Figure 1:

Visual representation of probabilities and thresholds used in Bayesian reasoning.

Panel (A) shows colored sections representing areas of probability with associated actions based on thresholds set by a provider. Each color next to the Fagan nomogram represents next steps based on thresholds set by the user. Green is below the test threshold, and no further testing is needed. Yellow is between the test and treat thresholds, and the clinician should consider getting further information via history, exam, or testing to move the probability into the red or green. The red is above the treat threshold, and the user should consider whatever management option they deem appropriate. The solid red line is an example case in which the pre-test probability is 50%, and a test with a LR of 10 is obtained, resulting in a post-test probability of 90%. Manipulation of this visual representation allows learners to see a map of how information affects their Bayesian reasoning. Panel (B) shows how groups of clinicians can recommend different actions, even when given the same information. For example, even if all of these clinicians agreed on a pretest probability of 70% (clinician 2’s probability), they still disagree on next steps. Clinician 1 would recommend treatment, clinician 2 might be considering treatment but still unsure, and clinician 3 would want more information. Complicating things further, most clinicians do not share exactly the same pre-test probability (as illustrated), which can lead to even more confusion. Helping learners understand these concepts and how to verbalize them can facilitate deeper learning of the diagnostic process.

© De Gruyter