J C Azzimonti Renzo | PubMed Health

PubMed2003

Failures of common measures of agreement in medicine and the need for a better tool: Feinstein's paradoxes and the dual vision method.

The main objective of the present report is to show failures in binary cases of the kappa, Yule and phi approaches when used for measuring agreement beyond chance. These methods were designed for studying the statistical concept of agreement, and for this reason they may fail when used in clinical situations. To solve these problems, or paradoxes, a new method for studying agreement is used--the dual vision procedure. The capacity of the new method to solve the paradoxes that other approaches cannot solve is demonstrated. The study of agreement is used to decide whether one clinical method can replace another one. An incorrect replacement could be dangerous for the patient. These failures caused by kappa, Yule or phi assessment are illustrated by using the clinical concept of agreement.

Decision Trees↗

PubMed2002

The agreement between two diagnostic methods in binary cases: a proposal.

This report can be considered as a resource in the analysis of agreement among raters, clinical tests, observers, judges or experts. The focus is on diagnostic methods. When the true results of diagnostic methods cannot be obtained, studying the agreement between them can reflect the difference between methods. Normal statistical procedures tackling this are not enough for deciding about the agreement from a clinical point of view. The clinical question is whether the new method agrees sufficiently with the old one. In binary cases, a solution for deciding about the agreement from a clinical viewpoint is introduced. The overall agreement is a duality and needs to be studied in two steps. In the first step, a condition for accepting the agreement is proposed: both methods need to have the same nosologic sensitivity and specificity. In the second step, another condition is proposed: the level of the agreement should be greater than a critical value, defined by the clinicians. When both steps show satisfactory results, the new method can replace the old one. The statistical procedures for testing both steps are presented.

Clinical Laboratory Techniques↗