Jump to content

COVID-19/Dan Polansky/2020-Q3Q4

Add links
From Wikiversity

Contribution of lockdowns to deaths

[edit | edit source]

There is the following interesting article:

It argues that it was the lockdowns that caused the spikes of deaths observed in excess death charts.

An objection would be: well then, why didn't the lockdown in Czechia cause a spike in all-cause deaths?

A related objection is against the chart from EuroMOMO that shows five graphs (Italy, Spain, Sweden, UK England, and UK Wales) but does not show other lockdowners such as France and Belgium: why? How would the graphs look for them? Would they suggest a time correlation between lockdowns and all-cause death spikes? What were the selection criteria that led to the choice of these 5 graphs, to the exclusion of other EuroMOMO charts?

Another interesting claim the article makes is that "COVID-19 has been circulating for at least a year", in part based on tests of sewage samples. This notion is a key input into the central claim of the article: if the virus was already circulating around the world during second half of 2019, and if it was the virus that was so very bad, it would have caused significant spikes back in 2019.

The article criticizes various UK policies during the lockdown in a relevant way, and is a worthy read for this alone. For instance: 'At the same time the government and the NHS were pressurising vulnerable older people to sign "do not attempt resuscitation" (DNAR) notices' and 'Professor Carl Heneghan observed: In order to free up space in hospitals, older patients were discharged into care homes without even being tested for the virus. [...]'

Caution: The above is not a peer-reviewed scientific article and an independent verification would be required. The ideas are intriguing, but I remain skeptical, that is, I don't know. --Dan Polansky (discuss • contribs) 11:24, 21 July 2020 (UTC)

Another relevant link:

--Dan Polansky (discuss • contribs) 09:32, 25 August 2020 (UTC)

An anecdote:

--Dan Polansky (discuss • contribs) 14:06, 22 September 2020 (UTC)

Quality of national covid reporting

[edit | edit source]

One thing that repeatedly strikes my attention is how good Swedish national covid reporting is. I mean the weekly reports by FOHM, the last of which is this.

The following desirable characteristics can be observed, some of which were introduced during the pandemic:

  • The reports are published weekly as immutable objects. Therefore, we can see how the state of knowledge and data developed over time.
  • Each weekly report is a single PDF.
    • It can be viewed easily on old devices without disruption; some of the "modern web" reports work poorly on older devices.
    • The reader does not have to wait until the "modern" and "brave new" Javascript code finishes fetching data and painting images or whatever it is doing.
    • No mouseover and clicks are required, only scrolling, possibly with help of search function.
  • Test positivity ratio chart is provided. Very important.
  • Weekly all-cause death chart in the EuroMOMO style is provided. Very important.
  • Daily confirmed case chart distinguishes mild cases from severe cases.
  • Daily confirmed case charts are provided per region, multiple charts fitting a single page, requiring less scrolling. These are also split between mild cases and severe cases.
  • New cases per last week per 100 000 pop are provided for multiple weeks in colored maps showing regional values.
  • ICU-hospitalization chart is provided.

Some of these characteristics apply to Czech reporting[1] as well: 1) Test positivity ratio is provided; 2) new cases in last 7 days per 100 000 people are provided in a colored map, on district level (good granularity); 3) hospitalization and icu-hospitalization chart is provided.

--Dan Polansky (discuss • contribs) 12:10, 8 August 2020 (UTC)

Efficacy of closing schools

[edit | edit source]

Impact of closing schools on spread of infection can be investigated. Multiple European countries first closed schools and then reopened them. Sweden closed some levels of schools but not all of them. All-cause death charts for 0-14y for 2020[2] show no elevated mortality for EuroMOMO-covered European countries; so much for impact on the children themselves. Still, in principle, there is a potential risk of spread from children to their parents and great parents, and that risk needs to be investigated. This little post just captures a single link at this points.

Links:

--Dan Polansky (discuss • contribs) 19:03, 13 August 2020 (UTC)

Summer spread and viral load

[edit | edit source]

What follows is very speculative and uncertain.

There could be smaller covid lethality (IFR) during summer spread. If so, it could have in principle the following possible explanations, and maybe others:

  • 1) The lower spread rate in humid summer compared to dry winter would lead not only to slower spread but also to lower viral loads per infected person. And the size of viral load would impact the lethality. The slower spread would be suggested by the following quote from another section on this page: "In fact, Shaman et al. showed that R0 must be understood to seasonally vary between humid-summer values of just larger than “1” and dry-winter values typically as large as “4” (for example, see their Table 2)."
  • 2) Human immunity could be better is summer; this was suggested by Michael Levitt as a research question.

The above would need to be properly empirically verified.

