RESUMO
In this study 595 lactating cows originating from 31 carinthian farms were investigated in accordance of liver fluke infection using individual and tank milk as well as individual blood and faecal samples. Two commercial ELISAs were used to test the milk and blood serum, and the results were compared with coproscopy and a commercial copro-antigen ELISA. Receiver operating characteristics (ROC) and two-graph operating characteristics (TG ROC) of tank milk results were conducted based on the individual milk to determine the minimum reliable in-herd antibody prevalence for the predominant condition in the investigation area. In 17.8% of the examined individuals located in 64.5% of the farms eggs were detected by coproscopy. The copro-antigen ELISA delivered 13.4% positive individuals from 54.8% of the farms. The milk ELISAs showed 42.7% (Euroclone) and 44.2% (Pourquier) positive cows on 90.3% of the farms. The blood samples were positive in 43% (Euroclone) and 45.2% (Pourquier) of the individuals from 90.3% to 96.8% of the herds, respectively. Based on the milk and the blood an average in-herd prevalence of 30-45% can be assumed. The serum and milk samples delivered correlating results with kappa values between 0.94 and 0.97, whereas the coproscopy and copro-antigen ELISA did not correlate well with the ELISA results. The two different ELISA tests highly correlated on individual and on herd level. Both showed a reliable minimum in-herd prevalence of â¼20%, meaning that one fifth of the individuals in a herd have to be positive to obtain a positive bulk tank milk result. In the investigated area a higher in-herd prevalence is expected, therefore the tank milk is useful as a monitoring tool and can be used as a basis for intervention strategies.