Evaluation of a five-antigen ELISA for diagnosis of tuberculosis in cattle and Cervidae

Abstract

Objective - To determine the validity of a 5-antigen ELISA for detection of tuberculosis in cattle and Cervidae. Design - Cross-sectional observational study. Sample Population - Serum samples collected from 5,304 cattle in 23 herds and 1,441 Cervidae in 12 herds. Procedure - Discriminant analysis was used to determine the linear combination of antigens that accurately predicted the true Mycobacterium bovis infection status of the most animals. The resulting classification functions then were used to calculate the percentage of animals that were correctly classified (ie, sensitivity and specificity). The kappa statistic was calculated to evaluate different combinations of test results. Results - Of the 23 cattle herds, 4 dairy and 2 beef herds were considered infected. Of the 12 Cervidae herds, 5 were considered infected. For cattle, the specificity and sensitivity of ELISA, using the discriminant function, were 56.4 and 65.6%, respectively. For Cervidae, the specificity and sensitivity of ELISA, using the discriminant function, were 78.6 and 70.0%, respectively. Clinical Implications - Results suggest that the 5-antigen ELISA would not be a good test for tuberculosis, especially in cattle, if used alone. However, when results of the ELISA and tuberculin test were interpreted in parallel, sensitivity of the combination was greater than sensitivity of either test alone. Similarly, when results of the 2 tests were interpreted in series, specificity of the combination was greater than specificity of either test alone.