Early dynamics of transmission and control of COVID-19: a mathematical modelling study

Adam J Kucharski; Timothy W Russell; Charlie Diamond; Yang Liu; John Edmunds; Sebastian Funk; Rosalind M Eggo

There are several other limitations to our analysis. We used plausible biological parameters for SARS-CoV-2 based on current evidence, but these values may be refined as more comprehensive data become available. However, by fitting to multiple datasets to infer model parameters, and performing sensitivity analyses on key areas of uncertainty, we have attempted to make the best possible use of the available evidence about SARS-CoV-2 transmission dynamics. Further, we used publicly available connectivity and risk estimates based on international travel data to predict the number of exported cases into each country. These estimates have shown good correspondence with the distribution of exported cases to date (23) , and are similar to another risk assessment for COVID-19 with different data (24) . We also assumed that the latent period is equal to the incubation period (i.e. individuals become infectious and symptomatic at the same time) and all infected individuals will eventually become symptomatic. However, there is evidence that transmission of SARS-CoV-2 can occur with limited reported symptoms (25) . We therefore conducted a sensitivity analysis in which transmission could occur in the second half of the incubation period, but this did not change our overall conclusions (Appendix, page 7). We also explored having a larger initial spillover event and also using different sources for flight connectivity data, neither of which changed the conclusions of the analysis. In our analysis of new outbreaks, we also used estimates of individual-level variation in transmission for SARS and MERS-CoV to illustrate potential dynamics. However, it remains unclear what the precise extent of such variation is for SARS-CoV-2 (13); if transmission were more homogenous than SARS of MERS-CoV, it would increase the risk of outbreaks following introduced cases. As more data becomes available, it will be possible to refine these estimates, and therefore we made an online tool so users can explore these risk estimates if new data become available (Appendix, page 4) Our results demonstrate that there was likely substantial variation in SARS-CoV-2 transmission over time, and suggest a decline in transmission in Wuhan in late January around the time that control measures were introduced. If COVID-19 transmission establishes outside of Wuhan, understanding the effectiveness of control measures in different settings will be crucial for understanding the likely dynamics of the outbreak, and the likelihood that transmission can eventually be contained.