The figure above shows a scatterplot of 2019 Global Health Index scores for more than 180 countries (on the X-axis, a higher score is meant to mean better prepared for a pandemic) versus COVID-19 deaths per 100,000 people. These data suggest that countries ranked in 2019 as better prepared for a pandemic have seen worse outcomes during COVID-19. Even restricting the data to countries with a GHSI Score >40 (not shown) indicates no relationship between perceived preparedness and COVID-19 death rates.
What is going on here? Did countries perceived to be better prepared for a pandemic actually see worse outcomes?
I welcome your comments.
The discussion THREAD is a new experimental feature here at The Honest Broker. Comments are open to everyone.
“In this paper, we explore the scientific literature suggesting that vaccination with an mRNA vaccine initiates a set of biological events that are not only different from that induced by infection but are in several ways demonstrably counterproductive to both short- and long-term immune competence and normal cellular function. These vaccinations have now been shown to downregulate critical pathways related to cancer surveillance, infection control, and cellular homeostasis. They introduce into the body highly modified genetic material.”
Innate immune suppression by SARS-CoV-2 mRNA vaccinations: The role of G-quadruplexes, exosomes, and MicroRNAs. Food and Chemical Toxicology. Volume 164, June 2022, 113008.
The better prepared countries had the resources to take more samples. The better prepared confused dying with COVID as opposed to from COVID. The better prepared nations had more people living with chronic aliments that would have killed them years before, such as diabetes or anyone of a number of autoimmune conditions. Living in cities in buildings with recycled air and not in single family homes or in rural areas. In short population density. The developed nation dwellers were probably exposed to fewer virus over the years and their immune systems were not properly primed. Our populations live longer and thus we had more frail elderly. That should give you a good start.
When we had our first lock down I signed up for an oniline course on Coursera in Epidemiology, just to refresh my basic knowledge about the subject, from which I have drawn confirmation on a couple of fundamental facts: first is that our knowledge on pathogens in general it's still very limited and,second, that stopping their circulation is impossible. So my take is on chaos and possibility.
Now you're playing the game of "Just asking questions" about whether an correlation is causation - without even mention of the importance of the myriad potentially confounding variables?
With an R^2 of 0.27 I’m not sure that it says that more prepared countries had worse deaths... I’m tempted to wager that if you put a confidence or prediction interval on that trend line, the model would contain both “nobody died” and “everybody is dead” for each point on the x axis!
I’m sure there are a lot of factors that could be modeled using these outcomes, but it would need to be a multi factorial analysis, including country demographics, health data reporting systems, healthcare system maturity, societal openness/rule following behavior, and a whole lot of other things that smarter people than me could think of. It is tempting to draw the line like this because lots of underdeveloped countries have terrible healthcare systems and are very young and had few deaths (or few reported) so there is a large cluster near (0,0) which biases the trend like crazy.
What the graph reflects is probably the demographics of the African subcontinent, a region with a younger and therefore a less COVID-prone population conjoined presumably with low levels of preparation.
Over the first year of the pandemic, I followed US data very closely, esp. at the county level (see https://resilientcommunities.home.blog/2021/04/19/1-ac-what-the-data-tell-us/). The one consistent predictor of both cases and deaths was population density; even the CDC admitted that, albeit sotto voce. One of the things I quickly concluded was that indices (e.g., the CDCs SVI, and its components) such as the GHSI had no predictive or explanatory value. Selection of parameters may be part of that, as well as simply averaging components.
We in the US did a horrible job of prevention. Frankly, the CDC should be ashamed. Our care of those sick was mediocre – perhaps only slightly better than average. Lack of tests early on contributed to that.
But what isn't considered enough are the components of wellness. Health care contributes only 10-20%. Over half relates to various aspects of lifestyle; the rest can be thought of as environmental. In the US, the vast majority of deaths occurred among the elderly or the immuno-compromised. I suspect (have no data to know one way or the other) that our poor showing re the coronavirus has much to do with our keeping alive those who would otherwise have long since died in other countries.
