Regular WUWT contributor Willis Eschenbach always goes to data when questions and issues arise, he has been plotting the official death rate data from the Coronavirus almost daily, and will continue to do so. I’ve dedicated a permanent WUWT page to this. We will continue to add to this page as needed and as Willis makes updates.
Note that it is now a menu item in the left most section of the WUWT Menu bar, right under the header image.
Friday’s graph:
See the full page of graphs here: https://wattsupwiththat.com/daily-coronavirus-covid-19-data-graph-page/
Willis,
Your data and curves are very useful. If all trends continue as shown, it is clear that except for some special cases, that the global death rate will roll over and drop dramatically over the next 1-2 months, and the rest is just extended treatment and cleanup for several more months. This assumes effort is maintained for a while to clean hands, cover coughs and sneezes, minimize contact. US deaths should not far exceed 1,000 by end of year! The country with greatest deaths likely will be Spain (5,000-7,000). If the world does not return to normal activity within 2 to 3 months, it will be a costly problem. If it does return to normal activity in less than 2 months, the economy will bounce back strongly, with little danger of more major issues.
Thanks Willis. You are quite a gift for us.
Willis: I’m not sure that the worldometers site is getting good data for some countries on severe COVID-19 cases. In Canada, the government web site says that ICU stats are unavailable but that the estimated number of severe COVID-19 cases is 3%. The CDC site says that as of March 16, the US had 4226 reported cases, 508 hospitalizations (~12%) and 121 in ICU (~3%). These numbers seem more plausible than those on worldometers.
In one of Willis’s graphs he asks , What does Germany know that the rest of the world doesn’t know? Well the answer could be NO 5G YET.
Is 2020 finally the year for German 5G?
https://www.dw.com/en/is-2020-finally-the-year-for-german-5g/a-51839533
re: “Well the answer could be NO 5G YET.”
Not more of this again?
THIS is how mankind will end – more and more ‘jmorpuss’ types running over and through the ‘system’ spreading stupidity (AND disinformation) since they KNOW NO BETTER! It’s the literal ‘Tower of Babel’ all over again, but this time the tower comes to you, inhabits your forum, posts idiocy, etc.
jmorpuss, you never have answered the question – what do you do for a living?
.
Apparently there is an army or troll bots and spam monsters roaming the interwebs…
Spreading rumors will get you on the organ donor short list in China…
It’s all part of the Corona Belt and Corona Road.
China figures are misleading. Deaths are compared with the total population whereas the outbreak in reality was ‘limited’ to the Hubei province with a population of 60 mio. That puts the curve way higher. More or less similar reasoning for Korea. Outbreak in other countries shown is nationwide and thus chart is correct (don’t know about Iran). Need another 5-10 days to accurately predict which curve US is following. If by then it would appear that it is way up it will be to late. Better prepare for the tornado…
Totally agree. Its an interesting chart, but its against per million of entire population, where it needs to be per million of the population of the effected geographical areas. Also , its against the day of first death, which is hugely sensitive, as the first death could easily be +/-15 days depending on the course if the illness in that particular person.
I am still just following deaths world wide and not trying to decipher infection rates. The daily increase is still more than 10% and trending up today. You can all work out what this would mean if the increase continues at more than 10% a day even for weeks. Eventually 10’s of thousands a day then 100’s. This is why extreme measures are being taken. The survivors can debate the final result eventually.
Per my comments on hospitalization for CV-19 (see Ethan Brand March 21, 2020 at 8:56 am above), I found a report that has the type of data I am looking for (this is for flu for 2006-2016). (“Specifically: What is the actual data for ventilator usage for CV-19 patients. Time and region. Duration. Demographics. Outcome.”).
I hope that some of this type of data is being collected now in real time…..I have sent one inquiry to the Penn Medicine folks, and now will try to track down how the data for the below report is garnered.
https://www.hcup-us.ahrq.gov/reports/statbriefs/sb253-Influenza-Hospitalizations-ED-Visits-2006-2016.jsp
Interesting take on why it is so bad in Italy..
https://uncoverdc.com/2020/03/20/why-italy/
See my earlier comment about close talkers, chronic huggers, and the touchy-feely types.
