The brilliant Harlan Krumholz @hmkyale and I have a new piece about coronavirus testing in @USATODAY.
usatoday.com/story/opinion/…
The reality is that this nation has "decided" not to devote sufficient resources to testing everyone.
Therefore: We're testing the wrong people. 🧵1/
usatoday.com/story/opinion/…
The reality is that this nation has "decided" not to devote sufficient resources to testing everyone.
Therefore: We're testing the wrong people. 🧵1/
What makes SARS-CoV-2 so hard to contain?
Symptom-free contagion.
So, we should simply ASSUME that people with suspicious symptoms are positive and NOT waste tests on them in many, many cases (some nuance to this, we recognize).
Instead, divvy up most of our tests in 3 ways...
Symptom-free contagion.
So, we should simply ASSUME that people with suspicious symptoms are positive and NOT waste tests on them in many, many cases (some nuance to this, we recognize).
Instead, divvy up most of our tests in 3 ways...
1. Use around 2000 tests per day (in a state like Massachusetts) to test a random sample of volunteers. This is similar to how you do election polling.
A good small random and representative sample of the population is better than a large non-random one.
A good small random and representative sample of the population is better than a large non-random one.
2. Routine widespread testing of all people in certain high-risk places. Nursing homes, detention facilities, etc.
Places where controlling spread is challenging, and where death rate (in particular nursing homes) is so high.
As for the rest of the tests?
SYMPTOM-FREE WORKERS!
Places where controlling spread is challenging, and where death rate (in particular nursing homes) is so high.
As for the rest of the tests?
SYMPTOM-FREE WORKERS!
We recommend testing workers with:
-a high number of public contacts
-who are SYMPTOM FREE.
Can be done on a volunteer basis. We can test the willing.
Example: if you found out that a grocery store employee had symptom-free disease, you just saved a lot of lives...
-a high number of public contacts
-who are SYMPTOM FREE.
Can be done on a volunteer basis. We can test the willing.
Example: if you found out that a grocery store employee had symptom-free disease, you just saved a lot of lives...
In some cases, you could shut that store down, deep clean it (wait 72 hours if you're worried about surfaces), and test all employees.
This "surgical strike" shutdown is more likely to be accepted by the public going forward.
And current testing strategies can't achieve this.
This "surgical strike" shutdown is more likely to be accepted by the public going forward.
And current testing strategies can't achieve this.
There's lots of interest in this type of approach.
But we need to roll it out now so that we can track symptom-free spread. That's what is DRIVING the pandemic and what usual things (like contact tracing) can't capture.
What are some places we should test? Mass transit is one.
But we need to roll it out now so that we can track symptom-free spread. That's what is DRIVING the pandemic and what usual things (like contact tracing) can't capture.
What are some places we should test? Mass transit is one.
We are interested in hearing other ideas.
Please share!
Above all, we need to intelligently develop a plan, not just test whoever shows up.
Yes, on-demand testing is good if we can do it.
But to save the SYSTEM, we need to think SYSTEMATICALLY.
Thx!
usatoday.com/story/opinion/…
Please share!
Above all, we need to intelligently develop a plan, not just test whoever shows up.
Yes, on-demand testing is good if we can do it.
But to save the SYSTEM, we need to think SYSTEMATICALLY.
Thx!
usatoday.com/story/opinion/…