If the above is true (an if!), excessive slowing of the infection during summer would miss on the opportunity to build collective immunity (also known in the Anglo-Saxon world as herd immunity), and push the infections to a season with higher lethality (IFR), i.e. fall and winter. Again, it would have to be shown that the virus has higher lethality when acquired in dry fall and winter than in humid summer. On the other hand, one has to watch for hospital overload, and if hospitals are being stretched, accelerating infections only because it is summer is bad.

One piece of evidence that hints at the above is the current development in Israel, where cases soared, while deaths increased only moderately, as per e.g. Worldometers[3], and the cumulative deaths per million pop remain low. For Israel, this seems unlikely to be a consequence of the law of diminishing returns of the infection since, unlike e.g. Spain, Israel did not see high initial deaths. But there is a catch: the infection positivity rate is similar to what it was during the first Israel peak per OWID[4] and there was increase in testing per OWID[5], and therefore, the increased testing could explain why confirmed cases have risen much more than deaths. This shows again the importance of looking at test positivity rate instead of looking at confirmed cases alone. Beware that this little piece of evidence is on its own very inconclusive, and the test positivity rate seems to weaken it considerably. --Dan Polansky (discuss • contribs) 12:23, 15 August 2020 (UTC)

Seasonality of coronaviruses is suggested by this:

The above article has a chart suggesting seasonality of test positivity rate. --Dan Polansky (discuss • contribs) 12:45, 15 August 2020 (UTC)

Weakness of the Neil Ferguson study

[edit | edit source]

A weakness of the Neil Ferguson study that was reported to influence UK government decisions are described here:

A weakness of the way estimates of IFR and from there resulting deaths were made are pointed out. And it is reported that Ferguson acknowledges that he received grants from the Vaccine Alliance (a pharma company which is now making Covid-19 vaccines), and the Bill and Melinda Gates Foundation. The grants from Gates are old news, but the unreliable way used by the study to calculate the denominator to be in IFR calculation is interesting. To quote: "[...] remember that it all comes back to the belief that finding six people with Covid on six flights [from Wuhan] was a good way to estimate how many people had the disease."

A relevant article:

--Dan Polansky (discuss • contribs) 17:25, 15 August 2020 (UTC)

Herd immunity threshold

[edit | edit source]

For the question of what percentage of population needs to gain immunity so that significant herd immunity sets in, there is an article:

The above is a preprint => caution required. --Dan Polansky (discuss • contribs) 07:45, 17 August 2020 (UTC)

Lawsuits

[edit | edit source]

In Canada:

In Czechia

--Dan Polansky (discuss • contribs) 10:09, 19 August 2020 (UTC)

Vaccine safety

[edit | edit source]

Some information on vaccine safety in general is provided by Robert Kennedy Jr. in the following debate:

  • Heated Vaccine Debate - Kennedy Jr. vs Dershowitz, Jul 23, 2020, youtube.com
    "The problem is, Anthony Fauci put 500 million dollars of our dollars into that vaccine; he owns half the patent [...] so you have a corrupt system and now they've got a vaccine that is too big to fail", Robert Kennedy Jr. speaking, about 37:35

More research and links required.

Other links:

Exemption from liability:

--Dan Polansky (discuss • contribs) 20:05, 23 August 2020 (UTC)

New sources:

  • Safety Review Underway of AstraZeneca’s Vaccine Trial, Sep 10, nytimes.com
    "A source familiar with the event, who spoke on the condition of anonymity, said in an interview on Tuesday that the afflicted volunteer had experienced symptoms consistent with a condition called transverse myelitis, or inflammation of the spinal cord."
    "Transverse myelitis is relatively rare, sparking symptoms in roughly 1,400 people each year in the United States, according to the National Institutes of Health. The condition can result in pain, muscle weakness, paralysis or bladder problems. Its root cause is often mysterious, although doctors believe the syndrome generally results when inflammatory responses in the body go awry, sometimes as a reaction to an ongoing or past infection, said Dr. Felicia Chow, a neurologist at the University of California, San Francisco."
  • Transverse myelitis and vaccines: a multi-analysis , 2009 Nov, pubmed.ncbi.nlm.nih.gov
    "We have disclosed 37 reported cases of transverse myelitis associated with different vaccines including those against hepatitis B virus, measles-mumps-rubella, diphtheria-tetanus-pertussis and others, given to infants, children and adults. In most of these reported cases the temporal association was between several days and 3 months, although a longer time frame of up to several years was also suggested."

