914 replies to this topic

[pamojja]

I can write us a python script to check these numbers daily.  A averaging metrics across 30+ countries is a calculation that shouldn't be done manually.

 

 

I guess we both can throw our calculations, manually or with python script, to trash.

 

 

The lack of scientific justification for the corona measures

25. April 2020 Christof Kuhbandner

Why the thesis of the epidemic spread of coronavirus is based on a statistical fallacy

Practically all over the world, we are experiencing an unprecedented situation: drastic measures are being taken all over the world to combat an epidemic that is obviously looming. In Germany, for example, so many basic rights have been restricted so comprehensively and comprehensively as never before in the history of the Federal Republic of Germany.

Numerous virological experts are advising politicians. One might therefore think that the need for dramatic interventions in our basic rights is well founded by sound science. However, if you, as a scientist experienced in research methods and statistics, look at the scientific basis of what justifies the drastic measures, doubts arise.

Virtually all the measures taken are justified by the fact that they are intended to prevent an increase in the number of new infections every day, in order to counteract the alleged exponential spread of the coronavirus. For example, at a press conference on 18 March, the president of the Robert Koch Institute (RKI) projected - based on the increase in new infections observed at that time - that there would be up to 10 million infected people in Germany in two to three months if it were not possible to effectively and lastingly reduce contacts between people over a period of several weeks.

The Leopoldina - the National Academy of Sciences - formulated a similar statement in its second statement: "Although the increase in registered new infections with SARS-Cov-2 in Germany has been slowing down for a few days now, the political measures for contact restriction, which were adopted on 22 March 2020 and are valid nationwide, must remain in place". And the Federal Minister of Health, Jens Spahn, said in the ZDF Heute Journal on 19th April: "If we have managed to bring down the number of new infections together in the direction of 3,000 to 4,000 per day, then we must also succeed in staying there, only then can we gradually return to a new normality.

 

The actual point in time of the decrease in daily new infections

In view of the fact that all measures taken are justified by the increasing number of new infections every day, we would like to take a closer look at these figures. To do so, we will first take a look at the typical graph of the increase in new infections, as it has been presented for a long time in the dashboard of the RKI (as of April 23rd):

 

 Abbildung_1.jpg   240.97KB   0 downloads

What you can first note..: The numbers seem to have been decreasing at least since April 3. But now we need to take a closer look. A first question is: What exactly is meant by the date in the above graphic? In this graph in the RKI dashboard, the date corresponds to the so-called reporting date - i.e. the time when the case became known to the public health department.

This is the first exciting point: It should actually be about the number of new infections per day, i.e. the time when a person became infected with the corona virus. But at the time when a case becomes known to the public health department, the person has not become newly infected. According to the RKI, an average of 5-6 days elapse between the time of infection - i.e. the actual time of the new infection - and the development of the first symptoms. As people do not go to the doctor immediately after the first symptoms, it often takes several days before a doctor is consulted who may then carry out a test, the result of which is often only available one or sometimes even two days later. The above graph therefore lags far behind the true time of the new infection.

This is exactly the reason why the RKI dashboard has a new graphic for a few days now. It shows the number of new infections per day according to the date of onset of the disease - i.e. the day on which the first symptoms were formed. The onset of the disease is currently known from 94,078 of the 145,664 cases confirmed by laboratory diagnosis. The following picture shows the chronological course of the new infections (the blue bars show the course of the new infections at the beginning of the disease):

 

 Abbildung_2.jpg   246.03KB   0 downloads

A decrease in the daily number of new infections is therefore in reality much earlier. In order to determine the exact time, the yellow bars can be included. The yellow bars correspond to cases where the start of the disease is unknown. They are therefore still fixed at the date of notification.

In order to estimate the onset of the disease, these cases can be assigned the most probable date of onset based on the cases for which the onset of the disease is known (technical term: "imputation"). In the daily situation reports of the RKI, this is done in order to better estimate the true course of new infections. Then the graph looks like this (the height of the grey bars shows the true course estimated with the help of imputation, fixed at the beginning of the illness, situation report of 22.4:)

 

 Abbildung_3.jpg   347.84KB   0 downloads

According to this, the number of new infections per day has been decreasing since at least March 19. However, it must be remembered that the date in this graph corresponds to the time of development of the first symptoms of the disease. As already described, however, there are another 5-6 days between the time of infection - the time of the real new infection - and the time of the formation of symptoms. The above curve must therefore be shifted back in time by another 5-6 days, and so the new infections have in reality already been falling since at least 13-14 March.

Translated with www.DeepL.com/Translator (free version)

 

[pamojja]

The actual increase in daily new infections

But now we need to look even closer. Over time, not only has the number of reported daily new infections increased, but also the number of coronavirus tests performed daily. However, if there is a high number of people who are infected but not detected because the number of tests is too low - which is the case with coronavirus according to several studies - this has striking consequences: Then, as the number of tests increases, more and more new infections are found - although the number of new infections may not have increased at all or, in reality, has actually decreased.

One can illustrate this with a simple everyday example: Let us assume that ten eggs are hidden in a garden every day (the true number of new infections). On the first day, the children are allowed to search for one minute and they find one egg, on the second day they are allowed to search for two minutes and they find two eggs, and on the third day they are allowed to search for four minutes and they find four eggs (increasing the number of tests over time). The children might now get the misleading impression that they are hiding exponentially more eggs (new infections) in the garden every day because they find exponentially more eggs every day. But of course this is a problematic interpretation, because in reality there were always the same number of eggs (new infections) hidden in the garden, and the increased number of eggs found (new infections) is only due to the increased number of search attempts (coronavirus tests).

