Misdiagnosis teaching and assessment tool

Ehud Zamir ezamir at UNIMELB.EDU.AU
Sun Jul 28 01:27:18 UTC 2013

I have recently conducted an experiment where real patients volunteered to be examined by senior ophthalmology residents in a mock clinical exam. Several cases were chosen because they had a history of (usually harmless) misdiagnosis, but were presented to the trainees just as they had presented to the doctor, with (erroneous) presumptive diagnoses from previous doctors or with other contextual/cognitive biases. In the vast majority of cases, trainees misdiagnosed them exactly in the same way it had been done by the original doctors, namely stumbled on the cognitive biases or contextual misleading information inherently present in the cases.  I found this a good way to teach about diagnostic error, as it is not just a list of theoretical biases. It is a very authentic simulation. Trainees are in fact led towards making a diagnostic error (similar to the original doctor) and made the error in a controlled environment, then received feedback and "learned the lesson". In addition to a good teaching model about diagnostic errors, I also believe it is a very useful method for assessment of diagnostic skills. A bit difficult logistically (finding patients with a history of misdiagnosis who still have the findings) but certainly feasible. If we value the skill of diagnosis, we have to test it in authentic conditions, and in the presence of proven misleading factors. Lessons learnt from misdiagnosis should be used for training and assessment in a concrete manner.

Ehud Zamir
Centre for Eye Research Australia 

More information about the Test mailing list