Twelve steppers are gonna be hit hard, now that I think about it…
They always hug each other, every night, everyone hugs everyone else.
They also ran out of protective equipment, which is especially bad for medical workers who came out of retirement to help (and several died for it).
Apparently, corruption is still rampant in Italy but the new Corona belt and Corona road initiative brings in a new kind of corruption.
Nice, unbiased reportage.
Willis,
Great work. Two suggestions: first, on your first chart where you show the actual cases at the end of the lines, maybe add the population in millions. Second a version of the worldometers table that showed all the totals scaled by population would be great
All this graph really measures is population. Smaller populations have corona throughout, while big populations have corona in only geographical parts. This is why, (with the exception of USA), the curves are stacked in descending order of population, with the smallest populations appearing worst hit. Taking this into consideration, it’s America that really has a problem much bigger than any other country.
Willis –
I suggest that you put an image of a grain of salt against the Chinese data
M
All this forecasting pretty much assumes ceteris paribus when it comes to treatment. In fact the discovery/invention of new medicines could make big differences. Anti virals to treat COVID-19 and vaccines to prevent are being worked on now but will take time. Faster more widespread testing for real time presence is possible and was started by UK govt who then backed off. Reasons unclear.(+Private self funded testing not available in UK). The current most promising drug option must be testing for antibodies and immunity. British govt has been saying this is coming down track fast and would be game changer.
One to watch is Switzerland
Total cases: 6863
New cases within last 24 hrs: 1248
Total deaths: 80
New deaths: 24
It has one of the highest rates of infection/ head population. It also has possibly the best health system in the world based around very expensive compulsory private insurance. They also tell the truth. Their rate of mortality is as accurate as we are going to get. Their testing and tracing will be state of the art.
I suggest that you add them to your graph Willis as a kind of control.
Cheers
M
M
Yesterday (or it might have been two days ago, with the time zone differences) Germany reported only two patients in critical or serious condition. The next day they reported 10 new deaths, and 23 in serious or critical condition.
I was just noting a report from Rome, Italy of a 34 year old man with no health problems at all who got sick four days ago, and then died without ever being in an ICU or in serious condition.
He got a fever after returning from Barcelona.
Re a separate issue: I am not noting any dramatically large or obvious diminution in the number of deaths in the countries in Europe that have for at least 5 days been treating everyone with one of the experimental medications, either chloroquine or remdesivir.
If we keep seeing daily deaths in the tens and hundreds, I am going to become increasingly worried.
Time to re-assess damage done by infinite human stupidity https://www.dailywire.com/news/stanford-professor-data-indicates-were-overreacting-to-coronavirus reference to Einsteins comments re human intelligence and the Universe LOL
There are a lot of factors why there are more serious / critical cases in some regions, such as progression of the illness, demographics, underlying health, available of required treatment, etc.
Italy has a high percentage of elderly people (many of whom have underlying issues), a lot of smokers and there is poor air quality in the north, where the outbreak occurred. Also, for the past couple of weeks, many hospitals have been far over capacity, resulting triage. Simply, there are not enough resources to treat all of the patients. Many who don’t get the treatment they need will die. This inflates the number of serious / critical cases and increased the mortality rate.
It’s very likely that the lack of resources will have the same effect elsewhere. I have close family in healthcare and our hospitals are already struggling to prepare for the expected case load. When the beds are full, there will be people who will not get treated, and some of them will die as a result, so you can expect that the percentage of serious / critical patients will increase.
Tracking the official numbers in the UK. This is what is actually happening.
https://www.bbc.co.uk/news/uk-51768274
Willis’ previous piece, extrapolating from the Diamond Princess, was absurd. To extrapolate more reasonably from the above numbers, do something like this. Though the parameters are subject to great uncertainty.
How to think about the numbers. You cannot rely on the confirmed cases since so far there has been very little testing of suspected cases, and no way to know whether how many mild cases followed by recoveries there have been.
To arrive at the likely numbers of cumulative infections in the population on a given day, multiply the cumulative number of deaths reported for that day by 750. That number may be wrong, and has large uncertainty. Propose a different one if you disagree.
The UK reported 233 deaths yesterday, so the cumulative number of cases in the population (if the 750 number is right) will be 175,000.
