Bovine Genital Leptospirosis (BGL) is a chronic reproductive syndrome characterized by genital infection by Leptospira spp. An accurate diagnosis of BGL is crucial to implementing proper control measures in field conditions. This study aimed to evaluate the reliability of serology by Microscopic Agglutination Test (MAT) for diagnosing leptospirosis in subfertile cows with genital infection. Of three herds, 93 non-pregnant cows with reproductive failures were submitted to the blood sampling (serology by MAT) and genital samples (lipL32-PCR). A total of 62/93 (66.6%) cows presented seroreactive to cutoff 100, while 45/93 (48.4%) cows were positive to cutoff 200, mainly against the Sejroe serogroup. In PCR analysis, 55/93 (59.1%) were positive. MAT results were compared with PCR (considered the standard), and test parameters and Cohen's kappa (ƙ) were calculated for the cut-offs 100 and 200. A ROC curve was performed for each cut-off of titers 100 to 1,600. The sensitivity and specificity of MAT100 were calculated at 66.6% and 33.3%, while for MAT200 the sensitivity was estimated as 35% and specificity as 54.5%. The accuracy of MAT was poor, being 54.8% in MAT100 and 42% in MAT200. Furthermore, the area under the curve of ROC analysis was low for all titers, and the correlation was poor for MAT100 and MAT200 (ƙ < 0). The results demonstrated that MAT is a limited technique to diagnose bovine genital carriers individually, and if only MAT is applied, genital carriers may pass undetected, impairing the control programs.
