Clinical decision analysis (CDA) is a quantitative strategy for making clinical decisions. The techniques of CDA are largely derived from the theory of signal detection, which is concerned with extracting signals from noise. The theory of signal detection has been used to study the detection of auditory signals by the human observer. This article provides a brief overview of CDA. More comprehensive discussions of CDA and its application to audiological tests can be found in Turner, Robinette, and Bauch (1999), Robinette (1994), and Hyde, Davidson, and Alberti (1990).

Assume that we are using a clinical test to distinguish between two conditions, such as disease versus no disease or hearing loss versus normal hearing. Most tests would produce a range of scores for each condition and therefore could be thought to contain some “noise” in their results. More important, there may be an overlap in the scores produced by a test for each condition, creating the potential for error. That is, a particular score could, with finite probability, be produced by either condition. CDA is well-suited to deal with this type of problem.

For the discussions and examples in this article, we will assume that we are trying to identify hearing loss. Of course, CDA can be used with diagnostic tests that are designed to identify a variety of diseases and conditions.