So if there is a high number of hidden but not found eggs (new infections) due to the low number of search attempts, you will automatically find more and more eggs (new infections) if you increase the number of tests, but this does not say anything about the true number of hidden eggs (new infections) per day. This example can be used to illustrate another interesting point: What would actually happen if more eggs (new infections) were actually hidden in the garden every day? Then one would have to find more eggs (new infections) than would be caused by increasing the number of tests alone. If, for example, ten eggs were hidden on the first day, 20 on the second day and 40 on the third day, not only two but four eggs would be found on the second day, and not only four but 16 on the third day. So if you double the number of tests, you will always find more than twice as many eggs.

There is now a relatively simple statistical method to determine the true course of the new infections: Simply divide the number of new infections found with a given test count by the test count. You can illustrate this with the example of eggs: The children could simply divide the number of eggs (new infections) found each day by the number of minutes of searching (number of tests). If the number of hidden eggs (new infections) remained the same, the value 1 would be obtained for all three days. If the number of hidden eggs (new infections) doubled every day, you would get the values 1, 2 and 4. This would then reflect the true course of the increase relatively accurately. In other words, this method estimates what would have happened if the same number of tests had always been used.

A note: This type of control works very reliably if the guidelines do not change which people are to be tested. If one suddenly switched from symptom-based to non-symptom-based testing, many more people who are not infected would suddenly be tested, so that the estimates would be distorted. However, this is not the case in Germany. Here, symptom-based testing was used throughout. For example, the National Association of Statutory Health Insurance Physicians (Kassenärztliche Bundesvereinigung) clarifies in a press release of 25 March

"'As before, the RKI recommends that only patients with typical symptoms of COVID-19 should be tested,' explained Gassen. In media coverage, the adjustment of the RKI criteria has sometimes been presented as an extension of the tests. 'But that is not the case', Gassen made clear. 'To test patients without symptoms is medically unsound.' Among the cases that are still to be clarified by laboratory diagnostics are people with acute respiratory problems who have had contact with a confirmed COVID-19 case in the past 14 days".

One can now see to what extent the observed increase in reported new infections overestimates the true increase due to the increase in the number of tests. Since reliable figures on the number of tests performed per week in Germany are only available to a limited extent, we will first look at the example of Italy, where the Ministry of Health publishes the official number of tests every day. The following figure shows the number of tests performed and the number of reported new infections from March 1 to April 20):

 

 Abbildung_4.png   530.15KB   0 downloads

 

As you can see, the number of tests was continuously increased. Relatively parallel to this, the number of reported new infections is increasing until 22 March. This means that, as in the example above, a larger part of the initial increase is due to the number of tests being increased from week to week. From then on, the number of tests continues to increase, while the number of new infections remains the same at first and eventually decreases.

So although more and more tests are being performed and more and more new infections are found, the number of new infections found remains the same and even decreases. Thus, the increase in the number of tests hides the fact that in reality the new infections decrease earlier and more strongly than is the case according to the reported new infections.

 

 

Percentage of new infections

If one now checks the observed numbers of new infections by the test number increase by dividing the number of new infections per day by the number of tests per day, the Y-axis corresponds to a growth factor. This expresses how much the number of new infections has increased compared to the value on the first day. So a value of 2 means a doubling, 4 a quadrupling, and so on.1

 

 Abbildung_5.png   609.13KB   0 downloads

 

As we can see, the reported figures on new infections dramatically overestimate the true increase. The reported figures give the impression that the number of new infections has increased up to 27 times, when in reality the number of new infections has only increased by a maximum of 4 times. Furthermore, it can be seen that the number of new infections per day actually starts to decrease earlier than the reported numbers of new infections seem to indicate.

The same can be done for Germany. In an article from the Robert Koch Institute, the total number of tests per week for calendar weeks 11-15, determined by a laboratory query, and the determined test capacity per day in calendar weeks 10-15 are mentioned. Since the number of laboratories participating in the laboratory survey varies, the figures are only conditionally reliable. However, both figures show that the number of tests has also increased extremely in Germany over time.

For a first rough estimate of the test number-related overestimation of the true course of the new infections, the average test capacity per day can be taken, for example, as an estimate of the increase in the number of tests over the weeks.2 Then the following picture emerges for the week-byweek increase in new infections in calendar weeks 10-15 (the Y-axis corresponds again to a growth factor, the date here corresponds to the reporting date):

 

 Abbildung_6.png   332.26KB   0 downloads

 

As can be seen, in Germany the reported figures on new infections overestimate the real increase even more dramatically than in Italy, due to the much stronger increase in the number of tests over the weeks. The reported figures give the impression that the number of new infections increased 41-fold from calendar week 10 (from 2 March) to 14 (until 5 April). In reality, however, the number of new infections has only increased up to calendar week 12 (until 22 March) and has only increased by a factor of 2.8. Since then it has been declining. A decrease in new infections is therefore in reality two weeks earlier than the reported figures for new infections seem to indicate.

 

Edited by pamojja, 27 April 2020 - 09:19 PM.

[pamojja]

Another factor: other pathogens

One might now think that a correction of the reported figures by the number of tests is sufficient to determine the true course of the new infections. But there is actually another hidden factor that artificially distorts the number of new infections upwards.

