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

I presented a rough estimate of a ROC curve to simulate how increasing the number of tests slow the spread of COVID-19. An important point that is easily overlooked is that knowing the TP rate relative the number infection cases is likely more useful than knowing the absolute number of tests or number of tests per capita, as is often reported in the news media. In fact, the relation between the number of tests and proportion of detected cases out of all positive cases is indicated by the TP rate. In the current crises, this is unknown. Nonetheless, with plausible values of the TP rate one can model the impact of increasing TP rates. With a modified SEIR model, I have shown that simulations of virus spread can produce valid estimates of currently infected persons and enable estimates of future growth under different scenarios.