How will # of new cases go down with more testing?

[Ottakee]

I keep seeing that one of the benchmarks for opening things up (slowly) is a 14 day decrease in number of new cases.

How is that going to work now with the great increase in testing?  Numbers made up...

   For example, April 1st, only those sick enough to be hospitalized are tested.....one set of numbers.....say 1000 test a day.    Then April 15, those with severe symptoms tested so 5,000 tested.    Now May 1st much wider testing so 10,000 tested.

Even if the number of severely ill that need hospitalization is going down, the number of new cases will continue to go up just because now we are testing so many more and catching those with mild symptoms and  no symptoms.  

How do they account for the fact that there were likely a huge number of positive cases back April 1 and April 15 but they were never tested ....and therefore never counted?

Are there any statistics out there are how many cases of covid are in the hospital and if that number is rising or falling or steady?

[Ausmumof3]

Well in the long run if you test every case and trace and isolate the contacts your actual number of new vases should go down.  But you’re right if you haven’t been catching all the cases and you increase testing on paper it will look like new cases go up.  
 

the benchmarks here for reopening were related to stable number of new cases with adequate testing.  

[Ottakee]

2 minutes ago, Ausmumof3 said:

Well in the long run if you test every case and trace and isolate the contacts your actual number of new vases should go down.  But you’re right if you haven’t been catching all the cases and you increase testing on paper it will look like new cases go up.  
 

the benchmarks here for reopening were related to stable number of new cases with adequate testing.  

And that is what we didn't have....and still don't.    This is comparing apples to oranges to bananas.

I am trying to find statistics about hospitalizations and vents needed over time.  

[Carrie12345]

Talking out of my rear for a minute.

In places that are, let’s use the catch all phrase “shut down”, there really shouldn’t be an increase in cases even with expanded testing because the transmission should be going down.

I’m not saying that’s what will happen, just that that’s what’s supposed to happen.

And then, when activities are supposed to increase, the expanding testing should be able to catch people before they transmit more.  But I’m also not saying that’s what will happen.

I’m on board with the importance of expanded testing. I’m having a hard time believing it’s going to help to the extent that they’re saying it’s supposed to help.

[Teaching3bears]

Number of hospitalizations and deaths is a more reliable indicator.

When the number of cases go down despite increased testing.

[teachermom2834]

Well, I would say the math is fuzzy for reopening. We are in a state that is reopening and the daily press conferences get into the numbers and how they are interpreting them. We have vastly upped our testing. We went from nearly  impossible (it seemed) to get tested to anyone can get tested even without an appointment or symptoms. 

Initially they said they were looking at the rate testing positive vs. the raw numbers. Now there have been spikes in numbers somedays and the report will be something like "we added 900 cases yesterday but 750 came from a prison so not representative of spread".  So they seem to be looking at general trends and special populations that skew results and not just raw numbers. 

So, while the benchmarks are there, they aren't being followed precisely at all. I think our state will continue to move forward with reopening even if the numbers increase as long as hospitals are not overwhelmed.

[Ausmumof3]

6 minutes ago, square_25 said:

I will say that this hasn't been the case from my observation. People's private behavior affects this a lot, and lots of people still congregate during a shut down. 

The Massachusetts outbreak basically got going through the shut down. And I know for a fact people weren't taking the restrictions seriously, until they were... 

Arguably they still aren’t based on this video 

 

[Lecka]

I have read the problem with looking at hospitalizations and deaths is that people might not be hospitalized for a time after infection, and then deaths have a longer lag.  So they are a good indicator but too late.  Except also there are deaths where they are never tested for coronavirus, so it’s not known if that is covering all the deaths.  
 

I’m also in an area with lower testing and I think they don’t care about this even though it’s obviously an issue if they are going from testing only people hospitalized to people who are just sick.  
 

I don’t know if they will make allowances or not.  
 

It seems like such a catch-22.  But hopefully testing increases and stays at a stable level and then this will not continue to be an issue!  In the meantime — yeah. 

