No. Regardless of their opinion of the work of others they cannot claim with any certainty that the excess deaths are not due to vaccination, and neither can we claim with certainty that the excess is entirely due to vaccination.
Vaccine harm, including death, is real and we must expect this to contribute to excess death since it is a new factor embedded within the population. To argue otherwise is idiotic.
What they are also glossing over is the fact that ASMRs in tables 1 - 4 are based on incomplete data. They are also ignoring issues using person years to calculate ASMR.
Right in this second I am interpolating the age stratified population count arrays from annual census estimates.
They are mid year estimates as well.
I was thinking I would subtract deaths as they occur and assume linear growth for births/immigration/individuals moving up one age tier and that way get half-way usable figures for 2022 by age.
Oh god. I just finished reading. Thank you for this. And thank you for not looking away when it hurts. I am working with a very heavy heart myself today.
I was thinking the same thing. Not just am I very glad to see this sense of community, but I was also thinking: After a decent meal things will look different.
Today is the anniversary of Sophie Scholl's execution. I like to think she lives on in all of us.
I've always been an empathetic fella, but I could usually switch these feelings off to some degree when needed. Nothing was ever too grim for me to hear or see. I don't quite understand why all of this is giving me such a hard time.
I'm still hoping you went wrong somewhere. My previous estimate was around 1/5th.
It's coming up to 1pm and my stomach is grumbling, so I'm off to fill it up! Once full I'll check every calculation from the beginning with the hope that I've made a right mess of things and that there is little difference.
Astonishing work as ever. Will take some time to digest.
Quick question: are the keys correct on all the graphs? The first and third have unvaccinated in red/orange and vaccinated in blue green, but the second is keyed as unvaccinated in blue/green and vaccinated as red/yellow. Might need a quick check between brews.
I don't follow your unvaccinated population chart. Have you stacked the areas? Also, axis label says percent uptake but has absolute numbers? More importantly, if you have less unvaccinated population than UKHSA (and presumably you used your population estimate for the ultimate deaths/100k) then the results would be even worse than you have displayed?
Why didn't you just use the dashboard numbers? If you do that, you get over 61 million for the population of England!
05_11 5,110,934
12_15 3,005,688
16_17 1,427,083
18_24 5,392,542
25_29 4,667,998
30_34 4,982,353
35_39 4,809,645
40_44 4,473,196
45_49 3,964,478
50_54 4,249,034
55_59 4,205,746
60_64 3,702,659
65_69 3,048,743
70_74 2,724,927
75_79 2,431,689
80_84 1,505,226
85_89 938,159
90+ 536,540
Total 61,176,640*
* Defined as "vaccine register population". Does this mean there are more people on the vaccine register than there are in the country? Many millions more??
Isn't it just? Or rather, it simply adds once again to the uncertainty which is 100 times more an enemy to those who support the official narrative and all the interventions it imposes. Remember when the MSM were all over these reports when they spuriously showed vaccine effectiveness, even for all things completely unrelated to COVID and didn't bat an eyelid? Now, as the ONS slowly but surely, walks back every claim, where are the MSM headlines??
I'm starting with the reasonable proposition that older people are more fearful, more compliant, and more stable: far more likely to have a fixed abode and be registered with a GP. Also far more likely to die than younger peeps. So, apart from a few crusty old farts like me who resisted the jabberwocky, older peeps are more like to have the jab, and also to die , than younger peeps. How else can the figures be so vastly different?
You have a very valid point, but I'm afraid we know too much about the pathophysiology of spike exposition to blame this all on such bias. The truth lies somewhere in between.
Notions like these are what kept me from going insane throughout the past years, but I'm afraid the picture is getting more and more dense.
Haha - you and me both! Age standardisation should soak up the differences we see but by how much is tricky to tell. After stage 2 of the work I'll wheel out an unpublished study I did on the health status of the unvaccinated... age and health are related but are still two different things!
The states which later vaccinated the most were clearly worst off in Q2'2020. I think this is very remarkable, because during every COVID wave after that, the most vaccinated states are better off than all others.
That being said, I think any benefit that does exist only serves to mitigate the carnage that was facilitated by these vaccines in the first place (in one word: Delta).
I still don't know what it all means, but I suppose age stratified deaths per 100k will shed some more light on this.
FYI I've abandoned my plan of taking people through from crude mortality to age-adjusted mortality to health adjusted mortality to realising we can't possibly hope to trap down all the factors necessary in order to make a balanced assessment. Instead of reading and thinking about the problem folk are just snatching charts and running, so I've locked them away behind a paywall. No supper for them!
