Abstract
Background: Bovine Tuberculosis (bTN) elimination programs are based on test-and-slaughter strategies. No diagnostic test used in the control of Bovine Tuberculosis (bTB), a zoonotic disease, is a perfect gold standard having a diagnostic sensitivity and specificity of 100%. Methods: To find relevant data to assess diagnostic accuracy of one variant of the tuberculin test, the Single Intradermal Comparative Cervical Tuberculin Test (SICCT) used in the diagnosis of bTB a systematic literature search was performed. To estimate diagnostic test accuracies of the SICCT, a Bayesian latent class analysis has been undertaken. Results: Only one eligible study was found. Assuming bacterial isolation has a specificity of 100%, specificity of gross lesion detection and histopathology were estimated as 90% and 97% respectively. Estimates of the sensitivity of the Single Intradermal Comparative Cervical Tuberculin Test (SICCT), gross lesion detection, histopathology and bacteriological isolation were 80%, 89%, 93% and 98% respectively. The design of this study precluded estimating the SICCT specificity, although official UK surveillance data indicates specificity greater than 99.67%. Conclusion: Studies to estimate the diagnostic performance of key tests for bTB are lacking, yet essential for an informed debate in regard to bTB public health policy.
Hartnack, S