This can be illustrated by the following consideration: In order to determine the true course of the new infections, the number of observed new infections was divided by the number of tests - in other words, one practically determines the percentage of positive coronavirus diagnoses in relation to the number of tests performed.

A closer look at the percentage of positive coronavirus diagnoses reveals something very interesting: In relation to the number of tests, only very few of the tested persons receive a positive coronavirus diagnosis. According to the Robert Koch Institute, the percentage of positive test results in the tests performed up to calendar week 15 was only 7.7% on average.

Now one must realize the following: Since in Germany, according to the guidelines of the Robert Koch Institute, only people with acute respiratory diseases are tested, people with a negative coronavirus test result are naturally not healthy. Instead, their disease symptoms are caused by other pathogens (except in cases where the coronavirus test incorrectly gives a negative test result).

Thus, the percentage of observed positive coronavirus diagnoses depends on another factor: the number of people suffering from acute respiratory diseases caused by other pathogens. If this number decreases, the percentage of positive coronavirus diagnoses received automatically increases, with the consequence that the impression of increasing new coronavirus infections is erroneously created. The following figure illustrates the mechanism.

 

 Abbildung_7.png   478.51KB   0 downloads

 

In Germany, the number of people suffering from acute respiratory diseases caused by other pathogens can be estimated on the basis of weekly visits to doctors for acute respiratory diseases. This number is determined in the weekly influenza reports of the RKI based on hundreds of doctor's visits.

From calendar week 10 to 12, the number of visits to the doctor per week increased slightly from 1.6 million to 1.8 million. Afterwards, this number starts to decrease rapidly, first to 1.1 million (week 13), then to 700,000 (week 14) and finally to 370,000 (week 15). The number of people suffering from acute respiratory diseases caused by other pathogens thus decreases relatively strongly - and this erroneously creates the impression of increasing new coronavirus infections.

With the help of the weekly number of visits to the doctor due to acute respiratory diseases, the course of the observed new coronavirus infections can now be corrected for this falsifying influence. Since persons with acute respiratory diseases correspond exactly to the persons who are tested for coronavirus according to the guidelines in Germany, the weekly percentage of positive coronavirus diagnoses can be used to estimate the weekly total number of patients with acute respiratory diseases who are infected by coronavirus or other pathogens. The following figure shows the correspondingly estimated true growth curve for new coronavirus infections compared to the growth curve of reported new infections (the date here again corresponds to the reporting date).

 

 Abbildung_8.png   320.69KB   0 downloads

 

The decreasing number of people suffering from acute respiratory symptoms caused by other pathogens thus in fact hides the fact that the decrease in new infections is in reality even more pronounced. To sum up, it can thus be said so far that the observed increase in reported new infections dramatically overestimates the true increase in new infections, and also hides the fact that the number of new infections is already falling in about two weeks earlier and more sharply than is the case according to the reported figures.

 

Edited by pamojja, 27 April 2020 - 09:26 PM.

[gamesguru]

This article seems to be translated in broken English.

 

It also relies on the assumption that the "ratio of tests administered to positive results" is really meaningful—which is far from certain.  For example, as testing capacity surges people with less symptoms may begin to be tested at leisure.  And this would clearly bring down the number of positives.  On the contrary, when testing supplies are very low they may be limited to the rich who are generally well.. then when an initial boost in supply comes, we start reporting tons more positives because the sick masses are now being test.

 

The trouble then with this one-size-fits-all assumption is it doesn't actually mean anything.  In the former case you would assume the infection is disappearing, while in the latter case you would assume it is exploding.  Neither case is it true.

[pamojja]

 

The actual date of the decrease in deaths

One might wonder how this fits in with the current reports that the number of deaths allegedly rose again so sharply in the last week (13-17 April). RKI Vice President Lars Schaade said at a press conference on 21 April: "Last week we also saw the largest increase in the number of deaths to date. On 16.4.2020 there were 315 newly reported deaths in one day."

Here there is a first very striking aspect: A closer look shows that the figures on the course of the deaths are also presented and interpreted very misleadingly. In order to be able to see the problem behind this, let us first look at the usual presentation of the course of the deaths. In the following figure, for example, we can see the course of deaths in Germany according to the data of the European Center for Disease Prevention and Control (ECDC, as of 23.4.) (Note 10:00 p.m.: According to the description of the data from the data download source NPGEO Corona Hub 2020, the reporting date corresponds to the date when the case became known to the public health department. However, it remains unclear whether this corresponds to the actual date of death. This aspect is currently being clarified and the article will be updated accordingly shortly).

 

 Abbildung_9.png   425.16KB   0 downloads

 

In view of such graphs, one might indeed think that the number of deaths has increased by 16 April and will start to decrease by 17 April at the earliest. But a closer look reveals that this is a very misleading representation. What the graph shows is the number of deaths newly reported to the RKI on a given day.

Now, however, there is a very large reporting delay of 14 days or more for deaths. In other words, hardly any of the deaths newly reported on a particular day occurred on that day, but in reality at some point in the 14 days before. For example, according to the publicly available data set of the RKI, only 5% of those who died on 16 April - the previous maximum - died on that day, and only 25% died in the week before.

