Bovine tuberculosis (bTB) is one of the major challenges facing the UK livestock industry. It has been estimated to cost the UK in excess of £100million per year, and has a devastating effect on individual farmers and the rural community. Researchers from the University of Warwick have used a variety of mathematical and statistical techniques to understand and predict the spread of this pathogen. Detailed statistical models are able to predict the chance of a positive test with great accuracy, opening the possibility of targeting testing. The mathematical model is more flexible and allows us to ‘experiment’ with a range of control options from vaccination to movement controls. We hope that these scientific methods with provide policy-makers with suitable tools that can inform the management of this disease.
Bovine tuberculosis (bTB) has been a long-standing problem in Britain. In the first half of the 20th century, 40% of British cattle were suspected to be infected, with contaminated milk being a major transmission route to humans. This route was eliminated with the pasteurisation of milk, and a ‘test-and-slaughter’ scheme was started in 1950 to eradicate infection among cattle. Incidence fell dramatically and in the 1970s only 0.22% of tests revealed infected animals.
However, over the past 20 years incidence has been steadily increasing, most notably in the southwest and west of England and the southwest of Wales. The low sensitivity of the standard diagnostic test for bTB leads to considerable ambiguity in determining the main transmission routes of infection, which exacerbates the continuing scientific debate. In turn this uncertainty fuels the fierce public and political disputes on the necessity of controlling badgers to limit the spread of infection.
Teams from the University of Warwick, under the leadership of Professor Matt Keeling, have brought cutting-edge mathematical and statistical tools to this problem. In close collaboration with Defra (Department of Environment, Farming and Rural Affairs), they developed a range of statistical models that could predict the risk of break-down (a new positive bTB test) in terms of farm history, location and cattle movements. This work highlighted to Defra the accuracy of the current targeting of testing, but the potential for significant improvements in border regions that are being invaded. The mathematical model was developed independently, and aims to capture the main mechanisms of transmission, incorporating within and between farm transmission, as well as transmission between cattle and transmission via an environmental reservoir.
We find that the majority of newly detected cases involve multiple transmission routes with moving infected cattle, reinfection from an environmental reservoir and poor sensitivity of the diagnostic test all playing substantive roles. This underpins our finding that controls that target only one of these routes are unlikely to be highly effective. We predict that very few control options have the potential to reverse the observed annual increase, with only intensive strategies such as whole-herd culling or additional national testing proving highly effective. This work highlights the complexity of bTB transmission and the need to adopt a more stringent set of controls.
Although the work to date represents state-of-the-art in terms of mathematical modelling, there are numerous factors that need to be addressed to make such models of immediate policy relevance. Of the greatest importance are understanding how farming communities would react to different control measures, and estimating long-term costs associated with each control.