Coronavirus Statistics: Cases, Mortality, vs. Flu

[DavidB66]

According to "10 doctors", it still counts.
….
But what exactly is considered an underlying health issue or condition in the context of the disease?
WFAA reached out to 10 doctors to find out, and they all responded the same way.
The term can be split into two categories.
The first is chronic conditions-- these are long-term medical diseases or illnesses like asthma, chronic bronchitis, diabetes, high blood pressure (even if it is controlled with medication), heart disease, lung disease, liver disease, COPD and cancer. These kinds of conditions also tend to be incurable.
...

Well, thanks for that! I'm not surprised, though. Even when hypertension is brought under control, it often does some irreversible damage before it is diagnosed. Also, whatever physical or physiological conditions underlie the hypertension may persist despite the treatment. And there is even a suspicion that some common types of medication make the lungs more vulnerable to COVID-19.

 

[Mendel]

The world is still at under 1 million cases (tested positive). If this undercounts the actual infected 1:10, the world is at 10 million.
https://www.worldometers.info/coronavirus/ currently has 810,123 in the world. Only 4 countries (USA, Italy, Spain, China) account for half of that sum, 4 further countries (Germany, Iran, France, UK) account for half of the rest, and all the other 190 or so countries combine for 183,750.

But the virus also visits other large countries with huge cities: India (1,251 cases so far), Brazil (4,681 cases), Egypt (656 cases), Russia (2,337), ... Suppose these countries tested as much as the 8 leading countries, and imagine coronavirus sneaks into Delhi, Sao Paulo, Cairo or Moscow: Each of these cities could easily match the world total today when all is said and counted.

I've just spent some time looking into Japan, since Tokyo is the largest city on the planet, and the Japanese don't have a lot of cases and don't test a lot: they say they're not undertesting, but we'll see. Japan is interesting because initially their curve looks like they had a lower transmission rate than the rest of the world (Financial Times, March 16):

Their current curve looks like they contained the outbreak twice (thin dark red line with daily dots):

I've been thinking on how to interpret These log/log diagrams that Oystein has introduced to this forum before, and rethinking it since I saw the method used in a youtube video. With a 7-day average and the logarithmic plots, it's basically destroying most of the variation that makes the curves recognizable, and makes it hard to compare countries. I found that by comparing these two graphs above, a lower rate of transmission means the trajectory on the log-log plot is shifted downwards. So that's one thing to watch out for when interpreting these: not the slope is important in the exponmential phase, because they're all similar, but how high up the chart the graph is: lower is better and means a slower rate of spread.

Japan looked even better initially until I reduced the total cases from the "all cases since the beginning" metric to only include the last 14 days in the total. That's a somewhat arbitrary cutoff, but it's equal to the incubation period, after 14 days, people should no longer be infecting others. And that means Japan's two dips don't give it a cushion of total cases against which the rate looks good because now we're looking at active cases. It also means that Korea and China "doubled back" to an earlier place in the state space, from which an outbreak could start anew if the virus breaches the containment again.

Japan also disproves the theory that this kind of diagram gives a good indication when a country has beat the virus: Japan looks like it did twice (and probably would have even on a 7-day sliding daily case average), but is now set to grow its epidemic further.

 

[Oystein]

I've just spent some time looking into Japan, since Tokyo is the largest city on the planet, and the Japanese don't have a lot of cases and don't test a lot: they say they're not undertesting, but we'll see. Japan is interesting because initially their curve looks like they had a lower transmission rate than the rest of the world (Financial Times, March 16):
View attachment 40109
Their current curve looks like they contained the outbreak twice (thin dark red line with daily dots):
View attachment 40110
...
Japan also disproves the theory that this kind of diagram gives a good indication when a country has beat the virus: Japan looks like it did twice (and probably would have even on a 7-day sliding daily case average), but is now set to grow its epidemic further.

Quick reply: Japan's first "escape" came when they had under 100 cumulative cases. That was a local event. The second came at well under 1000 cases - this could be regional.

Some countries have loops.

 

[Bruno D.]

The world is still at under 1 million cases (tested positive). If this undercounts the actual infected 1:10, the world is at 10 million.
https://www.worldometers.info/coronavirus/ currently has 810,123 in the world. Only 4 countries (USA, Italy, Spain, China) account for half of that sum, 4 further countries (Germany, Iran, France, UK) account for half of the rest, and all the other 190 or so countries combine for 183,750.

But the virus also visits other large countries with huge cities: India (1,251 cases so far), Brazil (4,681 cases), Egypt (656 cases), Russia (2,337), ... Suppose these countries tested as much as the 8 leading countries, and imagine coronavirus sneaks into Delhi, Sao Paulo, Cairo or Moscow: Each of these cities could easily match the world total today when all is said and counted.

Unfortunately Brazil is not doing a good job. Per lack of testing kits, they are restricting who gets tested ... a lot! In the biggest city, Sao Paulo, they are testing only people that are in danger of going to ICUs. Health experts there are estimating 1:11 underreporting instead of 1:10.

Good source for testing per country: https://ourworldindata.org/coronavirus-testing-source-data

 

[Mendel]

Unfortunately Brazil is not doing a good job. Per lack of testing kits, they are restricting who gets tested ... a lot! In the biggest city, Sao Paulo, they are testing only people that are in danger of going to ICUs. Health experts there are estimating 1:11 underreporting instead of 1:10.

