[Considerations concerning diagnostic certainties and cut-off values for a bulk milk ELISA for Mycobacterium avium ssp. paratuberculosis]

Abstract

Serological tests for the examination of individual samples from single animals are evaluated based on their ability to detect true positives above a defined threshold value. If results are obtained not from an individual but from a bulk sample this concept usually is adopted such that the threshold is set to allow the detection of a single positive sample within the pool. In conjunction with the development of a diagnostic paratuberculosis ELISA for the examination of bulk milk samples it is discussed which interpolations of this concept are justified when defining the true status of a herd based on the test parameters and the seroprevalence within the herd. Here, bulk milk from up to 50 animals each and the corresponding individual samples of 4241 dairy cows from 28 herds in the state of Brandenburg are investigated, and results are subjected to different evaluation approaches. Based on epidemiological considerations and test parameters a "critical prevalence" is defined which then serves as basis for the deduction of a cut-off value to be used for bulk milk samples. Finally, the practical relevance of this approach is demonstrated by suggesting an initial scheme for paratuberculosis classification of dairy herds with respect to possible control measures.