--Dan Polansky (discuss • contribs) 07:28, 12 September 2020 (UTC)

Anthony Fauci

[edit | edit source]

Here are some links and direct quotes from Anthony Fauci serving as supporting evidence for whatever conclusions one might want to draw:

--Dan Polansky (discuss • contribs) 11:15, 24 August 2020 (UTC)

More:

--Dan Polansky (discuss • contribs) 10:48, 24 September 2020 (UTC)

Chinese pro-lockdown propaganda and operation

[edit | edit source]

Relevant links:

As always, use the above with your critical faculties enabled. --Dan Polansky (discuss • contribs) 10:01, 25 August 2020 (UTC)

Ethics of interventions that kill some while saving others

[edit | edit source]

One can ask the following ethical questions:

  • 1) Let us have three people who are going to die unless supplied by organs for transplant, each person needing a different organ, and let there be lack of organs in the organ banks. Is the state allowed to randomly select a person to be killed against their will and used for organs to save the three people?
  • 2) Let us have a state-level intervention that the state can make that is going to kill 100 people while saving 300 people, on expected statistical everages. Let the state not know the identities of these people, but think to know the approximate quantities. Is the state allowed to proceeed with the intervention?
  • 3) If the above 2) is tentatively accepted for the sake of discussion, what is the right unit of death accounting? Is the state allowed to count all deaths as equal or should the state account in terms of lost life-years? That is to say, if the state thinks to know that the intervention is going to kill about 100 people at the age of about 35 while save about 300 people at the age of about 80, is the state still allowed to proceed with the intervention?

The above questions are relevant both to mandatory vaccination (not to voluntary one) and the lockdowns. A large state knows that the lockdown is very likely to lead to at least one additional suicide and one additional heart attack death resulting from missed doctor visit; is the state allowed to proceed with the extraordinary intervention of lockdown in spite of that? Of course, the quantitative relations would be different from those in the examples above.

There seems to be an ethical assymmetry between intervention and non-intervention. However, per Rachels 1975 mentioned by SEP (see FR below), actively drowning someone would be no worse than letting someone drown; this seems very dubious to me, subjectively speaking. SEP has a further discussion.

Further reading:

  • Doing vs. Allowing Harm in Stanford Encyclopedia of Philosophy (SEP), plato.stanford.edu - has extensive bibliography serving itself as further reading

--Dan Polansky (discuss • contribs) 07:55, 27 August 2020 (UTC)

More reading:

--Dan Polansky (discuss • contribs) 14:11, 22 September 2020 (UTC)

Using science as a guide

[edit | edit source]

Various politicians and media talk of being guided by scientists. That is fundamentally unscientific: the medium of science are not scientists but rather peer-reviewed scientific articles, or if under time pressure, at least preprints. The mass of evidence and arguments present in the articles is what constitutes science, sometimes bad science.

There is another consideration: science is concerned with descriptive facts about the world, and action to be taken or to be refrained from never depends on descriptive facts alone; action needs values, risk-aversion vs. risk-friendliness, and more, and these cannot be supplied by scientific research. Science is in fact a relevant input, and very important too, but it alone does not suffice. --Dan Polansky (discuss • contribs) 13:27, 27 August 2020 (UTC)

Sources supporting using all-cause deaths and excess deaths to learn about the covid

[edit | edit source]

The following sources support the notion that all-caude deaths and excess deaths calculated from them are useful in studying death impact of the covid:

  • HMD[6], mortality.org:
    "In response to the COVID-19 pandemic, the HMD team decided to establish a new data resource: Short-term Mortality Fluctuations (STMF) data series. Objective and internationally comparable data are crucial to determine the effectiveness of different strategies used to address epidemics. Weekly death counts provide the most objective and comparable way of assessing the scale of short-term mortality elevations across countries and time."
    To reinforce the credentialism card: HMD is a collaboration between organizations from Germany, the U.S., and France: Max Planck Institute for Demographic Research; University of California, Berkeley and INED, Paris.
  • Excess Deaths Associated with COVID-19, www.cdc.gov - has blue-bar all-cause death graphs for the whole U.S. and also for U.S. states via "Select a jurisdiction" and also specifically for New York City, assuming dashboard Weekly Excess Deaths was selected
    "Estimates of excess deaths can provide information about the burden of mortality potentially related to the COVID-19 pandemic, including deaths that are directly or indirectly attributed to COVID-19."
  • Comparisons of all-cause mortality between European countries and regions: January to June 2020, ons.gov.uk - has weekly age-standardised mortality rates in 2020, which includes Montenegro, Serbia, Wales and Northern Ireland, and shows comparison between the selected country and England
    "Analysis of all-cause mortality allows us to examine the impact of the coronavirus pandemic not only from deaths involving COVID-19, but also excess deaths that have occurred as a result of the wider impacts of the coronavirus on healthcare systems and society. Given the widely differing practices between countries in recording and reporting deaths relating to COVID-19 it is not possible at this time to conduct accurate international comparisons of deaths involving COVID-19 specifically."
  • Sweden's FOHM[7]
    It publishes all-cause death charts in its weekly covid reports, e.g. report for week 34[8], Figur 7. Antal dödsfall i Sverige i alla åldrar samt förväntat antal dödsfall per vecka 2016-2020, till och med vecka 32.
  • EuroMOMO[9]
    A support is weak in that I did not find a statement on use of all-cause deaths for covid analysis. There is at least this: "During the current COVID-19 pandemic, information on its mortality impact is of major concern."
  • Deaths in New York City Are More Than Double the Usual Total, April 10, 2020, nytimes.com
    A support is weak since this is merely mainstream media, a poor type of source for science. At least, one of the authors, Josh Katz, has some meaningful credentials: "Mr. Katz studied political science and philosophy at Drew University and earned his master's degree in statistics from N.C. State University."[10]