Good source for testing per country: https://ourworldindata.org/coronavirus-testing-source-data

First, note that people are complaining about not getting tested everywhere. You can't test everyone with a cold, there aren't enough tests for that, in any country. Patients are selected on whether they are actually at risk of having been infected, or when they have serious clinical symptoms: if your lungs show signs of a viral pneumonia, for example, in a chest X-ray. Many people feel entitled to a test and complain when the doctor decides that there is no necessity, and the doctor may say that the test are scarce, which is true. It doesn't mean that there is undertesting!

Your "good source" for testing numbers shows

The column "Cutoff date" corresponds to the date for which the "Total Tests" figure is calculated. The column "Source publication date" corresponds to the date when the figure was reported (e.g. date of press release).

Country or territory	Total tests	Date	Source	Source Date	Remarks
Brazil	2927	13 Mar 2020	Media report in Folha de S.Paulo	16 Mar 2020, 10:29 local time	Media report – not attributed to any official source.

Note that the cutoff date is March 13th.
https://www.who.int/docs/default-so...0314-sitrep-54-covid-19.pdf?sfvrsn=dcd46351_8 is the WHO report from March 14th. It shows 98 cases for Brazil. Thus, 3% of all tests were positive (possibly as much as 6% if tests are run twice). That seems perfectly ok.
Do you have more recent numbers?

I looked a bit at the Brazilian response to the epidemic, and what I'm seeing is positive: companies are working to develop and deploy blood tests, and people are using 3D printers to make face shields. (And Bolsonaro encouraging people to break the social distancing laws?)

 

[qed]

Had we turned a corner? Sadly no.

 

[qed]

Is it correct/obvious to conclude that countries with a lower number of infections per 1 million of their population, are doing better than those with a higher number?

And hence, is it correct to conclude that currently the USA is doing better than Italy, Spain, Germany, France, Switzerland, Belgium, the Netherlands, Austria, Israel, Norway, Ireland, Luxembourg and Iceland, say.

 

[Mick West]

Is it correct/obvious to conclude that countries with a lower number of infections per 1 million of their population, are doing better than those with a higher number?

Not really, as we are different points in the curve. You could easily say someone with stage 2 lung cancer is doing better than someone with stage 4. But if it develops into stage 4, then they really did the same.

At some point Italy had the exact same death/population as the US does now.

 

[Mendel]

Is it correct/obvious to conclude that countries with a lower number of infections per 1 million of their population, are doing better than those with a higher number?

And hence, is it correct to conclude that currently the USA is doing better than Italy, Spain, Germany, France, Switzerland, Belgium, the Netherlands, Austria, Israel, Norway, Ireland, Luxembourg and Iceland, say.

No. Rember the "lag behind Italy" trackers? Where we looked how far along the path that Italy had taken we all were?
The rate of spread determines how well a country is doing. As long as it's up, you'll eventually get to the point where the other countries are now.
Any country that adds more than 10% of its total cases each day is doing badly, in my estimate.

 

[Oystein]

Also, it is probably not prudent to compare large area countries with small countries like Luxemburg or Iceland, where an outbreak in the one larger city at once represents the majority of the people. In the USA, only some regions are farther into the curve, others are only just starting. This smears out the rates - and yet USA is growing at alarming rates.

 

[Mendel]

Is it correct/obvious to conclude that countries with a lower number of infections per 1 million of their population, are doing better than those with a higher number?

And hence, is it correct to conclude that currently the USA is doing better than Italy, Spain, Germany, France, Switzerland, Belgium, the Netherlands, Austria, Israel, Norway, Ireland, Luxembourg and Iceland, say.

Italy took 43 days to go from 100 cases to 100 000.
Spain did it in 36 days.
The US only needed 31 days.
Source:
Source: https://docs.google.com/spreadsheets/u/0/d/1g_YxmDfQx7aOU2DKzNZo9b-NTk62Bju6X3z6OuCa6gw/htmlview#
, Tab "cumulative cases". This is the data underlying https://informationisbeautiful.net/visualizations/covid-19-coronavirus-infographic-datapack/?

 

[Mendel]

The spreadsheet in my previous post has a tab, "average deaths by day". (They used to have a graph for that, but people apparently used it to conclude Covid-19 was harmless since the average rate wasn't that high yet.) Since we had well over 6000 Covid-19 deaths worldwide yesterday (ECDC), Covid-19 now kills faster than Aids, Malaria, and influenza combined (on average).

In Germany, the Covid-19 deaths yesterday exceed the lung cancer average. (RKI: 141 deaths; Wikipedia:Todesursache 41495 in 2007)

 

[DavidB66]

For people trying to follow and compare international statistics, I note that the latest daily update from the UK Department of Heath includes a note explaining what is and isn't covered by the death figures. As noted previously, with a few exceptions they cover only people dying in hospital.
https://www.gov.uk/guidance/coronavirus-covid-19-information-for-the-public
However, there is now a new source of information to supplement the daily updates. The Office for National Statistics will be publishing a weekly report, on every Tuesday, giving figures for all death registrations in the relevant week in which COVID-19 is mentioned as a cause of death. This will include people dying outside hospitals, such as nursing homes, etc. It may include some cases where COVID-19 has been diagnosed but not tested, and of course it may omit some cases where the patient was in fact infected but the infection was not diagnosed. The first report was issued on 31 March. As the ONS explain, there is an inevitable time lag between death, registration of the death, and collating of the information, so it will take some time to get the complete picture. For more details see the ONS document here https://www.ons.gov.uk/news/statementsandletters/deathsrelatingtothecoronaviruscovid19

 

[Agent K]

U.S. projections from the University of Washington.
https://covid19.healthdata.org/projections

 

[Jeff Semenak]

So, does this peak mean the majority of people were infected in April? Long after stay home advisories and school closures? How did all these new cases get infected?

