RÉSUMÉ
In this paper we evaluate serology as a tool to monitor Trichinella-free pig herds. Indoor, industrial-raised fattening pigs in the Netherlands are practically Trichinella-free, and were used as a negative reference cohort. A positive cohort was not available but we used sera from an endemic region in Argentina to model a plausible distribution of serological responses (as OD levels) in positive sera, employing the difference between the endemic sera and the negative Dutch sera. We describe a method for correcting for variation among ELISA plates using on-plate reference sera, and demonstrate how to apply these corrections to a collection of test sera from pig farms. The positive and negative reference distributions can be used to estimate fractions true and false positives, necessary for defining appropriate cutoffs to be used for classifying positive and negative animals. Based on this analysis, the serological test was shown to lack the predictive power required for its large scale deployment. The properties of the serological test were also compared to the conventional digestion assay, which is highly specific but considerably less sensitive.