We've been playing around with @AdamJKucharski @cmmid_lshtm #COVID19 contact tracing model ( https://www.medrxiv.org/content/10.1101/2020.04.23.20077024v1) to see what levels of symptomatic case testing and timeliness of case finding are required to keep the reproductive number <1 as we relax physical distancing. 1/

Without lots of distancing, we need to be testing a lot of symptomatic cases to keep R low (blue -> R<1 in these figures). Right now, we think the proportion of the population distancing is high (~0.7, bottom row), so levels of testing don't matter too much. 2/

With less physical distancing (middle rows in the figure), we need to be testing at least half of symptomatic index cases to keep R low. The amount varies depending on the reach and timeliness of case finding. But regardless, it's a lot of testing. 3/

This is a model, there are assumptions, etc. etc. and the precise values shown here should not be taken as gospel, but the message is clear: without expanded testing, being really good at contact tracing isn't going to be enough as we start to open up. End/