Determination of the sensitivity and specificity of bovine tuberculosis screening tests in dairy herds in Thailand using a Bayesian approach

Background: The objective of this study was to determine the sensitivity (Se) and specificity (Sp) of bovine tuberculosis (bTB) screening tests including a single intradermal tuberculin (SIT) test, interferon gamma (IFN-γ) assay, and a commercial ELISA test (M. bovis Ab) in dairy cattle, under field conditions, using a Bayesian approach. Results: The study population consisted of 128 dairy cows from 25 bTB-infected herds in Chiang Mai and Chiang Rai provinces, Thailand. A single-population Bayesian model was implemented assuming conditional dependence between the SIT test and IFN-γassays. The 95% posterior probability interval (PPI) of the SIT test (severe interpretation) Se ranged from 75.3 to 95.2% (median = 87.6%), while the Sp was slightly lower (median = 83.6%, PPI = 74.2-92.8%). The IFN-γassay Se was moderate and the 95% PPI ranged from 38.6 to 74.4% (median = 55.7%) with higher Sp (median = 93.5.4%, PPI = 87.0-98.1%). The M. bovis Ab ELISA Se was low, with 95% PPI ranging between 30.0 and 71.2% (median = 47.4%); however, the Sp was high (median = 90.9%, PPI = 84.5-95.5%). Conclusion: The SIT test sensitivity was similar to that demonstrated in other regions and can, therefore, be used effectively as part of control programs in this area. The IFN-γand M. bovis Ab ELISA assays can be applied as supplementary techniques. The test performance of these tests when used as single tests without confirmation, however, are expected to continue to challenge disease eradication efforts.