Decision analysis in the evaluation of diagnostic tests.

Abstract

Decision analysis was used to evaluate the clinical usefulness of 2 diagnostic tests: one for canine heartworm disease and the other for bovine traumatic reticulo-peritonitis. Several clinically relevant measures of test performance are introduced, including expected utility, risk profile, testing band, threshold analysis, and the relative cost of misdiagnosis. One of the principal benefits of decision analysis of diagnostic tests is that the technique can be used to determine how changes in underlying assumptions will affect clinical decisions. If clinicians can identify and assign values to relevant variables, then decision analysis can provide clinically meaningful guidelines for interpreting the results of diagnostic tests. To take advantage of these techniques, clinicians must become comfortable with quantitative expressions for test performance, risk, and prognosis.