Bayesian Estimation of the True Bovine Brucellosis Prevalence in Vaccinated and Non-Vaccinated Ecuadorian Cattle Populations, and the Sensitivity and Specificity of a Competitive and Indirect ELISA Using a New Synthetic Antigen

Abstract

Bovine brucellosis (bB) is a zoonosis mainly caused by the Brucella abortus species in cattle. Bovine brucellosis can present with either a range of clinical symptoms, including spontaneous abortions in the last trimester of pregnancy, retained fetal membranes, and decreased milk production, or it can be asymptomatic. In Ecuador, vaccination against bB with S19 and/or RB51 is not mandatory and is the responsibility of the farmer. As serology is a convenient method for detecting antibodies against Brucella, evaluating the diagnostic performance and discriminative ability of such tests in various epidemiological settings is required. To estimate and compare the diagnostic sensitivity (Se) and specificity (Sp) of two screening tests, a new competitive (cELISA) and an indirect ELISA based on a new synthetic antigen (iELISA), a randomized, stratified, cross-sectional, serological survey was performed on the cattle population (3299 bovine sera from 223 farms) in continental Ecuador. A Bayesian approach was used to evaluate the two tests by estimating their respective diagnostic Se and Sp, as well as the true prevalence of bB in different sub-populations (non-vaccinated, vaccinated with S19 or RB51). The Se of both tests was similar across Bayesian models, with values around 94%. In contrast, the Sp of the iELISA, ranging between 97 and 98%, was significantly higher than that of the cELISA, which was approximately 94–95%. The true prevalence of bB was 1.63% (95% CrI: 0.56–2.54) in non-vaccinated cattle, decreased to 0.97% (95% CrI: 0.005–2.54) in S19-vaccinated cattle and was 2.75% (95% CrI: 0.50–5.32) in RB51-vaccinated cattle. The results of this study suggest that, with similar Se and higher Sp, the iELISA based on an innovative synthetic antigen (which is more standardizable) should be recommended as a possible screening test for bB in Ecuador. Also, the proposed approach suggests insights into the quality of the vaccination campaign and highlights the need for refining the Ecuadorian national brucellosis control program.

Keywords

• bovine brucellosis
• Ecuador
• Bayesian approach
• ELISA
• test performance
• synthetic oligosaccharides