I didn’t want to write this article, but looking at my sparkling
3-year old daughter -my purest love- is the only thing that got me through.
I’m sitting here in New York right now, 7500 miles from Wuhan (a lesser-known city internationally, with twice the population and a fifth the density). I’ve been there once in my young adult journeys, but it’s in every way, a world apart for me. But now, a highly lethal virus has launched itself from that relatively unknown location, in late December 2019. It has ricocheted in and out of countries with relative ease, and is at last closing in on my own house. Arriving like a slow, but certain, unwanted FedEx package.
Along the way, we have seen a mixture of mortality data. Interesting, noisy, highly misleading. We’ve seen cruises, maxed-out on unwelcome infected passengers, anxiously docked anywhere they can. We’ve seen government officials and health organizations completely caught off guard, and scrambling to arrive at a playbook. The stock market collapsed -30%... so far. And now, through many attempts (mostly futile) to do anything to stop this, it’s now here. Charging up my doorstep. 90 days later. 7500 miles away from its Wuhan origin.
There are multiple things I wanted to cover in this article. Please read whatever sections you see fit. Perhaps share the insights with as many people as you can; and in doing so you may unintendedly rescue millions lives. Recall the South Korea disaster, which was ignited as a result of one foolish patient (of 31). There are fools everywhere, in our family and social circles.
1] give an overview of what statistical ideas we had at the start, in mid-January
2] give a chronology of the statistical understanding that was quickly ramping up since that time
3] show the types of policy mishaps that has created the disaster that is before us
4] likely the most important section: show what I consider the changes in mortality risk from outside China
5] give general advice about where things will go from here, based on our actions
We established a free hotline a few weeks ago, and fielded a few dozen calls from around the world. We also established a web calculator using mostly the China experience data, for individuals to predict their own individual death and hospitalization risk this year. The risk factors shown below have evolved very remarkably however, for younger individuals.
Part 1 – starting in mid-January
Over the years, I have studied patterns in the seasonal flu, as well as in previous outbreaks, and wanted to wait a couple weeks to see the experience data before commenting on the coronavirus. By late January, less than 10 days after China’s national health institute started reporting data, a few distressing things were apparent to me. The first was the China data was fake. It was too smooth to reflect a natural situation. The second was that the government apparatus was designed to reduce testing and suppress the true infection and death statistics. By a wide margin; and as we’ll see later, China isn’t the only one doing this. The third thing was China was losing control of this virus: as the days passed and even their low-ball, artificially-smoothed data was still growing irreversibly faster up the curve.
Part 2 – the statistics since mid-January
A couple weeks went by, and America celebrated the annual Super Bowl event. While people enjoyed the memorable, Jennifer Lopez and Shakira combined shake-a-thon, and Nate Silver’s yet-again failed game prediction, I was too busy worried about the fact that on this same hallmark night we finally saw our first ex-China, coronavirus death.
On February 6, I stated that the highly questionable China statistics was somewhat moot now, as the time for global containment was now effectively impossible. We were looking at a pandemic. I was explicitly worried about deaths connected with the virus leaking from outside of the Hubei quarantine. And instead, the world was still fixated on the fake Chinese trending data, which by that point misled people into thinking there was overall progress. And I was instead worried about where we would now be today: here, and everywhere else on earth (here, here).
On February 9, there was a swelling rebuttal from many people in the media (and to this day from President Trump) concerning how the coronavirus panic was worse than the virus itself. That also the seasonal flu vaccine kills many more people. Says who?
I stayed on message. We’ll get back to those death statistics however, further below.
Also on February 9, I also predicted the infection cases everyone have been staring at, which were again fake, would be forced to “soon explode higher”.
On February 10, I projected for example that the China cases (at that time 42 thousand) would explode by more than +50% in just one week. This seemed exaggerated to many, since at the time, these cases were “looking” as if they were growing just a 1-2% a day.