I agree: and one of the reasons I respect your style of work is that it is not tablets of stone, more of a collaborative or collegiate work in progress, which is why I'm happy to pay for the joy of contributing to a scholarly debate - this is so fast moving and the ONS and other data is shape-shifting every time we re-examine it. Some people want absolute certainties, well, that would be much less fun, and also not reflect best public health practice,
Lunch has cheered me up, so I have summarised my plans of taking this work to stage 2 (age standardisation) and stage 3 (health standardisation) in a final section called 'hope'. The basic idea is to try and trash the difference we see with crude mortality.
On the final graph, shouldn’t you draw the combined vaccinated and unvaccinated line to check it shows the seasonal patterns we expect. Honestly I can’t believe there’s that much difference.
That's a nice idea - I'll add another slide tomorrow. As I get stuck into age standardisation these differences will melt away, quite by how much is anyone's guess.
The proportion of the population who are unvaccinated seems quite difficult to get at, especially when the ONS appear to have no idea how many people there are in this country!! Your model suggests that just over 80% of the adult population are ever vaccinated but the BBC in a poll commissioned for their Unvaccinated programme in mid 2022 found that 664/2558 or 25.95% were unvaccinated. At the time Norman Fenton said that his statistical analysis of these figures showed that it was very unlikely that the proportion of unvaccinated could be less than 20%. This of course would make the blue and orange lines on your final chart even further apart.
It sure is! I've just been hammering away for 5 hours with several spreadsheets and discovered in excess of 100% uptake for the 65-69y and 75-79y age groups. In short we can't trust the dose data, but I'll be writing all this up for part 2, which is going to require another day or so.
When people went into walk-in vax centres, how were they accounted for? did you have to be registered with a GP? or could you just get jabbed and walk away with no jab record?
No it’s recorded, they have a database which are coded with those who took and those who remained unvaxxed. The official numbers are vague because they don’t want to inform the public of how many really are here.
A great question Dave. I saw a sobering post by a whistleblower who revealed just how sloppy data entry and record keeping was. I'm expecting jabbed folk with no record and records with no jabbed folk. Unless a centre manager can enlighten us I shall expect the worst. This really is criminally poor for such a momentous event.
John Dee you should know they harassed, bullied and badgered my friend trying to get her to vaccinate her autistic son.
What was outrageous and disgusting was how they sent a letter to her 14yr old daughter offering her a tenner to go get vaxxed, they’ve kept better track of those who haven’t taken anything.
Outrageous. I think things gotten so bad so quickly that it went over most people's heads. Effectively they went bed Sunday and woke up Monday morning to full-blown fascism; something they didn't recognise because it was writ huge across society.
The main issue with trying to determine the health of the unvaccinated is "confounding by contraindication" (CBC). This is way worse than health user biases, which is already wtf:
(Chinese study says education, healthy body weight, healthy diet, exercise, sleep, regular physical examination, hand washing, using sanitizers, wearing masks, social distancing, decreasing social gathering activities, and abstaining from smoking and drinking were all predictors of vaccination receipt. )
The issue with CBC is not just that the sickest people fear vaccines the most and avoid them. It is also that people who are correctly able to predict their susceptibility to vaccine injury also avoid vaccines. Examples:
1) I have autoimmunity. I don't want it to get worse. I will skip vaccines.
2) My family member had a reaction. I don't want it.
3) The first dose gave me a reaction. I don't want more.
So we have:
1) Some cases of sick people going into the unvaccinated group.
2) Some cases of people - sick or not - successfully preventing their own vaccine injuries, making vaccines look safer than they would be under coercive circumstances.
3) Some cases of vaccine harm that has already occurred causing vaccine avoidance, thereby making say 1 dose look relatively safer than dose 2.
Susceptibility to vaccines injuries does run in families. For example, a couple recent cases were published where siblings (in one case twins) both got myocarditis.
I've done a lot of research on CBC in vaccine-autism studies. Adjustment is basically impossible unless you have a huge volume of literature on vaccine hesitancy to mine that lets you figure out like 13 different parameter values and that describes every single type of contraindication a person may consider. Even then you need to make like 10 assumptions and do a lower-bound estimate. I can conservatively estimate MMR-autism cohort studies are off by a factor of at least 1.3, but probably more.