--Dan Polansky (discuss • contribs) 13:50, 2 September 2020 (UTC)

Tomas Pueyo

[edit | edit source]

Tomas Pueyo seems to be a major contributor to the disproportionate fear of and disproportionate response to the covid pandemic. Some links follow; I may add more in the way of analysis later:

  • Coronavirus: Why You Must Act Now, Mar 10, medium.com
  • Coronavirus special: Are we doing enough? at Channel 4 News, Mar 13, youtube.com
    A speaker opposing Pueyo: "There's two things, there's two strategies with a new virus with a new epidemic, there's two strategies. One, you can stamp out every single case in the world, every single case in the world, and then the virus, then you're free. You can stop the epidemic without achieving the herd immunity but you must catch every single case in the world. But with a mild disease, that's incredibly difficult. [...] The only other way [...] is achieving herd immunity. [...] As soon as you release the lockdown, it comes back. [...]", about 14:20
  • Tomas Pueyo, twitter.com
  • Thaddeus Michaels: Why did the whole world lose its nerve?, Aug 23, hectordrummond.com

--Dan Polansky (discuss • contribs) 11:46, 4 September 2020 (UTC)

New links:

  • 1/10 Tomas Pueyo (@tomaspueyo): Panic, Profit, and Denial. by @OBusybody, Sep 6, twitter.com
    Item 4/10 in the thread shows impact on selected media.
    Item 5/10: 'He didn’t stop there. He then wrote the “Hammer and Dance” that suggested extremely severe lockdowns were the only way to prevent 10,000,000 deaths in the U.S.' This is confirmed by the screenshot and is in Chart 3. Infections and Deaths if We Do Nothing in the US in the Hammer and the Dance article, where it says in the chart "Total Dead: >10,000,000".
    A post by @alisa252525 on Sep 6: 'Here’s the Chief Public Health Officer of Canada using the “hammer & dance” to underpin the Canadian goverment’s approach to Covid.'
  • Coronavirus: The Hammer and the Dance, Mar 19, medium.com
    In Chart 3, it indicates peak hospitalizations of 13,914,256, nearly 14 million; by contrast, covidtracking.com[11] shows peak current hospitalizations for the U.S under 60,000. One can admit that there was intervention via lockdowns in multiple U.S. states, though.

--Dan Polansky (discuss • contribs) 08:15, 12 September 2020 (UTC)

IHME projections

[edit | edit source]

Pdfs of IHME projections are being catalogued at W:Talk:COVID-19_pandemic_in_the_United_States#IHME_projections_for_the_U.S.. IHME is Institute for Health Metrics and Evaluation, founded by Bill & Melinda Gates Foundation. --Dan Polansky (discuss • contribs) 08:42, 6 September 2020 (UTC)

Report pdfs:

A page from which report pdfs can sometimes be found:

A current projection for Sweden, Sep 21:

A current projection for US, Sep 21:

--Dan Polansky (discuss • contribs) 08:15, 21 September 2020 (UTC)

Let us note some discrepancies already observable:

  • IHME_covid_briefing_USA_080520_0.pdf, Figure 14, projects for the reference scenario for the U.S. that daily deaths are going to be at about 1000 per day throughout September. By contrast, per covidtracking.com, 7d avg of daily deaths went under 1000 on Aug 22, reached about 750 on Sep 10, and is about 800 on Sep 20. The same projection's reference scenario sees daily deaths to be above 1000 per day through October and November. Neatly, their universal-mask-use scenario goes much lower, so they can always explain away the discrepancies by claiming that it was the considerable mask use that drove the deaths down, despite hard science lacking on such claim. Baring that stratagem, their projection is refuted in the downward direction, and we can hope it is going to continue to be refuted, noting the continuing smooth downward trend in U.S. hospitalizations.
  • briefing_US_091820.pdf, Figure 13 (search for Fig 13), cannot be refuted yet; still, it looks like a major overestimate: the reference scenario projects new daily deaths to exceed 2000 in December and peak at about 3500 by the end of December. The "Continued SD mandate easing" scenario is the worst-case one and looks like a massive hyperbole, reaching 8000 new deaths per day by the end of December.