But of course they never were. On February 13, China "shocked" the world when they were forced to significantly revise upward their coronavirus case counts, per the statistical errors I previously cited. And slightly exceeding the +50% explosive spike I had warned about just a couple days prior.
On February 16, a month ago, I had outlined what this pandemic would look like. Multiple global outbreaks; institution of massive martial law. Including many intra-country border closings, and a catastrophic recession.
Contrast this with various officials stating the virus would reach peak containment in February. And then, a couple weeks before it happened, I stated that we’d ultimately suffer multiple drops of worse than -3000 points on the Dow Jones Industrial Average (Trump’s preferred report card).
The market doesn’t drop like that by accident. Having predicted it, I have a sense as to why. It was because there is too much incorrect information floating around and participants were unhappy (here, here).
On February 29 the world again receives from President Trump that we’re doing "really well", that we should go about our business, and that only 15 Americans have coronavirus. The spring weather will help greatly, and everyone everywhere simply needs to calm down. But back to the 15 Americans, and zero deaths. I had argued by then we were already in the thousands instead. Everyone must decide on their own, how this wild difference in experiences has went on to be useful for people.
Fast forward more. On March 4, by this date we have reached the point where ½ the countries on earth have seen confirmed coronavirus cases. The World Health Organization still had not labelled this a pandemic, but was now fundraising under the mantra of “do more research and provide aid”.
On March 7, my personal investigation of major New York stores showed no Purell, facemasks, and other necessities. Knowing some of the store managers personally, I also discovered there was not going to be any warehouse inventory to ship to the retail stores for some time. Soon after this hysteria, we were all told a blatant lie that these products are simply not helpful. Then why do they get used by people all over the world, including in the United States during previous crises?
On March 10, there was not much more that could be done. I again conveyed that the United States would now be locked into an inescapable exponential curve. There would be no other choice but for policy makers to eventually come clean, as the brutality of nature would soon reveal itself mercilessly, on the ill-prepared.
We are cataclysmically headed towards thousands of “confirmed” deaths by mid-summer. Many dozens times more, as we’ll note below, which are unconfirmed. Stretching into all ages, and health categories. Also implying having more deaths than all other nations combined, excluding only China. It will not only be extraordinarily painful, but it will ingrain an everlasting scar on our American psyche.
We’ll note that at the time of this writing, we have more than 60 American deaths, and this is well within this cataclysmic forecasted trajectory we just described.
Part 3 – policy errors that got us here
The entire industrial world, but for South Korea, has so far shown gross incompetence in preparing and responding to an extreme outbreak. Corners were cut, over decades, including with (but not exclusive to) the current administration.
Look at South Korea, a shining beacon of what a national response should look like. They made the fewest errors. They were simply the most prepared. Fluid. Efficient. Leaving less to chance, their leadership example saved more lives outside of Korea than any other country’s leadership.
On the other extreme, we have the United States, which we’ll show further below. Late, ad-hoc, stock-market driven, and somewhat incomprehensible. We could have done better, but it would have been a tragedy on some scale regardless.
We should have been prepared for this (as we do for say natural disasters), instead of now preparing for the next hypothetical pandemic. It will take many years to better align stockpiles to deal with epidemics. If we can do it with oil reserves, we can do it for facemasks, disinfectants, and ventilators.
Many administrations, and various American people, shoulder some responsibility for the lack of focus on this. That same lack of discipline was shown on a micro-level this weekend, where crowds gathered nationally when they should have been social-distancing. This threat is also from within. Recall that the 1918 Spanish flu, despite the name, first started inside the United States.
