Comparison of two serum and bulk-tank milk ELISAs for diagnosing natural (sub)clinical Dictyocaulus viviparus infection in dairy cows

Lungworm antibody ELISAs developed in Germany (DE) and The Netherlands (NL) were compared using four sets of serum (S) and bulk-tank milk (BTM) samples from adult dairy cows. The samples originated from 37 farms with or without a suspected clinical lungworm infection during August–October 2010 (Dataset 1), from cows excreting lungworm larvae or not during August–October 2010 (n = 59) or May–June 2011 (n = 164) (Dataset 2), from 305 farms in a national survey during October 2010 (Dataset 3), and 14 zero-grazing farms during February–April 2011 (Dataset 4). During August–October 2010, covering the second half of the grazing season, the NL-S and NL-BTM ELISA outperformed the DE-S and DE-BTM ELISAs in terms of sensitivity. For at least the NL-S and DE-S ELISA the opposite was found during May–June 2011, covering the end of the winter housing period and the early grazing season. Of the 305 farms in the survey 62.6% were found positive with the NL-BTM ELISA, whereas 2.6% was found positive with the DE-BTM ELISA. ODR values for the zero-grazing farms indicated that a cut-off value of 30% for the DE-BTM ELISA might be more appropriate than the currently used 41%. Results suggest that the NL ELISAs also respond to lungworm antigens other than Major Sperm Protein as used in the DE ELISAs. It is concluded that the generally higher sensitivity of the NL-BTM ELISA makes it better suited for large-scale prevalence studies and herd health monitoring programmes than the DE-BTM ELISA, positivity of which is more associated with the presence of clinical lungworm infection. Reducing the cut-off value of the DE-BTM ELISA from the original 49.3% to the current 41% or the possibly more appropriate 30% increased its sensitivity for detecting lungworm infections, but did not lead to similar sensitivity estimates as found for the NL-BTM ELISA