The first 100 days: Modeling the evolution of the COVID-19 pandemic

Abstract

Highlights A new compartmental model, forced-SIR, quantifies COVID-19 pandemics impact. Effectiveness of COVID-19 intervention measures is linked to models parameters. Application of the model to 10 countries reveals a wide range of COVID-19 impacts.

Multiple Epidemic Wave Model of the COVID-19 Pandemic: Modeling Study

“…The choice of the 18 countries also aimed to represent parts of the world more heavily or less heavily impacted by the disease, as well as more typical cases. Here we defined the impact as the total number N T of infected individuals during the first 120 days of the pandemic, as predicted by the FSIR model [ 18 ]; this number is scaled by the population of the country ( N P ). In particular, we have included six countries in which the impact was small, China, Australia, Greece, Cyprus, Tunisia, and Japan for which ( N T /N P )<1000 infected per million; six countries in which the impact was moderate, Israel, Denmark, Germany, France, Canada, and Portugal for which 1000<( N T /N P )<3000 infected per million; and six countries in which the impact was large, Sweden, Switzerland, United Kingdom, Italy, the United States, and Spain for which ( N T /N P )>3000 infected per million.…”

“…The multiple-wave FSIR model can identify multiple waves (subepidemics), specifying each one by only three parameters, t 1 , ∆ t , and N ′, all of which are obtained by directly fitting the reported data of daily populations of infected individuals. Each of these parameters can be assigned a physical meaning, which help quantify certain generally held views; a detailed discussion of the meaning of these parameters can be found in [ 18 ]. Moreover, the quantitative picture that emerges from the values of these parameters produces a rather accurate picture of the severity of the epidemic in the various countries, and the effect of the intervention measures if and when any were taken.…”

“…Recently, the forced-SIR (FSIR) model was proposed by the authors and used to describe the evolution of the COVID-19 pandemic in a representative set of countries [ 18 ]. This model treated the evolution of the infected population as a single wave (single peak wave).…”

A quantitative framework for exploring exit strategies from the COVID-19 lockdown

“…As some relevant examples, we mention the works of [17] , [18] (while similar methods have been used in nonlinear engineering and mathematical physics problems [19] , [20] , [21] , [22] ). In the recently very highly active front of COVID-19 modeling, some of the efforts have been directed at modeling the early stages of the pandemic [23] ; others have focused on designing a pandemic response index (to quantify/rank the response of different countries) [24] or towards quantifying the response of different regions within a country, e.g., the states within the USA [25] . Similarly, models have focused on cruise ships [26] , on cities [27] , as well as states/provinces [27] , [28] , [29] , [30] , but also various countries [15] , [31] , [32] , [33] , [34] , [35] , aside from the prototypical examples of Wuhan, China [36] , and some among the hard-hit Italian provinces [37] .…”

The Reference Model: An Initial Use Case for COVID-19

“…The outbreak of the COVID-19 pandemic has resulted in a plethora of computational attempts to model the disease. Those attempts include: an attempt to provide a vulnerability risk score to patients according to their individual characteristics [ 1 ], attempts to model temperature effects on the virus [ 2 ], attempts to compute undocumented cases as well as duration parameters [ 3 ], attempts to predict peak active cases using two contrasting models and their combination [ 4 ], prediction of daily new cases and death in Brazil by using a statistical function [ 5 ], a well publicized prediction model that can also asses hospital resources [ 6 ], a model using machine learning techniques to fit data to multiple locations worldwide by an independent modeler with open source implementation [ 7 ], a sophisticated engine using graphical user interface and high performance computing (HPC) that can include mobility information [ 8 ], an attempt to model 100 days using a traditional susceptible-infected-recovered (SIR) approach applied to multiple countries worldwide [ 9 ], another application of the SIR model for Cape Verde Islands [ 10 ], and another differential equation based urban model that focuses on transmission through travel in different modes of transportation [ 11 ]. The US Centers for Disease Control and Prevention (CDC) has also collected and assembled an ensemble model listing many models [ 12 ].…”