[Carrie12345]

46 minutes ago, square_25 said:

I will say that this hasn't been the case from my observation. People's private behavior affects this a lot, and lots of people still congregate during a shut down. 

The Massachusetts outbreak basically got going through the shut down. And I know for a fact people weren't taking the restrictions seriously, until they were... 

Yes, that’s why I was trying to distinguish between “supposed to” and is/will.

[Bootsie]

https://covid19.healthdata.org  is now providing an estimate (with a range) for the number of people infected.  Part of the model includes the impact of increased testing.  

Texas, for example, had a spike in cases at the end of last week, but they also increased daily testing by 25% in about 3-4 days.

[Pen]

I think it is supposed to go down or be steady over 14 days. If there’s an initial spike with more testing, then from that spike it should show steady or lowering from the new testing level.

Afaik all Veterans nursing home residents in USA have been tested. All nursing home residents and staff in my state and some others are apparently now in process of being tested.  

The nursing home tests may reveal some of the highest levels and potentials for death and spread to emergency responder and medical staff. 

An attempt at random sample testing in my state is underway to help guide opening.  

And testing has been extended to anyone in essential services with any symptoms, which may initially spike levels, but then should be a big help for contact tracing. 

[Pen]

1 hour ago, Ausmumof3 said:

Arguably they still aren’t based on this video 

 

 

I think mainstream and social media are fueling this by making it into a thing to do.  Like with toilet paper.  

[maize]

My state is using the numbers of new hospitalizations as one measure; our positive test numbers have gone up as more testing becomes available but our number of new covid-19 hospitalizations has been declining, suggesting strongly that actual infections are declining.

[vonfirmath]

1 hour ago, Bootsie said:

https://covid19.healthdata.org  is now providing an estimate (with a range) for the number of people infected.  Part of the model includes the impact of increased testing.  

Texas, for example, had a spike in cases at the end of last week, but they also increased daily testing by 25% in about 3-4 days.

 

My husband's group has been brought back to full time (they were working half time) and told to stand ready to come in on the weekend if they get a bunch of COVID tests in and need more help entering them.

[catz]

At least locally, they are talking about hospital cases and ICU use trending downwards.  They expect overall numbers to go up as testing expands.  That is not trending down here yet.  They are also watching to see what percentage of tests done are positive.  In countries handling this well, it is not higher than like 2-3% I believe.

If people are tested and know they have it, tracing and smaller scale quarantines can be put into place.  

Edited May 5, 2020 by FuzzyCatz

[Bootsie]

2 hours ago, HeighHo said:

With more testing, the asymptomatic people who aren't in confined quarters will be identified and quarantined  When they are quarantined,  they won't be able to spread the virus as far as they are doing now and  it will prevent the upward trend that would otherwise happen when restrictions are loosened. My state has preliiminary numbers on this from police/fire/etc although the data isn't being publicly sh

I don't know if there is any way to tell if a state has increased testing if it is testing more asymptomatic people to attempt to get those individuals isolated, or if the increased testing is due to more people who think they have symptoms having access to test. 

[ktgrok]

3 hours ago, Pen said:

 

I think mainstream and social media are fueling this by making it into a thing to do.  Like with toilet paper.  

I really want to think that adults in this country are able to think for themselves enough to know that congregating in big groups during a deadly pandemic is a bad idea - even if bobby jones and sally sue on facebook say they are going. 

[vonfirmath]

3 minutes ago, Ktgrok said:

I really want to think that adults in this country are able to think for themselves enough to know that congregating in big groups during a deadly pandemic is a bad idea - even if bobby jones and sally sue on facebook say they are going. 

 

But if you don't believe it is a deadly pandemic? Then why worry about congregating.

 

[ktgrok]

8 minutes ago, vonfirmath said:

 

But if you don't believe it is a deadly pandemic? Then why worry about congregating.

 

Sure, but that's different than what I was responding to which seemed to be saying that people are going to these protests because facebook and the media are making them popular. 