 

[Mendel]

So, does this peak mean the majority of people were infected in April? Long after stay home advisories and school closures? How did all these new cases get infected?

a) The incubation time is up to two weeks, and then people still need to develop pneumonia.
b) Social distancing measures aren't total quarantine, like in Wuhan.
c) People still meet for work, on public transports, families...
d) Germany has had social distancing measures for three weeks, we're now at linear growth (R=1, approximately the same number of new cases last week as the week before). If that continues, we 're going to have a peak hospital bed use in a week or so as people getting dismissed from the hospital equal the number of admissions (and we might actually be already at that point), but what we are really looking for is to get the transmission rate below 1, which means declining numbers of infections. It's difficult.

 

[Agent K]

So, does this peak mean the majority of people were infected in April? Long after stay home advisories and school closures? How did all these new cases get infected?

Not necessarily. The projection of needed hospital beds predicts how many total hospital beds are needed on a given day, which counts patients who have been hospitalized for a while. The peak occurs when discharges and deaths start to outnumber new needed hospital beds.

Edit: Also, "flattening the curve" delays the peak by lowering the basic reproduction number, R0.

 

[Bruno D.]

First, note that people are complaining about not getting tested everywhere. You can't test everyone with a cold, there aren't enough tests for that, in any country. Patients are selected on whether they are actually at risk of having been infected, or when they have serious clinical symptoms: if your lungs show signs of a viral pneumonia, for example, in a chest X-ray. Many people feel entitled to a test and complain when the doctor decides that there is no necessity, and the doctor may say that the test are scarce, which is true. It doesn't mean that there is undertesting!

Covid-19 is contagious even if you are not showing symptoms, that's why it's so dangerous. Your argument of "testing is only needed on a selected number of patients" is what led the pandemic to where it is right now. If a country had enough test kits to test the whole population (no country has it), then the whole population should be tested. Testing for treatment purposes only is the wrong approach. Some countries didn't have enough kits due to lack of preparedness, money or care, but those that had it were able to contain the spread much better.

South Korea had much less restricting testing criteria from the beginning, as they had the test-kits to spare, and the virus there is much more contained there than anywhere else.

If the doctor says that they cannot test a patient because tests are scarce, then it's undertesting per definition.

Chart from the same source - https://ourworldindata.org/covid-testing

Your "good source" for testing numbers shows ...

As they explain in the same link I posted:

Given the large problems with data availability, we did a manual review of data across national reports, and included the most recent estimates that we could find as of 20 March 2020, 18:00 GMT

I used it to illustrate how little testing was being done in Brazil. 2927 tests on Mar/13th, with 98 confirmed cases and community spread already being confirmed, is a terrible number.

If you have a better source for real-time testing per million people than this one, please let me know ...

Do you have more recent numbers?

5 minutes googling ...

... government has distributed just 54,000 tests for a population of about 210 million. Last month the health ministry stopped releasing the numbers of suspected cases ...
... This week, Edmar Santos, health secretary for Rio de Janeiro, said the state could have 50 to 100 infected people for each of its 1,074 confirmed cases ...
... "You have a mountain of other deaths that are either directly or indirectly associated with the epidemic and they are not notified." ...
... In São Paulo state, Brazil's most populous with 44 million people and 4,048 cases, the government laboratory handling tests has a backlog of 16,000 ...
... Cemeteries in São Paulo are burying 30 to 40 people a day with coronavirus symptoms but, in most cases, no test result ...
... "This number being released is totally unreal. This makes people think everything's normal," Lima said, while her sister Christiane, coughed incessantly in the background. ...

https://www.theguardian.com/global-...oronavirus-tally-ignores-a-mountain-of-deaths

So, yes, undertested and underreported.

I looked a bit at the Brazilian response to the epidemic, and what I'm seeing is positive: companies are working to develop and deploy blood tests, and people are using 3D printers to make face shields. (And Bolsonaro encouraging people to break the social distancing laws?)

Sorry, it hurts me to say that no, Brazilian response not positive at all. They were motivated 1 month ago, but they lacked since that. Bolsonaro is denying that it's serious, he is against social distancing, and he says that the leftist main stream media is using the covid-19 scare to hurt the economy and hurt his presidency.

I don't know why you picked my answer to criticize like that, but my point is not that Brazil is doing a specially poor job (I'm Brazilian btw), Brazil is doing as poorly as most countries; my point was that it's undertested basically everywhere, and I used Brazil as an example because it was in Oystein post and I had more information on that.

 

[Mendel]

54000 test for 9000 cases means you can test 6 negatives for each positive. If you manage to use that wisely, it can work out.

Covid-19 is contagious even if you are not showing symptoms, that's why it's so dangerous. Your argument of "testing is only needed on a selected number of patients" is what led the pandemic to where it is right now.

From your source:

widespread testing helped Taiwan, South Korea and Germany blunt the epidemic's rise.

I'm from Germany, I know how we have been testing because the guidelines are public. We are also screening respiratory illnesses:

In der 13. KW sind in zwei von insgesamt 121 untersuchten Sentinelproben (1,7 %) SARS-CoV-2 identifiziert worden.

https://influenza.rki.de/Wochenberichte/2019_2020/2020-13.pdf
This says that in the week 21-27 March, 121 samples of persons with symptoms were tested for all kinds of viruses, and only 1.7% had coronavirus. You need 60 times more tests than cases if you want to test everyone with symptoms.According to the article you linked, we are doing the most testing per capita in the world, but even we don't have that.

