Pooled Testing Okayed in U.S. as Way of Boosting Covid Detection

[Yo_Mama_Been_Loggin]

Source: Bloomberg

(Bloomberg) -- The U.S. Food and Drug Administration has authorized so-called “pool testing” for Covid-19, a move aimed at broadening checks for the coronavirus and using fewer testing resources.

Quest Diagnostics Inc. will be able to test samples containing as many as four individual swab specimens, the agency said on Saturday in an emergency-use authorization.

The samples collected are then tested in a pool or “batch” using one Covid-19 test, rather than running each individual sample through its own test. If the pool is positive it means that one or more of the individuals tested may be infected, so each of the samples in that pool is then tested again, individually.

The authorization “is an important step forward in getting more Covid-19 tests to more Americans more quickly while preserving testing supplies,” FDA Commissioner Stephen Hahn said in a statement.

Read more: https://www.msn.com/en-us/money/companies/pooled-testing-okayed-in-us-as-way-of-boosting-covid-detection/ar-BB16Uh1i?li=BBnb7Kz

[rzemanfl]

1. This pandemic started when again? Is any other country so fucked up that they batch test people?

Government by lying fucks led to this.

[drray23]

22. other countries actually do it because its more efficient.

You can test a large number of people quickly that way.

[rzemanfl]

24. Okay, then why didn't we start doing that in say, March?

Other countries don't have "leaders" who lie and say anyone who wants a test can get one.

[soothsayer]

2. False negatives gonna mess up a whole group

[FBaggins]

4. Actually... it should help

Consider the possible distributions of four tests

None are infected and none test as infected - No change

One is infected and tests as infected - No change... and other three have a confirming test

One is infected but gets a false negative - No change (same chance of a false negative in solo test)

More than one is infected - better results... because any false negative would essentially get a second test

None are infected but results show false positive - everyone gets a confirming test.

[soothsayer]

6. 1-4 are infected, everyone gets a false negative?

[FBaggins]

10. If 1 in 4 is infected and they get a false negative

Then three out of four received correct negative results... and the one who is infected received the same results he would have received in a solo test.

[Xipe Totec]

13. With a positivity rate of 25%; what's current in Florida,

a pool of 4 samples has a probability of returning a positive result of 1.0 - (1.0 - 0.25)^4 = 0.683, meaning it will return a positive result in 68% of the cases.

This means that in 68% of the cases they will have to perform individual sample tests after performing the pooled test, so 5 tests instead of 4.

The average number of tests per sample will then be (5*0.683 + 1*0.316)/4 = 0.93

So they will perform 93 tests instead of 100, for every 100 samples processed.

Big Fucking Hairy Deal.

On the other hand, if the positivity rate was around 5%, which is where we would be, if the pandemic was under control,

1.0 - (1.0 - 0.05)^4 = 0.185, meaning 18.5% of the pooled tests will return a positive result.

(5*0.185 + 1*0.814)/4 = 0.343

So they would perform 34 tests instead of 100, for every 100 samples processed.

[drray23]

23. yes exactly.

Countries that are using this to mass test are doing it to contain the virus and have a small positivity rate. As you point out, once it gets high enough, you lose the benefits of the method.

[soothsayer]

31. That's a helpful explanation, thanks!

[Rebl2]

8. That was my thought

[CreekDog]

25. you can test it twice to reduce the false negative rate

so 4 people, 2 tests.

[FBaggins]

3. Clever if it's reliable

Even at a positivity rate of 10%... that could more than double the number of tests that they can perform.

[Xipe Totec]

14. See post #13 on the effects of positivity.

[rockfordfile]

19. I don't think so.

[CreekDog]

26. yep, they should do this

absolutely. it's about controlling the virus through testing and identifying people or groups of people that have it. whatever helps do that faster or more efficiently is more important than everyone getting their own individual test like it's sacrosanct. it's not. controlling the virus is sacrosanct, not ownership of individual test results.

sheesh, i can't believe people are up in arms.

they're going to start testing our sewage for hot spots (see Yosemite NP), another excellent idea.

[Mr.Bill]

5. I could see that being used in institutional settings where people are tested regularly

such as the military, prisons, schools, etc. It would not be difficult to retest those four people because you already have contact with them at the same location.

[jimlup]

7. This makes sense as a way to allocate scare critical resources...

But why are we having this conversation. We should be months ahead on this and instead we are months behind. I blame the WH.

[BadgerKid]

9. Our hospital is attempting batches of 100.

If the infection rate is going to explode, 4 isn’t going to cut it - at least in populated areas.

[FBaggins]

11. That must be something different

I suspect they're referring to the number of tests that they collect before sending them in for evaluation.

If it were a "pooling" like the OP discusses... then virtually every pool of 100 would be a positive hit and all 100 would need to be run individually.

I think the lowest positivity rate nationally was a bit over a month ago at ~5%. Running 100 tests in a pool would virtually guarantee a positive result.

If the infection rate is going to explode,

Just the opposite. This strategy is only useful with a comparatively low positivity rate. The higher it goes... the more likely that you need individual tests.

[Pobeka]

12. Your final sentence is what I heard an epidemiologist saying.

... that this batch testing only makes sense with infection rates substantially below what we generally have in the USA. The implication is this is being using politically, so you can claim we have ramped up the total number of tests on individual people, when in fact far fewer tests are being given.

Which is NOT to say there is no value in this type of testing.

But the sterile protocols have got to be sooo much stricter because a single false positive on a batch of N samples means you have to test another N samples, which would not be necessary with perfect protocols and the true negative result. There's a lot of places for samples to be contaminated in the chain of custody from original swab to the final lab test...

[CreekDog]

27. In Florida, 1 out of 59 people have tested postive

In the Bay Area, about 1 out of 225. In California overall, about 1 in 150.

So while 1 in 100 might be too much, 1 out of 50 or 25 might prove useful in case of workplaces or households.

[Xipe Totec]

15. With a positivity rate of 25% and a pool of 100 tests,

1.0 - (1.0 - 0.25)^100 = .999999999

Meaning 99.99999% of the pooled tests will return a positive result and you will then have to perform 101 tests for every 100 samples.

So long as the percent of tests coming back positive is high, pooled samples are useless.

You have to bring down the percent of positive cases first in order for pooled testing to be useful.

[CreekDog]

28. but lots of places like San Francisco have positivity rates of 3%

in New York, it's much, much lower, like 1%.

pooled testing makes sense.

actually, relatively few places have 25% positivity, so you could allocated individual testing to those places and do more pooled where the rates are far lower.

[cstanleytech]

16. For the math illiterate such as myself is this a good or bad change?

[caraher]

18. As the math literate have shown, it depends

If the rate of positive tests is low there's a significant savings in testing. If it's high, the advantage is negligible.

So it really doesn't help hot spots, but might save resources in places where the virus is comparatively under control.

[DallasNE]

17. Too Late For Pooled Testing

It is a great idea when positive tests are running at 5% but in Florida the rate is closer to 30%. If you pool 20 swabs at 5% you will have you will have maybe 1 in 4 with 1 or more positive tests where you then have to individually test all 20. That mean that out of 80 people you would need to do 21 tests, which is a nice savings. In a 30% environment you would likely have about a 95% change of having a positive pool test so if you would have 20 pools of 20, or 400 people, you would to individually test 380 plus the 20 pool tests for an identical 400 tests so the cost of waiting is expensive. (I did not use an actual math formula to my numbers will be off some but the point is still the same. And that means you needed to be doing this back in April, not August).

[CreekDog]

29. you said it works if positivity is running at 5%, there's tons of places where that's the case

i don't get you.

you just dump on the whole idea because selected places make it infeasible? as if we can allocate testing resources based on the efficacy?

[DallasNE]

33. Sure, Use It Where Still Feasable

Yes, it would still work in North Dakota, Montana and Wyoming but it is too late for places like Florida, Texas and California. In other words, this is just another lost opportunity that didn't have to happen but did because of a gross leadership vacuum at the top.

[Yavin4]

20. This is actually bad because it will slow down getting results to people.

They are introducing a new step in the process which does nothing given the high positive rates from the individual tests in some states.

For example, given the positive rates in FL, there's a strong probability that the pools will show at least one positive result, so you will then have to re-test the four samples individually to find the positive result(s).

As opposed to testing each individual right off the bat and getting the results to them as soon as possible.

[CreekDog]

30. that's false because they can find out if their group has a positive

say, it's one household. they get an immediate result that someone in the household has it, they quarantine and wait for individual results.

you telling me that's bad?

the current wait would have them wait longer for ANY result because only individual results were allowed.

[Yavin4]

32. Given the positivity rate, it's highly probable that you will be in a group sample with a positive

result. There's no real difference between taking an individual test and quarantining yourself until you get the results.

[iluvtennis]

21. Either China or South Korea or both have been doing pooled testing since

the beginning. Wonder why it took the US so long to get on board?

p.s. read about pooled testing back in March. Will find link and post