[Matryoshka]

2 hours ago, FuzzyCatz said:

At least locally, they are talking about hospital cases and ICU use trending downwards.  They expect overall numbers to go up as testing expands.  That is not trending down here yet.  They are also watching to see what percentage of tests done are positive.  In countries handling this well, it is not higher than like 2-3% I believe.

If people are tested and know they have it, tracing and smaller scale quarantines can be put into place.  

This here too.  We're looking at hospitalizations going down before reopening, not just cases, as there's a 2-week or more lag between the latter and the former, and the former is independent of how many tests are done, as hospitialzations are usually going to be a certain % of total actual cases whether you know about them or not.

[Pen]

23 minutes ago, Ktgrok said:

I really want to think that adults in this country are able to think for themselves enough to know that congregating in big groups during a deadly pandemic is a bad idea - even if bobby jones and sally sue on facebook say they are going. 

 

I would like to think so too.  

But I am also aware of how easily many people can be swayed—and then as a certain critical mass is reached it becomes a “happening”.   

And I have a relative in a field that sometimes  deals with ...  rabble rousing as a non technical term. 

 

[kbutton]

47 minutes ago, Ktgrok said:

I really want to think that adults in this country are able to think for themselves enough to know that congregating in big groups during a deadly pandemic is a bad idea - even if bobby jones and sally sue on facebook say they are going. 

 

17 minutes ago, Pen said:

I would like to think so too.  

But I am also aware of how easily many people can be swayed—and then as a certain critical mass is reached it becomes a “happening”.   

And I have a relative in a field that sometimes  deals with ...  rabble rousing as a non technical term. 

I really hope people do keep their heads. FB is not reassuring even after blocking or snoozing people. 

The phrase, "It's an economy if you can keep it," comes to mind. 

Regarding both of these, our governor has been reiterating daily that our ability to open and stay safe is tied to people using their heads and following guidelines. He sounds very much like someone really good at talking to people with ODD, lol! He has 8 kids, so I'm guessing he's reaching back to the their teen years. 🙂 He hasn't been flippant or thrown up his hands. He's been very level and authoritative. 

[JennyD]

I don't know about state benchmarks, but as others have mentioned, the positivity rate seems perhaps more important that the actual number of cases.  (Note that this article is from several weeks ago so the numbers it cites are outdated.)

Edited May 5, 2020 by JennyD

[Karen A]

@Ottakee, MI stats for hospitalizations are here:

https://www.michigan.gov/coronavirus/0,9753,7-406-98159-523641--,00.html

I didn't search around long, and didn't find a chart that compares hospitalization/ICU/vent utilization day to day, but you could keep a chart of your own numbers.

I live in IL, which has really increased testing in the last week or so, and the percent of positive tests daily has dropped from about 20% to 15% with expanded testing. Our number of new cases per day is NOT decreasing despite what I think is fairly good social distancing compliance in my county. Which is disheartening. 

[vonfirmath]

5 minutes ago, JennyD said:

I don't know about state benchmarks, but as others have mentioned, the positivity rate seems perhaps more important that the actual number of cases.  (Note that this article is from several weeks ago so the numbers it cites are outdated.)

 

Our governor didn't use this rate. But in a recent interview he did (2 days ago) he did address the "more than 1K new cases in the state, 50 new deaths" and one of the things he addressed was the percentage of people tested who test positive.  So 900 people testing possible out of 22K tested is about 4%.  It needs to be put into context and they are looking at keeping the numbers below 10% in terms of keeping Texas open.

 

[vonfirmath]

4 minutes ago, square_25 said:

I think you really have to be judicious. For example, the positivity rate will go down if testing goes up, even if cases don't go down. 

 

But if you test a bunch more people and find you are already capturing most of the covid positives under the previous conditions of testing, isn't that a good thing? You don't have a lot of hidden infection happening. -- That's what people used to say "Yes, the numbers seem low. But we aren't testing, so there could be a bunch out there. We need to test more"

 

 

Edited May 5, 2020 by vonfirmath

[vonfirmath]

1 minute ago, square_25 said:

If you're adding 1000 new cases every day and doubling your testing, your positivity rate will be halved without an actual drop in growth. 