First thing that comes to mind is that countries with higher GHSI score are also better at collecting and reporting data about COVID-related deaths than countries with a lower score, which could explain the correleation. Your comment that when you only include countries with a score above 40 correlation disappears further supports this idea.
Here is a handy conspiracy theory: The drug companies used their influence in developed companies to discourage the use of generic off-label medications such as HCQ and Ivermectin, while pushing the use of useless medications still under copyright (remdesevir). At the same time, the developed countries never came out with an effective treatment plan because it would have involved generic medications that would have limited the profits of drug companies. Wait and see what happens and put them on a respirator if necessary is not a treatment plan. Meanwhile less developed countries threw everything including the kitchen sink at the disease in hopes that something would work, and in some cases it did.
Personal interaction, number of people you physically interact with, is much higher in an African or Asian village than in developed nations. Most Americas and Europeans live relatively isolated lives. Viral spread rate should have been lower here. The effect noticed in the data is probably due to poor data collection in less developed nations.
Greater and more frequent interactions with others due to greater freedoms. Greater freedoms mean more wealth. More wealth means more abundant transportation means. More freedoms means more rights to meet others if feel like doing so.
We Americans physically interact much less than if we lived crammed into a tiny hut in an African village. We have a huge house and personal transport. They have crowded homes and either walk with everyone else or jam onto a bus.
My understanding is that people of advanced age and/or elevated BMI were/are most at risk. I don't have data on this, but I expect that variations in personal risk factors between countries could easily swamp any country-specific pandemic preparations.
In the "worse-prepared" countries the old, infirm and unfit have been culled (for want of a better word) on a regular basis since time immemorial. The population was already "hardened" in effect. The effect of Covid on these countries was hardly any worse than any of the many previous (e.g. yearly flu) epidemics.
In affluent western countries (and some non-western countries with socialised healthcare systems) the old, infirm and unfit were kept alive for a long time past their natural expiry date - almost artificially.
The populations of these countries were therefore "soft" and ripe for culling.
Better prepared may simply mean better able to count Covid deaths. Less prepared countries have worries greater than Covid to focus on. This may be a reporting problem, not Covid.
Yes, a valid hypothesis. But even among countries that would appear to be well prepared to count cases and deaths, the GHSI does not differentiate outcomes.
better prepared => better overall healthcare => better care for chronically ill => more vulnerable population for this particular respiratory virus => more deaths among vulnerable population.
The statistics of epidemics are messing, whether the 1855 yellow fever at Norfolk/Portsmouth, Virginia (my area of interest) or the current pandemic. While the scale of conversation is often national, what would one do with death statistics for a town like Bristol, Virginia/Tennessee? The deaths on one side of the border would be reported to a different county and state health department and might be subjected to very different suppression or amplification.
I agree with "Andy in Tx" that excess deaths would perhaps be more helpful and with "Gijs van Soest" that some control might be needed for numbers from nations with more comprehensive health statistic tracking mechanism in place before the pandemic versus the less prepared.
Woudl be good to see the countries named... but I suspect it would be "better prepared" = richer, older populations, more obese (although "better prepared" is only relative... better than what? having no plan..?). less well prepared = poorer, younger, healthier. There was much anquish in the press about how covid would "sweep through africa" and the "poorest would be hardest hit..." didn't happen. If you have survived malaria, ebola, dysentry, etc etc, Covid was never going to be an issue.
Like most pan-epidemics, it was spread-by mice in dirty hospitals at the solar minimum. They had no understanding of the history of epidemics; we were lucky it wasn't more deadly. Compare 20th century epidemics to the solar cycle.
COVID deaths is a pretty noisy and probably meaningless measure - excess deaths would be much better. How deaths from/with/during COVID got scored differed too much for that number to be real. I suggest redoing the graph with excess deaths.
A quick look at the index components also suggests it isn't the greatest measure. Lots of stuff (inequality, for example) isn't really about preparedness at all. If the sub scores are available, drilling down and making an index of the relevant things and testing that might be better.
.27 R2 is fairly low, doesn’t explain much variability. It would seem the index isn’t very predictive of preparedness. Without more information, can’t go much beyond that.