In the slow response, Americans were reminded (perhaps unfairly) of our response to the Americans of Puerto Rico (eventually rebuilding just 6 homes, distributing paper towels, and claiming only 64 people died in that hurricane disaster.) We’ll need the government to mandate mechanisms for organizations to be better prepared for future epidemics, as we currently do for things such as fire or other natural disasters. We should have well supplied sanitation equipment for example, instead of operating without it, in an economically optimized way that assumes only the best-case outcomes. Safety should be the focus; not real asset prices increases. There will be many rich people now frightened for their life in the ensuing months; this would protect their interests too. Less of a long-term focus to dominate other countries in an economics arms race. Let’s lower our hubris, politics, and ignorance. Let’s accept we don’t know everything, and learn from the South Koreans.
Policy is not only, of course, a one-country topic. China also clearly thwarted information about the disease details, which it had a unique awareness of. But how is it life-saving for anyone to simply ban Chinese nationals in January from travelling to the United States? Why not ban all citizens (including Americans) to and from China? Why not broadly emphasize the need for Americans to prepare for this eventuality? That’s a different style, for whatever it is worth.
Instead, for an additional seven weeks, we had 75,000 planes flying around daily: in, out, and throughout America. Full of infected and contagious, non-Chinese passengers. It was so obviously seeding an increased calamity ahead. Who or what do you think is causing these outbreaks we see now have around the country? They weren’t from the original 15 from Wuhan. This isn't magic.
No one was tested among these millions of passengers throughout February time. Instead TSA agents and travel staff, were infecting passengers. And visa versa. Our ill-fate was sealed, in a way a more robust policy would have helped.
In generic terms, corrective behavior that is too late... is simply that. Too late. There is so much more wreckage here, which we’ll eventually witness, and then in hindsight vow to never allow to happen again.
Part 4- mutating mortality rates
The initial China data suggested, with a superior >70 thousand cases, showed that the largest discriminatory factor among the deceased, was simply age. The elderly. Nothing else.
They tended to be the vast majority of who contracted and died from this disease. Assumptions were then made about why this made sense, including by Dr. Fauci. Pre-conditions, lack of occupation, low-immunity, etc. It was all hindsight looking, and as we’ll see here, all wrong.
Just as one should be careful extrapolating from the fake China case trends discussed in February, there is an ocean of missing data still from China. They didn’t test many people relative to their population. Their population is not representative of other ethnic populations. They way they test or not test, is also different.
Too many ways things can be misleading, despite a supersized >70 thousands cases (more than all other country’s combined).
There are two charts I would like to focus on to show some insight. The first we’ll show the mortality of the country population (vertical axis) by the testing of the country population (horizontal axis). The bubble size is for the size of the country population.
We show the United States, China, and 7 other major countries in this pandemic so far. Including multiple European countries which were infected with coronavirus after the United States first was (January 31).
Look where South Korea is on the chart. See that mortality exists, if you test for it. But also see the %s on both axis. Less than 1% or so of all cases are being tested, and a large fraction of deaths during this time, without a test, get miscategorized as from another cause.
Now let’s peek at this information from another perspective. Let’s look at confirmed coronavirus deaths (again on the vertical axis), relative to the conditional life expectancy of retirees (65+ years of age).
Despite doing a good job for our infants, life expectancy for America’s elders isn’t setting any records. Mortality is high. We should have been testing more. Instead we all know that there is a pandemic. And yet we are testing the least of any of the 9 other countries on this chart. If we consider ourselves exceptional, then what’s the excuse? There is no excuse.
Now the worst statistical error seen during this catastrophe so far is everyone quoting samples as populations. We sample 0.04% of the population, and then take that number and make it out official number of cases and deaths.
It’s the same as say sampling 100 people outside a polling booth, and then stating a total of only 100 people in the country have voted. It’s absurd. In America this sampling is even lower, at 0.01%. Congratulations on being made a fool.
In the end, we are purposefully flying blind. A great nation, unaware what tragedy lies ahead. We can guess that perhaps 5%-10% of the population will be infected, but then leave it to the risk statistician to determine how this works out to one’s individual death chance, for 2020.
