ABSTRACT
The notation P(A|B) is math shorthand read aloud “Probability of A given B,” which means the probability of observing event A given that we have also observed event B. If A and B have nothing to do with each other, then P(A|B) is equal to P(A), the overall probability of event A. Even an analysis of the entire patient population seen by a health system is conditional on the patient belonging to the health system. The p-value is a conditional probability, where A is the chance of observing a test statistic of a certain value or greater, and B is the condition where the null hypothesis is true and the results are just due to random variations. We discuss how Bayes’ theorem describes conditional probabilities and relates sensitivity to the positive predictive value.