Matching does nothing to fix this. Nor do "cumulative antigen exposure" studies.
Example: Children who have an older sibling with autism have a 20-fold risk of ending up with autism, relative to children who have an older sibling without autism. Such children also only have an 80% MMR vaccination rate, whereas the overall vaccination rate is 90%. Thus people at higher risk of the adverse outcome are also concentrated in control groups.
There are only a few things that can done afaik:
1) Confirm all the unaddressable biases point in the same direction (i.e. makes vaccines look better). Then if you still get signal of harm, you can be confident it is real.
2) Take some other matched studies that find decreased non-covid mortality among the vaccinated, assume it's all bias (or try to prove that assumption somehow), and then just use that as the fudge factor to correct population level data. Or do sensitivity analysis across a reasonable range of values that fudge factor may take. (One CDC study says 60-70% lower non-covid mortality fudge. Another study says 30% fudge. )
3) Or do something similar to what Ethical Skeptic has done for the US. Regress county-level vaccination rate against all-cause mortality *IN 2020* (or maybe 2019?). Since vaccines cannot time travel, any apparent benefit is the trajectory of the bias. Then see if you can interpolate the ACM of a 0% vax county and a 100% vax county. Divide the two. That is the correction factor.
That is a fabulous summary of the impossible situation we are in. I've started with crude mortality and will be showing how this gets bounced around as we try to compensate. The pitfalls of age standardisation will go through the mangle first, and then I'd like to introduce the results of a study that never got published, that uses what we may call 'health standardisation'. I'll then be asking what else we need to consider, and by then I hope folk will realise this is a bottomless pit. The aim, I guess, is not so much to evaluate the vaccine (I can't possibly do that) but 'expertism'. If folk start to snarl at the phrases 'experts say...' or 'a new study shows...' then I'll have done my job. Thank you for contributing to my wicked scheme!
David, would I be allowed to paste your text into my next article along with my response? I thought I could take folk through a few things when it comes to understanding mortality figures but it seems they don't read what I've written or stop to think - they snatch what they want and run off with it. If this is a sign of the times I may well stop all work on the ONS deaths by vaccination status datafile and produce something less volatile.... or perhaps put this behind a paywall.
Thank you kindly! I thought it prudent to hit readers right up front with how tricky this all is before I start producing slides and tables. Also, I appear to have discovered a major discrepancy in ONS' revised file leading to 80,067 missing deaths. I'm not keen on wasting time on bad data so shall probably let the dust settle and continue with part 4 of my further considerations series. Sometimes plain buttered toast is as good as one covered in chilli beans.
One thing that might still shine through with this data is vaccinated death spikes around vaccination waves. They supposedly fixed the vaccine misclassification issue in this data. And merging unvaccinated population count time series from some other data source can patch the denominator issue. Basically, what Fenton tried to due in 2021, but this time in may work. Maybe he's already doing it.
Nice one! I've been putting in a fair few hours behind the scenes and am sitting on tables of vaccinated and unvaccinated populations by age band and month. This was tedious work because you have to dredge-up all the monthly archived files on dose distribution then reconcile coding frame changes. It's done now, thank goodness, and whilst I was at it I produced monthly population estimates for England by age band. That required a bit of modelling and was far more enjoyable! Perhaps I'll pen an article on this work because the graphs are quite pretty.
The states which later vaccinated the most were clearly worst off in Q2'2020. I think this is very remarkable, because during every COVID wave after that, the most vaccinated states are better off than all others.
I think maybe I've used some misleading terms. I'm wondering if something like this can be extrapolated to guess at the difference between a would-be 100% vaxxed population and a would-be 0% vaxxed population in 2020 or 2019
This still doesn't really illuminate what happened I think.
I prefer plotting correlations and regression slopes over time. I know these charts aren't pretty to look at, but when it comes to understanding what's going on, I think they are worth their weight in gold. Oh wait, they don't have a weight.
Before we beat ourselves up over what might appear to be a monstrous cock-up, let us also remember what brought us here in the first place.
I think we all started out as concerned citizens, naturally worried about the news reports that a deadly new virus from China (yes, China, and Bats)! was about to slay us all.
I'm not just a bat-lover (I was a front-line international cave explorer in exotic places like Papua New Guinea and Japan long before I trained in Public Health Management) , I also have a good nose for bat-shit.
The problem we faced in early 2020 was one of exceptionalism and media sensation: we were at the start of a storm of bat-shit.