For examples of this bias, as we see large differences across nations. For example, only 20% of the Chinese deaths were less than age 60. But outside of China, where the smaller county datasets are now gleaming insights, the mortality % spikes to nearly 35% are less than age 60. So much for Dr. Fauci’s and everyone else’s not-very-useful educated guess about pre-conditions, lack of occupation, and low-immunity. My own death chance suddenly rose by >60%! So did that of all my school classmates. Let that sink in for a moment.
Hence we need to go back to my pair of charts above to better understand the interlocked dependencies of how this works, particularly as this infection starts killing Americans in the weeks ahead.
Applying my vast statistics experience with these cluttered mortality data, it is imperative to align and blend the tables with some prudence. In addition to the oft-discussed now, missing data, we have the complication again of differences in policy and enforcement, from governments around the globe. We have the non-trivial issue of erroneous/false test results, which again lead to many more dead people not classified as coronavirus deaths. And those with the disease not being identified at all, etc. And different tests altogether, with no universal standard protocols to follow.
For a variety of reasons, the medical categorization of country death cause statistics, or even the isolation of symptoms such as lung disease, pneumonia, and flue, are not the same around the world. This makes coronavirus or other disease analysis, somewhat dubious, frustrating, and skewed.
Again how will the mortality shift risk that we are seeing outside of China, for example from the countries in the charts above, impact the risks to Americans? When looking at the underlying risk factors (here) that cause longevity differences among people, it is my estimation that our own chance of infection and death in the US will resemble somewhat what we are seeing outside of China.
Not because of any ill-reliability of Chinese data. But because the genetic make-up of the United States is far more heterogenous across the various ethnicities, and the ex-China data provide a critical understanding into the population-weighted mortality risk that is missing from the large but uncertain China data set. Additionally, looking at Europe and the Middle East, also factors in a better representation of other currently available factors such as execution of social distancing, climate matching, and cultural matching. The charts above show it is our better estimate at this time.
Part 5- can anything really get done now?
It is always best to live a life that practices the best behavior. In trying to prevent both infection, and transmission. Sadly, it is still my view that we are now guaranteed a catastrophic pandemic. Tens of millions of infections, chaos in every hospital, and hundreds of thousands of unnecessary funerals and mourning families.
Panic, and prepare.
Avoid people, and stay hygienic at home with loved ones.
Adopt elders in your community to tend for, and pray for those who won’t make it.
During this distressing time, you may find my book (on sale for $1 this week) a resource to help. Please share this article with all your family and friends who could benefit from the advice.
I’m sitting here in New York right now, 7500 miles from Wuhan (a lesser-known city internationally, with twice the population and a fifth the density). I’ve been there once in my young adult journeys, but it’s in every way, a world apart for me. But now, a highly lethal virus has launched itself from that relatively unknown location, in late December 2019. It has ricocheted in and out of countries with relative ease, and is at last closing in on my own house. Arriving like a slow, but certain, unwanted FedEx package.
Along the way, we have seen a mixture of mortality data. Interesting, noisy, highly misleading. We’ve seen cruises, maxed-out on unwelcome infected passengers, anxiously docked anywhere they can. We’ve seen government officials and health organizations completely caught off guard, and scrambling to arrive at a playbook. The stock market collapsed -30%... so far. And now, through many attempts (mostly futile) to do anything to stop this, it’s now here. Charging up my doorstep. 90 days later. 7500 miles away from its Wuhan origin.
There are multiple things I wanted to cover in this article. Please read whatever sections you see fit. Perhaps share the insights with as many people as you can; and in doing so you may unintendedly rescue millions lives. Recall the South Korea disaster, which was ignited as a result of one foolish patient (of 31). There are fools everywhere, in our family and social circles.