--Dan Polansky (discuss • contribs) 09:25, 21 September 2020 (UTC)

Countries with 2nd case waves but almost no accompanying death waves

[edit | edit source]

Countries with 2nd case waves but almost no accompanying death waves can be seen e.g. here:

Countries shown by Ivor Cummins above include France, Germany, Ireland, Netherlands, Spain, Switzerland, and UK; US is a bit different case because of a different covid phase of its souther region. However, there seems to be no guarantee that this will work for countries that did not see significant 1st death spike.

The charts can get double checked via Worldometers and Our World in Data (OWID).

For each chart, the case wave vertical dimensions are often not directly comparable between the two waves since often, the testing would have increased dramatically between the spikes; one should check test positivity rates to see figures adjusted for testing increases, available e.g. in OWID.

Test positivity rates for multiple European countries:

A link that is not directly relevant to the thread but very interesting and mentioned by Ivor Cummins:

  • The Swine Flu Panic of 2009, 12 Mar 2010, spiegel.de
    "Swine flu kept the world in suspense for almost a year. A massive vaccination campaign was mounted to put a stop to the anticipated pandemic. But, as it turned out, it was a relatively harmless strain of the flu virus. How, and why, did the world overreact?"

--Dan Polansky (discuss • contribs) 11:33, 10 September 2020 (UTC)

Let's compare the development of confirmed covid cases to test positivity rate e.g. for the U.K.:

  • Confirmed covid cases per Worldometers[14] reached 2500 new cases per day in September, whereas in April it peaked at nearly 5000 new cases per day (7d moving avg). The chart suggests there is a major 2nd wave.
  • Test positivity rate[15] in September is under 1% and is a tiny fraction of the peak in April, which was over 30%.

We see above that the nominal 2nd wave of confirmed cases is solely due to hugely increased testing and there is in fact only tiny increase of infections (as opposed to confirmed cases). Again, one should always look at test positivity rate alongside confirmed cases. --Dan Polansky (discuss • contribs) 11:47, 10 September 2020 (UTC)

Infection fatality rate (IFR)

[edit | edit source]

Infection fatality rate is the ratio of the count of the infected who died from the infection to the count of all infected. The notion of infected is distinct from the one of confirmed case. A related notion is case fatality rate (CFR), on which I have notes at COVID-19/Dan Polansky#Case fatality rate; the value of IFR is easily an order of magnitude smaller than the nominal CFR, as follows from these notes. The IFR is fundamentally uncertain since the number of people who were ever infected during an epidemic is not known with anything like certainty.

The number of infected people can be estimated from serological studies, but that may lead to underestimates: not all infected people have to develop seroimmunity: some may develop other kind of immunity instead. And underestimates of the number of infected people lead to overestimates of IFR.

Relevant sources:

  • A systematic review and meta-analysis of published research data on COVID-19 infection-fatality rates, July 7, 2020, medrxiv.org - a preprint
    "Based on a systematic review and meta-analysis of published evidence on COVID-19 until May, 2020, the IFR of the disease across populations is 0.68% (0.53-0.82%). However, due to very high heterogeneity in the meta-analysis, it is difficult to know if this represents the true point estimate."
  • The infection fatality rate of COVID-19 in Stockholm – Technical report, 2020, folkhalsomyndigheten.se
    "Our point estimate of the infection fatality rate is 0.6%, with a 95% confidence interval of 0.4–1.1%."
    Comment: This is an individual study; the meta-analysis above covers a multitude of studies.
  • High SARS-CoV-2 Seroprevalence in Children and Adults in the Austrian Ski Resort Ischgl, August 22, 2020, medrxiv.org
    "Over the previous two months, two COVID-19-related deaths had been recorded, corresponding to an infection fatality rate (IFR) of 0.25% (95% CI 0.03-0.91)."
    Comment: This is an individual study; the meta-analysis above covers a multitude of studies, although not this one.
  • Coronavirus: Why death and mortality rates differ, Apr 2, bbc.com
    "One example is the H1N1 pandemic of 2009, known as swine flu. Early case fatality rate estimates were inflated by a factor of more than 10. Even 10 weeks into the epidemic, estimates varied widely between countries, coming in between 0.1% and 5.1%. When medics later had a chance to go through case documents and evaluate cases, the actual H1N1 case death rate was far lower, at 0.02%."

--Dan Polansky (discuss • contribs) 05:27, 12 September 2020 (UTC)

Seasonality

[edit | edit source]

For seasonality, see #Summer spread and viral load above.

