The purpose of this research was the interlaboratorial validation of the polarization fluorescence assay (PFA) for the serodiagnosis of bovine brucellosis, verifying the relative sensitivity, the relative specificity and the reproducibility of the test in four Brazilian laboratories. Serum samples from 1,389 bovines were selected and submitted to the rose Bengal (RBT) and 2-mercaptoethanol tests in one of the laboratories. The same samples were tested by the complement fixation (CFT) test and by the PFA in the four laboratories participating of the research. The reference population (golden standard) used to evaluate the PFA was the combination of the results of RBT, CFT and 2-ME. TG-ROC analysis was used to obtain the cut-off that provided the best combination of sensitivity and specificity. The agreement between laboratories was obtained by the kappa statistic and Pearson correlation coefficient (r). The PFA cut-off values were from 85.2 to 93.6. The sensitivity of the PFA assay varied from 91.7% to 97.3%, and the specificity values varied from 82.6% to 98.3%. When comparing PFA results from the four laboratories, the kappa values was between 0.69 and 0.95, which indicates, in most situations, excellent reproducibility, and the correlation coefficient varied from 0.76 to 0.99. The results showed that the PFA had a good performance, with high sensitivity and specificity. Compared to the conventional tests, the PFA has the advantages of being easy and quick to perform, and it is not prone to the occurrence of prozone, as the CFT or the 2-ME, nor to the occurrence of anti-complementary effect, as the CFT.
bovine brucellosis; serological diagnosis; fluorescence polarization assay