Instead, besides treatment tests on people who present with pneumonia, we tested people who have symptoms and a link to an infection source, or people who have symptoms and have high risk (e.g. doctors). This supports contact tracing and containment. Contact tracing and containment is how South Korea managed its epidemic. Note also that Taiwan is an island, and South Korea is a peninsula witha fairly tight border to the North; while Germany has open borders and many Germans went skiing in the Alps in Italy or in Austria next to the Italian border and brought Covid home. In my town, 5 such vacationers developed symptoms on the way home, their group was tested, they were identified and quarantined.

You can't use the tests to test everyone who is fearful and afraid, it just is not possible. You need to use them wisely, or you don't have enough where they can really make a difference: where you need to identify and contain a local spread one infection at a time.
The "per capita" test rate is indicative of potential, but what you want is a good "per case" test rate that allows you to keep track of the people whom your identified cases might have infected, and who these might have infected, etc. I don't know if there are data on this, but my impression is that this should be possible if your tests come back 5%-10% positive. So your 15% is cause for concern, but not the end of the world.

The problem with ramping up testing is that you need expensive lab equipment and medical technicians to run it. A lab machine can do 96 tests in 4-6 hours if you have the chemicals, but at a certain point, you need to buy more machines and train more people to operate them. You can't just demand more tests and expect them to happen. Germany has many independent labs; the labs have been learning this process since January and have been increasing their capacities, but we're starting to run out of chemicals.

What you need to achieve in Brazil is to keep the infection rate down so that quarantine measures (based on tests) have a chance to contain the spread again and again. Contact tracing is the only way to let people know they might be infectious before they have symptoms; that is the path to containment, and containment is the way out of this crisis.
(The other way is to push the transmission rate from R=5.2 to R=0.58 and wait a few weeks, as Wuhan has done with their strict quarantine measures, and today China uses an app to do contact tracing and quarantine. It appears to be working.)

Keep a diary on whom you were in contact <6 feet unprotected, and for how long. If you develop Covid-19, go back two days from the onset of symptoms; some of the people you had contact with for more than 15 minutes will have been infected by you.

If social distancing reduces the contacts everyone has, you do not need as many tests to do thorough contact tracing, because there aren't as many contacts for each case. These measure reduce the number of test per case. Do it well and you can beat the virus!

 

[Mendel]

For people trying to follow and compare international statistics, I note that the latest daily update from the UK Department of Heath includes a note explaining what is and isn't covered by the death figures. As noted previously, with a few exceptions they cover only people dying in hospital.
https://www.gov.uk/guidance/coronavirus-covid-19-information-for-the-public
However, there is now a new source of information to supplement the daily updates. The Office for National Statistics will be publishing a weekly report, on every Tuesday, giving figures for all death registrations in the relevant week in which COVID-19 is mentioned as a cause of death. This will include people dying outside hospitals, such as nursing homes, etc. It may include some cases where COVID-19 has been diagnosed but not tested, and of course it may omit some cases where the patient was in fact infected but the infection was not diagnosed. The first report was issued on 31 March. As the ONS explain, there is an inevitable time lag between death, registration of the death, and collating of the information, so it will take some time to get the complete picture. For more details see the ONS document here https://www.ons.gov.uk/news/statementsandletters/deathsrelatingtothecoronaviruscovid19

I looked.

Of the deaths registered in week 12, 103 mentioned novel coronavirus (COVID-19), which is 1.0% of all deaths.
This is lower than the figures reported by the Department of Health and Social Care (DHSC) as it takes time for deaths to be reported as it takes time for deaths to be reported.
If we analyse the data by date of death and look at registrations after 20 March, then 181 deaths resulting from COVID-19 occurred by 20 March, which is higher than the figures the DHSC publish as it includes deaths related to COVID-19 that took place outside of hospitals and those not tested for COVID-19.
This number is different from the count of deaths published on the GOV.UK website because of different reporting methods and timing: Office for National Statistics (ONS) weekly deaths figures are based on deaths registered in the stated week, and we have counted all deaths where COVID-19 was mentioned on the death certificate as "deaths involving COVID-19"; the GOV.UK figures are based on deaths occurring to date, among hospital patients who have tested positive for COVID-19, and include deaths that have not yet been registered.

https://www.ons.gov.uk/peoplepopula...nalfiguresondeathsregisteredinenglandandwales
(scroll down, there's a section titled "Important Notes")

The ECDC data for March 14-March 20 has 134 cases for Great Britain; 167 if their numbers lag a day. We know that 103 must be lower than the number of deaths coming from the hospital system on the day because registration lags; and 181 must be high because it includes deaths that became known after the daily hospital number was published. The numbers we have (134 or 167) fit in that interval, and I wouldn't expect the number of non-hospital deaths to make too much of a difference, possibly 10% if that?

The daily numbers are useful if you're tracking the progress of the epidemic; the ODS numbers will be useful if you're tracking the exact mortality. But either way, there isn't some conspiracy-level number of deaths missing from the statistic, the hospital-only numbers are fairly close to the real deal.

 

[Leifer]

Among the reported #s of completed tests, we need to remember......
that each "positive but recovered" person, often receive at least 3 COVID tests if diagnosed as positive.
One to confirm infection.
Two or more to verify it has passed (no longer shedding, and sufficiant presence of antibodies,)

 

[Mendel]

Among the reported #s of completed tests, we need to remember......
that each "positive but recovered" person, often receive at least 3 COVID tests if diagnosed as positive.
One to confirm infection.
Two or more to verify it has passed (no longer shedding, and sufficiant presence of antibodies,)

Do you have a source for that?
I'm certain that the lab tests are rt-PCR everywhere, so it is absolutely impossible to test for presence of antibodies, and I'd have thought the infectiousness "times out" as the decrease in viral load in the throat and nose has been documented, but I don't really know.

