Simulation model of within-herd transmission of bovine tuberculosis in Argentine dairy herds.

Transmission of bovine tuberculosis was quantified in three dairy herds located in south Santa Fe Province, Argentina. Using estimates of Mycobacterium bovis transmission (beta) and a Reed-Frost simulation model, the prevalence of tuberculosis infection in the study herds over time was investigated. The Reed-Frost model was modified by incorporating randomness in both beta and the incubation period (alpha) of M. bovis. The mean estimated herd beta was 2.2 infective contacts per year and did not differ significantly between the study herds. Modeling alpha as Poisson distributed (mean 24 months) best fit the observed prevalences. Infection was predicted by the model either to spread quickly (<10 years) within a herd and reach a high prevalence (>50%), or to persist at a low prevalence (<5-10%). The model was robust, predictions were realistic and the mean beta estimated was consistent with previous studies of bovine tuberculosis.