Testing and Contact Tracing - What Do We Need?

[capjak]

Anyone else really tired of the nonstop non-specific reporting on testing - we have enough versus not enough - without anyone saying anything about what IS enough! What does more of that accomplish? Most of the reports I’ve seen on testing have been useless, only spurring confusion (and partisanship).

And I’ve yet to see anything specific on contact tracing, how many people have any idea what it entails, and who’s going to to do it?

After reading many useless reports, I found this - but when will journalists actually quantify what’s needed and how to work toward it? Stop wasting time with vague information.


https://www.theguardian.com/us-news...ond-historic-failure-of-leadership-from-trump

If you believe World meter data.

Iceland has 350,000 people and completed 41,000 (.041 million) tests for (120,000 tests per million) people. I would not use Iceland as a comparator.

South Korea has 52 million people and has done 546,000 tests (10,659 test per million pop). US has done MORE per capita than S.Korea.

GERMANY is opening/loosing restrictions on Monday...
Germany 83 million people and 1.7 million tests (20,600 tests per million pop).

The US has 327 million people and has completed 3.6 million tests (10,825 per million people) more per capita than South Korea and almost 2 million more tests than any country.

But that is not the whole testing picture.....

New York has done 29,000 tests per million people 30% more than Germany but I would definitely not open them up....so testing alone is not the answer....

If you believe S.Korea is/was successful in containing the virus than you would conclude that the US already has the testing capacity to contain a outbreak.

The government has several levels of monitoring to identify outbreaks in different locals

But can the US effectively tract people that test positive and than trace everyone that they came into contact with over the last 14 days or so is a much larger concern/question.

Will everyone in the US allow a phone app to trace everyone they have been in contact with (maybe 50% I guess) Would people in the US if tested positive but no illness shelter at home (probably).


https://www.worldometers.info/coronavirus/country/us/

Finally the King of each state will determine when to loosen restrictions

 

[sengsational]

True Random Sample from Various Environments

I've been pretty much ignoring all of the bellowing about testing, but I'll throw my two cents in because I don't see anyone calling for what I see as a logical approach.

You pick several spots [defined geography] that are in various stages of outbreak (high, medium, low,
"none"). You use census or voter or some list that has "everybody" to make a random selection of people from that defined geography. You compel everyone who was randomly selected to get tested (yeah, yeah, please don't go there, it won't happen the way I want anyway). Test would include both active disease as well as antibody testing for everyone randomly sampled.

Now, given the limitations of the testing, you have an indication of what the disease "looks like" in the population within each of these different levels of outbreak. So now, as the disease progresses, you find one of these historical sample data sets that match on other parameters (like hospital admittance) and you "know" what the actual prevalence of the disease is in the community, at that time. And you know it without testing "everyone".

 

[Aerides]

A recent harvard study recommends a minimum of 152 tests per 100k population, roughly 3x the rate we're at currently.

The average now is 20% positive, which indicates that we have a lot of untested positives in the wild. Ideally, we'd be under 10% positive to have confidence in the real infection rates, and where to target different social protocols.

To a layperson, 20% positive sounds low, but to the data scientists, that's very high.

https://www.nytimes.com/interactive...testing-states.html?smtyp=cur&smid=tw-nytimes

 

[Midpack]

FWIW, the US has now passed South Korea and Taiwan in tests per million people according to: https://www.worldometers.info/coronavirus/#countries

Even Germany's ~20,000 tests per million people is only 2% tested (probably really <1% due to medical workers and infecteds getting tested multiple times).

You’re right, I’m wrong there. I was conflating aggressively getting on testing early (as SK, Germany, HK, Singapore and Taiwan) versus numbers of tests. Hard to keep facts straight for me at least.

I guess we’re seeing how much harder it is get on the testing/contact tracing bandwagon later. But I don’t see how we reopen and contain this at manageable levels without more testing and contact tracing.

I don’t doubt we need way more testing, I haven’t heard a single healthcare professional state otherwise. I’m just sick of talking heads saying we’re not doing enough testing - without also saying what they think we need. Are we 2X off, 10X, 100X or what? Without quantifying, it’s just mindless partisanship.