Now take the average time to death. Assume 17 days. And take the mortality rate. Assume 2%. It could be higher or lower, so you have to figure out what you think is reasonable given the international statistics.
Assuming the above numbers the forecast would be that by about April 12 there would be around 3,500 cumulative deaths in the UK.
No matter what the government has just done. That will affect the spread, and that in turn, assuming their latest measures slow or stop the spread, will have an effect on death rates after mid-April.
The measures should have an effect. On Friday the Prime Minister basically put the country into total lockdown. Probably in the nick of time.
The measures should have an effect. On Friday the Prime Minister basically put the country into total lockdown. Probably in the nick of time.
We won’t know for sure for a couple of weeks. The new cases & deaths are already in the pipeline. There will be at least 1500 deaths even if further infections stopped on Friday.
I would love to see graphs of past pandemics placed over this one. Something tells me this isn’t as unusual or as unprecedented as it is being made out to be.
Current total numbers are not reason to panic, what the most of people are concerned about it speed of propagation trough population and most of all the rapid exponential rise in number of casualties. When projected forward, the fact there is no effective treatment or vaccine, with impact of economy of the major industrial countries, there is a very good and valid reason to be very concerned indeed. I hope I am wrong, and that in six months or a year all this will be well over, but for the moment
we are now living in the post-normal democracy.
Now I’m hearing people claim that the current pandemic is not caused by a flu. They claim it isn’t a flu because the incidence and mortality is too high. I retained my hard copies of both Taber’s and Steadman’s cyclopedic medical dictionaries from when I was an SF medic and instructing in the course. This is influenza by any definition of the word. The mortality rate has absolutely nothing to do with determining if a viral infection is influenza or not. Causative agent, transmission mode(s) and signs and symptoms/pathology are what determine the definition.
“This is influenza by any definition of the word.”
Coronavirus:
Virus classification
(unranked): Virus
Realm: Riboviria
Phylum: incertae sedis
Order: Nidovirales
Family: Coronaviridae
Subfamily: Orthocoronavirinae
Influenza:
Virus classification
(unranked): Virus
Realm: Riboviria
Phylum: Negarnaviricota
Class: Insthoviricetes
Order: Articulavirales
Family: Orthomyxoviridae
They are both viruses.
Different phylum.
The second broadest grouping in viral taxonomy is the phylum, so these two are almost as far apart as two viruses can get.
Not quite, but almost.
Both viruses, both riboviria.
Beyond that…different.
If taxonomic classification is among “any definitions of the word”, this is not influenza.
It would be more true to say it is a bad cold, but it is not that either.
It is a novel corona virus.
Not the flu, and not influenza.
I have no idea what you are trying to say here.
Rewrite virology?
What I am saying is that according to the common cyclopedic medical dictionary references of Taber’s and Steadman’s, this would be an influenza even if the etiology is not an influenza virus. Transmission, signs & symptoms, and pathology, supportive care, are virtually identical even if the microbiological classifications are different. In those important aspects the microbiology is a distinction without a practical difference. That said, when it comes to developing or discovering a treatment the microbiology is where that distinction becomes practical in it’s application.
It may well be the case that this virus is spread in an identical manner as the flu.
But I have noticed that front line medical doctors and others working closely in the care of COVID patients are becoming ill and in some cases dying, in numbers I have never been aware of occurring with any other illness in recent times, and certainly not in cases of pneumonia or garden variety flu.
So a route of transmission may exist that is not known or not properly taken account of.
One possibility is the oral fecal route, one manifestation of which could be aerosol particles from toilet flushing.
In particular, I have wondered, could it be the high pressure toilets used in commercial and institutional settings are infecting people after someone who is carrying the virus uses the bathroom?
This is a speculation, but there is strong evidence to support the presence of virus in the gut and hence bowel movements of COVID carriers, and very possibly asymptomatic ones.
There is also a wealth of evidence that aerosol plumes are generated whenever a toilet is flushed, and that these aerosols can spread many meters and persist for at least several hours, as well as the virus being present in the bowl for several hours and be aerosolized by subsequent flushing.
This may be completely false, but some other method for transmission equally unappreciated or considered my be true.
Something is causing ti to spread so quickly, and to infect people who are taking precautions.
I believe SARS is strongly suspected to have been spread this way in at least one apartment building in Asia during that outbreak.