 

I guess that gets back to the question: Are we trying to eliminate the virus or slow it down. For elimination you want an actual drop in growth. For the latter, it becomes more of a balancing act.

 

 

[catz]

1 minute ago, square_25 said:

I think you really have to be judicious. For example, the positivity rate will go down if testing goes up, even if cases don't go down. 

Well, that is not necessarily true.  Positivity went up in April across the US in general despite 25x more testing according to the article linked.  When everyone has a quick and easy means to a test, we will have more meaningful numbers.  We aren't there yet.  The positivity rate is over 20%.  That is astronomically high and we have a long way to go there.

The other number experts are watching is the R0 (R-naught).  Ideally that would drop to less than one, or be as close to one as possible.  I think the data is far from simple and I really think the average person doesn't get it.  

[catz]

Just now, square_25 said:

The problem is that to estimate R0, you really need to know the number of cases, and that is something we only have a vague handle on! 

Exactly, another reason we need to keep expanding testing!

[JennyD]

11 minutes ago, square_25 said:

I think you really have to be judicious. For example, the positivity rate will go down if testing goes up, even if cases don't go down. 

 

This point is well-taken.  It does seem like the virus's ability to spread generally far exceeds our abilities to expand testing, which is why I still think the positivity rate is helpful. But yes, if you suddenly double your testing (by, say, bringing another lab online) that number becomes less useful for a while.

 

[Lecka]

I just read an article from Washington Post (I subscribed to an email from them) that has someone saying that excess mortality can be looked at to balance with testing and try to prevent seeing what looks like a huge increase or positive rate, when it is not representative.  
 

But then it does say it takes longer and testing is important.  
 

Anyway — I can see more now how they can look at a lot of information and take it together. 

[ScoutTN]

5 hours ago, vonfirmath said:

 

I guess that gets back to the question: Are we trying to eliminate the virus or slow it down. For elimination you want an actual drop in growth. For the latter, it becomes more of a balancing act.

Doesn't seem like eliminating it is even remotely possible in the US. 

[Lecka]

They did not mention any of that, sigh.  

[ElizabethB]

13 hours ago, Ottakee said:

And that is what we didn't have....and still don't.    This is comparing apples to oranges to bananas.

I am trying to find statistics about hospitalizations and vents needed over time.  

Most states actually have contact tracing, those that didn't have much got more help from the CDC.

"LANSING, MICH. As part of efforts to continue expanding COVID-19 testing and contact tracing of potentially exposed Michiganders, the Michigan Department of Health and Human Services (MDHHS) again expanded its testing criteria to include all essential workers still reporting to work in person, whether they have symptoms or not, and launched a large-scale effort with more than 2,000 volunteers to expand contact tracing capacity."

https://www.michigan.gov/som/0,4669,7-192-47796-526523--,00.html

[ElizabethB]

5 minutes ago, square_25 said:

 

I don't know if any states have ENOUGH contact tracers, though. I think it's usually much smaller scale. Any idea how many most states have? 

2,000 would be quite good for Michigan, although I'm a bit worried about the "volunteer" bit. 

Idaho and Utah have enough.  CDC sent some, they hired more.  Idaho had a robust team that was able to keep up with all but the worst week (but fairly low Covid case rates compared to many states) but just hired more temporary tracers in case of increase after any of the opening phases.

You would have to google each state and see, some people think it's creepy and invasive, it's not highly advertised.  That's how I found Michigan's plan.  I Googled "Michigan health department contact tracing" with and without Covid/Coronavirus as search terms.

[Lecka]

I read some coverage about Utah and Idaho a couple of weeks ago, that they had a chance to stay at “containment” instead of “mitigation.”  
 

It sounded like the numbers of contact tracers needed could be fairly low because they were getting on top of it while numbers were low and I think they said they had good testing?  

