Performance of a commercial serum ELISA for the detection of antibodies to Neospora caninum in whole and skim milk samples

Abstract

Control of Neospora caninum infection in cattle depends on specific, ante-mortem detection of infected animals and limiting their use as breeding stock or by culling. The objective of the present study was to determine appropriate cut-off values and diagnostic performance of a milk ELISA test using whole and skim milk in a commercial serum ELISA test (IDEXX Neospora Ab). Serum and milk samples were obtained from a total of 475 lactating cows from two herds with and two herds without a previous history of N. caninum-associated abortion. Overall seroprevalence determined by the ELISA was 18.3%. Compared to serum ELISA values, correlation and overall performance assessed by receiver operating characteristic analysis was higher when either whole or skim milk samples were diluted 1:2 compared to undiluted or 1:5 diluted samples. Diagnostic performance for analysis of whole and skim milk was compared at cut-off values that achieved a desired operating characteristic of at least 95% specificity. For whole milk diluted 1:2 and a cut-off of 0.14 (S/P ratio), sensitivity and kappa values were 74.7% (95% CI 64.3–83.4) and 0.70 (95% CI 0.61–0.78), respectively. For skim milk diluted 1:2 and a cut-off of 0.30, sensitivity and kappa values were 77.0% (95% CI 66.8–85.4) and 0.72 (95% CI 0.64–0.80), respectively. Using the selected cut-offs, the IDEXX Neospora Ab Test is equally suited for the analysis of whole and skim milk as a screening tool in neosporosis control programs.