Efficient use of sentinel sites: detection of invasive honeybee pests and diseases in the UK

Optimising risk-based surveillance for early detection of invasive plant pathogens

Emerging infectious diseases (EIDs) of plants continue to devastate ecosystems and livelihoods worldwide. Effective management requires surveillance to detect epidemics at an early stage. However, despite the increasing use of risk-based surveillance programs in plant health, it remains unclear how best to target surveillance resources to achieve this. We combine a spatially explicit model of pathogen entry and spread with a statistical model of detection and use a stochastic optimisation routine to identify which arrangement of surveillance sites maximises the probability of detecting an invading epidemic. Our approach reveals that it is not always optimal to target the highest-risk sites and that the optimal strategy differs depending on not only patterns of pathogen entry and spread but also the choice of detection method. That is, we find that spatial correlation in risk can make it suboptimal to focus solely on the highest-risk sites, meaning that it is best to avoid ‘putting all your eggs in one basket’. However, this depends on an interplay with other factors, such as the sensitivity of available detection methods. Using the economically important arboreal disease huanglongbing (HLB), we demonstrate how our approach leads to a significant performance gain and cost saving in comparison with conventional methods to targeted surveillance.

Emerging infectious diseases of plants continue to devastate ecosystems and livelihoods worldwide. By linking a mathematical model of pest spread with a computational optimisation routine, this study identifies where to look for invasive pests if we wish to detect them at an early stage; this method improves upon conventional methods of risk-based surveillance and is robust to model misspecification.

Bayesian epidemiological modeling over high-resolution network data

Mathematical epidemiological models have a broad use, including both qualitative and quantitative applications. With the increasing availability of data, large-scale quantitative disease spread models can nowadays be formulated. Such models have a great potential, e.g., in risk assessments in public health. Their main challenge is model parameterization given surveillance data, a problem which often limits their practical usage. We offer a solution to this problem by developing a Bayesian methodology suitable to epidemiological models driven by network data. The greatest difficulty in obtaining a concentrated parameter posterior is the quality of surveillance data; disease measurements are often scarce and carry little information about the parameters. The often overlooked problem of the model's identifiability therefore needs to be addressed, and we do so using a hierarchy of increasingly realistic known truth experiments. Our proposed Bayesian approach performs convincingly across all our synthetic tests. From pathogen measurements of shiga toxin-producing Escherichia coli O157 in Swedish cattle, we are able to produce an accurate statistical model of first-principles confronted with data. Within this model we explore the potential of a Bayesian public health framework by assessing the efficiency of disease detection and -intervention scenarios.