On frequent wide scale rapid testing programs as foundation to slow #COVID19

“Total cost, in tens of billions, would still be far less than the many trillions of dollars of... continued economic loss”

True!

(One comment about the headline...
1/x)
forbes.com/sites/williamh…
Headline says as much as 1M Americans getting COVID daily. This is based on a back of the envelope calculation

We know that only 10% of cases may be diagnosed. So if we have 100k diagnosed daily, then the thinking is that x10 = 1 million

But this misses a crucial piece

2/x
Even though 100k+ new cases DETECTED daily, and we can multiply by 10 for underdiagnosis and get 1M... this is not NEW cases happening daily. This is prevalence of virus positivity daily....

To convert to daily INCIDENCE (Ie new infections daily) takes another step

3/x
People remain positive for 10-30 days average. So those cases accumulate daily. 10 people today + 10 tmrw + 10 the next day and so on for 20 days.... Prevalence may be 200 positives in the community, but new cases daily is still just 10 new cases/day

Prevalence builds up

4/x
So even for an Acute respiratory infection (vs a chronic infection) we must be sure to consider incidence (true new events daily) vs prevalence (number positive daily). These are different and with a 10-30 day positivity, this building of prevalence vs incidence is important

5/5
And this is the most bare bones consideration.... SO many more nuances that must be achieved to fully estimate the true number. This thread was more to point out a most basic piece about prevalence vs incidence....

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Michael Mina

Michael Mina Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @michaelmina_lab

9 Nov
Wow! Good news! Pfizer press release says vaccine 90% effective at preventing symptomatic #COVID19!

If results hold, this is a very good efficacy for a vaccine, higher than I was expecting, by far.

From this, we can be cautiously optimistic!

1/

nytimes.com/2020/11/09/hea…
We do have to be cautious in interpreting these early results (from a press release) and we don’t know if same level of protection will persist long-term or if this is driven a lot by early effects more than robust immune memory

Nevertheless, this is a positive data point

2/
A couple of other items we will need to look out for

Even it protects from symptomatic disease, does it also protect from onward transmission?

The assumption is it would largely do this (and this would underpin herd immunity). But blocking transmission isn’t a sure bet

3/
Read 6 tweets
8 Nov
Frequent testing is needed to detect people early before infecting others. They must be highly scalable and for buy in, very convenient - i.e. home use

10M Rapid tests/day (i.e. antigen bc scalable / simple) can form the foundation to build back economy

theatlantic.com/health/archive…
Frequent tests alone won’t be the end all be all of this pandemic. Not by a long shot. These can greatly reduce odds of transmission - when used en masse, odds of onward transmission (i.e. Rt) can plummet.

To work, also need huge social/economic safety nets, + masks/distancing.
Although new administration will not officially start till January, we have options now. @JoeBiden and team led by @vivek_murthy can start now to engage manufacturing and begin planning/building these at scale w promise of payment once in office, as one example.
Read 5 tweets
7 Nov
Adoption of strategies that aim to allow people to test themselves, privately,simply, at home, 1-2x/week could help suppress outbreaks quickly. Especially in context of other public health measures.

~10M/day to achieve herd effects and drop incidence
healthaffairs.org/do/10.1377/hbl…
We don’t need perfect compliance, at all. We just need decent compliance - we can achieve this. Like herd immunity, we don’t need this to be everyone, just enough people to drop R<1. I estimate 10% of people in a community would test at home per day. So need not have perfection
We still need reporting so public health officials can keep track

No problem... we work w Google, Apple and whoever else to make voluntary reporting easy as a FaceTime call. If I can see my twin brother 2000 miles away w the push of a button, I should be able to report a result
Read 6 tweets
5 Nov
A new article in ⁦@nytimes⁩ states rapid tests falter in asymptomatics

Warning of their use for asymptomatic screening

Says: among random asymptomatic #COVID19 screens, Rapid Tests caught only 32% of PCR +ve people

This is misleading...

1/

nytimes.com/2020/11/02/hea…
Tests have to be matched to their purpose. If doing asymptomatic screening - you are looking for INFECTIOUS people.

Importantly, MOST (~70%+) of the time someone is PCR +ve, they are POST-infectious!

2/
Since antigen tests meant to detect viable/live virus, we only EXPECT them to be positive about 30% of the time of PCR

Thus, finding 32% of positive PCR tests in random asymptomatic screening is absolutely the EXPECTED result for a test looking for infectious people

3/
Read 14 tweets
3 Nov
NEW terrific study on rapid tests

The importance of this figure canNOT be overstated!

Rapid tests caught 80% of PCR positives but all missed had very low viral loads

Thus, as expected it will successfully capture ppl most likely to be contagious

1/

medrxiv.org/content/10.110…
The specificity of the PanBio, BinaxNow and SD Biosensor tests (the three leading manufactured rapid tests in the world) are looking very good!

In this paper, specificity was 100% in >400 samples... thus >99.2%

Now multiple studies showing very high specificity!

2/
The sensitivity metrics AND specificity metrics are now completely in alignment with what we have proposed for frequent rapid testing that can control outbreaks without vaccines.

medrxiv.org/content/10.110…

3/
Read 5 tweets
3 Nov
Can contact tracing work for #COVID19?

Much effort assumes contact tracing works for COVID - surveillance testing assumes so

I’ve had major concerns - this paper reinforces it.

For fast pandemics, we cant act only on what has been successful before

1/

jamanetwork.com/journals/jamai…
In this paper, 791 #COVID19 positive people were contact traced AND universal testing was recommended for their close contacts.

Ultimately only 120 people were discovered.

2/
If COVID has an R of ~1.3… those 791 people perhaps led to >1000 additional cases.

If 120 were discovered... ~900 were not.
And this was an organized operation.

3/
Read 13 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!