Using random testing to manage a safe exit from the COVID-19 lockdown

Markus Mueller; Peter Derlet; Christopher Mudry; Gabriel Aeppli

Even if policy measures should eventually not be taken in a region-specific manner, it is very useful to study a regionally refined model of epidemic dynamics. Indeed a host of literature exists that studies epidemiological models on lattices and analyzes the spatial heterogeneities. [17, 18] In certain circumstances those have been argued to become even extremely strong. [19] In the present paper, we will content ourselves with a few FIG. 6. Time after which a significant positive growth rate is confirmed in the worst case scenario for which the growth rate jumps to k 1 = 0.23 after reboot. An intervention will be triggered in 3-4 days, since in the case that such a strong growth must be suspected, one should apply a small confidence parameter α ≈ 1. Results are shown for r = 15 000 and r = 20 000 tests a day. The circles are the mean values, the vertical lines indicate the standard deviations for the first intervention time.