Infectious disease spread modelling for evaluating control strategies

Livestock infectious diseases can lead to poor animal health and welfare, increased use of antibiotics and significant production losses. However, this is not consistent with sustainable agriculture. It is therefore important to continuously work on developing new knowledge in the interface between animal health and welfare, preventive measures, and the economy. In addition, certain infections can be transmitted from animals to humans in, for example, foodborne outbreaks, and these zoonotic infections can be of major public health concern. There are thus many reasons why it is essential to control and reduce the risk of spreading diseases in animal populations.

Why do we observe a higher prevalence of disease in one region compared to another? What is the most efficient strategy to eradicate an infection from a population? How can we optimize surveillance for a disease? My research involves using disease spread modelling to study these types of questions. A disease spread model is a mathematical model where the biology and epidemiology of a disease has been described in a mathematical language. Commonly, the population is divided into discrete health states called compartments, together with rules for how individuals move from one compartment to another. The SIR model is a classic example of a compartment model, it contains the three compartments: Susceptible, Infectious and Recovered (or Removed) animals. Such a model has, for example, been used to predict the spread of Foot and mouth disease in cloven-hoofed animals.

Nowadays, in the computer age of statistical inference, disease spread modelling is an integral part of studying livestock diseases and evaluating mitigation strategies prior to implementation in the real world. These models have become increasingly sophisticated and can now also include real livestock data, for example, animal movement data. At the same time, it is important to keep in mind the famous quote “All models are wrong, but some are useful” attributed to the British statistician George Box. Meaning that, regardless of the complexity of a model, it is a simplification of reality based on assumptions. However, if a model captures central features of an epidemiological system, it is critical to our understanding of how an infectious disease spread at the animal and population level, as well as giving important insights into how to control the spread most efficiently.

In my docent lecture, I will present how the epidemiology of an infectious process can be captured in a disease spread model, using Shiga-toxin producing Escherichia coli O157 (STEC O157) in cattle as an example pathogen. I will illustrate, from the bottom-up, how the biology of the fecal-oral spread of STEC O157 between cattle in one farm, to local spread between farms in a region, to the national level spread via livestock movements, can be transformed into a disease spread model. Additionally, I will demonstrate the model’s response to various control strategies. I will also highlight knowledge gaps and discuss ideas for future research.