Right now, though, I'm watching a TED talk about physicians' mistakes by Brian Goldman. As ever, Dr Goldman delivers a good talk, and in particular reveals one of his own mistakes.
I enjoyed the talk, but I must say that I have a very different approach to my own learning and clinical decision-making. Dr Goldman says that he always strove to learn "everything" and memorize as much as possible so he'd make the right diagnoses and plans.
My way of going about learning and practice is quite different - I am concerned mainly with knowing enough to make good decisions and avoid bad ones. The further I get into training, the more I become aware of the limits of my own knowledge and the necessity with not becoming too comfortable with what I do know. I don't think it's really true that medical knowledge changes with quite the speed that is often described - there really is no substitute for a good history and physical examination, and - crucially - taking the time to actually do one, along with looking for additional history (e.g. old charts) if available. As one of the ICU fellows was telling me last week, being a good physician is not so much about knowing the most, but rather taking the time to be thorough and careful. Good decisions come from good information and the simple question "what else could this be?".
But you can still make a call that turned out to the wrong decision in retrospect. We can't know everything that will happen and many, many conditions do not "declare themselves" earlier or clearly enough that they cannot ever be missed. The important thing is to make decisions that address the worst case scenario while not overcalling the situation so much that you subject a patient to ultimately unnecessary tests or interventions. That's not always an easy balance to make, and you won't be immune from "good faith" mistakes, but it's the safe approach.