 

[Leifer]

Do you have a source for that ?

There may (edit: will) be progress during and after this is solved. A quick antibody test, could be relied upon if figured-out.
https://www.fiercebiotech.com/medte...-serological-antibody-blood-test-for-covid-19

 

[Mendel]

There may (edit: will) be progress during and after this is solved. A quick antibody test, could be relied upon if figured-out.
https://www.fiercebiotech.com/medte...-serological-antibody-blood-test-for-covid-19

From your source:

The FDA granted its first emergency authorization for a rapid antibody blood test for COVID-19, developed by Cellex, allowing people to gauge a person's immune response to the novel coronavirus.

This EUA was granted April 1st, so this testing is not included in test numbers before that. I looked at it more closely:

Furthermore, the samples tested came from people with acute infections, and it's unclear if the test is sensitive enough to show antibodies in people who are no longer sick.

The UK rejected it and similar tests:

None of the antibody tests ordered by the government is good enough to use, the new testing chief has admitted.

John Newton said that tests ordered from China were able to identify immunity accurately only in people who had been severely ill and that Britain was no longer hoping to buy millions of kits off the shelf.

https://www.thetimes.co.uk/article/...-antibody-tests-but-they-don-t-work-j7kb55g89

I have looked at treatment guidelines for Covid-19 and haven't seen any mention of people getting tested to be cleared, I don't recall that being done on any infectious disease I had, but maybe that's routine in the US and nobody mentions it, that's why I asked for a source. What I do know is that your nasal swab may come up negative when you check into the hospital with pneumonia, so they'd want to get something out of your lungs for the test, and test that, if they didn't do that in the first place, so I can see two tests happening for some people, but I'm sceptical of claims that 3 test are required, and 2 of them being required when you are no longer sick.
I'm now assuming you don't remember where you heard that from, and that it's not reliable.

 

[DavidB66]

The Office for National Statistics in the UK has just released its latest weekly bulletin on death registrations, for the week up to 27 March, including a discussion of COVID-19 cases.

https://www.ons.gov.uk/peoplepopula...landandwalesprovisional/weekending27march2020

I haven't read this in detail, but it appears that COVID-19 deaths outside hospital, which are not recorded in the Department of Health daily reports, are only a small proportion of the total:

Of deaths involving COVID-19 in Week 13, 92.9% (501 deaths) occurred in hospital with the remainder occurring in hospices, care homes and private homes.

As previously, caution is needed because of the time lags in registration of deaths.

 

[Mendel]

The Office for National Statistics in the UK has just released its latest weekly bulletin on death registrations, for the week up to 27 March, including a discussion of COVID-19 cases.

https://www.ons.gov.uk/peoplepopula...landandwalesprovisional/weekending27march2020

They seem to have worked through quite a bit of backlog over the past week (or maybe some doctors remembered to send their registrations in), it extends over more than 3 weeks:

 

[deirdre]

but I'm sceptical of claims that 3 test are required, and 2 of them being required when you are no longer sick.
I'm now assuming you don't remember where you heard that from, and that it's not reliable.

that happened with the cruise ship passengers we put in military quarantine. Diamond Princess specifically. (looked for 5 mins but there are so many diamond princess articles, I haven't found one discussing that yet. but honest that is what was being reported.. 1 person got their 2 negative tests 24 hours apart and was released but still infected their family on return home)

But I doubt that happens for hospital patients. hospital patients are likely told to go home and discuss it with their regular doctor. Articles I have read rearding "when am I no longer infectious?" tell you to discuss it with your primary doctor and stay home (isolating from family) until he tells you you are probably ok.

I just read an article yesterday of a nurse who had it and she said she insisted on a test to see if she was negative because they aren't (she says) testing before medical personnel come back to work.. although it's a big country and it could just be her hospital or state.

edit add: march 19, nurse in colorado

I received conflicting information. Initially, it was that I had to have two negative COVID-19 tests that were 24 hours apart. A few days later, the public health department said I just need to quarantine for 14 days and don't need to be retested. As a healthcare provider, I want to get retested just to make sure I'm not a carrier.

(The Colorado Department of Public Health told 9NEWS, "Currently, Colorado health care workers who test positive or who were potentially exposed to COVID-19 are excluded from work for 14 days, following CDC guidance.")

https://www.9news.com/article/news/...ioner/73-2f1c2070-7d2d-47e9-aede-9e1cbdde60f5

 

[deirdre]

the key phrase to google is "24 hours apart"

that happened with the cruise ship passengers we put in military quarantine. Diamond Princess specifically. (looked for 5 mins but there are so many diamond princess articles, I haven't found one discussing that yet. but honest that is what was being reported.. 1 person got their 2 negative tests 24 hours apart and was released but still infected their family on return home)

If patients have been tested positive for coronavirus, the same requirements apply and patients must have two consecutive negative tests 24 hours apart.

https://www.mysanantonio.com/news/l...out-new-requirements-for-leaving-15137964.php

Centers for Disease Control and Prevention (CDC) officials announced on Sunday the release of an individual from isolation in a facility in San Antonio, Texas, after the person twice tested negative for novel Coronavirus. The person subsequently tested "weakly positive" and was taken back into quarantine. Officials said the patient had contact with others during the short release

https://www.breitbart.com/border/20...-coronavirus-quarantine-later-tests-positive/

 

[Mendel]

This is the current CDC guideline:
https://www.cdc.gov/coronavirus/2019-ncov/hcp/disposition-hospitalized-patients.html

Discontinuation of transmission-based precautions for patients with COVID-19:
The decision to discontinue Transmission-Based Precautions should be made using a test-based strategy or a non-test-based strategy (i.e., time-since-illness-onset and time-since-recovery strategy). Meeting criteria for discontinuation of Transmission-Based Precautions is not a prerequisite for discharge.

