Risk Factors Influencing the Specificity of Skin Tests for Bovine Tuberculosis in a Low-Prevalence Region

Abstract

Background: Skin tests are the main official diagnostic method for granting, maintaining, and regaining the officially tuberculosis free (OTF) herd status in the European Union. Although highly specific, repeated testing may lead to the culling of animals not infected with bovine tuberculosis (bTB) and the suspension of OTF status, a problem especially relevant in low-prevalence regions where most herds are OTF. Identifying risk factors that affect the specificity of the single (SIT) and comparative (CIT) intradermal tuberculin tests is therefore essential to optimize control measures. Materials and methods: We analyzed 119,465 herd tests (comprising 3,588,477 skin tests) from 23,893 OTF herds tested over 5 years (2014-2018) to estimate the animal-level reactor probability and herd test specificity while accounting for relevant covariates. Results: Median reactor probabilities were low (SIT: 0.050%; CIT: 0.034%). Dairy herds had higher SIT-reactor odds compared to beef herds, whereas CIT showed the opposite pattern. Reactor risk increased over time and was higher in summer compared to spring for SIT, while a reduced risk was observed in cooler seasons for both tests. Herd test specificity remained high (median values >97% for SIT and CIT) but declined with increasing herd size, particularly for SIT. Over 5 years, the median cumulative probability of detecting ≥1 reactor in a herd was 10.9%, with higher values in dairy and large herds. Conclusion: Although skin tests maintain near-perfect animal-level specificity under field conditions, herd-level performance was influenced by several covariates, particularly herd type and size. In dairy herds, using CIT when cross-reactions are suspected and bTB-infection has been ruled out, could mitigate OTF suspensions in low-prevalence settings