Short communication: Performance of a blocking antibody ELISA bulk-tank milk test for detection of dairy sheep flocks exposed to border disease virus

The aim of this study was to investigate the test characteristics of a blocking antibody ELISA applied to bulk-tank milk (BTM) samples for the detection of dairy sheep flocks positive for antibodies to border disease virus. In 161 flocks recruited in 2009 and 2010, the antibody inhibition percentage (AIP) in BTM was compared with the prevalence estimate of antibody-positive ewes obtained from an age-representative sample of 45 milking ewes. A strong negative exponential relationship (R2 = 0.89) was found between AIP in BTM and seroprevalence level. Using receiver operating characteristic analysis, the best AIP decision threshold in BTM to discriminate between low (<10%) and high (≥10%) antibody-positive flocks was 65%. Diagnostic performance estimates based on observed seroprevalence levels and Monte Carlo simulations showed that this threshold value was associated with high sensitivity and specificity (91.9 ± 5.5% and 95.9 ± 1.6%, respectively), whereas the 80% decision threshold recommended in dairy cows yielded lower specificity (83.6 ± 2.0%). Results obtained from the same flocks during 2 subsequent milking campaigns showed that the 65% AIP cut-off value was associated with fewer false-positive results and is preferred. Testing of BTM samples could be a powerful tool in inferring border disease virus seroprevalence in a flock and in Pestivirus control schemes in dairy sheep flocks.