Yes. I want to know if the weird patterns in that COVID plot can be seen for another pandemic virus. In particular, it seems to me some specifics of high-income lifestyles should have contributed to the rapidity of spread of SARS-CoV-2. It may be that some influenza viruses are, also, spread by the factors that allowed SARS-CoV-2 to get going early - transmissions by aerosol (while authorities were fixated on droplets) and by asymptomatic infected people.
Low GHSI score countries had insufficient tools to qualify ( and therefore quantify) deaths as covid (lack of Antigen or pcr test). In contrast, some high GHSI countries attributed death to covid for every death testing positive, even if they were run over by a bus! I also agree with @jesse age related adjustment.
I am an obese American and offer that the C19 deaths were exaggerated. Excess deaths may offer better trends. Btw while obese i am still quite active and survived Delta with 3 days of downtime and got J&J jabbed and got B.4 last Xmas (w/zero dt). I think the conclusions are that we need better data in order draw anything meaningful out. I believe i contracted delta at the office (i work for a large Company that has a campus of 5 bldgs each with ca 10 floors... so there were a few thousand of us crammed in)
I can think of one very obvious reason: the vaccine does not work! See https://www.sciencedirect.com/science/article/pii/S027869152200206X
“In this paper, we explore the scientific literature suggesting that vaccination with an mRNA vaccine initiates a set of biological events that are not only different from that induced by infection but are in several ways demonstrably counterproductive to both short- and long-term immune competence and normal cellular function. These vaccinations have now been shown to downregulate critical pathways related to cancer surveillance, infection control, and cellular homeostasis. They introduce into the body highly modified genetic material.”
Innate immune suppression by SARS-CoV-2 mRNA vaccinations: The role of G-quadruplexes, exosomes, and MicroRNAs. Food and Chemical Toxicology. Volume 164, June 2022, 113008.
The better prepared countries had the resources to take more samples. The better prepared confused dying with COVID as opposed to from COVID. The better prepared nations had more people living with chronic aliments that would have killed them years before, such as diabetes or anyone of a number of autoimmune conditions. Living in cities in buildings with recycled air and not in single family homes or in rural areas. In short population density. The developed nation dwellers were probably exposed to fewer virus over the years and their immune systems were not properly primed. Our populations live longer and thus we had more frail elderly. That should give you a good start.
When we had our first lock down I signed up for an oniline course on Coursera in Epidemiology, just to refresh my basic knowledge about the subject, from which I have drawn confirmation on a couple of fundamental facts: first is that our knowledge on pathogens in general it's still very limited and,second, that stopping their circulation is impossible. So my take is on chaos and possibility.
Jesus -
Now you're playing the game of "Just asking questions" about whether an correlation is causation - without even mention of the importance of the myriad potentially confounding variables?
What has happened to you, Roger?
And here I thought my conspiracy theory, below, was an original idea. https://www.zerohedge.com/markets/anthony-fauci-aids-covid-19-pharma-love-story
more prepared countries count a lot more deaths as COVID deaths, all things being equal
wide variety in how deaths are counted
With an R^2 of 0.27 I’m not sure that it says that more prepared countries had worse deaths... I’m tempted to wager that if you put a confidence or prediction interval on that trend line, the model would contain both “nobody died” and “everybody is dead” for each point on the x axis!
I’m sure there are a lot of factors that could be modeled using these outcomes, but it would need to be a multi factorial analysis, including country demographics, health data reporting systems, healthcare system maturity, societal openness/rule following behavior, and a whole lot of other things that smarter people than me could think of. It is tempting to draw the line like this because lots of underdeveloped countries have terrible healthcare systems and are very young and had few deaths (or few reported) so there is a large cluster near (0,0) which biases the trend like crazy.
What the graph reflects is probably the demographics of the African subcontinent, a region with a younger and therefore a less COVID-prone population conjoined presumably with low levels of preparation.
Roger:-
Over the first year of the pandemic, I followed US data very closely, esp. at the county level (see https://resilientcommunities.home.blog/2021/04/19/1-ac-what-the-data-tell-us/). The one consistent predictor of both cases and deaths was population density; even the CDC admitted that, albeit sotto voce. One of the things I quickly concluded was that indices (e.g., the CDCs SVI, and its components) such as the GHSI had no predictive or explanatory value. Selection of parameters may be part of that, as well as simply averaging components.