 

[donheff]

Once you get testing capacity somewhere in the region of the minimum necessary, you will need the contact tracers to do something about positive tests. If you are going to have even a partial reopening, you will need to be able to quickly follow up on positives and do localized lockdowns and quarantines to prevent a wider spread. Hopefully we get an effective treatment sometime this year that A) we know how to manufacture and B) can be rapidly distributed widely to hospitals. Not seen anything like it yet.

Less than shutting down the economy.

Really, we don't need to test the population. What we need is the ability to test suspected cases and anyone they have been i contact with. We do not have that.

I highlighted. One is most contagious before they have symptoms. Thus, we would have to test everyone. As symptoms get worse, the contagion goes down. The issue with this virus, that makes it different than other viruses, E bola (much more deadly) or swine flu or SARS is how quickly it passes from one to another. The contagion virus is in the nasal, upper respiratory system, which is more readily passed from person to person.
(My source is Dr. Peter Attia and his many interviews)

Fauci and others acknowledge that large numbers of infected are symptom free. It is important to identify that proportion through some decent random samples of serological tests. Then they can come up with more realistic mortality estimates.

But we don't have to test everyone to fight the disease. Fauci and the rest are emphasizing what Brewer points out is going on elsewhere around the globe and is recommended by CDD: sentinal testing in selected environments, testing symptomatic individuals, tracing and testing contacts of positives. Sure the unsymptomatic will be out there spreading the virus but when that results in local symptomatic cases, test, and trace them. Beat the epidemic back to levels we can deal with.

 

[capjak]

A recent harvard study recommends a minimum of 152 tests per 100k population, roughly 3x the rate we're at currently.

The average now is 20% positive, which indicates that we have a lot of untested positives in the wild. Ideally, we'd be under 10% positive to have confidence in the real infection rates, and where to target different social protocols.

To a layperson, 20% positive sounds low, but to the data scientists, that's very high.

https://www.nytimes.com/interactive...testing-states.html?smtyp=cur&smid=tw-nytimes

We are already well above the harvard study... currently we have tested 1,000 per 100k population, of course that varies state to state but as a nation we are well above what Harvard study says is needed. ......

20% positivity rate in a population that has cover-19 symptoms (you have to have symptoms in order to be tested in many areas) indicates that of the people that have Covid-19 like symptoms the majority have something other than Covid-19 or some of the home-brew tests are not that great.

 

[USGrant1962]

We are already well above the harvard study... currently we have tested 1,000 per 100k population, of course that varies state to state but as a nation we are well above what Harvard study says is needed. ......

20% positivity rate in a population that has cover-19 symptoms (you have to have symptoms in order to be tested in many areas) indicates that of the people that have Covid-19 like symptoms the majority have something other than Covid-19 or some of the home-brew tests are not that great.

That study is a testing rate - 152/100K per day.

 

[Rianne]

But we don't have to test everyone to fight the disease. Fauci and the rest are emphasizing what Brewer points out is going on elsewhere around the globe and is recommended by CDD: sentinal testing in selected environments, testing symptomatic individuals, tracing and testing contacts of positives. Sure the unsymptomatic will be out there spreading the virus but when that results in local symptomatic cases, test, and trace them. Beat the epidemic back to levels we can deal with.

Anyone with any kind of pre-existing condition:
-Current or past chemotherapy
-Obesity
-Heart, kidney, liver disease

-Lung disease
-Auto immune disease
-Diabetes
All of these makes one more vulnerable and susceptible to a more serious outcome of this virus. What % of the American population has one or more of these diseases? I will not interact "normally" again. I consider myself relatively healthy, perfect weight, exercise daily, eat very well (never processed food) but I have kidney disease. This virus would take me down in a week. Not like the flu or a cold. That's my take away. The medical community is still trying to figure this thing out. They do not know, it changes daily. It is unlike anything we've seen before. It will re emerge. No question. Even testing is not accurate.

We do not have accurate testing. Many false negatives and false positives. And the virus could still be present after many weeks of recovery. Anti body testing is a start.

 

[Aerides]

We are already well above the harvard study... currently we have tested 1,000 per 100k population, of course that varies state to state but as a nation we are well above what Harvard study says is needed. ......