And it is reported that a large percentage of COVID patients have gastrointestinal symptoms, in particular diarrhea. The people that do have these symptoms are said to have a higher incidence of severe disease and a higher incidence of mortality. And unconsolidated fecal matter is known to produce a far more concentrated viral plume when a toilet is flushed.
The GI aspect alone may differentiate this virus from flu in at least this one important way.
Maybe.
Very likely, but that is only my opinion.
At the risk of causing people not to read my comment due to it being too darn long, i am going to raise another point here that touches on this: Small differences in viral genome…occasionally tiny differences, can have a huge effect on the virulence of a pathogen.
Similarly, a single tiny difference in the genome of a patient, in some cases that are well know it can be one single base pair substitution out of the the full 6.6 billion base pairs in our diploid set of genetic information, is sufficient for one person to not respond to a viral cure, when people with an alternate base pair are almost always cured.
Specifically, it was discovered years ago that if someone has a cytosine base at position rs12979860 on the IL28B gene in their DNA, they have a far better chance of having interferon and ribavirin cure them of hepatitis C.
If both of your #19 chromosome have a C at this position, that person was shown to be twice as likely to be cured of the virus my interferon combination therapy.
Having just one C conferred some advantage, with people having both be Thymine faring the worst of all.
( I should note that since since was first discovered, there have been rapidly increasing understandings of exactly what is going on with this SNP, and in fact the gene in question shown to be a completely separate gene, now called IFNL3, for interferon lambda. At the time however the closest known gene was that of Interleukin 28. IFNL is one of the most recently discovered genes in the human genome.)
So a difference of as little as one base pair substitution in a persons entire genome can have a drastic effect on how one is able to fight off a virus.
Similarly, tiny variations in the base pairs encoding a viral RNA genome can have a drastic effect on whether a virus is able to be quickly shrugged of by a person’s immune system, or if they develop a lifelong viremia…known as being chronically infected, and have their liver gradually destroyed day by day over many decades, eventually resulting in fibrosis, cirrhosis, and/or hepatocellular carcinoma…liver cancer.
There are 6 major and dozens more minor subgroups of the hepatitis C virus (HCV), called clades or genotypes of the virus.
Some are easily cured, genotype 3 for example, while some are very resistant, such as genotype 4.
(Amazingly some people who have been chronically infected for decades, have cleared the virus while taking a placebo during clinical trials for one of the powerful antivirals. This can greatly complicate the clinical trial and approval processes. It is rare though.)
Many, perhaps most, people who have hep C are completely unaware of it.)
Small pox can be prevented in close to 100% of people by being inoculated with vaccinia virus, a closely related but distinct virus which is harmless to people. Even after 100 years of viral mutation, this was true and eventually the world was rid of the virus by vaccination campaigns.
And yet even smaller variations in the outer proteins of the influenza virus cause vaccine against one strain to be useless against other strains, as probably almost everyone here knows.
The study of the subjects of virology, immunology, and epidemiology are among the most interesting and yet devilishly complex topics in all of science, IMO.
There is so much we have learned that so many are not aware of, even people like myself who might have been considered well informed ten years ago.
And there is so much more to learn, as we are all finding out right now, eh?
I can give references for any assertions made here if anyone would like to verify or read more.
Any search for any of these words or phrases will turn up an immense body of information.
The SARS outbreak suspected of being caused by dry floor drains allowing sewage gases to back-draft from the sewer vent stacks was in Hong Kong Amoy Gardens complex. Read that right after the first patient in Washington State was found to have the virus in his feces.
Made a vow to avoid public restrooms at that point and haven’t broken it yet.
Right there with yah!
It was easy for me to make up my mind to stay the hell out of them places.
Not to be morbid about it. But since Coronavirus kills those with existing medical problems and the elderly, a percentage of those people would die with or without the virus.
So if Coronavirus kills 86,000 Brits, a tragedy, It is not a increase of 86,000 deaths to the normal yearly fatality rate. It is hard to know how to quantify this number. But some of those who contract the virus would have passed in any case.
Willis,
I read your commentary on the perma-post nCOV stats page. Thanks for the work keeping that up!
However, I would note that the lockdowns aren’t being ordered by the US federal government – it is being ordered by the various state governments.
