Countries highly exposed to incoming traffic from China were expected to be at the highest risk of COVID-19 spread. However, COVID-19 case numbers (infection levels) are negatively correlated with incoming traffic-level. Moreover, infection levels are positively correlated with population-size, while the latter should only affect infection-level once herd immunity is reached. These could be explained if a low-virulence strain (LVS) began spreading a few months earlier from China, providing immunity from the later emerging known SARS-CoV-2 high-virulence strain (HVS). We find that the dynamics of the COVID-19 pandemic depend on the LVS and HVS spread doubling-times and the delay between their initial onsets. We find that LVS doubling-time to be $T_L\sim1.59\pm0.17$ times slower than the HVS ($T_H$), but its earlier onset allowed its global wide-spread to the levels required for herd-immunity. In countries exposed earlier to the LVS and/or having smaller population-size, the LVS achieved herd-immunity earlier, allowing less time for the spread of the HVS, and giving rise to lower HVS-infection levels. Such model accurately predicts a country's infection-level ({ \rm R^{ 2}=0.74}; p-value of { \rm 5.2\times10^{ -13}}), given only its population-size and incoming-traffic from China. It explains the negative correlation with incoming-traffic ($c_{ exp}$), the positive correlation with the population size (n_{ pop}) and their specific relations (${ \rm N}_{ { \rm cases}}\propto n_{ pop}^{ { \rm T_{ L}/{ \rm T_{ H}}}}\times c_{ exp}^{ { \rm T_{ L}/{ \rm T_{ H}-1}}}$). We find that most countries should have already achieved herd-immunity. Further COVID-19-spread in these countries is limited and is not expected to rise by more than a factor of 2-3. We suggest tests/predictions to further verify the model and biologically identify the LVS, and discuss the implications.
Categories: q-bio.PE physics.soc-ph