctor, how is she doing? Is she going to be okay tonight? You can be honest with us”. 

I could see what the harried sister was not  able to put a voice to. I thought I would cut through the sisters’ fears and help them out with an honest answer.

“Are you asking me if she is going to die tonight?” I volunteered.

“Yes”, they nodded in affirmation, probably relieved that I could see through the veiled question.

“No she is not”, I said. “Thank god”, the sisters resounded together in great relief.

Immediately after, a sickening feeling ran through my stomach. I looked around. The nurse who was by the bedside gave me a hidden look of incredulity. I can’t believe you just said that, his eyes seemed to scream. The intern who was working with me walked out of the room. Later I learnt, he had just told the sisters he wasn’t too sure of the prognosis.  

It felt like a weird contortion of the Murphy’s law: the one time you say something isn’t going to happen for sure, it is going to happen. Was my patient going to die just because I had predicted confidently that she was not going to die?  

I had my reasons to feel confident about the patient. A few hours earlier in the ER, she had been in a rather bad shape. She was not breathing well, and appeared to be anxious and confused. We got her upstairs to the ICU, put  her on a breathing machine and she started doing much better- she calmed down, her blood pressure was holding up well. The reversal was very visible. I felt happy.

Calmed by what was to be my ignominious hubris, the sisters decided to leave the hospital and go home. An hour later, I had to call them to say that their sister wasn’t doing well. An hour more or so later, I had to examine the patient with the the two sisters and the rest of the other siblings  by the bedside, and declare her dead.

Talk about eating your own words.

Such is the treachery of the art and science of prognosticating patient outcomes. In order to deal with the complexity of such an arcane endeavor, doctors often resort to defensemanship: an art of ambivalating, avoiding sticking the neck out  and commiting oneself to an answer.  That way you may not really be right about your answer, but then you are never really wrong either. Sadly, the common understanding is that this is the best strategy for the doctor. There is hardly any incentive for the doctor to rightfully predict patient outcomes, whereas there may be a huge amount of liability associated with getting your prognostication wrong.  

Despite all the numbers that you memorise from texts, generous experience and a dollop of common sense, it isn’t very difficult to make an idiot out of yourself trying to predict how a particular patient is going to fare in the immediate or the long term. You sure have tools to calculate average outcomes, but they are just that, at best a rough shot at predicting outcomes averaged over a population. There may be little corelation with the individual patient at hand.

Then how do you prognosticate immediate outcome on a ciritcally ill patient at an acute care setting. Is it even a wise idea to prognosticate outcomes in a critical care setting? Is the standard, this is the ICU, so anything can happen answer still the best we can come up with?

The reason I stuck my neck out and volunteered a definitive answer  with this patient was two fold: first, the patient was really doing better and there was little reason to believe based on clinical experiene as well as the available data that this patient was going to die in the next 12 hours.  Second, always saying I can’t really predict the outcome almost felt dishonest, when I know majority of the times a patient like this is not going to die in a situation like this. So if a more straight forward and honest answer was going to ease the sisters’ anxiety greatly, why not volunteer with an answer?  

It is easy to rationalise this experience as a vital element of learning. Such experiences however, are double edged swords I believe. True, in the future, before I venture out with a definitive answer, I will swallow my words three times over and probably chicken out with an “anything can happen”. I will not get my answer wrong, and thus avoid the ignominy of it, but far more often than that, I will fail to get it right, and thus lose out on a chance to provide some solace to the patient and the family and provide a positive approach around a practise of medicine that is increasingly being usurped by the self defeating black art of the negative and the defensive.

The sisters were understandably shaken, but then they were immensely thankful for the vigil that we’d provided and the care that we had given through their sister’s final hours. They did not hold me up for venturing with an answer- they knew the impossibility of mastering this art. Much like I was able to decipher the intentions of their words, the sisters were able to decipher the intentions of my words- they were grateful that I had volunteered to say them aloud.