A multiagent coronavirus model with territorial vulnerability parameters (preprint)

ABSTRACT

We developed a simple and user-friendly simulator called MD Corona that is based on a multiagent model and describes the transmission dynamics of coronavirus for a given location considering three setting parameters population density, social-isolation rate, and effective transmission probability. The latter is represented by the Corona Protection Index (CPI) - a measurement of a given territory's vulnerability to the coronavirus that includes characteristics of the health system and socio-economic development, as well as infrastructure. The dynamic model also relies on other real epidemiological parameters. The model is calibrated by using immunity surveys and provides accurate predictions and indications of the different dynamic mechanisms. Our simulation studies clearly demonstrate the existence of multiple epidemic curves in the same city due to different vulnerabilities to the virus across regions. And it elucidates the phenomenon of the epidemic slowing despite a reduction in social-distancing policies, understood as a consequence of ''local protection bubbles''. The simulator can be used for scientific outreach purposes, bringing science closer to the general public in order to raise awareness and increase engagement about the effectiveness of social distancing in reducing the transmissibility of the virus, but also to support effective actions to mitigate the spread of the virus.