Pyramid of Likelihood (Bayesian Reasoning)
Definition
The probability of a diagnosis depends not only on how well the presentation fits but also (and perhaps more importantly) on the prior probability (pre-test probability) of that disease.
How to Apply
Common conditions are common and rare conditions remain rare. Imagine a pyramid – at the bottom you have a common presentation of a common condition and at the very top – an atypical presentation of a rare condition. Often, as clinicians, we get stuck in the middle part – an atypical presentation of a common condition is still statistically more likely than a common presentation of an atypical condition. For example, an atypical presentation of a pneumonia is more likely than a common sounding presentation of a PE.
Pre-test probability is very important, think before I even examine the patient – how likely is this disease? This is why age, sex, epidemiology & risk factors are the starting point of the problem representation framework!
Never let an atypical presentation tempt you into overthinking or over-diagnosing a rare disease simply because it fits the pattern better.
The Acute Mind 