1] give an overview of what statistical ideas we had at the start, in mid-January
2] give a chronology of the statistical understanding that was quickly ramping up since that time
3] show the types of policy mishaps that has created the disaster that is before us
4] likely the most important section: show what I consider the changes in mortality risk from outside China
5] give general advice about where things will go from here, based on our actions
We established a free hotline a few weeks ago, and fielded a few dozen calls from around the world. We also established a web calculator using mostly the China experience data, for individuals to predict their own individual death and hospitalization risk this year. The risk factors shown below have evolved very remarkably however, for younger individuals.
Part 1 – starting in mid-January
Over the years, I have studied patterns in the seasonal flu, as well as in previous outbreaks, and wanted to wait a couple weeks to see the experience data before commenting on the coronavirus. By late January, less than 10 days after China’s national health institute started reporting data, a few distressing things were apparent to me. The first was the China data was fake. It was too smooth to reflect a natural situation. The second was that the government apparatus was designed to reduce testing and suppress the true infection and death statistics. By a wide margin; and as we’ll see later, China isn’t the only one doing this. The third thing was China was losing control of this virus: as the days passed and even their low-ball, artificially-smoothed data was still growing irreversibly faster up the curve.
Will this turn out to be true, or false?— Statistical Ideas (@salilstatistics) January 28, 2020
“China has been working very hard to contain the Coronavirus. The United States greatly appreciates their efforts and transparency. It will all work out well. In particular, on behalf of the American People, I want to thank President Xi!”
Part 2 – the statistics since mid-January
A couple weeks went by, and America celebrated the annual Super Bowl event. While people enjoyed the memorable, Jennifer Lopez and Shakira combined shake-a-thon, and Nate Silver’s yet-again failed game prediction, I was too busy worried about the fact that on this same hallmark night we finally saw our first ex-China, coronavirus death.
Philippines: 2 cases of coronavirus infections,— Statistical Ideas (@salilstatistics) February 2, 2020
1 of those is now passed (first ex-China)
= 50% mortality rate
please don't panic... until AFTER the Super Bowl pic.twitter.com/pAyYTaPTXv
On February 6, I stated that the highly questionable China statistics was somewhat moot now, as the time for global containment was now effectively impossible. We were looking at a pandemic. I was explicitly worried about deaths connected with the virus leaking from outside of the Hubei quarantine. And instead, the world was still fixated on the fake Chinese trending data, which by that point misled people into thinking there was overall progress. And I was instead worried about where we would now be today: here, and everywhere else on earth (here, here).
On February 9, there was a swelling rebuttal from many people in the media (and to this day from President Trump) concerning how the coronavirus panic was worse than the virus itself. That also the seasonal flu vaccine kills many more people. Says who?
I stayed on message. We’ll get back to those death statistics however, further below.
live a life where you avoid dying from EITHER— Statistical Ideas (@salilstatistics) February 9, 2020
coronavirus OR the seasonal flu!
if you are in the improbable situation of being cornered inside a burning theatre... that's not the right time to be worried about the millions dying from cancer
Also on February 9, I also predicted the infection cases everyone have been staring at, which were again fake, would be forced to “soon explode higher”.
Exigeons: "You don't believe in structural breaks?"— Statistical Ideas (@salilstatistics) February 10, 2020
this is a camouflaged catastrophe. the number of cases are not just continuing to increase at high rate of thousands of new cases daily... but this rate will soon continue to explode higher (not implode)https://t.co/egBPPUpjXe
On February 10, I projected for example that the China cases (at that time 42 thousand) would explode by more than +50% in just one week. This seemed exaggerated to many, since at the time, these cases were “looking” as if they were growing just a 1-2% a day.
But of course they never were. On February 13, China "shocked" the world when they were forced to significantly revise upward their coronavirus case counts, per the statistical errors I previously cited. And slightly exceeding the +50% explosive spike I had warned about just a couple days prior.
On February 16, a month ago, I had outlined what this pandemic would look like. Multiple global outbreaks; institution of massive martial law. Including many intra-country border closings, and a catastrophic recession.
