Test Selection Measures

Abstract

Usually, when a patient is being investigated, only a small subset of all available tests is performed. Most selection methods for making this choice fail to account for the risk and cost of the test. By attempting to approximate a decision-analytic ideal, via the concept of quasi-utility, the authors developed the information-to-cost ratio and related measures, which balance the utility of the information gained against the price paid. Measurement of the relative importances of diseases is used to further refine the method. Key words: test se lection; diagnosis; decision analysis. (Med Decis Making 1989;9:133-141)