Explaining the Bomb-Like Dynamics of COVID-19 with Modeling and the Implications for Policy

Abstract Understanding the transmission dynamics of the severe acute respiratory syndrome coronavirus 2 (SARS-COV-2) is critical to inform sound policy decisions. We demonstrate how the transmission of undetected cases with pre-symptomatic, asymptomatic and mild symptoms, which are typically underreported due to lower testing capacity, explained the “bomb-like” behavior of exponential growth in the coronavirus disease 2019 (COVID-19) cases during early stages before the effects of large-scale, non-pharmaceutical interventions such as social distancing, school closures, or lockdowns. Using a Bayesian approach to epidemiological compartmental modeling, we captured the initial stages of the pandemic resulting in the explosion of cases and compared the parameter estimation with empirically measured values from the current knowledgebase. Parameter estimation was conducted using Markov chain Monte Carlo (MCMC) sampling methods with a Bayesian inference framework to estimate the proportion of undetected cases. Using data from the exponential phase of the pandemic prior to the implementation of interventions we estimated the basic reproductive number (R0) and symptomatic rates in Italy, Spain, South Korea, New York City, and Chicago. From this modeling study, R0 was estimated to be 3·25 (95% CrI, 1·09-29·77), 3·62 (95% CrI, 1·13-34·89), 2·75 (95% CrI, 1·04-22·44), 3·31 (95% CrI, 1·69-20·55), and 3·46 (95% CrI, 1·01-34·41), respectively. For all locations, 3-25% of infected patients were identified with moderate and severe symptoms in the early stage of the COVID-19 pandemic. Our modeling results support the mounting evidence that potentially large fractions of the infected population were undetected with asymptomatic and mild symptoms. Furthermore, a significant number of models of transmission that do not account for these asymptomatic cases may lead to an underestimation of R0 and, subsequently, policies that do not sufficiently reduce transmission to contain the spread of the virus. Detecting asymptomatic transmission can help slow down the spread of SARS-CoV-2. Author Summary The spread of SARS-CoV-2 has led to a global pandemic that is still spreading across countries. We fitted a mathematical model to reported infections in Spain, Italy, South Korea, New York, and Chicago before any large-scale interventions, such as lockdowns and school closures, and found that undetected infected individuals drove the accelerated pace of transmissions. Due to the limited capacity in testing in many of the five locations, undetected cases were most likely asymptomatic and mild to moderate symptomatic infections. Given the explosive nature in the number of cases during the early phase of the pandemic and the latest serological surveys, our study suggested that most active cases were undetected. Other cohort studies have shown that a significant proportion of cases reported little or no symptoms. We also showed that early detection of asymptomatic and mild symptomatic cases can lead to a slower spread of SARS-CoV-2 as evident in South Korea. Policies targeting symptomatic individuals, such as travel restrictions on affected areas or quarantines of sick individuals, are not as effective because they neglect asymptomatic transmission events.