'Immunising' physicians against availability bias in diagnostic reasoning: a randomised controlled experiment.
There is uncertainty about the effectiveness of cognitive debiasing in reducing bias that can contribute to diagnostic error. Instead of focusing on the process of reasoning, this study examined whether an intervention directed at refining knowledge of a cluster of related disease can ‘immunize’ physicians against bias. Ninety-one internal medicine residents in Brazil were randomized to one of two sets of vignettes (reflecting diseases associated with either chronic diarrhea or jaundice) and compared/contrasted alternative diagnoses. After residents encountered one case of a disease, non-immunized residents twice as likely to give that incorrect diagnosis to a different (but similar) disease, resulting in a 40% decrease in diagnostic accuracy between immunized and non-immunized physicians.