Résumé :
|
[BDSP. Notice produite par INIST HJGGR0xa. Diffusion soumise à autorisation]. Measures of validity of diagnostic procedures are commonly employed that quantify the ability of the procedures to distinguish individuals with and without a certain disease. These measures include, among others, the sensitivity and specificity, the likelihood ratios, and the positive and negative predictive value. In many clinical situations, however, the primary purpose of diagnostic procedures is to distinguish between two or more alternative diagnoses in addition to or rather than confirming or discarding a single diagnosis. This article presents a variety of measures that may be used to quantify the ability of a diagnostic procedure to distinguish between two among several alternative diagnoses. The measures are introduced algebraically, and their use is illustrated with hypothetical and empirical examples.
|