Development and Bayesian analysis of a competitive inhibition ELISA for caprine brucellosis

Abstract

Caprine brucellosis, caused by Brucella melitensis, is a zoonotic disease responsible for abortions in goats and is endemic in several countries. Control measures include vaccination, serological assays, and culling of infected animals. Sensitive, specific, automatable, and low-cost serological assays are essential for use in endemic regions. This study aimed to develop and validate a competitive inhibition enzyme-linked immunosorbent assay (ciELISA) to detect anti-B. melitensis antibodies in goats. The ciELISA was standardized with pre-dilutions of ISaBmS (1/8 and 1/300) in negative goat serum. Validation was done using 1254 serum samples from non-vaccinated goats: 843 from brucellosis-free flocks and 411 from infected flocks in Argentina. Serum samples were tested with the ciELISA, buffer plate antigen (BPA) test, fluorescence-polarized antigen (FPA) test, and complement fixation test (CFT). The optimal ciELISA cutoff was determined using a receiver operating characteristic (ROC) curve, with CFT as the gold standard. Sensitivity and specificity were evaluated using the Bayesian Latent Class Model (BLCM). The ciELISA showed sensitivity of 98.29 % (95 % CI: 96.4–99.6) and specificity of 98.62 % (95 % CI: 97.6–99.6). The kappa statistic (κ) values between ciELISA and other tests were: BPA (κ = 0.94), FPA (κ = 0.79), and CFT (κ = 0.95). The ciELISA demonstrated excellent performance, being rapid and objective, making it suitable for detecting anti-B. melitensis antibodies in goat serum.