Harvard is recommending that be a daily number of tests, not a cumulative total. The current US average is about 45 per 100k daily, so a ways to go.

 

[audreyh1]

We are already well above the harvard study... currently we have tested 1,000 per 100k population, of course that varies state to state but as a nation we are well above what Harvard study says is needed. ......

20% positivity rate in a population that has cover-19 symptoms (you have to have symptoms in order to be tested in many areas) indicates that of the people that have Covid-19 like symptoms the majority have something other than Covid-19 or some of the home-brew tests are not that great.

The problem is compounded by false negatives so I don’t think so many strongly symptomatic people can have something other than Covid-19, but that doesn’t mean we shouldn’t keep tracing the positives.

 

[capjak]

That study is a testing rate - 152/100K per day.

Thanks....Didn't read the article

 

[brewer12345]

Fauci and others acknowledge that large numbers of infected are symptom free. It is important to identify that proportion through some decent random samples of serological tests. Then they can come up with more realistic mortality estimates.

But we don't have to test everyone to fight the disease. Fauci and the rest are emphasizing what Brewer points out is going on elsewhere around the globe and is recommended by CDD: sentinal testing in selected environments, testing symptomatic individuals, tracing and testing contacts of positives. Sure the unsymptomatic will be out there spreading the virus but when that results in local symptomatic cases, test, and trace them. Beat the epidemic back to levels we can deal with.

You better hope we can do that. Pulling a Wuhan in the US will result in armed conflict.

 

[Silver]

I believe that testing is important for oversight of the spread, and I agree with the method described by the CDC. Testing doesn't relate to treatment of the virus...only the symptoms do that.

I think the ultimate key to any return to any "normal" is the discovery of a treatment that limits the symptoms, and a successful vaccine. I would like to hear more concrete "news" about where the WORLD is on both treatment and vaccine. We know every country is working on it...let's just hope whomever discovers it first will share their intel. Honestly, I doubt we will be first. I'm thinking Germany...but who knows.

 

[USGrant1962]

Harvard is recommending that be a daily number of tests, not a cumulative total. The current US average is about 45 per 100k daily, so a ways to go.

So this might be a partial answer to Midpack's original question of how much "more testing!!!" is enough. And NY is approaching that number.

 

[capjak]

The problem is compounded by false negatives so I don’t think so many strongly symptomatic people can have something other than Covid-19, but that doesn’t mean we shouldn’t keep tracing the positives.

I never stated we shouldn't trace the positives?

According to CDC they estimate there have been 31 Million to 45 million symptomatic flu illnesses, which can be similar to covid-19 symptoms.

There is % that is actually Covid-19 but not tested/confirmed. As Flu season goes away more and more of these symptomatic cases will be due to Covid-19. That is one of the surveillance tools they will use to help with outbreaks of Covid-19.

 

[Doribe]

I think it is fairly obvious that you and people like you would be among the last to be tested. Where more robust testing and tracing would be important are the areas where people live cheek by jowl, such as NYC.

+1

 

[capjak]

I think the ultimate key to any return to any "normal" is the discovery of a treatment that limits the symptoms, and a successful vaccine. I would like to hear more concrete "news" about where the WORLD is on both treatment and vaccine. We know every country is working on it...let's just hope whomever discovers it first will share their intel. Honestly, I doubt we will be first. I'm thinking Germany...but who knows.

Agree that would be a game changer....

FYI....The US used to discover more drugs by far (2010 data) not sure now that most pharma is somewhat global research..
April 01, 2020
Every Vaccine and Treatment in Development for COVID-19, So Far
https://www.visualcapitalist.com/ev...hyONkeF30f_r3JEljyI0nNX1y9AYLUiQXU7XQzdW0HL4o

 

[Chuckanut]

Don't we have some sort of mechanism setup to take the Census? Can't it be modified to at least help with tracking?

This isn't rocket science, just a lot of time on the phone asking the right questions and entering the information into a computer. I am sure we have plenty of experts who can give people the right questions to ask.

“You go to war with the army you have, not the army you might want or wish to have at a later time.” ― Donald Rumsfeld.

 

[Midpack]

Don't we have some sort of mechanism setup to take the Census? Can't it be modified to at least help with tracking?