Beyond that, the following is relevant:

  • Viral Issue Crucial Update Sept 8th: the Science, Logic and Data Explained! by Ivor Cummins, Sep 8, 2020, youtube.com
    At 18:55, discussion of seasonality begins. "Coronaviruses are sharply seasonal, super seasonal, like influenza." Refers to Hope-Simpson work on seasonality, according to which the peak severity months for mild northern regions differs from those for the tropical regions. The 2nd death bump in the U.S., the smaller one, is driven by southern states, and approximately matches the severity months for tropical regions by Hope-Simpson.

--Dan Polansky (discuss • contribs) 20:51, 13 September 2020 (UTC)

Let us again look at the following source that has a chart suggesting seasonality of test positivity rate:

The chart reveals the seasonality of coronaviruses is strong but not perfectly consistent between the viruses: CoV 229E is low in Jul to Dec, while Cov NL63 and Cov HKU1/OC43 reach high test positivity rates in Nov and Dec; Cov HKU1/OC43 have the highest values of all months in Dec. Quoting Figure 1: "Seasonal variation in the fraction of positive CoV tests in Stockholm, Sweden. Panel A shows test results between 2010 and 2019. Panel B shows aggregated data for all years. All CoVs show a marked decline in summer and autumn, with HKU1/OC43 peaking January–December, and NL63 and 229E peaking in February–March." --Dan Polansky (discuss • contribs) 08:41, 7 October 2020 (UTC)

Exponential analysis of US hospitalizations

[edit | edit source]

Questions:

  • For how long did the growth of current US hospitalizations appear to be exponential?

First, let's plot the hospitalizations from covidtracking.com:

Let us further plot 100 times the ratio of the value of the day to the value of the previous day, from 7-day moving average of the above hospitalizations, where the minimum for the y-axis is set to 100, and therefore, the parts where the hospitalizations decline are cut off underneath the x-axis:

We can observe that in March, the growth appeared to be sharply exponential, but then the base of the exponential growth went down rapidly, it plateaud for a few days at the beginning of April, and the went down again. In July, the base of the exponential growth was much less dramatic. Many people observing the above chart in March could have been scared by the apparent exponential growth of hospitalizations.

A note on the rectangular bump in July: this is likely caused by a one-off jump in hospitalizations, which then got reflected into the 7-day moving average as 7 higher values, as per Note A, 'Jul 10: Florida started reporting hospitalizations of people with a “primary diagnosis of COVID-19.”'. Therefore, the actual daily increase rate of current hospitalizations in July peaked at about 2.5% per day. However, the rate would probably have been higher in individual states since what we see is an aggregate.

There is a good chance that the initial March growth was caused by the growth of the number of tests; this could be corroborated if we had all-cause hospitalizations. That is to say, some of the new hospitalizations would have been not covid-caused but rather covid-positive. To support this hypothesis but not conclusively prove it, let us plot the growth ratio for 7-day moving average of daily tests from the same source:

The above chart seems consistent with the hypothesis that the March huge growth of hospitalizations was very much contributed by the growth of tests. Again, all-cause hospitalizations would provide further confirmation if available. --Dan Polansky (discuss • contribs) 12:11, 14 September 2020 (UTC)

Exponential analysis of Arizona hospitalizations

[edit | edit source]

Let us do an exercise similar to what we did in #Exponential analysis of US hospitalizations, but for Arizona. Again, the data is from covidtracking.com.

And growth ratio multiplied by 100:

Beware that Arizona is more of a tropic or subtropic climate and can therefore have slower growth, as per Hope-Simpson charts mentioned in #Seasonality. We can see that the growth was never higher than 6% per day and that an observer extrapolating the growth beyond the actual time period could get worried or even scared of bad hospital overload, which did not take place. --Dan Polansky (discuss • contribs) 12:44, 14 September 2020 (UTC)

We can plot a 5-smoothed growth ratio multiplied by 100:

To calculate the 5-smoothed ratio, we calculate the (val_n / val_(n-5)) ^ 1/5, that is, we ask what the daily ratio would have to be over the period of 5 days to yield the growth given by the ratio of the day value and the value five days ago. The peak growth rate so obtained is 5% growth. --Dan Polansky (discuss • contribs) 06:57, 19 September 2020 (UTC)

Excess deaths, their baseline

[edit | edit source]

Question:

  • What is the proper baseline against which to calculate excess deaths?

Tentative answer: A proposed baseline is the weekly minimum over selected recent years; by contrast, mortality.org charting offers the average.

What speaks against taking the average is that the years from which the average is taken include excess deaths for these years. By contrast, the minimum over these years captures the year in which the excess deaths were very low.

An alternative tentative characterization of the baseline is as the deaths that would occur under elimination of respiratory viral infections and other one-off events such as heats.

EuroMOMO seems to use a sine-shaped baseline distinct from mere average; double check required.

The proposed minimum is quite likely a crude approximation. One objection is that even the minimum year has some excess deaths above the ultimate baseline. However, the minimum determined on a weekly basis could approach the baseline quite well. More research into competent sources is required.