We in the US did a horrible job of prevention. Frankly, the CDC should be ashamed. Our care of those sick was mediocre – perhaps only slightly better than average. Lack of tests early on contributed to that.
But what isn't considered enough are the components of wellness. Health care contributes only 10-20%. Over half relates to various aspects of lifestyle; the rest can be thought of as environmental. In the US, the vast majority of deaths occurred among the elderly or the immuno-compromised. I suspect (have no data to know one way or the other) that our poor showing re the coronavirus has much to do with our keeping alive those who would otherwise have long since died in other countries.
First thing that comes to mind is that countries with higher GHSI score are also better at collecting and reporting data about COVID-related deaths than countries with a lower score, which could explain the correleation. Your comment that when you only include countries with a score above 40 correlation disappears further supports this idea.
Here is a handy conspiracy theory: The drug companies used their influence in developed companies to discourage the use of generic off-label medications such as HCQ and Ivermectin, while pushing the use of useless medications still under copyright (remdesevir). At the same time, the developed countries never came out with an effective treatment plan because it would have involved generic medications that would have limited the profits of drug companies. Wait and see what happens and put them on a respirator if necessary is not a treatment plan. Meanwhile less developed countries threw everything including the kitchen sink at the disease in hopes that something would work, and in some cases it did.
Personal interaction, number of people you physically interact with, is much higher in an African or Asian village than in developed nations. Most Americas and Europeans live relatively isolated lives. Viral spread rate should have been lower here. The effect noticed in the data is probably due to poor data collection in less developed nations.
Greater and more frequent interactions with others due to greater freedoms. Greater freedoms mean more wealth. More wealth means more abundant transportation means. More freedoms means more rights to meet others if feel like doing so.
We Americans physically interact much less than if we lived crammed into a tiny hut in an African village. We have a huge house and personal transport. They have crowded homes and either walk with everyone else or jam onto a bus.
Is this using reported deaths or excess deaths? If the former, data from LICs will be undercounts
Average age and low physical fitness correlate with affluence, which correlates with HSI.
As others have said, I would suggest that the correlation between preparedness and deaths is virtually non-existent.
Fitting a straight line to this data (while standard practice) only shows that the hypothesis is not worth following.
My understanding is that people of advanced age and/or elevated BMI were/are most at risk. I don't have data on this, but I expect that variations in personal risk factors between countries could easily swamp any country-specific pandemic preparations.
In the "worse-prepared" countries the old, infirm and unfit have been culled (for want of a better word) on a regular basis since time immemorial. The population was already "hardened" in effect. The effect of Covid on these countries was hardly any worse than any of the many previous (e.g. yearly flu) epidemics.
In affluent western countries (and some non-western countries with socialised healthcare systems) the old, infirm and unfit were kept alive for a long time past their natural expiry date - almost artificially.
The populations of these countries were therefore "soft" and ripe for culling.
as someone has suggested, we need age-adjusted stats. Kenya will have fewer deaths than Germany
Better prepared may simply mean better able to count Covid deaths. Less prepared countries have worries greater than Covid to focus on. This may be a reporting problem, not Covid.
Yes, a valid hypothesis. But even among countries that would appear to be well prepared to count cases and deaths, the GHSI does not differentiate outcomes.
As others have suggested,
better prepared => better overall healthcare => better care for chronically ill => more vulnerable population for this particular respiratory virus => more deaths among vulnerable population.
The statistics of epidemics are messing, whether the 1855 yellow fever at Norfolk/Portsmouth, Virginia (my area of interest) or the current pandemic. While the scale of conversation is often national, what would one do with death statistics for a town like Bristol, Virginia/Tennessee? The deaths on one side of the border would be reported to a different county and state health department and might be subjected to very different suppression or amplification.
I agree with "Andy in Tx" that excess deaths would perhaps be more helpful and with "Gijs van Soest" that some control might be needed for numbers from nations with more comprehensive health statistic tracking mechanism in place before the pandemic versus the less prepared.