Now, reflecting back, we may be at the start of a different storm: one of recrimination as the tide turns
Yes, personally I wish to see those b'stards exposed and tortured for what they did, but I also have to admit that every day, in every way, the world is killing people, and that we need to step back a bit: if we care about humanity then we also must care about the many millions of ghastly deaths and injustices perpetrated globally due to the boring old horsemen of the apocalypse that Covid has somewhat obscured: famines in North Korea, Africa and the Middle East, War in Syria, Congo and Ukraine, mass forced migration from the poor world towards the developed world, but mainly the endless drive of the rich countries to steal the wealth of the poorest, through bribery and a fountain pen.
So: lets avoid exceptionalism: the world is indeed becoming a stinking rotten midden, but Covid is just a symptom of that sickness, and not the cause.
Quite. The COVID circus has eclipsed a great deal, and not just in healthcare. It has usefully shown just how corporately captured our legacy and social media, regulators, experts and governments are.
No. Regardless of their opinion of the work of others they cannot claim with any certainty that the excess deaths are not due to vaccination, and neither can we claim with certainty that the excess is entirely due to vaccination.
Vaccine harm, including death, is real and we must expect this to contribute to excess death since it is a new factor embedded within the population. To argue otherwise is idiotic.
What they are also glossing over is the fact that ASMRs in tables 1 - 4 are based on incomplete data. They are also ignoring issues using person years to calculate ASMR.
A good article to read is this one...
https://open.substack.com/pub/boriquagato/p/the-new-uk-ons-data-is-out-and-its?utm_source=share&utm_medium=android
Profs Fenton & Neil are also worth following if you want the truth...
https://open.substack.com/pub/wherearethenumbers/p/the-latest-ons-data-on-deaths-by?utm_source=share&utm_medium=android
Oh my god, such coincidence.
Right in this second I am interpolating the age stratified population count arrays from annual census estimates.
They are mid year estimates as well.
I was thinking I would subtract deaths as they occur and assume linear growth for births/immigration/individuals moving up one age tier and that way get half-way usable figures for 2022 by age.
I don't think there are any coincidences my dear chap! All seems strangely guided...
Yes. And that is one of the very few things that continuously gave me comfort throughout these past months.
Oh god. I just finished reading. Thank you for this. And thank you for not looking away when it hurts. I am working with a very heavy heart myself today.
I'm glad we've teamed up - this is hard, hard stuff to bear. Maybe we should get that kettle on and have a decent lunch!
I was thinking the same thing. Not just am I very glad to see this sense of community, but I was also thinking: After a decent meal things will look different.
Today is the anniversary of Sophie Scholl's execution. I like to think she lives on in all of us.
I've always been an empathetic fella, but I could usually switch these feelings off to some degree when needed. Nothing was ever too grim for me to hear or see. I don't quite understand why all of this is giving me such a hard time.
I'm still hoping you went wrong somewhere. My previous estimate was around 1/5th.
I've got to have that meal now.
It's coming up to 1pm and my stomach is grumbling, so I'm off to fill it up! Once full I'll check every calculation from the beginning with the hope that I've made a right mess of things and that there is little difference.
There's a German saying that goes "Die Hoffnung stirbt zuletzt."
Hope dies last.
Astonishing work as ever. Will take some time to digest.
Quick question: are the keys correct on all the graphs? The first and third have unvaccinated in red/orange and vaccinated in blue green, but the second is keyed as unvaccinated in blue/green and vaccinated as red/yellow. Might need a quick check between brews.
Right, I better check!
Bottoms! I'll go change the second key...
Sorted - the labels got mixed up. Thanks for spotting that blooper!
I don't follow your unvaccinated population chart. Have you stacked the areas? Also, axis label says percent uptake but has absolute numbers? More importantly, if you have less unvaccinated population than UKHSA (and presumably you used your population estimate for the ultimate deaths/100k) then the results would be even worse than you have displayed?
Why didn't you just use the dashboard numbers? If you do that, you get over 61 million for the population of England!
05_11 5,110,934
12_15 3,005,688
16_17 1,427,083
18_24 5,392,542
25_29 4,667,998
30_34 4,982,353
35_39 4,809,645
40_44 4,473,196
45_49 3,964,478
50_54 4,249,034
55_59 4,205,746
60_64 3,702,659
65_69 3,048,743
70_74 2,724,927
75_79 2,431,689
80_84 1,505,226
85_89 938,159
90+ 536,540
Total 61,176,640*
* Defined as "vaccine register population". Does this mean there are more people on the vaccine register than there are in the country? Many millions more??