If one now draws a graph of the course of the deaths based on the actual date of death (according to the report to the public health department), a completely different course is shown (status 23.4.). (Note 22:00: According to the description of the data of the data download source NPGEO Corona Hub 2020, the reporting date corresponds to the date when the case became known to the public health department. However, it remains unclear whether this corresponds to the actual date of death. This aspect is currently being clarified and the article will be updated accordingly in the near future if necessary).

 

 Abbildung_10.png   555.68KB   0 downloads

 

As can be seen very impressively, the number of coronavirus deaths has actually been falling since the beginning of April. One must still take into account the delay in reporting. But this graph looks unchanged for more than a week, and even if you add the estimated reporting delay of the next 20 days (the yellow bars) to the respective days, this does not change. So the usual way of presenting the deaths in terms of the newly added deaths per day hides the fact that the number of deaths has actually been falling for three weeks.

The timing of the decrease in deaths is in itself well consistent with the above estimate of the true course of the new infections. According to studies, the average time between the time of infection and the time of death is about 24 days, which fits relatively well with the above estimates.

 

Problem with death rates

But there is a striking second point. In order to see this, one must look at the curves for new infections and deaths in comparison. In order to be able to compare the two curves well, the following graph shows the number of new infections over the current mortality rate brought to the level of the number of deaths. In practice, this means that we draw the curve of new infections in such a way that it shows how many people should die from them at some point in the future according to the mortality rate. Then the following picture appears (note 22:00: according to the description of the data of the data download source NPGEO Corona Hub 2020, the reporting date corresponds to the date when the case became known to the public health department. However, it remains unclear whether this corresponds to the actual date of death. This aspect is currently being clarified and the article will be updated accordingly in the near future if necessary):

 

 Abbildung_11.png   401.04KB   0 downloads

 

Here a very peculiar pattern emerges: the number of new infections and the number of deaths increases and decreases almost exactly in parallel. From a biological perspective, this is actually impossible, because studies show that there are 18 days between the development of the first symptoms of the disease and the time of death. So even if one assumes that a test is performed eight days after the symptoms have developed, the death curve should still lag 10 days behind the curve for new infections.

The only reasonable explanation for the absence of a time lag between new infections and deaths may be that many of the deceased were tested for coronavirus shortly before or after death. This implies two possibilities:

Option A: The deceased did die of coronavirus. This would mean that exactly the same growth curve would have been found in the new infections 10 days before, if they had been tested accordingly. In other words: If the coronavirus test had been developed 10 days earlier and then the testing had been started 10 days earlier, a comparable increase in new infections would have been found - dramatically distorted upwards by the number of tests.

Possibility B: The deceased persons caught the coronavirus only shortly before death and in reality did not die of it at all. This would mean, however, that both curves actually show the same thing: A dramatic increase in new infections, distorted by the number of tests.

 

[pamojja]

 

On overburdening the health care system

One important final point: One can still ask how the finding that the number of new coronavirus infections has only risen slightly in reality fits in with reports that intensive care units are overcrowded or with pictures from Italy or New York in which coffins of deceased people had to be stacked in churches or refrigerated warehouses. This has even led experts to believe that such scenarios could occur in many countries if no countermeasures are taken against the transmission of the coronavirus.

The situation in Germany can be considered here first: According to the current situation report of the Robert Koch Institute (as of 22.4.), 12,686 (40%) of the total 31,885 available intensive care beds at the 1,230 clinic locations covered are currently free. This means that there is currently, at least on average, no overload of intensive care units in Germany.

With regard to the number of deaths, there is no indication that Germany currently has a particularly high number of deaths. A comparison with earlier years is interesting here. A few days ago, the Federal Statistical Office published figures on the total number of deaths in Germany up to at least mid-March compared to previous years. Here the following picture emerges:

 

 Abbildung_12.png   261.48KB   0 downloads

 

In Germany, therefore, the maintenance of measures cannot be justified by the threat of overburdening the health system. For example, similar reports appeared in the media in March 2018. In the newspaper Die Welt, for example, an article about a hospital dated March 1, 2018, stated that patients were lying in the corridors of both the emergency room and the wards. Several were given oxygen to help them breathe better. People over 60 years of age with previous illnesses were particularly affected. Due to the large number of flu patients, planned operations had to be postponed.

In contrast to Germany, the total number of deaths to date this year in Italy or New York is actually significantly higher than the usual level at this time of year, and is similar to the high flu epidemics in 2017/18, for example - although it should be noted that the situation in New York is a local phenomenon in the USA. According to the National Center of Heath Statistics, the current total number of deaths to date in all other states is even below the usual level.

With regard to the images from Italy and New York, however, it is important to be aware of another often overlooked aspect: in both Italy and New York, the percentage of positive coronvirus diagnoses, even in the peak phase, was only about 20 (Italy) to 40 (New York) percent. However, the people who received a negative coronavirus test result were not healthy because of this but suffered from other pathogens. Pictures of overcrowded intensive care units or of many coffins of deceased people can therefore possibly be misleading, because a larger part of them may not have been caused by the coronavirus at all.

This is indeed confirmed by data from the USA from the National Center of Health Statistics (as of 23.4.). Of the 8,072 coronavirus deaths in New York City up to and including April 18, according to the diagnosis, only 2,911 (36%) died of pneumonia. Even if one does not count the pneumonia deaths caused by the flu virus, a total of 4,722 people died of pneumonia in New York City during the same period. Accordingly, a relatively large proportion of deaths from pneumonia in New York are not due to the coronavirus.

