Review Article
José Antonio Rold&aacut
Abstract
Sensitivity and specificity are classic parameters to assess and to compare the precision of binary diagnostic tests in relation to a gold standard. Another parameter to assess and to compare the performance of binary diagnostic tests is the weighted kappa coefficient, which is a measure of the beyond-chance agreement between the binary diagnostic test and the gold standard, and it is a function of the sensitivity and the specificity of the diagnostic test, the disease prevalence and the relative loss between the false positives and the false negatives. In this study, we carry out a review of the weighted kappa coefficient, its estimation for a single diagnostic test and the hypothesis tests to compare the weighted kappa coefficients of two or more diagnostic tests, both when the gold standard is applied to all of the subjects in a random sample and when the gold standard is only applied to a subset of subjects in a random sample. The results were applied to different examples.