I fully agree that the US federal response isn’t great at all, either, but it isn’t clear to me that the federal government can even prevent the state governments from locking down – or if they had the authority, that the largely blue states that are locking down would let a Trump-led federal government in.
Nobody at any level of government seems to even be considering a South Korea style mobilization: tens of thousands, probably more, of people needed to ubiquitously track and check.
And then there’s the US health care system. Do the private owners and reapers of profit of the US health care system – from insurance companies to pharma to hospitals etc want the heavy hand of the US federal government laid upon them?
Just pointing out that there are a lot of structural issues – not just failure to do more.
Thank you Willis,
This is excellent!
Looking at the Chinese data for deaths per 10 million people, versus the data from other countries, I struck by how smooth the Chinese curve is. These is one slight inflection point, but other wise the deaths reported from China look more like the plot of a model than the plot of raw data.
I have wondered about the accuracy of the infections and deaths reported by China based on videos from Wuhan that I have seen. A recent one showed 30+ people queued outside a Wuhan hospital last week as China was reporting no new cases of infection. Now, looking at your graphs I have even greater concerns about what China is reporting.
China implemented lockdowns in stages. First Wuhan was in lockdown and later other cities were locked down. Wuhan built several new hospitals in record time. If these efforts were effective at slowing the spread of infection, and treating those who were infected, then their data should have some bumps as the impacts of these efforts became evident. Their smooth data does not show this. In my opinion, this indicates that China is releasing data that shows the result which they want to achieve.
“…The US government is about to spend a trillion dollars of your and my tax money to prop up the economy whose wheels have just been taken off by the insane shelter-in-place orders of the US government…”
The US gov’t didn’t implement any “shelter-in-place orders.” Some states basically have. Other states have instituted policies that aren’t to the degree of “shelter-in-place.”
More followup on my quest for real time/dynamic data for ICU/Ventilator CV-19 demand.
See previous “Ethan Brand March 21, 2020 at 3:35 pm”
Some progress in my quest for data:
See:
https://www.cdc.gov/mmwr/volumes/69/wr/mm6912e2.htm?s_cid=mm6912e2_w
Title of the report:
“Severe Outcomes Among Patients with Coronavirus Disease 2019 (COVID-19) — United States, February 12–March 16, 2020”
One key paragraph:
“This report describes the current epidemiology of COVID-19 in the United States, using preliminary data. The findings in this report are subject to at least five limitations. First, data were missing for key variables of interest. Data on age and outcomes, including hospitalization, ICU admission, and death, were missing for 9%–53% of cases, which likely resulted in an underestimation of these outcomes. Second, further time for follow-up is needed to ascertain outcomes among active cases. Third, the initial approach to testing was to identify patients among those with travel histories or persons with more severe disease, and these data might overestimate the prevalence of severe disease. Fourth, data on other risk factors, including serious underlying health conditions that could increase risk for complications and severe illness, were unavailable at the time of this analysis. Finally, limited testing to date underscores the importance of ongoing surveillance of COVID-19 cases. Additional investigation will increase the understanding about persons who are at risk for severe illness and death from COVID-19 and inform clinical guidance and community-based mitigation measures.*”
Please note the very important statement above “data is missing….9%–53% of cases”. Please carefully read the report before making any tentative analysis. For example:
“Data on these cases are preliminary and are missing for some key characteristics of interest, including hospitalization status (1,514), ICU admission (2,253), death (2,001), and age (386).
Based on the incomplete, but perhaps well distributed data (?), we can derive one of the real time (up to a week ago) ICU utilization:
As of March 16, 2019 the report includes 4226 reported cases of Covid-19. Out of the 2253, “121 patients known to have been admitted to an ICU”.
For the reported ICU cases, we have 121/2253 = .054, or 5.4%. I will strongly speculate that any patient requiring a ventilator will be in ICU, so we can perhaps bound the ventilator usage at 5.4% for reported cases.