Woudl be good to see the countries named... but I suspect it would be "better prepared" = richer, older populations, more obese (although "better prepared" is only relative... better than what? having no plan..?). less well prepared = poorer, younger, healthier. There was much anquish in the press about how covid would "sweep through africa" and the "poorest would be hardest hit..." didn't happen. If you have survived malaria, ebola, dysentry, etc etc, Covid was never going to be an issue.
I just posted up a map that shows preparedness levels by country: https://twitter.com/RogerPielkeJr/status/1569695284766117888?s=20&t=kzo3UMFL9WQwU07am1KsnQ
Like most pan-epidemics, it was spread-by mice in dirty hospitals at the solar minimum. They had no understanding of the history of epidemics; we were lucky it wasn't more deadly. Compare 20th century epidemics to the solar cycle.
COVID deaths is a pretty noisy and probably meaningless measure - excess deaths would be much better. How deaths from/with/during COVID got scored differed too much for that number to be real. I suggest redoing the graph with excess deaths.
A quick look at the index components also suggests it isn't the greatest measure. Lots of stuff (inequality, for example) isn't really about preparedness at all. If the sub scores are available, drilling down and making an index of the relevant things and testing that might be better.
Yes, excess deaths would also be worth regressing here.
.27 R2 is fairly low, doesn’t explain much variability. It would seem the index isn’t very predictive of preparedness. Without more information, can’t go much beyond that.
Agreed!
I'd like to see a scatterplot of illness due to the most recent pandemic form of Influenza vs GHSI.
In theory (but maybe not in practice) the GHSI is supposed to be an indicator of preparedness for any pandemic or public health emergency
Yes. I want to know if the weird patterns in that COVID plot can be seen for another pandemic virus. In particular, it seems to me some specifics of high-income lifestyles should have contributed to the rapidity of spread of SARS-CoV-2. It may be that some influenza viruses are, also, spread by the factors that allowed SARS-CoV-2 to get going early - transmissions by aerosol (while authorities were fixated on droplets) and by asymptomatic infected people.
How about better preparedness correlating with higher rates of co-morbidity?
Another interesting hypothesis
Low GHSI score countries had insufficient tools to qualify ( and therefore quantify) deaths as covid (lack of Antigen or pcr test). In contrast, some high GHSI countries attributed death to covid for every death testing positive, even if they were run over by a bus! I also agree with @jesse age related adjustment.
Wealthier countries may have a more complete registration of cause of death, thus counting more COVID victims.
How much is just a spurious correlation of wealthy countries are A) listed as better prepared and B) have older populations?
Need to get an age adjusted excess mortality rate for comparison. The age effects are way too non-linear for a straight comparison to be meaningful.
Yes, agreed. Multiple comments along these lines ... but have a look at just the (perceived to be) best prepared countries: https://twitter.com/RogerPielkeJr/status/1569631903573250048?s=20&t=kzo3UMFL9WQwU07am1KsnQ
Perhaps the index doesn't work as designed, which is another possibility
A couple of possible explanations:
- the countries with better preparation are rich countries with a lot of aged population, the population most vulnerable to the illness
- the countries with better preparation had better means of policy implementation, and they did implement solutions that were counter-productive
Yes, I agree with you and other here about the importance of different demographics. However, when I look at only the top-ranked countries in the index (with less variation in age profiles) there remains no relationship, see: https://twitter.com/RogerPielkeJr/status/1569631903573250048?s=20&t=kzo3UMFL9WQwU07am1KsnQ
I am an obese American and offer that the C19 deaths were exaggerated. Excess deaths may offer better trends. Btw while obese i am still quite active and survived Delta with 3 days of downtime and got J&J jabbed and got B.4 last Xmas (w/zero dt). I think the conclusions are that we need better data in order draw anything meaningful out. I believe i contracted delta at the office (i work for a large Company that has a campus of 5 bldgs each with ca 10 floors... so there were a few thousand of us crammed in)
No need to withdraw! I am stating a fact and not criticizing anyone (except self). Forget to mention im 61 yo. All the best to you and yours