Quantitative assessment of the probability of introducing bovine brucellosis into English cattle herds by imported live cattle

Abstract

A stochastic simulation model was developed to estimate the quarterly probability (PIntro) of introducing bovine brucellosis into English cattle herds, by at least one imported live cattle (potential carrier of Brucella abortus). The probability of spread after introduction was not included and imports from several countries were considered. Information used to parameterise model´s inputs was obtained from the literature, legislation and analysis of several national datasets (2013 to 2016), which contained information on imported cattle and testing schemes used in the English cattle population. Exporting countries were divided according to official brucellosis status “J” into: Officially Brucellosis Free (OBF), Non-Officially Brucellosis Free (Non-OBF) and in OBF-Validation (during the first five years of OBF status). The entire English cattle population was divided into eight strata “S” by combination of laboratory testing data and herd type. The only risk mitigation measure considered was the testing for antibodies on animals older than one year and imported from Non-OBF countries. Probabilities of introduction at herd and stratum level were combined into the overall national PIntro. Two scenarios were run. In the baseline scenario, the between herds prevalence BHP(OBF) in OBF countries was set according to information from EFSA and from the EU legislation. In the alternative scenario only the former was used and BHP(OBF) was set very low/negligible. For Non-OBF and OBF-Validation countries, the BHP was set with distributions based on the literature. In the baseline scenario, between 2013 and 2016, the quarterly median PIntro ranged from 1.3% to 5.5%. For the last year considered, the median of the quarterly medians PIntro was 2.8% (median of 5th percentiles = 0.4%; median of 95th percentiles =10.7%). Therefore, on average, at least one introduction could be expected each (approximated) 36 surveillance periods (9, 281), so each ≈ 9 years (2; 70). According to the alternative scenario, the PIntro was very low and on average at least one introduction could be expected each ≈ 125 years.