Now for some reality checks:
The current reported cases in the USA is 38,167 (worldometer 1428 EST 3.22.2020). A grossly conservative estimate of current ICU cases would be .052 x 38167 = 2052. Note that this is likely grossly conservative (off by a factor or 2 or more) as the 38167 includes new daily cases…and we have increased testing over the past several days. From what I understand, it takes on the order of 5-10 days from testing positive to going into ICU…IF you are that sick. Worldometer reports 708 serious cases (as of 3/21 as far as I can tell). The number of reported cases 5 days ago was 9259 (on 3/18). 9259 x .052 = 482…hmm this approximately “smells” plausible.
Important observation: The 5.2% ICU rate is derived from data before March 16. Due to testing bottlenecks, it may be reasonable to conclude that the reported cases were a smaller subset of all actual infections than today or tomorrow…due to more readily available testing.
More speculation: (AGW adherents please pay attention :)). I am going to further build on this very tentative foundation to speculate for future numbers. Going from maybe decent informed guesstimate to speculative guesstimate. Each step away from any good data gets more crappy really fast.
I have seen values (ala Diamond Princess) that perhaps 50% or so of actual infections show no symptoms. This number appears to be reasonable on my reading of data the past month or so. No symptoms (<<>>), no test (usable approximation…conservative at any rate).
IF we surmise that ALL CV-19 infections with symptoms are tested (likely grossly conservative), we could cut the ICU per infection to 2.5%.
I previously noted “Penn Medicine – COVID-19 Hospital Impact Model for Epidemics”, https://penn-chime.phl.io/“. This model defaults to 2% ICU, 1% Ventilator of total infections.
Ok, I like this. When you can independently derive a target value and get reasonably close to another independently derived value my confidence goes up.
Working conclusion:
Approximately 1% of total Covid-19 infections may require use of a ventilator.
Note that this has a very important implication for overall death rate. Out of 2001 cases with outcome data, 44 died, a 2.2% rate, again divide by 2, and you get a 1.1% death rate. Note that in reality you are probably put on ventilator before progressing to death…so I would speculate that the death rate is actually well below 1%…which seems to be well supported by data coming in around the world. Note that this is using speculative TOTAL infections (a guessed value at this time).
At any rate, I conclude that using a rate of approximately 1% require ventilator of total infections (NOT reported infections) is reasonably supportable.
This analysis may provide a reality check on ICU/Ventilator demand as the virus infections progress.
For example, if we were to speculate that in 20 days we have a 20% daily rate, we would 36 x the current number of infections, ie about 1, 368,000 reported cases…x 2 for actual cases = 2.7 million infections (20 raised to 1.2 then times 38,167). This might require 27000 ventilators. Given that the numbers being tossed around is to order and have on hand 10’s of thousands of ventilators, we can see why those numbers may be viable.
I vote for:
Warm weather peak (ie virus because much less viable as temps go up)
Anti Malaria drug efficacy (and any thing else that we can dream up)
Social distancing (or draconian South Korea case followup)
Virus mutation (toward a weaker virus).
Being wrong (ie my approximation is 5-10 times too high).
The next few weeks are really critical. I hope the heck the curves start to go down. Next week would be really nice…..
Ethan Brand
We can vote on those things?
Can we protest the elections if the vote does not go the way we want?
Correlation vs. causation. I assume the COVID19 deaths refer to people who have died who tested positively for COVID19 either while alive or post-mortem. So there is a correlation. To what extent does that correlation reflect causation? Suppose there is a person so critically ill that they are under hospice care, and their care worker unintentionally infects them with COVID19, and they die. Does that prove that COVID19 was the principle cause of death? To what extent are doctors citing COVID19 as cause of death when there are other comorbidities involved? If someone with advanced heart disease contracts COVID19 and then dies of a heart attack, is the cause of death COVID19 or heart disease? Such a person would appear in the graphs, would they not? I would say something to check is whether the overall death rate of a population has increased. In the USA the number of COVID19 deaths is so small as to be no more than statistical noise, unless one focuses on a single municipality. Northern Italy might be a good example.
we are seeing the same flawed statistics of the Global Warming crowd played out in real time … some journalist takes 3 days of data and a ruler and forecasts that 112% of the population will get corona and 57% will die … in 2 years …
Thanks for adding Hubei and CA, not that Chinese “data” are to be trusted.
It looks as if US and RoK curves just might be on incipient stability track, but need at least a week without accelerating losses to know. Italy also appears to be stabilizing, but more than two days required.