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

Here, using preliminary online data reports, I derive rough approximations of TP rates from numbers of tests relative to a population to produce a ROC (Receiver Operating Characteristic) curve, which, to my knowledge, has not been presented yet even in such a rough form. I then modify a standard SEIR model (R, Epidynamics package) to use TP rate to decrease the pool of infected individuals. Using model parameters published elsewhere, a grid search of TP rates and beta (time to contact) values shows that, in regions of parameter space that match empirical data, increasing TP rate can have critical impacts in slowing spread of the virus. TP rate can increase with large increases in testing or by improving the prediction on whom to test for. In fact, public health agencies should be gathering data for such endeavor which could be done independently of any need for material resources of testing kits.