State-based targeted vaccination

Abstract

Vaccination has become one of the most prominent measures for preventing the spread of infectious diseases in modern times. However, mass vaccination of the population may not always be possible due to high costs, severe side effects, or shortage. Therefore, identifying individuals with a high potential of spreading the disease and targeted vaccination of these individuals is of high importance. While various strategies for identifying such individuals have been proposed in the network epidemiology literature, the vast majority of them rely solely on the network topology. In contrast, in this paper, we propose a novel targeted vaccination strategy that considers both the static network topology and the dynamic states of the network nodes over time. This allows our strategy to find the individuals with the highest potential to spread the disease at any given point in time. Extensive evaluation that we conducted over various real-world network topologies, network sizes, vaccination budgets, and parameters of the contagion model, demonstrates that the proposed strategy considerably outperforms existing state-of-the-art targeted vaccination strategies in reducing the spread of the disease. In particular, the proposed vaccination strategy further reduces the number of infected nodes by 23–99%, compared to a vaccination strategy based on Betweenness Centrality.