Fig. 2: Nowcast of the transmission dynamics of COVID-19 using Ct distribution.

a Nowcasting \({R}_{t}\) using the Ct-based method over four representative weeks. Gray bars represent the number of laboratory-confirmed local cases by date of reporting, black lines and shaded areas indicate the mean and 95% CrIs for incidence-based \({R}_{t}\), while dots and vertical lines represent mean and 95% prediction intervals for Ct-based \({R}_{t}\) estimated from Eq. (7) (same for b, c). In this panel, \(n=\) 19, 33, 59, and 74 daily values from top to bottom panels sequentially. b, c Comparison of incidence-based \({R}_{t}\) and Ct-based \({R}_{t}\) over the training period (July 2020–August 2020, b) and the testing period (November 2020–March 2021, c). \(n=57\) daily values for training period in panel b and \(=146\) for testing period in panel c (same for d). d Distributions of incidence-based \({R}_{t}\) under various intervals of Ct-based \({R}_{t}\). Boxes represent the IQR and median of incidence-based \({R}_{t}\) under the corresponding interval of Ct-based \({R}_{t}\) over training (purple) and testing (pink) periods, lower whiskers represent the minimum and upper whiskers represent either the maximum or the largest values that are within the distance of 1.5 times the IQR of all incidence-based \({R}_{t}\,\) under various Ct- based \({R}_{t}\) intervals, dots represent values beyond the lower and upper whiskers.