Quantifying the risk of spread of bovine viral diarrhoea virus (BVDV) between contiguous herds in Ireland

The control of bovine viral diarrhoea virus (BVDV) mainly focuses on the identification and restriction of persistently infected (PI) animals. However, other transmission pathways can also result in new breakdowns, including the movement of animals pregnant with PI calves (Trojan animals) and the spread of infection between contiguous farms. Contiguous spread is likely an important problem in the BVD eradication programme in Ireland, given the spatial distribution of residual infection, and the highly fragmented nature of land holdings on many Irish farms. In this study, we seek to quantify the risk of BVD spread between contiguous herds in Ireland. Multivariable logistic models were used to estimate the risk of a herd having BVD positive calves in January to June 2014 (the study period) when contiguous to a herd that had at least one BVD positive calf born in 2013. The models included risk factors relating to the study herd and to neighbouring herds. Separate multivariable models were built for each of four "PI-neighbour" factors relating to the presence of BVD+ animals and/or the presence of offspring of PI breeding animals. In total, 58,483 study herds were enrolled. The final model contained the province, the log of the number of calf births born during the study period, the number of cattle purchased between January 2013 and January 2014, and with a two-way interaction between the number of animals of unknown BVD status in the study herd and the PI-neighbour risk factor. When the number of PI-neighbour herds was used as the PI-neighbour risk factor, the odds ratio (OR) associated with the number of PI-neighbour herds ranged from 1.07 to 3.02, depending on the number of unknown animals present. To further explore the risk associated with PI-neighbour factors, the models were repeated using a subset of the study herds (n=7440) that contained no animals of unknown status. The best fitting model including "any PI-neighbour" as the PI-neighbour factor and also contained the log of the number of calf births born during the study period and the number of cattle purchased. The OR associated with "any PI-neighbour" was 1.92 (95% C.I. 1.37-2.70). This study provides the first quantitative information on the risks posed by the presence of BVD+ animals in neighbouring herds and also highlights the importance of clarifying the BVD status of animals that have not yet been tested in the context of the Irish eradication programme.