Links:

--Dan Polansky (discuss • contribs) 11:20, 15 September 2020 (UTC)

Role of population density in death outcomes

[edit | edit source]

Intuitively, one might think that the population density of a country would impact death outcomes, especially deaths per million. Whatever impact there is, it is not so simple.

It matters what portions of the population live in how densely populated areas. This will be made obvious by considering a country consisting of two equally large regions, one with zero population and one evenly populated. The overall population density is half the population density of the populated region, and yet, the only thing that matters for death outcomes is the population density of the populated region. This very artificial example does not exactly match real world, but drives the point home: even if there are large areas that are very sparsely populated, there can be relatively small areas that are very densely populated, and these can make relatively high portion of the overall population.

Generally, the factor that matters for a country is a mapping from population density to subpopulation size. Instead of having a complete mapping, one can start with observations of the form "X percent of the population lives in areas with population density over Y", and if well chosen, a single such observation for a country can give an interesting idea that corrects the non-differentiated impression created by the overall population density.

Considerations like those above are required when considering outcomes in e.g. U.S. and Sweden, where there are large sparsely populated areas. Alternatively, one can consider individual U.S. states as more homogeneous than the U.S. as a whole and perform analyses on the individual state level. --Dan Polansky (discuss • contribs) 12:10, 15 September 2020 (UTC)

Gompertz function

[edit | edit source]

Michael Levitt thinks some aspects of growth of the covid are well captured using Gompertz function. If so, the function would describe the totals, not the deltas; the deltas would be described using the derivative.

Let us plot an example Gompertz function using a = 100, b = -12, c = -0.22 for the formula G(x) = a * exp(b * exp(c * x)):

And let us also plot growth ratio for the above function, that is, the ratio of the value to the previous value:

We can see that the Gompertz curve above never grows exponentially; rather, the exponential base of growth is decreasing all the time.

Links:

--Dan Polansky (discuss • contribs) 20:05, 15 September 2020 (UTC)

Lord Sumption

[edit | edit source]

Over time, UK's Lord Sumption has presented compelling reasoning, whether in fact conclusive or not. His articulate speeches are worth the attention.

Links

  • Lord Sumption Speaks against Hysteria-Driven Government Coronavirus Policy. by Peter Hitchens, Mar 30, mailonsunday.co.uk
    "The real problem is that when human societies lose their freedom, it's not usually because tyrants have taken it away. It's usually because people willingly surrender their freedom in return for protection against some external threat. And the threat is usually a real threat but usually exaggerated. That's what I fear we are seeing now. The pressure on politicians has come from the public. They want action. They don't pause to ask whether the action will work. They don't ask themselves whether the cost will be worth paying. They want action anyway. And anyone who has studied history will recognise here the classic symptoms of collective hysteria."
    "Hysteria is infectious. We are working ourselves up into a lather in which we exaggerate the threat and stop asking ourselves whether the cure may be worse than the disease."
  • Lord Sumption : UK September lockdown not enforceable + Internal market bill., Sep 13, youtube.com

--Dan Polansky (discuss • contribs) 08:09, 17 September 2020 (UTC)

Excess deaths from influenza

[edit | edit source]

Flu (influenza) seems to be a major cause of excess death in fall and winter. Not all years are equally badly impacted; see e.g. COVID-19/All-cause deaths. Sources use the term ILI to refer to influenza-like illness, an illess whose clinical manifestation is like that of influenza, but whose virus is not necessarily an influenza virus.

Links:

--Dan Polansky (discuss • contribs) 10:06, 17 September 2020 (UTC)

Projections from covid19-projections.com

[edit | edit source]

covid19-projections.com offers covid projections, as the name suggests. Covered is US, US states, and EU states including Czechia. Their disclaimer says the projections are optimized for the US.

Links:

--Dan Polansky (discuss • contribs) 08:43, 21 September 2020 (UTC)

UK open letters

[edit | edit source]

Open letters appeared in the UK, and here are links:

--Dan Polansky (discuss • contribs) 08:42, 22 September 2020 (UTC)

Michael Levitt

[edit | edit source]

Michael Levitt, a Nobel laureate for chemistry and a biophysicist, has for a long time provided a counterweight to the semi-orthodoxy of covid-related publishing in mainstream media. Let's collect some resources.

Links:

--Dan Polansky (discuss • contribs) 12:48, 23 September 2020 (UTC)

Vitamin D

[edit | edit source]

Taking vitamin D and vitamin D deficiency were mentioned in the media in relation to covid. Let's collect some sources.

Links:

--Dan Polansky (discuss • contribs) 12:40, 25 September 2020 (UTC)

When did the covid start to spread

[edit | edit source]

When did the covid start to spread? Was it in December 2019 or earlier?

