Thread by @caraebrook

Excited to share our pre-print on optimizing #COVID19 control with asymptomatic surveillance testing, an inspiring collaboration with the #UCBerkeley SARS-CoV-2 testing lab @igisci.
medrxiv.org/content/10.110…
(with @graham_northrup, @ajehrenberg, @jadoudna, @BootsLab)
(1/

We built a model of #COVID19 dynamics in a university environment to explore the efficacy of behavior- and surveillance testing-based NPIs—and to advise our own campus lab on optimal use of testing resources
(2/

As has been shown previously by @morde, we demonstrate the effectiveness of group size limits to deter superspreading, assuming a negative binomial R0. The group size limit of 12 imposed by the Berkeley Public Health Dept is a good one; limits of 50+ are not effective at all
(3/

As has been shown by others (@michaelmina_lab
and colleagues, @CT_Bergstrom), we find asymptomatic testing to be the most effective when prioritizing frequent testing with rapid turnaround time over test sensitivity—another win for antigen test advocates.
(4/

Critically, we show how even testing regimes with slow (7+ day) turnaround times are useful when combined with contact tracing such as the NotifyCA app rolled out today on UC campuses. Testing is vital to halting downstream chains in transmission.
(5/

Together, combined behavior-based & testing-based NPIs limit variation in daily case counts. More predictable epidemics are less likely to spiral out of control. Every case avoided reaps compounding benefits in cases saved later. This is vital to remember now more than ever!
(6/

We’re using these guidelines to combat #COVID19 at #UCBerkeley. Visit our Github Repo to use our model in your university or small community environment. Let’s keep up the hard work and stay focused in the months ahead! (7/7)
github.com/carabrook/Berk…