1. Test-based strategy.
   • Resolution of fever without the use of fever-reducing medications and
   • Improvement in respiratory symptoms (e.g., cough, shortness of breath), and
   • Negative results of an FDA Emergency Use Authorized COVID-19 molecular assay for detection of SARS-CoV-2 RNA from at least two consecutive nasopharyngeal swab specimens collected ≥24 hours apart (total of two negative specimens) [1]. See Interim Guidelines for Collecting, Handling, and Testing Clinical Specimens for 2019 Novel Coronavirus (2019-nCoV).
2. Non-test-based strategy.
   • At least 3 days (72 hours) have passed since recovery defined as resolution of fever without the use of fever-reducing medications andimprovement in respiratory symptoms (e.g., cough, shortness of breath); and,
   • At least 7 days have passed since symptoms first appeared

When a Testing-Based Strategy is Preferred
Hospitalized patients may have longer periods of SARS-CoV-2 RNA detection compared to patients with mild or moderate disease. Severely immunocompromised patients (e.g., medical treatment with immunosuppressive drugs, bone marrow or solid organ transplant recipients, inherited immunodeficiency, poorly controlled HIV) may also have longer periods of SARS-CoV-2 RNA detection and prolonged shedding of infectious recovery. These groups may be contagious for longer than others. In addition, placing a patient in a setting where they will have close contact with individuals at risk for severe disease warrants a conservative approach.
Hence, a test-based strategy is preferred for discontinuation of transmission-based precautions for patients who are

• Hospitalized or
• Severely immunocompromised or
• Being transferred to a long-term care or assisted living facility

If testing is not readily available, facilities should use the non-test-based strategy for discontinuation of Transmission-Based Precautions or extend the period of isolation beyond the non-test-based-strategy duration, on a case by case basis in consultation with local and state public health authorities.

So basically, you have a choice: when the symptoms are gone, you can either
a) take a swab, wait 24 hours, take another swab, send both in for testing, wait for the result, or
b) wait 3 days
In practice, there's probably not much of a difference? So most people would probably go with the "wait it out" strategy?

I'm also assuming that the double swab uses only one test, but I don't know how that is handled in practice.

 

[Agent K]

L.A. county projected that 95.6% of its residents would be infected by August if stay-at-home measures are lifted.
https://covid19.lacounty.gov/wp-content/uploads/04.10.20-Daily-Briefing.pdf

Does that mean they have an idea of how many people are actually infected compared to the number of confirmed cases?

By April 10, they counted 8,430 cases, 2,043 hospitalizations, and 241 deaths. 20% of hospitalized patients needed a ventilator.
So the ratio of deaths to hospitalizations is 12%, which is similar to the flu, but the big question is what percent of infections are ever confirmed.
24% of those who test positive are hospitalized, which means they had to be really sick to get tested.
Indeed, on March 20, it was reported that

Los Angeles County health officials advised doctors to give up on testing patients in the hope of containing the coronavirus outbreak, instructing them to test patients only if a positive result could change how they would be treated...
A front-line healthcare provider who was not authorized to speak to the media and requested anonymity said county doctors are interpreting Thursday's letter and other advice coming from senior L.A. County public health officials to mean they should only test patients who are going to be hospitalized or have something unique about the way they contracted the virus.
https://www.latimes.com/california/story/2020-03-20/coronavirus-county-doctors-containment-testing

According to UCLA's dashboard, 91% of its COVID-19 tests were negative, and they probably required a negative flu test first. But those patients must have been very sick to get tested for COVID-19, so how many of them are hospitalized? For every hospitalized patient who tests positive, are there nine who tested negative? Were they false negatives? What made them so sick if it's not the flu or COVID-19?

https://www.uclahealth.org/coronavirus

 

[Mendel]

@Agent K
1) You don't have to be "very sick" to get tested. Fever and a cough is enough. But you need to hospitalized, work in a hospital, belong to a high-risk-group, or be essential. The CDC hasn't advocated teasting contacts for 3 weeks now, which indicates to me that they have given up on containment.

Clinicians should use their judgment to determine if a patient has signs and symptoms compatible with COVID-19 and whether the patient should be tested. Most patients with confirmed COVID-19 have developed fever1 and/or symptoms of acute respiratory illness (e.g., cough, difficulty breathing).

https://www.cdc.gov/coronavirus/2019-ncov/hcp/clinical-criteria.html

2) Hospitalization rates differ by hospitalization strategy. Early on, we hospitalized more people to isolate them; now we do more home isolation. The hospitals in our Covid-19 register report that over 80% of patients in the ICU receive ventilation, but not all Covid-19 patients are necessarily in the ICU.

3)

24% of those who test positive are hospitalized

This means for every hospitalized patient, there are 3 positive cases who are not in the hospital, and 36 who had symptoms but not from Covid-19. They may or may not be hospitalized (e.g. if they have pneumonia).

