Dynamic estimation of epidemiological parameters of COVID-19 outbreak and effects of interventions on its spread.

Abstract

Background: A key challenge in estimating epidemiological parameters for a pandemic such as the initial COVID-19 outbreak in Wuhan is the discrepancy between the officially reported number of infections and the true number of infections. A common approach to tackling the challenge is to use the number of infections exported from the originating city to infer the true number. This approach can only provide a static estimate of the epidemiological parameters before city lockdown because there are almost no exported cases thereafter.Methods: We propose a Bayesian estimation method that dynamically estimates the epidemiological parameters by recovering true numbers of infections from day-to-day official numbers. To illustrate the use of this method, we provide a comprehensive retrospection on how the COVID-19 had progressed in Wuhan from January 19 to March 5, 2020. Particularly, we estimate that the outbreak sizes by January 23 and March 5 were 11,239 [95% CI 4,794–22,372] and 124,506 [95% CI 69,526–265,113], respectively.Results: The effective reproduction number attained its maximum on January 24 (3.42 [95% CI 3.34–3.50]) and became less than 1 from February 7 (0.76 [95% CI 0.65–0.92]). We also estimate the effects of two major government interventions on the spread of COVID-19 in Wuhan.Conclusions: This case study by our proposed method affirms the believed importance and effectiveness of imposing tight nonessential travel restrictions and affirm the importance and effectiveness of government interventions (e.g., transportation suspension and large scale hospitalization) for effective mitigation of COVID-19 community spread.Significance for public healthIn fighting global pandemic such as COVID-19, an important early task for understanding the spread is to closely monitor the infection size and assess the disease epidemiological parameters. The in- sights gained from the epidemiological parameter estimation enable public health practitioners to dynamically monitor the temporal spread trend and to quantitatively analyze the effectiveness of new public health policies. In this paper, we aim to address a key technical challenge potentially arising from the under-reporting issues in pandemic early periods, and critically re-examine the COVID-19 situation at the initial epicenter Wuhan city as a practically relevant case study. Methodological development for modeling dynamic evolution involving parameter estimation therefore has important public health applications and is expected to have significant impact on modeling practice for understanding future epidemic events well beyond COVID-19.