Clinical inferences and decisions--I. Diagnosis and Bayes' theorem

Abstract

Clinical decisions are based upon inferences derived from test evidence where data is collected as a means of hypothesis testing. In medicine, the initial expectation of a clinical state is rarely expressed in the classical scientific language of a null hypothesis where alternative outcomes are assumed to be equally probable, because clinical experience reinforces the maxim that common events occur commonly. The collection of test evidence in clinical practice, therefore, is directed towards overturning prior likelihoods for a clinically pathological state which are far from mere chance expectations. The extent to which any test evidence can modify such prior expectations not only depends on its relevance to the clinical state in question (i.e. the hypothesis) but also is largely influenced by the inherent error rates in the test itself. Clinical decision models should reflect these facts. Using examples from the field of ophthalmology, this paper presents a normative model using Bayes' theorem of conditional probabilities which provides a rational framework upon which to base or appraise clinical decisions. Parts II and III of the series will expand this clinical application of decision theory to show how it may be used in the absence of hard test evidence and also where a different emphasis or utility may be placed on false positive or false negative errors.