Slaughterhouse detection of lesions compatible with bovine tuberculosis represents a key passive surveillance component in Santa Catarina, Brazil, yet subsequent trace-back investigations often fail to identify infected farms. This study developed a quantitative framework to prioritize epidemiological investigations by estimating the probability of infection associated with each farm connected to PCR-confirmed cases. Using official movement records and historical diagnostic data, we reconstructed the lifetime contact networks of slaughtered cattle presenting confirmed Mycobacterium bovis lesions (n = 502). For each sentinel animal–farm interaction (n = 1452), infection probability was estimated through a non-homogeneous Poisson process incorporating exposure duration and the time-weighted average herd size as determinants of infectious pressure. After evaluating stochastic variability through Monte Carlo simulation, a deterministic model using the mean infectious-pressure parameter was applied to classify farms into high-, medium-, and low-risk categories. Model performance was assessed using validated field diagnostic outcomes within a three-year temporal window. High-risk farms represented most validated contacts (58%) and demonstrated a relative risk of 3.48 compared with lower-risk category. These findings indicate that a standardized risk-based classification can substantially improve the prioritization of trace-back investigations, offering a practical decision-support tool to enhance bovine tuberculosis surveillance and contribute to eradication strategies in Santa Catarina.
Lourenço et al. (Thu,) studied this question.