(1/2)— Statistical Ideas (@salilstatistics) February 16, 2020
our likely scenario is frightening:
* #coronavirus outbreak in at least one Western country but not directly from people in Hubei province
* explosive outbreaks in at least one other proximate location: ðŸ‡ðŸ‡°, 🚢, 🇸🇬, etc
Contrast this with various officials stating the virus would reach peak containment in February. And then, a couple weeks before it happened, I stated that we’d ultimately suffer multiple drops of worse than -3000 points on the Dow Jones Industrial Average (Trump’s preferred report card).
The market doesn’t drop like that by accident. Having predicted it, I have a sense as to why. It was because there is too much incorrect information floating around and participants were unhappy (here, here).
On February 29 the world again receives from President Trump that we’re doing "really well", that we should go about our business, and that only 15 Americans have coronavirus. The spring weather will help greatly, and everyone everywhere simply needs to calm down. But back to the 15 Americans, and zero deaths. I had argued by then we were already in the thousands instead. Everyone must decide on their own, how this wild difference in experiences has went on to be useful for people.
Fast forward more. On March 4, by this date we have reached the point where ½ the countries on earth have seen confirmed coronavirus cases. The World Health Organization still had not labelled this a pandemic, but was now fundraising under the mantra of “do more research and provide aid”.
On March 7, my personal investigation of major New York stores showed no Purell, facemasks, and other necessities. Knowing some of the store managers personally, I also discovered there was not going to be any warehouse inventory to ship to the retail stores for some time. Soon after this hysteria, we were all told a blatant lie that these products are simply not helpful. Then why do they get used by people all over the world, including in the United States during previous crises?
this week i have traveled to and examined major stores around the New York area (@homedepot, @target, @walmart and @wholefoods) versus my #coronavirus preparation checklist, and have spoken with the store managers there as well to assess recent evisceration in warehouse inventory pic.twitter.com/q8WdDdahZx— Statistical Ideas (@salilstatistics) March 7, 2020
On March 10, there was not much more that could be done. I again conveyed that the United States would now be locked into an inescapable exponential curve. There would be no other choice but for policy makers to eventually come clean, as the brutality of nature would soon reveal itself mercilessly, on the ill-prepared.
We are cataclysmically headed towards thousands of “confirmed” deaths by mid-summer. Many dozens times more, as we’ll note below, which are unconfirmed. Stretching into all ages, and health categories. Also implying having more deaths than all other nations combined, excluding only China. It will not only be extraordinarily painful, but it will ingrain an everlasting scar on our American psyche.
We’ll note that at the time of this writing, we have more than 60 American deaths, and this is well within this cataclysmic forecasted trajectory we just described.
revised low-growth trajectory still is monumental tsunami of 🇺🇸 deaths!— Statistical Ideas (@salilstatistics) March 11, 2020
30 by mid-March, >ðŸ‡ðŸ‡°ðŸ‡¹ðŸ‡¼ðŸ‡¯ðŸ‡µðŸ‡¸ðŸ‡¬ combined
85 by mid-April, >any nation ex-🇨🇳🇮🇹🇮🇷
250 by May, >all European nations combined ex-🇮🇹
750 by June, 3rd highest of any nation
2200 by mid-June, >all nations combined ex-🇨🇳
Part 3 – policy errors that got us here
The entire industrial world, but for South Korea, has so far shown gross incompetence in preparing and responding to an extreme outbreak. Corners were cut, over decades, including with (but not exclusive to) the current administration.
Look at South Korea, a shining beacon of what a national response should look like. They made the fewest errors. They were simply the most prepared. Fluid. Efficient. Leaving less to chance, their leadership example saved more lives outside of Korea than any other country’s leadership.
On the other extreme, we have the United States, which we’ll show further below. Late, ad-hoc, stock-market driven, and somewhat incomprehensible. We could have done better, but it would have been a tragedy on some scale regardless.
