RESUMO
Serologic diagnosis of ovine contagious agalactia (Mycoplasma agalactiae) with the enzyme-linked immunosorbent assay (ELISA) developed by Agence Française de Sécurité Sanitaire des Aliments (AFSSA) may produce a few false-positive (FP) and false-negative (FN) results. When the prevalence of disease is low, these erroneous results may generate problems for eradication schemes. To prevent this, 2 commercial ELISAs were compared with the AFSSA ELISA. Flocks of known status were selected and classified into 4 categories: true positive (TP), FP, true negative (TN), and FN; 20 sheep per flock were submitted for blood sampling. A flock was considered positive when at least 1 out of 20 sera was positive or 2 sera were doubtful. In the flock, the diagnostic sensitivity of the 3 kits was very good (100%), and the diagnostic specificity showed an improvement from 46% (AFSSA test) to 88% and 92% (commercial tests). Considering individual animals, very few positive ewes were detected within TN or FP flocks; the proportion of positive ewes varied greatly from one kit to another (48% to 82%) within TP flocks. The kinetics of antibody response in sheep experimentally infected with various field strains of M. agalactiae were quite similar with all 3 ELISAs. The agreement between the 3 tests, assessed using the kappa value, varied from moderate to good (respective values of 0.56, 0.61, and 0.86). The 2 commercial ELISAs showed better performances, probably because of a superior analytical sensitivity, and are a good alternative for the serodiagnosis of contagious agalactia in sheep.