Diagnosis and control of Brucella abortus infection in domestic water buffalo (Bubalus bubalis) of Trinidad, West Indies

Standard plate agglutination (SPAT), Card, buffered plate agglutination (BPAT), and standard tube agglutination (STAT) tests were evaluated for brucellosis screening in cattle and water buffalo (Bubalus bubalis) of Trinidad. Bayesian statistical methods utilizing iterative simulation techniques were used to estimate accuracy of these tests without a gold standard. Cattle sensitivity and specificity estimates, respectively, were SPAT: 66.7, 98.9; Card: 72.7, 99.6; BPAT: 88.1, 98.1; and STAT: 80.2, 99.3. Corresponding buffalo estimates were SPAT: 51.4, 99.3; Card: 90.4, 99.4; BPAT: 96.3, 90.7; and STAT: 75.0, 98.8. The best screening test for both species was BPAT because it had the highest estimated sensitivity. Diagnostic accuracy of BPAT was found to be different between species with sensitivity being higher in water buffalo (95% probable) and specificity higher in cattle (90% probable).; Receiver-operating characteristic (ROC) curve analysis was performed to evaluate use of the brucellosis competitive enzyme-linked immunosorbent assay (c-ELISA) as a confirmatory test for cattle and water buffalo of Trinidad. Iterative simulation modeling without a gold standard was employed to estimate these curves. Areas under ROC curves for cattle and buffalo were 0.94 and 0.98, respectively, and not statistically different. The most appropriate cutoff value was 35% inhibition based on peak Youden index (sensitivity + specificity − 1) and high specificity. This cutoff yielded estimates of sensitivity and specificity, respectively, of 83.9% and 96.2% for cattle and 88.0% and 97.4% for water buffalo.; Thirty-two brucellosis-free domestic water buffalo were enrolled in a randomized field trial to assess effectiveness of RB51 vaccination on prevention of B. abortus infection under natural exposure conditions. Animals were classified as positive if they demonstrated persistent agglutination titers or had B. abortus isolated from specimens collected at completion of study. Six of 14 (43%) vaccinated buffalo completing the study were classified as positive for Brucella infection. Two of 13 (15%) control animals were classified as positive for Brucella infection. The difference between proportions of Brucella -positive animals in each treatment group was not statistically significant (p = 0.2I). RB51 vaccination failed to protect water buffalo from B. abortus infection using recommended procedures for cattle immunization.