This is in no way intended to imply that overburdening of health care systems, as in Italy or New York, should be regarded as unproblematic. On the contrary, one of the highest goals of any society should be to help sick people as much as possible. But it is an open question to what extent this overload can really be attributed to the coronavirus or possibly other pathogens.

 

[gamesguru]

Not surprising Germany is following similar trend to China and Korea.  They have been very proactive, among the first to isolate cases and issue social distancing, before many cases or deaths were even reported.

[pamojja]

 

Conclusion

At the end of a closer look at the figures on the course of coronavirus infections, the following can be stated:

1. the reported figures on new infections dramatically overestimate the true spread of the coronavirus. The observed rapid increase in new infections is almost exclusively due to the fact that the number of tests has increased rapidly over time. Thus, at least according to the reported figures, there has never actually been an exponential spread of the coronavirus.

2) The reported figures on new infections hide the fact that the number of new infections has been declining since about the beginning of March.

3. the number of deaths has also been falling since the beginning of April, which is hidden by the misleadingly common presentation of new deaths per day. Moreover, the course of the death curve may only reflect the course of the curve of new infections, which is dramatically distorted upwards by the number of tests.

At this point we would like to return to the beginning of the article and recall that all of the drastic measures taken are justified by the fact that they are intended to prevent a rapid increase in the number of new infections. After a closer methodological examination of these figures, it is very clear that none of the measures taken can really be scientifically justified:

On the one hand, the number of new infections has never increased rapidly in reality, and on the other hand the number of new infections has been declining since about the beginning to mid-March - this was only concealed by the fact that the number of coronavirus tests has increased so much over time and that too little attention was paid to the time lag between the actual time of infection and the time of testing. In particular, none of the measures taken can explain the decline, because the first measure (cancellation of large events with more than 1,000 participants) was not taken until 9 March. Similarly, there are no signs of an overload of intensive care units or a higher number of deaths in Germany compared to previous years, so that none of the measures can be justified.

From this perspective, it seems questionable if virologists like Christian Drosten from the Charité are currently fanning fears in the media of a second wave of infection, because he assumes that the coronavirus could again spread exponentially if the measures were to be withdrawn. Such statements are possibly misleading, given that the alleged exponential increase in new infections during the alleged first wave of infection was only an artificial effect of the fact that the number of tests was increased so much.

It would appear to be one of the highest duties of any scientist to finally set these points right in public, in order to relieve people of their probably unnecessary great fears and to eliminate the extremely negative side effects of the probably unnecessary drastic interference with our fundamental rights.

 

One final remark

Finally, I would like to explain why I am going public with such a contribution. As head of a chair of psychology, we are aware of the possible extreme side effects of the drastic measures taken. I would like to illustrate this briefly with a single example:

There have been major studies on how many additional suicides were caused by the world financial crisis of 2008-2010 - which, according to experts, will be outnumbered by the impending economic crisis by worlds: In the USA, Canada and Europe alone (not counting other poorer countries that depend on our purchasing power and may also suffer accordingly), this was 10,000 additional suicides in the years 2008-2010. When you realize that behind every suicide there are many more people who are similarly burdened but do not commit suicide, it becomes clear how much suffering the measures taken can cause.

Here it is important to mention that such suicide rates in response to economic crises are not automatic and that countermeasures can be taken. But if you take the past as a heuristic, which could possibly happen in the future - because even then people tried to take countermeasures - you have to expect extremely drastic side effects. And that was just one example of the possible side effects.

 

When a scientist who is very experienced in research methods then notices that there may be problems with the scientific basis underlying the measures taken, he usually does not seek publicity. Instead, one tries to contact the relevant experts to point out these possible problems. Since the beginning of April I have sent several emails to several virologists without having received an answer to my questions until today, which is absolutely understandable considering the current workload of these people. In such a situation, however, the only way out is to go public.

The problem is that one is then often very quickly dismissed as a "conspiracy theorist", which tempts some people not to look for the public. Here, however, it's just like that: All reported analyses are based on the official figures, and every person can simply check that for himself and think things through accordingly.

One final sentence: this is in no way a question of trivialising the suffering of those affected. The ultimate goal of any society must be to help these people in the best possible way. It is a question of questioning the scenario, adopted by many, of an epidemic spread of the coronavirus with several million infected people. Because if this scenario were not to threaten in reality, many people would experience such great fears for no real reason, and measures would be taken for no real reason, the possibly dramatic negative side effects of which cannot yet be assessed.

Note: Some parts of the text have been changed, the final remark has been added.

 

 

[gamesguru]

The observed rapid increase in new infections is almost exclusively due to the fact that the number of tests has increased rapidly over time. Thus, at least according to the reported figures, there has never actually been an exponential spread of the coronavirus.

 

that's just like his opinion man.  not very well substantiated either

[gamesguru]

However, if you, as a scientist experienced in research methods and statistics, look at the scientific basis of what justifies the drastic measures, doubts arise.

... the new infections have in reality already been falling since at least 13-14 March.

 

utterly flummoxed by this dreadful level of naivete, which runs contrary to all of conventional wisdom and sensibility.  It is ar tho the authors, all, had entered a drunken haze in the cellar and had their heads dunked in cold water beforth producing these most egregious articles.

[Hebbeh]

https://nypost.com/2...by-coronavirus/

 

Top Manhattan ER doc commits suicide, shaken by coronavirus onslaught

 

The head of the emergency department at a Manhattan hospital committed suicide after spending days on the front lines of the coronavirus battle, her family said Monday.