4) Deaths lag behind because people take some time to die. As of April 13, they are at 320/2358=13.6%. LA's case fatality rate is 3.4% right now, which is what I remember Wuhan having.
Data show that more people need to be ventilated than with a normal flu, didn't I already post that?

5) The "test only symptoms" strategy is similar to what Wuhan did, and leads to the 80%/20% mild/severe split. You are likely not testing the ~20-40% who are asymptomatic (as data from the Diamond Princess shows). If younger people are asymptomatic more often, then we may have more asymptomatic people than that. Studies are ongoing.

 

[Agent K]

@Agent K
1) You don't have to be "very sick" to get tested. Fever and a cough is enough. But you need to hospitalized, work in a hospital, belong to a high-risk-group, or be essential.

When almost a quarter of those testing positive are hospitalized, that tells me that they're very sick when they're tested. The L.A. Times reported

They are not planning to test patients who have the symptoms but are otherwise healthy enough to be sent home to self-quarantine — meaning they may never show up in official tallies of people who tested positive.

Well, it's not that bad, since not all patients who tested positive were hospitalized, but still almost a quarter. And if they're stressing the hospitals, then what about the 91% who had similar symptoms but tested negative? If a quarter of them can't breathe, then they to be hospitalized and given oxygen whether they have COVID-19 or not.

 

[Mendel]

When almost a quarter of those testing positive are hospitalized, that tells me that they're very sick when they're tested. The L.A. Times reported

They are not planning to test patients who have the symptoms but are otherwise healthy enough to be sent home to self-quarantine — meaning they may never show up in official tallies of people who tested positive.

Well, it's not that bad, since not all patients who tested positive were hospitalized, but still almost a quarter. And if they're stressing the hospitals, then what about the 91% who had similar symptoms but tested negative? If a quarter of them can't breathe, then they to be hospitalized and given oxygen whether they have COVID-19 or not.

Your assumptions are bad.

Americans are older on average, which means you have somewhat more severe cases, but 75% vs 80% is close. You wrote that 20% of hospitalized patients are on ventilators, 20% of 25% is 5%, again close to the Wuhan value of 4.7%.
5% of the positive patients have critical cases. Your idea that everyone who got tested is already at death's door is wrong.

Secondly, while we know that 25% of the cases who test positive are severe, it does NOT follow that 25% of those who don't are as well. Hypothetical example: 2 people have ARDS, 8 people have pneumonia, 390 people are old or have multiple risk factors and also fever and a slight cough. We test them and find 2/8/30 positive and 360 negative. All of the negative tests merely have a cold, but we needed to test them so we could have an eye on the 30 who are in danger.
(And consider that cases usuall get critical days after they get tested.)

Btw, people who test negative and don't have the virus are true negatives. I don't know why you mentioned false negatives.

The new guidance says, if you're not going to do anything differently, don't test. If you're going to send them home either way, no test. If you're going to wheel them in the ICU and put them on a ventilator either way, don't test. The LA guidelines pretty much say "no matter what your test result was, isolate anyway".

9. I have symptoms and got tested for COVID-19 but my results are negative is there anything I should be doing?
If you are symptomatic but have negative test results for COVID-19, we still recommend that you stay home for at least 3 days (72 hours) after recovery, which means your fever has resolved without the use of fever- reducing medications and there is improvement in your respiratory symptoms (e.g., cough, shortness of breath), AND at least 7 days have passed since your symptoms first appeared.
It is important to note that if you were a contact to a suspected or positive case when you got tested you must remain in quarantine for the full 14 days even if your test results were negative.

http://publichealth.lacounty.gov/media/Coronavirus/GuidanceTestResults.pdf
The difference is that with a positive test, you're supposed to notify your contacts.

I obviusly don't know what the doctors will actually do about this guidance, that remains to be seen. But I think it sends a signal that tests are still somewhat scarce.

 

[Agent K]

Your assumptions are bad.
View attachment 40324
Americans are older on average, which means you have somewhat more severe cases, but 75% vs 80% is close. You wrote that 20% of hospitalized patients are on ventilators, 20% of 25% is 5%, again close to the Wuhan value of 4.7%.
5% of the positive patients have critical cases. Your idea that everyone who got tested is already at death's door is wrong.

Secondly, while we know that 25% of the cases who test positive are severe, it does NOT follow that 25% of those who don't are as well.

Many people have been denied tests because they're not sick enough, like this one in New York
https://www.businessinsider.com/coronavirus-patient-experience-new-york-emergency-room-test-2020-4

When the doctor arrived, she told me there was nothing she could do to treat my symptoms. The fact I could speak full sentences was a good sign, she said. In fact, she added, I was an example of why young people with mild cases needed to stay home. Since New York City hospitals are reserving tests for people with the most severe illnesses, many of the city's reported cases are acute patients...
"We unfortunately just aren't able to test everyone, even though we wish we could," Megan Coffee, an infectious-disease clinician in New York City, told me. "Right now we're having to focus on everybody who needs intubation and critical care."
...
In the emergency room, I asked the doctor what it would take for my case to no longer be considered mild — or even to warrant a test. The doctor said I would have to struggle to breathe while seated. (I certainly felt that way, but what more was there to say?) The hospital could do a chest X-ray, the doctor added, but it would only confirm what they already knew: I had the virus.

That's New York, but China also counted mild cases as those that don't require hospitalization, including mild pneumonia. They're probably the tip of the iceberg of undiagnosed milder cases.

 

[Mendel]

@Agent K , you are equating test availability in New York at the time of that anecdote (NY had 7845 cases then) with test availability in Los Angeles --> bad assumption.

