Dynamic Modeling to Identify Mitigation Strategies for Covid-19 Pandemic

Patrick Jenny; David F Jenny; Hossein Gorji; Markus Arnoldini; Wolf-Dietrich Hardt

Obviously, the scenario that the whole susceptible population is tested at once would lead to a trivial disease free state. However in the absence of such capacity, individuals would be tested at different schedules. Let us consider a situation where each person is tested once every N days. Therefore we focus on the set of days D = {1, ..., N}. An individual is infected at some random time x. To characterize x, we assume that the likelihood of getting infected does not vary much during these days. This is exactly the case, if testing is applied at an intensity such that R 0 = 1; see Fig. 7c . Hence x becomes uniformly distributed in the interval [1, N] . The incubation time, which is the time x s between infection and the onset of symptoms, is assumed to obey a log-normal distribution [12] . We assume an incubation time of x s = 5.84 ± 2.98 days [8] . Moreover, we suppose that the testing result becomes available 1/2-day after the test is taken, and that the results are free from false negatives and false positives. The problem of testing frequency hence is reduced to finding N satisfying