Links:

  • Novel coronavirus may have been spreading since August — study, June 9, 2020, dw.com
    "A study by Harvard Medical School has shown that SARS-CoV-2 may have been spreading in China in August 2019. That would be months earlier than when the outbreak is thought to have started in the central city of Wuhan."
  • French expert: Covid has exposed the limits of the human species by UnHerd, Sep 24, 2020, youtube.com - interview with Jean-Francois Toussaint
    "The first is that we have seen that this virus probably started to go around the world in the summer of 2019, getting out of China with the military games and with the touristic destination of many people getting out of China in the fall of [20]19, then it needed some special condition to make the explosion that we have observed in February in Iran, in March in Italy, and then Spain and France and everywhere", at about 6:35
  • COVIDView: A Weekly Surveillance Summary of U.S. COVID-19 Activity, cdc.gov
    Chart NSSP: Percentage of Visits for Influenza-Like Illness (ILI) and COVID-19-Like Illness (CLI) to Emergency Departments has Covid-Like Illness (CLI) going back as far as 201940: 2019, week 40

--Dan Polansky (discuss • contribs) 09:10, 26 September 2020 (UTC)

General death statistics

[edit | edit source]

According to this article the average all-cause daily death count for the usa was 7,567 in 2017. There were 22,929 deaths in the usa due to covid in September 2020 according to the ECDC, I believe, which translates into 764 daily.

Does this mean that COVID deaths accounted for about 10% of total deaths in September in the usa? Ottawahitech (discuss • contribs) 18:17, 7 October 2020 (UTC)

For comparing death toll between years in the U.S., one is best served by COVID-19/All-cause deaths/United States. There are bar charts on the year basis but also on season basis.
Thus, for weeks 1-34, you can compare e.g. 2020 with 2019 and see that for these weeks 2020 saw 2106700 deaths while 2019 saw 1863208 deaths, resulting in 2020 having seen 13% more deaths than 2019 in the U.S.
When one does the same on seasonal basis, and therefore take prev-year weeks 40-53 and the cur-year weeks 1-34, one gets that 2020 saw 10% more deaths in 2020 in the U.S.
The additional 13% or 10% deaths are very likely to be directly or indirectly caused by covid, where indirectly includes collateral damage caused by anti-covid measures and other collateral damage. --Dan Polansky (discuss • contribs) 08:49, 10 October 2020 (UTC)

Sustainable measures growing population immunity vs. hammer and the dance

[edit | edit source]

Sweden's Anders Tegnell justified the choice of the Sweden's mitigation policy as one that has to be sustainable over a longer series of months. The virus was to stay, could not be suppressed globally and we had to learn how to live with it for many months to come without causing grave collateral damage. The virus was not to be ignored, proportional measures were to be taken (e.g. temporarily closing some levels of schools), but virus suppression attempt with grave collateral damage was not an acceptable option. By taking less drastic proportional measures, the country would see population immunity ("herd immunity") build up neither too fast nor too slow, and eventually provide a spread protection layer that protects both the vulnerable and the non-vulnerable portions of the population; this may not have been emphasized by Anders Tegnell, but was likely to happen given the moderate measures.

By contrast, a Thomas Pueyo's article proposed to first drive the virus near the ground via the "hammer" of drastic measures, lockdowns, and then keep the spread rate very low using the "dance" of tracking, tracing and isolating the positive individuals, banning large gatherings and other measures.

On a superficial analysis, the hammer seemed to have the desired effect in some countries, e.g. in Germany, in bringing the spread rate near the ground. However, how many countries can successfully implement the "dance" part, that is, tracing and isolation and other very moderate measures for many months, until a vaccine arrives? Even in Germany, as of Oct 10, new daily cases are slowly rising, slow but steady. What happens in December, January, and February, drier months for which some scientific reports indicate high seasonal rate of spread of coronaviruses and flu viruses? Can Germany with its culture of precision and strict adherence to rules keep the spread rate low through these dry months? Does Germany intend to interleave the dance part with more lockdowns, causing more collateral damage?

Without a hope for a vaccine, for how long would one be expected to maintain the dance part and what would one be waiting for? Would one be waiting for better cures?

In 2009 in Sweden, vaccine against pandemic 2009 H1N1 influenza, Pandemrix, was widely administered, resulting in a certain increase of narcolepsy cases in children. This is a reminder that vaccines are not guaranteed to be 100% without harm.

See also COVID-19/Dan Polansky#Sustainable mitigation, featuring links to sources deliberating on sustainability of measures.

Links:

--Dan Polansky (discuss • contribs) 14:01, 10 October 2020 (UTC)

Retrieved from "https://en.wikiversity.org/w/index.php?title=COVID-19/Dan_Polansky/2020-Q3Q4&oldid=2656029"