Clinical decision making part II: Why diagnostic procedures vary in performance

As outlined in Part I of this two-part series, the consequences arising from Type 1 and Type 2 errors in clinical decision making may be serious. Thus, it is helpful to consider what is the predictive value of a positive result, and also a negative result, regardless of whether the method of diagnosis is clinically based (signs and symptoms) or laboratory based. The positive predictive value (PPV) and negative predictive value (NPV) of a diagnostic method can vary depending on the background prevalence of the condition. This overview of probability, errors and consequences in clinical decision making also briefly introduces Bayes' Theorem, which is the mathematical basis for evaluating the "conditional probability" of an event, e.g what is the probability that the individual has the disease given that they have tested positive (PPV)? Similarly, what is the probability that the individual does not have the disease given that they have tested negative (NPV)? The theorem is a useful tool for physicians and researchers wishing to study the process and performance of medical diagnoses and related decisions.