CMStatistics 2021: Start Registration
View Submission - CMStatistics
B1389
Title: A one-sided testing procedure for comparing predictive values of two diagnostic tests simultaneously Authors:  Kanae Takahashi - Hyogo Medical University (Japan) [presenting]
Kouji Yamamoto - Yokohama City University (Japan)
Abstract: Diagnostic tests are important for the early detection and treatment of disease in medicine. The positive predictive value (PPV) and the negative predictive value (NPV) describe how good the test is at predicting abnormality, and these are used for quantifying the diagnostic ability of the test. The PPV is the probability of disease when the diagnostic test result is positive, and the NPV is the probability of no disease when the diagnostic test result is negative. There are several methods to compare each of the PPVs and NPVs in paired designs, and a global two-sided testing procedure to simultaneously test the equality of PPVs and NPVs has already been proposed. However, in some cases, it may be necessary to evaluate the superiority of at least one of PPVs and NPVs. We propose a one-sided testing procedure that assesses the superiority of at least one of PPVs and NPVs by applying an approximate likelihood ratio test proposed.