Abstract:
|
Although the estimations of the sensitivity and specificity of an IVD (in-vitro diagnostic) test are based on a simple 2 × 2 table, difference aspects of testing algorithm (e.g., retest, gray zone, pool testing, multiplex) could lead to complexity in the study design and analysis. Complexity due to retest and gray zone (GZ) are discussed. In general, the gray zone is defined as Index value or signal over the cutoff value being in a specified range. Although gray zone is often used in the study, the final interpretation of the assay may or may not involve a GZ. In this paper, we will discuss two examples where a GZ is used for two different purposes. In the first example, the GZ is used so that a retest may be done in the study with the purpose to refine the cutoff value. In the second example, the GZ is simply the retest zone. Although refining the cutoff value based on the current data is against the basic principle, there should be a balance between practicality and basic principle. Therefore, further research is needed to assess the bias and to see how such bias may be adjusted.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.