 

“She tried to do her job, and it killed her,’’ Dr. Philip Breen told the New York Times of his physician daughter, Dr. Lorna Breen, who had been medical director of the NewYork-Presbyterian Allen Hospital amid the pandemic.

 

The battle-weary ER doctor, 49, was only the latest city healthcare worker to take her own life.

 

Two days earlier, a Bronx EMT witnessing the virus’s ruthless toll fatally shot himself with a gun belonging to his retired NYPD cop dad.

 

Tragic rookie paramedic John Mondello, 23, worked out of EMS Station 18 in The Bronx, which handles one of the biggest 911 call volumes in the city.

 

Lorna died Sunday in Charlottesville, Va., where she’d been staying with her family, the father told the Times.

 

Philip Breen said his daughter had gotten sick with the virus while on the job at one point, but then returned to work after about a week and a half of recuperating. Still, the hospital sent her home again, and her family brought her to Virginia.

 

She had no history of mental illness, he said. But when they last spoke, she told him how excruciating it was to have to continually watch contagion patients die, including some even before they could be taken from the ambulance.

 

“She was truly in the trenches on the front line,’’ Philip Breen told the Times.

 

“Make sure she’s praised as a hero,’’ he added. “She’s a casualty just as much as anyone else who has died.’’

 

Reached later by The Post, Philip Breen, his voice cracking, said he was too distraught to talk further.

 

Mental-health professionals told The Post that PTSD from the pandemic is becoming a very real crisis.

 

“The group that is most at risk are the front-line health care workers,’’ as well as the people who lost loved ones, said Stanford University Professor Debra Kaysen, head of the school’s International Society for Traumatic Stress Studies.

 

An ICU doctor who works in the city said Monday that the onslaught of virus patients can be almost too much to handle for anyone at times.

 

For a while, “it felt like we were standing under a waterfall and couldn’t get a breath for air,’’ she said. “Now it feels busy but not in a way that’s suffocating.

 

“I was in a really low place. But I feel hopeful that I’m starting to come out of it, finally.”

 

Still, “It’s just very depressing because people in the ICU aren’t really coming out of it, and I don’t think my patients are going to live,’’ the doctor added.

 

She admitted that she has mixed feelings about the people who clap outside her hospital and others to honor healthcare workers during the pandemic.

 

“The clappers make me cry whenever I hear them,’’ she said. “But also it’s weird — because none of us feel like heroes because we feel so defeated by this disease.”

 

 

    

[gamesguru]

But it is an open question to what extent this overload can really be attributed to the coronavirus or possibly other pathogens.

 

Yeah, I thought this whole thing was caused a mutated E. coli, gahhh

 

In all seriousness, I am awaiting your kind response to posts 355 and 358.

 

Until then it would seem the ideas against lock-down have no substance.  I also ask that you please stop quoting articles in full, or if you do, I kindly request you at least make sure they are less opinionated and sensationalist.

 

Also how does the article in post 361 eliminate the need for an automatic calculation thru python scripts?  If anything it just shows the disagreement among us and the necessity for a trustworthy sort of calculation, on a scale which is fast exceeding our manual capabilities.  Or, how am I wrong here?

Edited by gamesguru, 28 April 2020 - 11:16 AM.

[gamesguru]

a useless cellar goblin defiantly howls at an inquisitive, intelligent man

 

[pamojja]

 

According to this, the number of new infections per day has been decreasing since at least March 19. However, it must be remembered that the date in this graph corresponds to the time of development of the first symptoms of the disease. As already described, however, there are another 5-6 days between the time of infection - the time of the real new infection - and the time of the formation of symptoms. The above curve must therefore be shifted back in time by another 5-6 days, and so the new infections have in reality already been falling since at least 13-14 March.

 

Translated with www.DeepL.com/Translator (free version)

 
This article seems to be translated in broken English.

 
Is that how carefully you read?
 
 

In all seriousness, I am awaiting your kind response to posts 355 and 358.

 
Since you don't even carefully read or understood this article, less had any rational responce to it. I simply can't bother with any answer to nothing than bias without balanced support of any data.

 

 

Until then it would seem the ideas against lock-down have no substance.

 
I already many times pointed out that there isn't any data, in support or against brude lockdowns. Compared to more sensible measures taken for example in Sweden.
 
First there are too few countries without brude lockdown, to get to any significant findings. Now with 152 affected countries, it is 25 days with lockdown, vs. 14 days without - it took in average from 1st covid death till plateauing. Adding now 17 days from first symptoms till death, and an other 5 for the time of infection to first symptom, we are now at 46 days in total for all countries to plateauing.
 
But again completely meaningless data for supporting either brude lockdowns (too late), or more sensible measures. Because of the self-selection of lockdowns by panic informed from worse-case models already multiple times corrected downwards.

 

But sure in a years time all the accumulted final death-toll could be in support of either.

 

 

Also how does the article in post 361 eliminate the need for an automatic calculation thru python scripts?  If anything it just shows the disagreement among us and the necessity for a trustworthy sort of calculation, on a scale which is fast exceeding our manual capabilities.  Or, how am I wrong here?

