By Dr Louise Matthews, University of Glasgow. This article first appeared in the Winter 2013 edition of the RSGS’s magazine, The Geographer.

Identifying the major sources of risk in disease spread is key to designing effective means of control. However, understanding of how disease spreads across species boundaries is typically poor, making the design and evaluation of control methods especially challenging for zoonotic pathogens.

E coli O157 is one such widespread zoonotic pathogen, which causes serious gastrointestinal illness in people. Infection can lead to death or lifelong kidney damage and is a major cause of acute renal failure in children. Cattle are the main reservoir for E coli O157, and harbour the pathogen in their gastrointestinal tract without suffering clinical disease. Although it can spread between people, the usual routes of infection are either by consuming contaminated food and water, or by contact with livestock faeces in the environment.

Vaccines for cattle have been developed, but their adoption is being hampered by delays in their licensing. These delays highlight the particular challenges to control planning that are posed by zoonotic infections. First, the medical and veterinary agencies have conflicting responsibilities. The bodies responsible for licensing vaccines in animals must typically show that a new control is not just safe, but that it improves the health of animals receiving it; this poses a problem for zoonotic pathogens that are benign in their reservoir hosts, and demands greater co-ordination from medical and veterinary agencies. Second, it is not easy to test a control measure used in animals against the key outcome – the reduction in human illnesses – and this lack of data on impact hampers effective decision making for policy change.

Our focus is on the understanding of disease ecology and cross species transmission dynamics needed to predict how E coli O157 cases in people could be prevented by vaccinating cattle. In the absence of direct data, mathematical models of transmission between animals and people provide a tool to predict the success of interventions. However, because of the difficulty in mapping the distribution of infection in the animal reservoir onto disease incidence in the human population, few such models for zoonotic infections exist. Because of variation in factors such as the infectiousness of cattle, pathogen strains, and infection routes, simple measures of the presence and degree of infection in animals may not always be a good predictor of risk to humans. Epidemiological models need to be able to (or ‘can be used to’) represent the range of possible outcomes arising from these sources of variation.

For E coli O157 the role of variation in transmission from cattle is a key issue. ‘Supershedding’ is the rare but epidemiologically important situation where some individuals are responsible for much more onwards disease transmission than most others. In the case of E coli O157, some cattle shed the pathogen in faeces at unusually high concentrations. This is important because, despite being relatively rare, supershedders appear to be the major source of deposition of pathogens into the environment, from where other animals and humans can become infected.

A recent study (see Further Reading) looked at the links between supershedding in cattle and transmission risk to humans, and showed that only the relatively rare supershedding events (rather than the more common low-level shedding seen in most infected animals) contribute significantly to human risk. This new understanding of animal to human transmission of this pathogen has important consequences for our assessment of the potential impact of cattle vaccines, which not only reduce the frequency of bacterial shedding by cattle, but also reduce the number of bacteria shed by infected animals when this occurs. Consequently, the benefit to people of cattle vaccination should be substantially greater than previously anticipated based on the impact on the frequency of shedding alone. The recent study indicates that vaccines producing a 50% reduction in shedding frequency in cattle could reduce human cases by nearly 85%, and concludes that vaccination of cattle, the major reservoir for E coli O157, could be an especially effective public health control against a serious disease.

For zoonoses such as E coli O157, where controls are available in the animal reservoir but the benefit is to the human population, the challenges to the design, evaluation and delivery of effective interventions for humans can be added to by conflicting responsibilities of veterinary and public health agencies. What this example has highlighted is the need for a One Health approach to policy that understands animal and human health to be fundamentally integrated, rather than treated as discrete issues to be dealt with by separate organisations with their own policy goals.

Further Reading: Matthews L, Reeve R, Gally DL, Low JC, Woolhouse MEJ, McAteer SP, et al (2013), Predicting the public health benefit of vaccinating cattle against Escherichia coli O157 (Proceedings of the National Academy of Sciences of the USA, 2013, 110 (40)).