Transmission Dynamics of COVID-19 and Impact on Public Health Policy

B Shayak; Mohit Manoj Sharma; Richard H Rand; Awadhesh Kumar Singh; Anoop Misra

is the author/funder, who has granted medRxiv a license to display the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.03. 29.20047035 doi: medRxiv preprint thereafter. The data ends on the 33 rd day (yesterday) with the number of cases at 9037. The raw data is given below, as Table 1 . An interesting feature of the data is that the growth rate rises to a maximum of about 700/day on days 08-10, then falls rapidly upto day 20 but thereafter becomes almost constant at 100/day for the next 13 days as the cases increase linearly with time. 01 02 03 04 05 06 07 08 09 10 11 Cases 204 346 602 763 977 1261 1766 2337 3150 3736 4212 Day 12 13 14 15 16 17 18 19 20 21 22 Cases 4812 5328 5766 6284 6767 7134 7382 7513 7755 7869 7979 Day 23 24 25 26 27 28 29 30 31 32 33 Cases 8086 8162 8236 8320 8413 8565 8652 8799 8897 8961 9037 This data has to be fitted to a curve for w(t), obtained from our model (7) . When fitting, the first thing to note is that South Korea has a population of 5,17,00,000 [53], in comparison to which the number of cases is negligible. This is because the bulk of the outbreak occurred in several localized regions which were very effectively cut off from mingling with other, uncontaminated regions. Hence, the initial x in this case will not be South Korea's entire population but an effective value which takes into account the heavily localized character of the affected regions, where mixing of population took place. We find that this effective value naturally comes out of an attempt to fit the known curve. Since Ï„1 and Ï„2 represent biological properties of the virus, we keep them fixed at 7 and 3 days respectively. South Korea has been extremely proactive at testing, so we set Ï„3 = 2. We keep the seeding period at 15 days, assuming linear growth of y and w during this phase. We choose the growth rate so that w becomes equal to 100 at the start of the free evolution and make the ad hoc assumption that at this time there are thrice as many undetected cases as reported ones (this assumption is harmless -changing the number from 3 to 10 makes almost no difference with respect to the trajectory of the disease). Thereafter, we find the following effects of varying parameters :