 
There are only a handful of countries driving the death-toll of this pandemic up now. More speficially: the EU with an average of at now 2.95 times the incidence of influenzia and pneumonia mortality averaged for 2017. Second the US at 1.75 times now. The rest of all the 153 by now affected countries, inlcuding all those tiny states with higher mortality too, is averaged at 0.47 times the influenzia and pneumonia mortality of 2017.I don't think that will change much in future anymore, since both players have already reached their plateau from rising deaths:
 
 Covid-19_new_cases_in_top_5_countries_and_the_world.png   108.65KB   0 downloads
Semi-log plot of daily new cases of Covid-19 (three-day average) in the world and top five countries (mean with deaths) By Chris55 - Own work, CC BY-SA 4.0, https://commons.wiki...?curid=89237234
 
More detailed for the US:
 
 Screenshot_2020-05-02 2020 coronavirus pandemic in the United States - Wikipedia.png   212.49KB   0 downloads

 

Though I'm not believing in models, the following at least comes up with a plausible graph based on real data. Though the rest of its thesis and many of its conclusions wont hold up.

 

 

When would COVID19 be over from a data science perspective and what’s the fatality rate?

Ali S. Razavian - intactive graph at site:

 Screenshot_2020-05-01 COVID-19 from a data scientist’s perspective.png   147.91KB   0 downloads

 

2 weeks later after first publishing its post, its peak already extented from 37 to 45 days.

Edited by pamojja, 02 May 2020 - 03:14 PM.

[gamesguru]

I simply can't bother with any answer

 

there isn't any data, in support or against brude lockdowns. Compared to more sensible measures taken for example in Sweden.

If there isn't any data in support or against, then surely you must be agnostic?  With all due respect, I read the articles, but I simply disagreed with certain assumptions being made.

 

There's nothing sensible about Sweden's approach.  They think they're out of the woods, just because the initial wave of fear attenuated the initial wave of virus?  Summer is also not coronavirus season, and so Sweden will likely either close schools or endure a large wave in the Fall.

 

They're at the slow precipitous to an exponential.  A locally linear curve, with ominous warning signs.

Along their quest for herd immunity, 1% of their people will die.  It's a talking point they rarely bring up

[pamojja]

The main assumption, that numbers having been already in decline before the lockdowns, is what you disagree with? - Show the data instead of insisting on your bias without.

 

And show me such a clear decline of mortality in any of equally badly affected country's curves, as available for Sweden:

 

 Screenshot_2020-05-02 2020 coronavirus pandemic in Sweden - Wikipedia.png   338.63KB   0 downloads

 

22 days in row in clear decline, after 24 days of straight increase!

 

Of course, you just will post the worldometer data again, which doesn't spread out deaths between Saturday and Monday when they happen, but when they are reported during the working-week. And again will wrongly feel assured of your believe, by confirming your bias.

[gamesguru]

Yes but Sweden is the one country that has seen a steady number of new cases.  What is suggests is Rt > 1, so the outbreak is still not under control.  Deaths will sadly follow, stay tuned.

 

Practically no one has seen a decline in new cases til a lock-down was realized.  Look at France, Italy, Spain... none of them saw a pre-lockdown decline.  The UK was one of the last to react, and is now being hit hardest!

 

What sort of reasons do you have for believing the numbers will go down on themselves?  Seems to be putting a lot of lives at risk.

[pamojja]

What sort of reasons do you have for believing the numbers will go down on themselves?

 

The declines in deaths of 152 countries now. While you base your believe on hope of confirmation of your bias in the future. And infection rates, which shows rather the rate of testing.
 

 What is suggests is Rt > 1, so the outbreak is still not under control.  Deaths will sadly follow, stay tuned.

 

Your still concerned with the infection rate impossible to know for months to come?

 

Singapore had 17,548 tested positive, only 17 deaths within 36 day.

 

This disease is too unpredictable from country to country, even region to region, to even start speaking of anything like a coherent reproduction number. Unless one wants to give one's fortune telling based on one's believe seemingly more weight. Of the 153 now affected countries there are 60 countries now with equal or below 17 deaths after in average 35 days of duration (of deaths).

 

It just never took off in so many places. Completely unpredictable.

Edited by pamojja, 02 May 2020 - 07:25 PM.

[gamesguru]

The declines in deaths of 152 countries now. While you base your believe on hope of confirmation of your bias in the future. And infection rates, which shows rather the rate of testing.
 

 

Your still concerned with the infection rate impossible to know for months to come?

 

Singapore had 17,548 tested positive, only 17 deaths within 36 day.

 

The UK and Sweden have not shown declines.  Neither has the US.  And you know something about the majority of countries that have seen declines?  They all locked down.

 

Infection rates show something, which contrasted against the % of positive tests.  If you ramp testing up by ten times and the same percent come back positive, it points to increased infections.  Because at first tests are limited to those with obvious symptoms, and so more positive % is expected.

 

Many of the positives in Singapore have come from refugee camps in the last few weeks.  It takes time for fatalities to catch up.  If the refugee camps are exclusively younger people, it may stay well under 1%.

[pamojja]

Do you really have nothing to add than blind believes not based on any real data?

 

Practically no one has seen a decline in new cases til a lock-down was realized.  Look at France, Italy, Spain... none of them saw a pre-lockdown decline.  The UK was one of the last to react, and is now being hit hardest!

 

You again and again look only at worst affected countries, for confirming your bias.

 

 

 

The UK and Sweden have not shown declines.  Neither has the US.  And you know something about the majority of countries that have seen declines?  They all locked down.

