The role of specific disease knowledge in counteracting bias in diagnostic reasoning

Diagnostic errors have been attributed to flaws in physicians’ reasoning associated with the use of heuristics. Experimental studies have provided evidence that the use of heuristics may eventually induce bias in reasoning, causing physicians to make diagnostic errors. However, this research suggests that among physicians at the same level of training some seem to be more susceptible to bias than others. The sources of these differences in ability to overcoming the influence of bias are unclear. The present study examines whether specific disease knowledge predicts susceptibility to bias in diagnostic reasoning. Physicians will diagnose the same set of clinical cases both under conditions that make them subjected to bias and not subjected to bias in a counterbalanced design. Physicians' knowledge of critical diagnostic features for the diseases presented in the cases will be evaluated and it will be analyzed whether differences in knowledge predict the frequency of bias-induced errors.

Overall, diagnostic accuracy will be lower on subjected-to-bias than on not-subjected-to-bias cases, but the difference in accuracy will be smaller in the group of physicians with higher knowledge of critical diagnostic features.
Overall, the frequency of the wrong diagnoses induced by the bias will be higher on subjected-to-bias than on not-subjected-to-bias cases but lower in the higher-knowledge than in the lower-knowledge group.
Overall, physicians will spend more time to diagnose subjected-to-bias than not-subjected-to-bias cases.
Overall, the rating of confidence in the diagnosis will be lower on subjected-to-bias than on not-subjected-to-bias cases, but the difference will be lower in the higher-knowledge group.
Overall, the diagnosis triggered by the bias will be mentioned more frequently when physician recall which diagnoses they had considered on subjected-to-bias than on not-subjected-to-bias cases, without differences between the two knowledge groups.

Outcomes measured in the knowledge evaluation phase and in the diagnostic performance phase, with 4-48 hours in between.

Each clinical case will be manipulated to become either subjected to bias or not subjected to bias. Bias will be induced by adding to the case salient distracting features, findings that though irrelevant to the case tend to catch physicians' attention because they are strongly associated with a particular disease that seems at first glance a plausible diagnosis.