Dynamics of Transmission and Control of COVID-19: A Real-time Estimation Using the Kalman Filter

The COVID-19 pandemic has become the center of attention for both researchers and authorities. In this paper, we propose and test a methodology to estimate the daily effective reproduction number (ℛ _t ) through the lens of the Kalman Filter and Bayesian estimation. Moreover, we apply our method to data from the current COVID-19 pandemic in China, Italy, Japan, and South Korea. We correlate our findings with the implementation of control measures in each of these countries. Our results show that China, Italy, and South Korea have been able to reduce ℛ _t over time. We find significant heterogeneity in the way ℛ _t decreases across countries. For instance, China reduced ℛ _t from its peak to below one in 19 days, while South Korea achieved the same reduction in 12 days. In contrast, it has taken Italy almost a month to reach similar levels. We hypothesize this is related to how strict, enforceable, and comprehensive are the implemented policies.