Estimating Neospora caninum prevalence in wildlife populations using Bayesian inference.

摘要: Prevalence of disease in wildlife populations, which is necessary for developing models and conducting epidemiologic analyses, often understudied. Laboratory tests used to screen diseases populations are validated only domestic animals. Consequently, the use these may lead inaccurate estimates prevalence. We demonstrate Bayesian latent class analysis (LCA) determining specificity sensitivity a competitive enzyme-linked immunosorbent assay (cELISA; VMRD (®), Inc.) serologic test identify exposure Neospora caninum (hereafter N. caninum) three southeastern Ohio, USA. True prevalence N. caninum was estimated range from 0.1% 3.1% American bison (Bison bison), 51.0% 53.8% Pere David's deer (Elaphurus davidianus), 40.0% 45.9% white-tailed (Odocoileus virginianus). The accuracy cELISA be close 96% 99% reported by manufacturer. Sensitivity deer, however, ranged 78.9% 99.9%. Apparent results not equal true populations. Even when species inhabit same community, two differed population. Variances some suggest differences epidemiology colocated LCA methods could as this example overcome constraints on validating species. ability accurately evaluate status population improves our understanding multihost pathogen systems at community level.