I told you what "the tip of the iceberg" is --- they tested everyone on the Diamond Princess, and there were similar studies, and 59% to 82% of cases have symptoms. The Diamond Princess cases with symptoms have a mortality rate in the older age groups that is similar to the rate reported from Wuhan.

You can be on the safe side and guess that wherever people are tested to CDC guidelines (gotta have symptoms), about half the cases are asymptomatic and are not in the statistic. This doesn't affect the forecasts substantially as to hospital use until you get to ~35% cases because that is when herd immunity will really kick in strongly and turn the curve around.

If they were missing a significant percentage of "mild" cases, then the number of severe cases and deaths would be proportionally much higher than in Wuhan, but it isn't.

 

[Mendel]

@Agent K
Information from Lower Saxony (a German state, population ~8 million):

Es gibt derzeit 8070 laborbestätigte Covid-19-Fälle (+119 im Vergleich zu gestern) in Niedersachsen. 223 Menschen, die mit dem Virus infiziert waren, sind verstorben.

In niedersächsischen Kliniken werden derzeit 945 mit dem Virus infizierte Patientinnen und Patienten behandelt: Davon liegen 709 Erwachsene und ein Kind auf Normalstationen, 235 Erwachsene benötigen intensivmedizinische Behandlung. Auf den Intensivstationen müssen 162 Erwachsene beatmet werden, davon fünf auf ECMO-Plätzen.

Geschätzte Zahl der Genesenen*: 3994 (49,5 % der bislang gemeldeten laborbestätigten Fälle)

https://www.niedersachsen.de/Coronavirus
8070 cases
223 deaths (CFR 2.8%)
3994 recovered (49.5%, estimated)
3853 active cases
945 hospitalized (24.5% of active cases)
235 in intensive care (ICU)(24.9% of hospitalized, 6.1% of active cases)
162 of ICU patients (68.9%) (17.1% of hospitalized) are ventilated, including 5 on ECMO

I don't know the current test rates, but the last numbers I have seen for Germany were well below 10% positive results. The percentages we are seeing here compare to Los Angeles; conditions in both cases are a health system that still has plenty of spare capacity.

 

[Agent K]

If they were missing a significant percentage of "mild" cases, then the number of severe cases and deaths would be proportionally much higher than in Wuhan, but it isn't.

Unless Wuhan was also missing a significant number of mild and asymptomatic cases, which it surely was.
What percent of the Diamond Princess cases were hospitalized?
Now we have another experiment in the USS Theodore Roosevelt, where almost 600 sailors tested positive and one died so far. They're surely younger and more fit than the Diamond Princess passengers.

 

[DavidB66]

In the UK, the Office for National Statistics has issued its second weekly report on deaths in England and Wales where the death certificate mentions Covid-19 as a factor. The report includes an analysis by a senior statistician.

https://www.ons.gov.uk/peoplepopula...cesinenglandandwales/uptoweekending3april2020

As previously, the ONS try to explain why there are differences between these figures and those published daily by the Department of Health. These are partly due to a difference of coverage. The DoH figures include all people who have died in hospital after a positive test for the Coronavirus. The ONS figures cover all registered deaths where Covid-19 is mentioned on the death certificate, whether or not they have been tested. Unlike the DoH figures, it will include people dying in care homes or in their own residence. However, the difference (so far) is not as great as might be feared: date for date, the ONS figures are only about 15 percent higher. This is significant, but not the massive undercounting that some people suspected.

The other major complication is the effect of time lags. The ONS figures take account of death occurring in a given week and registered up to a week later. [That is, up to the end of the following week, which could be nearly 2 weeks after the date of death.] The latest report covers the week up to 3 April, based on deaths in that week registered up to 11 April. It is therefore different from the total of the daily 'death counts' released by the DoH during that week, which include some deaths occurring in previous weeks, but exclude some deaths occurring in that week but not yet reported to the DoH. The net effect in the latest report is that total deaths reported in the ONS figures for the week are considerably higher than the DoH figures (by more than the 15 percent difference already mentioned). However, I think this is to be expected whenever the number of actual deaths per day is still rising and there is a time lag in reporting. If the true number of deaths per day starts falling, one would expect the time lag to have the opposite effect, as the daily DoH reports will include some figures from an earlier stage of the epidemic when the death rate was higher.

 

[Mendel]

What percent of the Diamond Princess cases were hospitalized?

IIRC all of the positive patients were moved off the ship and hospitalized?
Home isolation isn't really an option when you are abroad.

And anyway, you have to examine the cruise ship statistics by age group. AFAIK none of the 1045 crew on the Diamond Princess died, though not all were infected. My prediction is that of the young sailors on the Roosevelt, around 3 may die, excluding those over 50 years of age.

 

[Agent K]

IIRC all of the positive patients were moved off the ship and hospitalized?
Home isolation isn't really an option when you are abroad.

And anyway, you have to examine the cruise ship statistics by age group. AFAIK none of the 1045 crew on the Diamond Princess died, though not all were infected. My prediction is that of the young sailors on the Roosevelt, around 3 may die, excluding those over 50 years of age.

I was just editing my post to say that at the time all patients were hospitalized, so the number of ICU patients and deaths were a better measure of severity. But I don't think that all the USS Roosevelt patients will be hospitalized.
For comparison, about 1% of people with the flu are hospitalized, and 0.1% die, so the ratio of deaths to hospitalizations is 10%, which is similar to COVID-19, or possibly lower because of the lag between hospitalizations and deaths. The question is what percent of people with COVID-19 are hospitalized. We know that it's about a quarter of those who test positive, but we don't know how many are never tested.