We should have been prepared for this (as we do for say natural disasters), instead of now preparing for the next hypothetical pandemic. It will take many years to better align stockpiles to deal with epidemics. If we can do it with oil reserves, we can do it for facemasks, disinfectants, and ventilators.
Many administrations, and various American people, shoulder some responsibility for the lack of focus on this. That same lack of discipline was shown on a micro-level this weekend, where crowds gathered nationally when they should have been social-distancing. This threat is also from within. Recall that the 1918 Spanish flu, despite the name, first started inside the United States.
In the slow response, Americans were reminded (perhaps unfairly) of our response to the Americans of Puerto Rico (eventually rebuilding just 6 homes, distributing paper towels, and claiming only 64 people died in that hurricane disaster.) We’ll need the government to mandate mechanisms for organizations to be better prepared for future epidemics, as we currently do for things such as fire or other natural disasters. We should have well supplied sanitation equipment for example, instead of operating without it, in an economically optimized way that assumes only the best-case outcomes. Safety should be the focus; not real asset prices increases. There will be many rich people now frightened for their life in the ensuing months; this would protect their interests too. Less of a long-term focus to dominate other countries in an economics arms race. Let’s lower our hubris, politics, and ignorance. Let’s accept we don’t know everything, and learn from the South Koreans.
Policy is not only, of course, a one-country topic. China also clearly thwarted information about the disease details, which it had a unique awareness of. But how is it life-saving for anyone to simply ban Chinese nationals in January from travelling to the United States? Why not ban all citizens (including Americans) to and from China? Why not broadly emphasize the need for Americans to prepare for this eventuality? That’s a different style, for whatever it is worth.
Instead, for an additional seven weeks, we had 75,000 planes flying around daily: in, out, and throughout America. Full of infected and contagious, non-Chinese passengers. It was so obviously seeding an increased calamity ahead. Who or what do you think is causing these outbreaks we see now have around the country? They weren’t from the original 15 from Wuhan. This isn't magic.
there are 75,000 planes, loaded with #coronavirus-infected humans, flying throughout all of America every day.— Statistical Ideas (@salilstatistics) February 26, 2020
and their deadly symptoms only appear well after those travelers then infect others. who will in turn, then infect you.
No one was tested among these millions of passengers throughout February time. Instead TSA agents and travel staff, were infecting passengers. And visa versa. Our ill-fate was sealed, in a way a more robust policy would have helped.
In generic terms, corrective behavior that is too late... is simply that. Too late. There is so much more wreckage here, which we’ll eventually witness, and then in hindsight vow to never allow to happen again.
Part 4- mutating mortality rates
The initial China data suggested, with a superior >70 thousand cases, showed that the largest discriminatory factor among the deceased, was simply age. The elderly. Nothing else.
They tended to be the vast majority of who contracted and died from this disease. Assumptions were then made about why this made sense, including by Dr. Fauci. Pre-conditions, lack of occupation, low-immunity, etc. It was all hindsight looking, and as we’ll see here, all wrong.
Just as one should be careful extrapolating from the fake China case trends discussed in February, there is an ocean of missing data still from China. They didn’t test many people relative to their population. Their population is not representative of other ethnic populations. They way they test or not test, is also different.
Too many ways things can be misleading, despite a supersized >70 thousands cases (more than all other country’s combined).
There are two charts I would like to focus on to show some insight. The first we’ll show the mortality of the country population (vertical axis) by the testing of the country population (horizontal axis). The bubble size is for the size of the country population.
We show the United States, China, and 7 other major countries in this pandemic so far. Including multiple European countries which were infected with coronavirus after the United States first was (January 31).
Look where South Korea is on the chart. See that mortality exists, if you test for it. But also see the %s on both axis. Less than 1% or so of all cases are being tested, and a large fraction of deaths during this time, without a test, get miscategorized as from another cause.
