Abstract
Taenia solium cysticercosis is an endemic zoonosis in many developing countries. Serological tests are the most appropriate diagnostic tools to understand the transmission dynamics of the parasite, but the performances of these methods in such a setting are not known. A south Ecuadorian human population living in an endemic area was tested using three common serological tests. Because none of them is a gold standard, a Bayesian Latent Class analysis was used to estimate the test characteristics. Two definitions of a case were considered to differentiate between prevalence of current infection and prior exposure to the parasite. Differences between the performances of the same test in function of the definition of a case were observed. This study shows that test results and prior information should be interpreted carefully in a Bayesian analysis framework, particularly when the latter is based on clinical studies.