 

There are now 152 countries affected by covid deaths. 34 countries nothing about a strict lockdown is published (strict, as in closing all non-essential shops and fines for not complying with self-isolation of the whole population).

 

Other than you believe, one of course can speak of declines from the day the highest daily mortality was reached, after which it will linger on a somewhat lower plateau, or already fastly declines like Sweden. 37 countries reached their highest point during the last 10 days only, and therefore some of them certainly will increase again. April 11 was the average day of peaks for all 152 countries. UK and Sweden included too.

 

Which however, wont change the total aggregate data of all 153 countries much again. At the most a few days. Which at the moment is 25 days (from first death till plateau) + 22 days (avg. from infection to death) = 47 days duration with brude lockdown.

 

And 16 days + 22 days = 38 days without brude lockdown. Both together averaging 46 days.

 

So far the data. Not biased believes.
 

Edited by pamojja, 02 May 2020 - 08:26 PM.

[gamesguru]

Youre the one with biased beliefs.  You scoffed at the idea of a Python program to show you wrong.  Now you're calling "hundreds" of countries into question, when most of them have only been recently affected, implemented early lockdowns and didn't experience much of a first wave, e.g. they're not really true outbreak comparisons.  Let's limit ourselves to the few dozen countries that have been hit hardest.

 

I don't want to make any sort of conclusions on a country that has reported only 40 cases and 6 deaths so far.  They are in the early stages.

 

Among those, you will see the ones not implementing lockdowns are the ones with more positives.  All nations are taking more testes, but only the ones with less lockdown measures are reporting steady number of new daily cases daily.

 

Just see how long some of these countries in the early stage can maintain no lockdown, which besides Nicaragua, I'm not aware of anyone like Sweden.

[Danail Bulgaria]

The countries, that were clever enough to close borders and implemented "lockdowns" are suffering the least. Thats what the time showed, and I think it is unbreakable fact.

[smithx]

Very good interactive simulations showing the effect of different scenarios including shelter in place and not doing that at all:

 

https://ncase.me/covid-19/

 

 

 

 

[pamojja]

Just see how long some of these countries in the early stage can maintain no lockdown, which besides Nicaragua, I'm not aware of anyone like Sweden.

 

As already pointed out with links to you, there is an incomplete wikipedia page. And in the case of countries missing, one finds a timeline at their individual pages. If and what kind of meassures have been taken. If you don't investigate you remain ignorant.

 

The countries, that were clever enough to close borders and implemented "lockdowns" are suffering the least. Thats what the time showed, and I think it is unbreakable fact.

 

Not true at all with no data to support your 'unbreakable fact'. Just your blind faith. Most of the countries impossing brude lockdowns, due to panic from becoming much worse, all suffer much. Certainly because of lockdowns too late implemented. But there are more countries with lockdown not suffering as much also implemented too late. As there are countries without, the majority not suffering at all. With the exception of Sweden, but not as bad as the worse with lockdown.

 

Very good interactive simulations showing the effect of different scenarios including shelter in place and not doing that at all:

 

https://ncase.me/covid-19/

 

If one still believes in models, all till now shown to have been plainly wrong. Based on assumptions purely speculative, like an at this point impossible to know R0.
 

[gamesguru]

impossible to know R0.

It's called Rt, the transmission rate, because it's a function of time.  And it can be calculated instantaneously.  It's basic calculus

[Danail Bulgaria]

...

Not true at all with no data to support your 'unbreakable fact'. Just your blind faith. Most of the countries impossing brude lockdowns, due to panic from becoming much worse, all suffer much. Certainly because of lockdowns too late implemented. But there are more countries with lockdown not suffering as much also implemented too late. As there are countries without, the majority not suffering at all. With the exception of Sweden, but not as bad as the worse with lockdown.

 

.....

 

There is the data:

 

https://www.worldome...gn=homeAdUOA?Si
 

And not only that source. Mnay other sources give the most deaths in USA, Italy, Spain, etc.

 

And Russia, Bulgaria and others, that introduce "lockdowns" are less affected.

* in times less affected

[gamesguru]

But there are more countries with lockdown not suffering as much also implemented too late. As there are countries without, the majority not suffering at all. With the exception of Sweden, but not as bad as the worse with lockdown.

 

That's because they didn't implement very good lockdowns at first, it barely affect the Rt value.  Italy had to tighten it several times to bring epidemic under control.  They are afraid to lift it now.

 

As for countries without a lockdown, the only one hit early by the virus was Sweden.  You can't compared to Nicaragua that is very, very early in its outbreak.

[pamojja]

I don't want to make any sort of conclusions on a country that has reported only 40 cases and 6 deaths so far.  They are in the early stages.

 

Not really. They are all evenly mixed it at diverent time length. The average first covid death of all 152 countries was March 18h. The average lockdown implemented on March 21st. The average peak at April 11. Now there are just 15 countries left not already 1 month in. Those 15 too already at their average 22nd day. No additional countries already for 2 weeks. Just now Tajikistan and Capo Verde joined, adding up to 154 countries.

[pamojja]

It's called Rt, the transmission rate, because it's a function of time.  And it can be calculated instantaneously.  It's basic calculus

 

Not if the total number of infected isn't known. As it isn't with covid.
 

[pamojja]

There is the data:

 

https://www.worldome...gn=homeAdUOA?Si

 

Which doesn't supports your claim. Have a closer look by not myopically only looking at worst countries all with brude lockdowns.