Modeling Transmission Dynamics of SARS-CoV-2













In response to the ongoing global outbreak of SARS-CoV-2, I have led and participated in several modeling studies addressing important and urgent questions in understanding the transmission of SARS-CoV-2.

A list of my press interviews and media coverage on COVID-19 is here.

Projecting COVID-19 spread in the United States

Together with colleagues at Mailman, we use a metapopulation SEIR model to prject the spread of COVID-19 cases in the United States at county level. Updated projections are posted on GitHub. Three online tools allow users to visualze the updated weekly projections of new COVID-19 cases, new infections, and available critical care beds in states and counties across the United States under a variety of social distancing and hospital response scenarios over a six-week period: (1) a data visualization tool that graph projections over time, (2) a mapping tool that charts county-level projections, and (3) animated maps. We also contribute to the collaborative efforts of COVID-19 forecast led by CDC.

Our early simulation of COVID-19 spread in the US was covered by the New York Times "Without Urgent Action, Coronavirus Could Overwhelm U.S., Estimates Say".

Undocumented infection fuels rapid spread of coronavirus in China

Estimation of the prevalence and contagiousness of undocumented novel coronavirus (SARS-CoV-2) infections is critical for understanding the overall prevalence and pandemic potential of this disease. Here we use observations of reported infection within China, in conjunction with mobility data, a networked dynamic metapopulation model and Bayesian inference, to infer critical epidemiological characteristics associated with SARS-CoV-2, including the fraction of undocumented infections and their contagiousness. We estimate 86% of all infections were undocumented (95% CI: [82%-90%]) prior to January 23, 2020 travel restrictions. Per person, these undocumented infections were 55% as contagious as documented infections ([46%-62%]) and were the source of infection for two-thirds of documented cases. These findings explain the rapid geographic spread of SARS-CoV-2 and indicate containment of this virus will be particularly challenging. See the research article published in Science.

Counterfactual simulations of COVID-19 spread in the United States

Assessing the effects of early non-pharmaceutical interventions on COVID-19 spread is crucial for understanding and planning future control measures to combat the pandemic. We use observations of reported infections and deaths, human mobility data, and a metapopulation transmission model to quantify changes in disease transmission rates in US counties from March 15 to May 3, 2020. We find that marked, asynchronous reductions of the basic reproductive number occurred throughout the US in association with social distancing and other control measures. Counterfactual simulations indicate that, had these same measures been implemented 1-2 weeks earlier, substantial cases and deaths could have been averted, and that delayed responses to future increased incidence will facilitate a stronger rebound of infections and death. This study, published in Science Advances, was reported by the New York Times "Lockdown Delays Cost at Least 36,000 Lives, Data Show".

Compound risks of hurricane evacuation amid the COVID-19 pandemic in the United States

In recent years hurricane evacuations in the United States have displaced millions of people from their homes. Amid the ongoing coronavirus disease 2019 (COVID-19) pandemic, such an evacuation—and the associated increase in human-to-human interactions—poses an additional risk of disease transmission. In this study, we use an epidemiological model to simulate a hypothetical hurricane evacuation from southeast Florida. We find that evacuation is likely to increase the total number of COVID-19 cases. However, directing evacuees to locations experiencing lower COVID-19 transmission rates and simultaneously minimizing human contact during evacuation could reduce the excess number of infections. Our results indicate that evacuation-induced COVID-19 infections can be minimized by optimizing evacuation plans based on real-time information about disease incidence and transmission. This study is published in GeoHealth.

Role of meteorological factors in the transmission of SARS-CoV-2 in the United States

Improved understanding of the effects of meteorological conditions on the transmission of SARS-CoV-2, the causative agent for COVID-19 disease, is needed. Here, we estimate the relationship between air temperature, specific humidity, and ultraviolet radiation and SARS-CoV-2 transmission in 2669 U.S. counties with abundant reported cases from March 15 to December 31, 2020. Specifically, we quantify the associations of daily mean temperature, specific humidity, and ultraviolet radiation with daily estimates of the SARS-CoV-2 reproduction number (Rt) and calculate the fraction of Rt attributable to these meteorological conditions. Lower air temperature (within the 20–40 °C range), lower specific humidity, and lower ultraviolet radiation were significantly associated with increased Rt. The fraction of Rt attributable to temperature, specific humidity, and ultraviolet radiation were 3.73% (95% empirical confidence interval [eCI]: 3.66–3.76%), 9.35% (95% eCI: 9.27–9.39%), and 4.44% (95% eCI: 4.38–4.47%), respectively. In total, 17.5% of Rt was attributable to meteorological factors. The fractions attributable to meteorological factors generally were higher in northern counties than in southern counties. Our findings indicate that cold and dry weather and low levels of ultraviolet radiation are moderately associated with increased SARS-CoV-2 transmissibility, with humidity playing the largest role. Check out the study published in Nature Communications. This is a collaborative study led by colleagues Yiqun Ma, Robert Dubrow and Kai Chen at Yale.

Burden and characteristics of COVID-19 in the United States during 2020

The COVID-19 pandemic disrupted health systems and economies throughout the world during 2020 and was particularly devastating for the United States, which experienced the highest numbers of reported cases and deaths during 20201–3. Many of the epidemiological features responsible for observed rates of morbidity and mortality have been reported4–8; however, the overall burden and characteristics of COVID-19 in the United States have not been comprehensively quantified. Here we use a data-driven model-inference approach to simulate the pandemic at county-scale in the United States during 2020 and estimate critical, time-varying epidemiological properties underpinning the dynamics of the virus. The pandemic in the US during 2020 was characterized by national ascertainment rates that increased from 11.3% (95% credible interval (CI):8.3 – 15.9%) during March to 24.5% (18.6 – 32.3%) during December. Population susceptibility at year’s end was 69.0% (63.6 – 75.4%), indicating that roughly one third of the US population had been infected. Community infectious rates, the percentage of people harbouring a contagious infection, rose above 0.8% (0.6 – 1.0%) before the end of the year, and were as high as 2.4% in some major metropolitan areas. In contrast, the infection fatality rate fell to 0.3% by year’s end. Find more details in the paper published in Nature.