New approximations, and policy implications, from a delayed dynamic model of a fast pandemic

Chandrika Prakash Vyasarayani; Anindya Chatterjee

In closing, we must acknowledge that in any lumped model of the sort we study here, spatial variations in parameters and infected population densities are not modeled. Such lumped models are averaged models. Thus, it is not really clear at a detailed spatial level what it means to reduce the average infection rate β by, say, 2%. If the average person engages in social distancing, benefits will be seen on average, although there could still be localized outbreaks within pockets where people engage in riskier behavior. In constrast, however, the speed and probability of being detected and quarantined is less up to individual members of the public, and more in the hands of public institutions. Such institutions are amenable to tighter quality measures. So, from the viewpoint of reliability, we believe that large scale testing and near-certain isolation or quarantining can be critically useful in containing pandemics like COVID-19.