[Ausmumof3]

3 hours ago, ScoutTN said:

Doesn't seem like eliminating it is even remotely possible in the US. 

Well you currently have R0 below 1 but I realise you can’t stay shut down forever 

[ElizabethB]

2 hours ago, square_25 said:

How many does Idaho have? 

I'm not sure exact numbers, but with the numbers they had, they were able do full tracing all except one week, the week with the most cases, that week they did just high risk cases.  The state is divided into several different health districts, each with its own team, they have epidemiologists in each district.  Here is the article where they talked about hiring 17 temporary workers to be prepared for any spikes with the opening phases.  The CDC also sent 5 people.  https://www.idahostatesman.com/news/coronavirus/article242446611.html

[ElizabethB]

Nationwide contact tracing, describes it, there is a firm hiring, online work. (Firm in DC but job done from home.)

https://www.idahostatesman.com/news/coronavirus/article242536066.html

"You can earn an average of $17 to $22 an hour Monday through Friday during eight to 12-hour shifts, the job posting said."

 

Edited May 6, 2020 by ElizabethB

[Farrar]

Here, they currently have nearly a hundred, are hiring so they have 200 within the next few weeks, and are planning to have 700 by... some time in the summer but I forget when. So that would be about 1 for every 1000 people, which is considerably better than 30 for 100,000.

Here, they will be employed directly by the government. As far as I could see from the job listing, they will not be work from home jobs, but I'm guessing it will depend. I would not trust a private company that sprang up in the wake of this so I'm very glad that our not-a-state is actually employing and overseeing this and not an unproven, out to make a buck corporation with no track record.

[Pam in CT]

re decreased hospitalizations vs decreased #positive test cases

On May 5, 2020 at 1:35 PM, Matryoshka said:

This here too.  We're looking at hospitalizations going down before reopening, not just cases, as there's a 2-week or more lag between the latter and the former, and the former is independent of how many tests are done, as hospitialzations are usually going to be a certain % of total actual cases whether you know about them or not.

 

On May 5, 2020 at 2:14 PM, square_25 said:

Same in NY. 

 

Same in CT.  Right now hospitalizations seem to be plateauing in my county, but are still increasing in the others in the state (I'm in the one closest to NYC).

[Pam in CT]

I may have missed if someone else posted this link, but I like the way this graph compares testing state-by-state.  It shows testing (green bar), confirmed cases (orange bar), and deaths (dark blue bar) all expressed as per 100K population... and so adjusts for relative state population size.  

Rhode Island has done WAY more testing per capita than any other state, even including NY/NJ/CT.

[lewelma]

Seems to me that the number of contact tracers required has more to do with the number of cases being followed than the size of the population.  NZ was able to contact tract a max of 80 cases/day with 200 contact tracers.  That is a LOT of contact tracers required for some American states that seem to have like 2000 new cases a day.  NZ has upped the contact tracing capability now that we are leaving lockdown to over 10,000 phone calls per day.  But in addition, contact tracing requires access to other databases to find people - like police, immigration, health.  When you are tracing people a positive person has come in contact with, clearly you call the people they know they have been near, but then you also have to figure out all the people in the restaurant they ate at, and all the people in the train when they went into work, and the shop clerk for the store they were in, etc.  This is a very time consuming process, and requires lots of people to both call and to investigate. 

[Terabith]

19 hours ago, Pam in CT said:

I may have missed if someone else posted this link, but I like the way this graph compares testing state-by-state.  It shows testing (green bar), confirmed cases (orange bar), and deaths (dark blue bar) all expressed as per 100K population... and so adjusts for relative state population size.  

Rhode Island has done WAY more testing per capita than any other state, even including NY/NJ/CT.

Yikes!  I've been kind of trusting that we were doing okay, what with our governor being a doctor and all, but only Arizona and Puerto Rico have done less testing per 100k population.

[ElizabethB]

19 hours ago, Pam in CT said:

I may have missed if someone else posted this link, but I like the way this graph compares testing state-by-state.  It shows testing (green bar), confirmed cases (orange bar), and deaths (dark blue bar) all expressed as per 100K population... and so adjusts for relative state population size.  