This isn't rocket science, just a lot of time on the phone asking the right questions and entering the information into a computer. I am sure we have plenty of experts who can give people the right questions to ask.

Yeah. I was thinking Google, FaceBook or another social media overlord could easily develop contact tracing for us (and sell that info too). They’re already tracking countless aspects of our lives including a form of contact tracing and movements of billions of users worldwide. Facebook reports 220 million users and Google has 259 million unique users in the US and they gather info on non-users too - population 327 million. All the development, infrastructure and staffing are already in place...

It came from Florida Rep. Kathy Castor during the House’s questioning of Facebook CEO Mark Zuckerberg last week, when she asked:

You are collecting personal data on people who are not Facebook users. Yes or no?

Yes, Facebook tracks both users and non-users across websites and apps, according to a post written by David Baser, Product Management Director.

It does so for three main reasons, he said:
- To provide its services to the sites or apps;
- To improve safety and security on Facebook; and
- To enhance its own products and services.

From the post: When you visit a site or app that uses our services, we receive information even if you’re logged out or don’t have a Facebook account. This is because other apps and sites don’t know who is using Facebook.

Facebook is far from the only online service to do this. Twitter, Pinterest and LinkedIn have similar Like and Share buttons, Google has a popular analytics service, and Amazon, Google and Twitter all offer login features, Baser said.

In fact, most websites and apps send the same information to multiple companies each time you visit them.

 

[Aerides]

Yeah. I was thinking Google, FaceBook or another social media overlord could easily develop contact tracing for us. They’re already tracking countless aspects of our lives including a form of contact tracing and movements of billions of users worldwide. Facebook reports 220 million users and Google has 259 million unique users in the US - population 327 million. All the development, infrastructure and staffing are already in place...

Guarantee they are already working on it, as well as building protocols to integrate with Uber, AirBnB, etc., (ie monetize it).

 

[MichaelB]

Google and Apple are working together on tracking https://www.theverge.com/2020/4/11/...ronavirus-tracking-app-covid-bluetooth-secure

Essentially, Apple and Google have built an automated contact-tracing system. It’s different from conventional contact-tracing, and probably most useful when combined with conventional methods.

 

[brett]

To my knowledge the only time we are doing tracking is when there is a specific outbreak in an institutions, workplace, etc. Other jurisdictions that have sophisticated tracking processes in place have been much more successful in containing this that those which do not. South Korea comes to mind.

We have opened, in our jurisdiction, drive up covid testing to the entire population based upon some on line screening and done via specific appointment times. Goal is to test anyone over 65 who wants it, anyone with symptoms regardless of age, with of course preference to those who present pressing symptoms such as difficulty breathing.

So far distancing and this process is working. We have been flat lining in terms of the curve. If there are no pressing or urgent symptoms it is taking four days to get the test appointment and another four days before the test results are emailed to you. They are apparently working to shorten those time frames.

I cannot imagine living in, or near, a jurisdiction that has not instituted some sort of stay at home and/or distancing regulations. Seems to me that we should listen to our public health/medical experts and not to those politicians who have no knowledge and expertise. Nor should be listening to those few with with an excess of opinion or bravado but with zero scientific or medical knowledge on which to base them/it.

 

[Midpack]

Guarantee they are already working on it, as well as building protocols to integrate with Uber, AirBnB, etc., (ie monetize it).

Google and Apple are working together on tracking https://www.theverge.com/2020/4/11/...ronavirus-tracking-app-covid-bluetooth-secure

Great. I’m sure social media or healthcare systems can ramp up contact tracing way faster than the public sector, especially as they have their hands full with other Covid-19 concerns.

Interesting read Michael B. And thanks for being a voice of reason in all this, as usual.

 

[cbo111]

As soon as we can deploy a reliable test that can be read rapidly (~5 min), it would be very useful at the major airports. I'm a bit of a hypochondriac even in good times, so being cooped up with 200+ germy travelers already bother me. But knowing everyone around me has been screened would probably get me back in the air this summer.

 

[Gumby]

It may be here

Abbott promises its test can show a positive case of COVID-19 in five minutes, while it takes about 13 minutes to ascertain negative results.

https://www.ctpost.com/news/coronav...ite-can-find-deadly-disease-in-5-15209361.php