Predicting Whom to Test is More Important Than More Tests - Modeling the Impact of Testing on the Spread of COVID-19 Virus By True Positive Rate Estimation

Paul F Rodriguez

It is not known how effective limited COVID-19 testing can be for slowing the spread of the virus, but there is some preliminary cases that can help give rough estimates of TP and FP rates from the number of testing per day. For example, South Korea health board reported to have been testing up to about 13000-15000 individuals per day, from when the maximum increase in cases was near 1000 per day at the end of February up to the end of March (South Korean Public Health Board). The new cases rate has dropped to about 100 per day (March 28 th ). The ratio of 15-to-1 ratio of tests-per-new-cases at the peak is much higher than in US states like New York and California currently, and these and others states are trying to rapidly increase these number (COVID Tracking Project). Studies suggest that COVID-19 can incubate for an average of 6.4 days (e.g. Prem et al., 2020) . In the standard SEIR epidemiological model new infected individuals (I) come from the exposed incubating group (E). Therefore, if one assumes that the COVID-19 test is near 100 percent accurate for candidates from both groups, and that results are produced in 1 day, then at the peak of new cases in South Korea there were about 6000 cases in group E and the TP rate was ~ 0.16. Of course, this is an approximation that does not take into account many other factors, such as travelers, individuals who have symptoms but do not have COVID-19, retesting, or the accuracy of the test itself, as opposed the accuracy of selecting whom to test.