Infectious disease surveillance in animal movement networks: An approach based on the friendship paradox

Abstract

The network of animal movements among livestock premises is an important topological structure for the spread of infectious diseases. The central focus of this study was to analyze strategies for selecting premises based on the friendship paradox (“your friends have more friends than you do”) – in which premises that neighbor randomly selected premises are sampled for surveillance or control – to determine whether these strategies are viable alternatives for the surveillance and control of diseases in scenarios with insufficient data on animal movement. To test the effectiveness of these strategies, we performed three sets of simulations. In the first set, we examined the risk of spreading an infectious disease using the cattle movement network of the state of Mato Grosso, Brazil. All tested strategies based on the friendship paradox have comparable performance to the hub control strategy (controlling premises that sold more animals) and superior performance to random sampling in terms of both reducing the risk of purchasing infected animals and the number of premises that need to be controlled. In the second and third sets of simulations, we observed that the friendship paradox strategies were more sensitive than the random sampling strategy to detect cases and disease, respectively. The survey of the entire animal movement network to identify animal premises with a key role in trade is not always possible, either because the data are insufficient or because informal trade is significant. If surveying the network is not possible, all approaches based on knowledge of the network become useless. As an alternative, knowing that there is a hidden movement network that follows rules inherent to all networks, such as the friendship paradox, can be used to our advantage. Strategies based on the friendship paradox do not assume knowledge of the animal movement network and therefore may be viable alternatives for the surveillance or control of infectious diseases in the absence of network information, thus optimizing the use of human and financial resources.