Thread by @sfiscience

Streaming now: @michaelmina_lab of @HarvardBiz gives today's SFI Seminar on "#Testing our way out of #COVID19: A bumpy but possible path forward."

facebook.com/santafeinstitu…

Feat. work with Former SFI Postdoc @DanLarremore.

Follow this thread for highlights...

Now may be the best window to squash #COVID19 before flu season begins, but pandemic surveillance—especially in the US—has been a mess.

What are the relative strengths of virus testing vs. serological testing?

@michaelmina_lab speaking at SFI right now:
facebook.com/santafeinstitu…

"Really baseline testing doesn't give you enough information. If it tells you there is 5% infection in a nursing home it doesn't tell you if the virus is increasing. But serological testing can. These are complementary tools."

@michaelmina_lab at SFI now:
facebook.com/santafeinstitu…

Lateral flow assays—paper #COVID19 tests that are easy to manufacture rapidly and at scale—don't create the same testing bottlenecks as #PCR.

Can less sensitive, but more frequent, testing outperform more accurate methods?

@michaelmina_lab at SFI now:
facebook.com/santafeinstitu…

"If you have a very fast that is very sensitive, it's probably not going to be very scalable, and it's going to be expensive. But what if we had very fast results that are inexpensive, but 100x less sensitive at the molecular level? Can that be a useful tool?"
- @michaelmina_lab

"[#COVID19] is a virus that, when it is infecting people, sometimes gets up to a trillion copies. This isn't #HIV. It doesn't really matter if you're catching every molecule. By focusing on a super-sensitive test, you can't do frequent testing."
- @michaelmina_lab at SFI now

"The quicker and more frequently you can test people, the more likely you are to detect infected people and remove them from the pool. It's the frequency, more than the sensitivity, that counts."

- @michaelmina_lab on work with @DanLarremore:

facebook.com/santafeinstitu…
#pandemic

Why don't we have rapid #COVID19 testing today? @michaelmina_lab of @HarvardChanSPH speaking at SFI today on how test approval criteria prioritized 90% sensitivity for low viral loads—orders of magnitude lower than what people are actually transmitting:
facebook.com/santafeinstitu…

In brief:

• Paper #COVID19 tests are less sensitive in part because they don't detect residual RNA from individuals no longer transmitting the disease.

• Capturing transmissibility *on time* matters more than very accurate but uselessly slow surveillance.

@michaelmina_lab:

(Correction: not @HarvardBiz, @HarvardChanSPH. Insert self-effacing joke about rapid but insensitive typo testing...)