Now let’s peek at this information from another perspective. Let’s look at confirmed coronavirus deaths (again on the vertical axis), relative to the conditional life expectancy of retirees (65+ years of age).
Despite doing a good job for our infants, life expectancy for America’s elders isn’t setting any records. Mortality is high. We should have been testing more. Instead we all know that there is a pandemic. And yet we are testing the least of any of the 9 other countries on this chart. If we consider ourselves exceptional, then what’s the excuse? There is no excuse.
Now the worst statistical error seen during this catastrophe so far is everyone quoting samples as populations. We sample 0.04% of the population, and then take that number and make it out official number of cases and deaths.
It’s the same as say sampling 100 people outside a polling booth, and then stating a total of only 100 people in the country have voted. It’s absurd. In America this sampling is even lower, at 0.01%. Congratulations on being made a fool.
what’s the #coronavirus mortality rate (choose nearest estimate):— Statistical Ideas (@salilstatistics) February 19, 2020
In the end, we are purposefully flying blind. A great nation, unaware what tragedy lies ahead. We can guess that perhaps 5%-10% of the population will be infected, but then leave it to the risk statistician to determine how this works out to one’s individual death chance, for 2020.
For examples of this bias, as we see large differences across nations. For example, only 20% of the Chinese deaths were less than age 60. But outside of China, where the smaller county datasets are now gleaming insights, the mortality % spikes to nearly 35% are less than age 60. So much for Dr. Fauci’s and everyone else’s not-very-useful educated guess about pre-conditions, lack of occupation, and low-immunity. My own death chance suddenly rose by >60%! So did that of all my school classmates. Let that sink in for a moment.
Hence we need to go back to my pair of charts above to better understand the interlocked dependencies of how this works, particularly as this infection starts killing Americans in the weeks ahead.
Applying my vast statistics experience with these cluttered mortality data, it is imperative to align and blend the tables with some prudence. In addition to the oft-discussed now, missing data, we have the complication again of differences in policy and enforcement, from governments around the globe. We have the non-trivial issue of erroneous/false test results, which again lead to many more dead people not classified as coronavirus deaths. And those with the disease not being identified at all, etc. And different tests altogether, with no universal standard protocols to follow.
For a variety of reasons, the medical categorization of country death cause statistics, or even the isolation of symptoms such as lung disease, pneumonia, and flue, are not the same around the world. This makes coronavirus or other disease analysis, somewhat dubious, frustrating, and skewed.
Again how will the mortality shift risk that we are seeing outside of China, for example from the countries in the charts above, impact the risks to Americans? When looking at the underlying risk factors (here) that cause longevity differences among people, it is my estimation that our own chance of infection and death in the US will resemble somewhat what we are seeing outside of China.
Not because of any ill-reliability of Chinese data. But because the genetic make-up of the United States is far more heterogenous across the various ethnicities, and the ex-China data provide a critical understanding into the population-weighted mortality risk that is missing from the large but uncertain China data set. Additionally, looking at Europe and the Middle East, also factors in a better representation of other currently available factors such as execution of social distancing, climate matching, and cultural matching. The charts above show it is our better estimate at this time.
Part 5- can anything really get done now?
It is always best to live a life that practices the best behavior. In trying to prevent both infection, and transmission. Sadly, it is still my view that we are now guaranteed a catastrophic pandemic. Tens of millions of infections, chaos in every hospital, and hundreds of thousands of unnecessary funerals and mourning families.
Panic, and prepare.
Avoid people, and stay hygienic at home with loved ones.
Adopt elders in your community to tend for, and pray for those who won’t make it.
During this distressing time, you may find my book (on sale for $1 this week) a resource to help. Please share this article with all your family and friends who could benefit from the advice.
How is that officials are expecting millions of cases in the US and China has had less the 80,000 and only 27 in the last 24 hours (WHO data)? I don't understand why we will have millions when China will likely have less than 100,000. Are the China numbers totally bogus?
ReplyDelete