Rhode Island has done WAY more testing per capita than any other state, even including NY/NJ/CT.

There are several states with low case numbers that are not doing much testing per capita but have enough tests to test people who need to be tested, the percent of tests that come back positive is a good measure of that, not really captured in that chart.  

For example, Idaho 2158 positive tests, 30890 total tests, 6.7% positivity rate.  Oregon 2989 positive, 70,458 total, 4.2% positivity rate.  You're wasting tests if you are testing people who don't need testing in low areas of cases, low areas of spread. 

[ElizabethB]

10 minutes ago, Terabith said:

Yikes!  I've been kind of trusting that we were doing okay, what with our governor being a doctor and all, but only Arizona and Puerto Rico have done less testing per 100k population.

AZ positivity rate is 8.9 percent, under 10 percent is considered to be you are testing everyone who needs tested, the low per capita numbers are tied to low infection rates and low number of cases, not low testing rates. (See my post above.)

Puerto Rico doesn't report enough data to calculate a positivity rate.

Virginia's positivity numbers are looking bad, though...but may be OK now but have been worse in the past. You need to look at recent daily/weekly percent positive vs. total testing numbers to get a current positivity rate.

Edited May 7, 2020 by ElizabethB

[Pam in CT]

re "wasting testing"

1 hour ago, ElizabethB said:

There are several states with low case numbers that are not doing much testing per capita but have enough tests to test people who need to be tested, the percent of tests that come back positive is a good measure of that, not really captured in that chart.  

For example, Idaho 2158 positive tests, 30890 total tests, 6.7% positivity rate.  Oregon 2989 positive, 70,458 total, 4.2% positivity rate.  You're wasting tests if you are testing people who don't need testing in low areas of cases, low areas of spread. 

I'm not understanding what the bolded means. We don't know what we don't know.  How do we figure out where the next Smithfield / Tyson / Birdseye hot spot is if we don't test until AFTER people get sick... which is AFTER they've already transmitted on to boatloads of other people... without testing a whole lot of people who aren't showing symptoms?  

Really not understanding how a low positive rate represents "wasted" tests.  

[ElizabethB]

13 minutes ago, Pam in CT said:

re "wasting testing"

I'm not understanding what the bolded means. We don't know what we don't know.  How do we figure out where the next Smithfield / Tyson / Birdseye hot spot is if we don't test until AFTER people get sick... which is AFTER they've already transmitted on to boatloads of other people... without testing a whole lot of people who aren't showing symptoms?  

Really not understanding how a low positive rate represents "wasted" tests.  

We can't test everyone, even S Korea and New Zealand weren't testing everyone, there is not enough reagent in the world.  If you test people who are sick and their contacts and monitor hot spots with testing, you get maximum bang for your buck.  If you test random people not likely to be infected, you are wasting the limited amount of reagent and swabs, there is a worldwide shortage of both right now.  The antibody tests use a different media and use blood, no swabbing, excess of those tests in not wasting anything.  

So, crowded nursing homes and meat plants are worth doing some random testing there, but just random testing of random people not that likely to be infected is wasting tests, or more specifically, wasting reagent and swabs that could be better used elsewhere.

 

Edited May 7, 2020 by ElizabethB

[Terabith]

I think it makes sense both to test people who are sick and their contacts AND to do random testing, because that's how we find hot spots and get real information about prevalence of virus.  Not test everyone, but truly random sampling can give us really useful information.  

[kiwik]

When you increase test rates you test a lot more possible cases compared to probable and the percentage of positive drops.  If you are getting under control there are fewer cases to find.  At out peak we were getting round 80 cases a day with 1000 to 2000 odd tests.  Now we are getting 1 or 2 cases a day with 6000 to 7000 tests.  These numbers don't seem big but we have a relatively small population.  At this point 3.5 % of the population has been tested (by number as I suspect some have been tested more than once).