SENSITIVITIES AND SPECIFICITIES OF 2 ELISA TESTS FOR DETECTING INFECTION WITH SARCOCYSTIS IN CATTLE OF WESTERN-AUSTRALIA

The accuracies of two ELISAs, one using antigen from merozoites of S. cruzi grown in vitro and the other using antigen from cystozoites of S . cruzi, for detecting infection of cattle with Sarcocystis, were eval uated by testing the sera of 303 cattle from 36 Western Australian her ds. The results were compared with those obtained by digestion of oeso phageal samples collected from the same animals. A similar proportion of infected animals were detected by the three methods. The sensitivit y of the assays for detecting infected cattle was comparable (98 and 9 5% for the assay using antigen from merozoites and cystozoites, respec tively), however the specificity (97%) of the assay which used antigen derived from merozoites was significantly higher (P < 0.001) than tha t (84%) which used antigen from cystozoites. When herds which had at l east five animals sampled were considered, the same infected and non-i nfected herds were detected by the ELISA employing antigen from merozo ites and the digestion methods (sensitivity and specificity of 100%). The sensitivity and specificity of the assay using cystozoite antigen were 100 and 67%, respectively. The kappa values for agreement beyond chance between the two ELISAs were calculated as 78% for the animal-ba sed data and 72.5% for the herd-based data. We conclude that because o f the high sensitivity and specificity of the ELISA using antigen deri ved from merozoites, this assay would be a useful and reliable tool fo r general sero-epidemiological studies into infection with Sarcocystis . (C) 1997 Elsevier Science B.V.