Evaluation of an enzyme-linked immunosorbent assay kit to detect Babesia bovis antibodies in cattle☆

Abstract The sensitivity and specificity of an enzyme-linked immunosorbent assay (ELISA) to detect antibodies; to Babesia bovis was evaluated in 1000 sera from Holstein heifers. Five hundred of them were from cattle naturally or experimentally infected with B. bovis and 500 from uninfected heifers born and raised in a region free of the vector of cattle babesiosis. Additionally, the ELISA was evaluated and compared with an indirect immunofluorescent antibody (IFA) test in 374 heifers inoculated with different kinds of B. bovis antigens in four trials. The cross-reaction was also evaluated in 50 heifers infected with Babesia bigemina and 50 heifers infected with Anaplasma marginale . The mean percentage positivity of negative sera in relation to the ELISA strong positive sera was 8%. The seropositive/seronegative cutoff point was set as twice the mean percentage positivity of negative cattle sera (= 16%). The sensitivity of the ELISA was 98% with a 95% confidence interval (CI) of 96–99%. The specificity was 95% (CI 93–97%). The agreement was 97% and the kappa value was 0.93. The predictive values of positive and negative results were 95% and 98% respectively. ELISA showed a similar sensitivity to that of the IFA test to detect antibodies to different B. bovis antigens. Its sensitivity ranged from 97.1% to 100% (CI 89–100%), while the sensitivity of the IFA test ranged from 92.8% to 100% (CI 83–100%). ELISA cross-reacted in 8% and 6% of the sera carrying B. bigemina and A. marginale antibodies, respectively, while the IFA showed 4% cross-reaction in each situation. The ELISA evaluated has the advantages of a proper sensitivity, objectivity and capacity to be adapted to test large number of samples in a short period of time. The results indicate that the ELISA is a suitable replacement for the IFA test to detect B. bovis antibodies in cattle sera, especially in epidemiological studies.