Transmission Dynamics of COVID-19 and Impact on Public Health Policy

B Shayak; Mohit Manoj Sharma; Richard H Rand; Awadhesh Kumar Singh; Anoop Misra

Dynamic modeling of COVID-19 along these lines has been attempted in several prior works [29] [30] [31] [32] [33] [34] . Kucharski et. al. [29] modelled the early phases of the outbreak in Wuhan to estimate the value of R0 there. The number they obtained was between 1.6 and 2.6. A similar study, determination of R0 as a function of time, has been performed by Feng et. al. [30] , whereas an extensive determination of parameters in the S-I-R model has been done by Rabajante [31] . Peng et. al. [32] use a similar model but with a different purpose -they predict the long-term course of the outbreak in Wuhan as well as in the rest of China. Chen et. al. [33] have used a more complicated model with 14 dynamical variables to explain the transmission of the virus from bats to humans and thereafter its spread in the human population. Neher et. al. [34] have introduced a parametric excitation into the basic S-I-R model to study the effects of seasonal temperature variation on the spread of the epidemic. As alternative approaches to the modeling, we would like to cite References [35, 36] which have used stochastic models to analyse the course of the disease in Wuhan and on the cruise ship Diamond Princess respectively. A mathematical analysis of the consequences of public health policy is Reference [37] which predicts the effects of various kinds of social distancing measures such as school closures and social isolation of the high-risk senior citizens. The models in this paper are based on individual-based stochastic influenza models [38] . As recently as the day before yesterday, a new study [39] has emerged which explores the effects of non-pharmaceutical interventions (social distancing etc) on the spread of COVID-19 in Singapore. The mathematical core of this work is a stochastic model, once again for influenza [40] . Another recent work [41] attempts to find the required hospital capacity in Chicago in the presence and absence of a lockdown, using the model [34] . Yet another contribution [42] predicts the effects of imposing a lockdown in India, using an in-house lumped-parameter model. This model is based on S-I-R and accounts for the variation in susceptibility across different age groups. It also accounts for quarantining. is the author/funder, who has granted medRxiv a license to display the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.03. 29.20047035 doi: medRxiv preprint One of the features indigenous to this new and dangerous Coronavirus strain is that there are time lags involved in several stages of the transmission dynamics. One major delay step is that patients become transmissible before showing symptoms of the disease [43] ; in other words, there is a time gap or latency period between the time a patient turns infectious and time that s/he manifests symptoms. A second delay step comes from the fact that, on account of the element of surprise associated with the pandemic, testing facilities in many countries and regions are inadequate relative to the number of cases taking place. For example, the testing rate in USA increased by a factor of ten from 03 March to 17 March [44] , with some states such as New York achieving the maximum testing rates. Similarly, in India, Russia and some European countries, the testing rate is also undergoing a sharp increase even as we write. A limitation in the number of testing facilities means that there is a waiting period involved between t