Jeffrey Shaman - February 17, 2021

Jeff Shaman, PhD, Professor, Department of Environmental Health Sciences, Mailman School of Public Health, Columbia University, Burden and characteristics of COVID-19 in the United States during 2020: implications for vaccine distribution

Dr. Shaman interested in the establishment of mathematical models to estimate epidemiological characteristics of COVID-19 with various parameters such as R0: Basic reproduction number, Rt: time-varying reproduction number, Re: effective reproduction number, Ascertain infectious rate, Community infectious rate, Case fatality rate, Infection fatality rate and Population susceptibility. To estimate everything observed, they developed an inference framework comprising a mathematical metapopulational model and a Bayesian inference algorithm to simulate the data. They collected mobility and confirmed disease cases. The model system runs at the county scale, but the model system would not fit well if there are few cases or observations, such as very small counties. The model provides daily estimates of characteristics. The produced fit in both national and New York metropolitan area scale showed good estimation of outbreaks in summer and fall/winter after the initial outbreak in spring 2020. Similar quality of fitting showed in different places such as Los Angeles, Miami, Phoenix and Chicago. They also cross-validate the system with CDC data in accumulated population

The estimated ascertainment rate was graphed nationally and in five metropolitan areas, including N.Y., Chicago, L.A., Phoenix and Miami. The national average ascertainment rate was 21% in 2020. Population susceptibility estimation showed how the susceptible population changed over time. They estimated 68% of people remain susceptible and not infected nationally at the end of the year. Community infectious rate estimated less than 0.5% in spring and increased up to 1% in Dec on a national scale. The infectious fatality rate was above 1% early on the outbreaks, but it decreased to 0.3% at the end of the year, which is ten times higher than seasonal flu estimation indicating a high initial surge of the disease was deadly.  Six Vaccination scenarios (S0-S5) was introduced along with the vaccination schedule. Additional 29 million infections are estimated until late July 2021 if the current restriction gets relaxed at the beginning of February and an additional 6 million infections if the restriction is lifted in March with vaccination. If the restriction strengthens until Feb, 9 million fewer infection is estimated and if strengthen until late July, 19million fewer infection is estimated with vaccination.