A perfect diagnostic test (with 100% sensitivity and specificity) is for the most part a theoretical concept. In practice, there are several diseases (e.g. tuberculosis, pneumonia, Alzheimer's disease) for which there is no perfect test that can detect the presence of the disease with certainty. This complicates estimation of disease prevalence and evaluation of new diagnostic tests.
Latent class models provide a solution for this problem by modeling the imperfect accuracy of all tests involved. My research program has focused on different aspects of applying latent class models - adjusting for conditional dependence, meta-analysis and sample size calculations. Some key publications are listed below: