More Testing versus "Bending the Curve"

[washunut]

By now we all have heard the experts tell us we have to shelter in place until we are able to "bend the curve". By this they mean that we show that the number of new cases has crested and is starting on a downward slope. We have also learned that the number of people who have actually been infected is higher by orders of magnitude than has been reported.

Thus it would seem logical that as we test more, clearly we will find more cases. While this is good,how will these experts know when we have actually been able to bend the curve? That is as we test more we will naturally increase new cases,so it is a lot like chasing your tail. More tests translating to more new cases does not mean we have not bent the curve.

Not sure that the models are robust enough to take this into account. A model is only as good as the assumptions input into it. Pretty sure our experts,politicians,media are unable to do anything better than parrot numbers literally. In some cases not sure some are even adequate in understanding the numbers literally.

 

[Helix]

Test as many as possible.

 

[ecofarm]

I follow deaths, not cases. I think everyone does. Tests serve general awareness, not analysis.

 

[Gaius46]

By now we all have heard the experts tell us we have to shelter in place until we are able to "bend the curve". By this they mean that we show that the number of new cases has crested and is starting on a downward slope. We have also learned that the number of people who have actually been infected is higher by orders of magnitude than has been reported.

Thus it would seem logical that as we test more, clearly we will find more cases. While this is good,how will these experts know when we have actually been able to bend the curve? That is as we test more we will naturally increase new cases,so it is a lot like chasing your tail. More tests translating to more new cases does not mean we have not bent the curve.

Not sure that the models are robust enough to take this into account. A model is only as good as the assumptions input into it. Pretty sure our experts,politicians,media are unable to do anything better than parrot numbers literally. In some cases not sure some are even adequate in understanding the numbers literally.

A couple of things.

First there are indications from testing in CA that there are many more infected than most people believe. While I personally have believed for a while that there are many more asymptomatic people I was thinking in the 5x-10x range while the CA data suggests more like 50x.

It’s extremely likely that lots more people have it than we expect but I think the magnitude is still open to debate.

What the experts really worry about is hospitalizations since that’s the scarce resource here. So I think you track hospitalizations more than cases.

Either that or I’m guessing the total new cases can be adjusted for the extra testing with a bit of math and you track that.

 

[Praxas]

By now we all have heard the experts tell us we have to shelter in place until we are able to "bend the curve". By this they mean that we show that the number of new cases has crested and is starting on a downward slope. We have also learned that the number of people who have actually been infected is higher by orders of magnitude than has been reported.

Thus it would seem logical that as we test more, clearly we will find more cases. While this is good,how will these experts know when we have actually been able to bend the curve? That is as we test more we will naturally increase new cases,so it is a lot like chasing your tail. More tests translating to more new cases does not mean we have not bent the curve.

Not sure that the models are robust enough to take this into account. A model is only as good as the assumptions input into it. Pretty sure our experts,politicians,media are unable to do anything better than parrot numbers literally. In some cases not sure some are even adequate in understanding the numbers literally.

So let's take what you said for example, here are some of the excerpts"

#1. Unknown amount of people that could have the virus and aren't tested yet.
#2. Curve bending may or may not be happening as well or as bad as officials think.
#3. Officials are only good at parroting the numbers.

So given 1-3, your solution is what? Re-open everything and hope for the best? Now, tell us how that is better in fighting COVID than the measures officials have put into place?

 

[TomFitz]

Test as many as possible.

I doubt that the Trumpster crowd has figured this out yet, but the crisis won’t be over with any degree of confidence until just about everyone is tested. And most likely tested again.

 

[bluesmoke]

Test, test, test. Only problem is, Trump is doing nada to make that happen to the degree it should be done quick enough. Oh, then there's the reliability of the test kit itself. For some reason, the CDC rejected the proven model. Oh, well. And the beat goes on...

 

[washunut]

A couple of things.

First there are indications from testing in CA that there are many more infected than most people believe. While I personally have believed for a while that there are many more asymptomatic people I was thinking in the 5x-10x range while the CA data suggests more like 50x.

It’s extremely likely that lots more people have it than we expect but I think the magnitude is still open to debate.

What the experts really worry about is hospitalizations since that’s the scarce resource here. So I think you track hospitalizations more than cases.

Either that or I’m guessing the total new cases can be adjusted for the extra testing with a bit of math and you track that.

Good point about hospitalizations versus cases. Perhaps Birx needs to explain what you did so ably.

 

[washunut]

I follow deaths, not cases. I think everyone does. Tests serve general awareness, not analysis.

Tests are a lagging indicator. I think the vast majority look to cases not deaths.

 

[Variant]

By now we all have heard the experts tell us we have to shelter in place until we are able to "bend the curve". By this they mean that we show that the number of new cases has crested and is starting on a downward slope. We have also learned that the number of people who have actually been infected is higher by orders of magnitude than has been reported.

Thus it would seem logical that as we test more, clearly we will find more cases. While this is good,how will these experts know when we have actually been able to bend the curve? That is as we test more we will naturally increase new cases,so it is a lot like chasing your tail. More tests translating to more new cases does not mean we have not bent the curve.

Not sure that the models are robust enough to take this into account. A model is only as good as the assumptions input into it. Pretty sure our experts,politicians,media are unable to do anything better than parrot numbers literally. In some cases not sure some are even adequate in understanding the numbers literally.

You can look at the data differently depending on how much testing is being done, say how many tests are coming up negative.

South Korea is doing a lot more testing per person and are finding more and more negative results. That is an indication that they are bringing the virus under control.

 

[washunut]

Test as many as possible.

Didn't say we should not. Just pointing out how that confuses the statistics as we look to bend the curve.

 

[ecofarm]

Tests are a lagging indicator. I think the vast majority look to cases not deaths.

Looking at cases is stupid. Deaths are solid data with invariable methods.

 

[Variant]

Looking at cases is stupid. Deaths are solid data with invariable methods.

If you accurately attribute the number of deaths to the virus, which takes a bit of time.

 

[Variant]

Didn't say we should not. Just pointing out how that confuses the statistics as we look to bend the curve.

Being able to identify positive cases and quarantine them will lower the amount of new infections faster than social distancing.

Being too focused on the overall numbers and not what is causing them is a mistake.

 

[ecofarm]

If you accurately attribute the number of deaths to the virus, which takes a bit of time.

For some there's no end to the conspiracy claims.

 

[washunut]

So let's take what you said for example, here are some of the excerpts"

#1. Unknown amount of people that could have the virus and aren't tested yet.
#2. Curve bending may or may not be happening as well or as bad as officials think.
#3. Officials are only good at parroting the numbers.

So given 1-3, your solution is what? Re-open everything and hope for the best? Now, tell us how that is better in fighting COVID than the measures officials have put into place?

Actually your summary was pretty good. Where you fail is in your conclusion. No where in my post can you read out the inference of my wanting to open things.

So let me say this a bit more clearly to those who are having a hard time with this. People need to calibrate the increase in testing with the numbers we look to understand how we are doing.

 

[snakestretcher]

Tests are a lagging indicator. I think the vast majority look to cases not deaths.

Indeed, and a test is only valid for the day of the test. If, for example, a test takes several days before a lab shows a negative for virus result, there is no way of knowing if the person tested became infected subsequently. A more useful test is that used to ascertain if an individual was infected and subsequently recovered.

 

[Helix]

Didn't say we should not. Just pointing out how that confuses the statistics as we look to bend the curve.

I'm fine with confusing the statistics, whatever that means. I want to know if the person breathing on me in my "integrated workspace" sardine can is sick or not.

 

[washunut]

I'm fine with confusing the statistics, whatever that means. I want to know if the person breathing on me in my "integrated workspace" sardine can is sick or not.

There are many times I see your posts and feel that there is so much aggression and anger you do not take the time to think what was posted.

I am NOT against more testing. Simply tried to point out that as we expand testing we need to understand what that means to the metrics people are looking at.

 

[Helix]

There are many times I see your posts and feel that there is so much aggression and anger you do not take the time to think what was posted.

I am NOT against more testing. Simply tried to point out that as we expand testing we need to understand what that means to the metrics people are looking at.

I've thought quite a bit about it. Test more people before opening up and la la laaing all over each other like the past couple months didn't happen.

 

[Gaius46]

Looking at cases is stupid. Deaths are solid data with invariable methods.

Deaths lag new cases by weeks up to a month or more so they tell us what’s happened in the past not where we are today.

As well deaths don’t directly tell us about whether we have resource problems.