1997 | OriginalPaper | Buchkapitel
Modeling Correlations Between Diagnostic Tests in Efficacy Studies with an Imperfect Reference Test
verfasst von : Yinsheng Qu, Alula Hagdu
Erschienen in: Modelling Longitudinal and Spatially Correlated Data
Verlag: Springer New York
Enthalten in: Professional Book Archive
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
In the Chlamydia trachomatis (CT) study, we compare five screening tests in terms of sensitivity and specificity for detecting CT in endocervical specimens. The five tests are correlated, and the reference test (cell culture test) is subject to error. The conventional method which ignores both the correlations between the tests and the misclassifications of the reference test cannot provide valid analysis. We propose a model by which one can evaluate and compare the sensitivities and specificities for correlated diagnostic tests with an imperfect reference test, or even without any reference test. We use a hybrid algorithm, which consists of the expectation maximization (EM) algorithm and the Newton-Raphson method, for its maximum likelihood estimation.