Oh crap, I forgot to update the axis label....and the graphing option has flipped to stacked!
Right, the area option is refusing to cooperate so I've replaced it with a boring line graph. The population thingy is a head-banger.
Isn't it just? Or rather, it simply adds once again to the uncertainty which is 100 times more an enemy to those who support the official narrative and all the interventions it imposes. Remember when the MSM were all over these reports when they spuriously showed vaccine effectiveness, even for all things completely unrelated to COVID and didn't bat an eyelid? Now, as the ONS slowly but surely, walks back every claim, where are the MSM headlines??
A great shame we can't go punching people.
I don't get this at all:
I'm starting with the reasonable proposition that older people are more fearful, more compliant, and more stable: far more likely to have a fixed abode and be registered with a GP. Also far more likely to die than younger peeps. So, apart from a few crusty old farts like me who resisted the jabberwocky, older peeps are more like to have the jab, and also to die , than younger peeps. How else can the figures be so vastly different?
You have a very valid point, but I'm afraid we know too much about the pathophysiology of spike exposition to blame this all on such bias. The truth lies somewhere in between.
Notions like these are what kept me from going insane throughout the past years, but I'm afraid the picture is getting more and more dense.
You are going to have to wait until part 2 when I roll out age-standardisation, then pull the rug from under this using a rather unique dataset.
Maybe I am a masochist, because I can't wait.
Haha - you and me both! Age standardisation should soak up the differences we see but by how much is tricky to tell. After stage 2 of the work I'll wheel out an unpublished study I did on the health status of the unvaccinated... age and health are related but are still two different things!
"age and health are related but are still two different things!"
I only fully realized this very recently.
Take a look at section 3.1 of this unpublished work I did for HART...
https://drive.google.com/file/d/1-UYepKXGuzkqITnbK5tWlFGJ4j3dNvuI/view?usp=share_link
I'm on it. I hope it's good news! ;)
I have been thinking a lot about this bias, but it's so hard to quantify.
I am still not completely certain about there being no benefit at all.
These are simple regressions across 52 jurisdictions, 1 per week:
https://substack.pervaers.com/USA_Misc/temp_final_rate_0p_weekly_excess_deaths.png
And for reference, US mortality, COVID mortality and vaccinations:
https://substack.pervaers.com/USA_Misc/allcause_covid_mortality_vaccinations.png
The states which later vaccinated the most were clearly worst off in Q2'2020. I think this is very remarkable, because during every COVID wave after that, the most vaccinated states are better off than all others.
That being said, I think any benefit that does exist only serves to mitigate the carnage that was facilitated by these vaccines in the first place (in one word: Delta).
I still don't know what it all means, but I suppose age stratified deaths per 100k will shed some more light on this.
FYI I've abandoned my plan of taking people through from crude mortality to age-adjusted mortality to health adjusted mortality to realising we can't possibly hope to trap down all the factors necessary in order to make a balanced assessment. Instead of reading and thinking about the problem folk are just snatching charts and running, so I've locked them away behind a paywall. No supper for them!
I agree: and one of the reasons I respect your style of work is that it is not tablets of stone, more of a collaborative or collegiate work in progress, which is why I'm happy to pay for the joy of contributing to a scholarly debate - this is so fast moving and the ONS and other data is shape-shifting every time we re-examine it. Some people want absolute certainties, well, that would be much less fun, and also not reflect best public health practice,
I'll drink to that!
Lunch has cheered me up, so I have summarised my plans of taking this work to stage 2 (age standardisation) and stage 3 (health standardisation) in a final section called 'hope'. The basic idea is to try and trash the difference we see with crude mortality.
On the final graph, shouldn’t you draw the combined vaccinated and unvaccinated line to check it shows the seasonal patterns we expect. Honestly I can’t believe there’s that much difference.
That's a nice idea - I'll add another slide tomorrow. As I get stuck into age standardisation these differences will melt away, quite by how much is anyone's guess.
The proportion of the population who are unvaccinated seems quite difficult to get at, especially when the ONS appear to have no idea how many people there are in this country!! Your model suggests that just over 80% of the adult population are ever vaccinated but the BBC in a poll commissioned for their Unvaccinated programme in mid 2022 found that 664/2558 or 25.95% were unvaccinated. At the time Norman Fenton said that his statistical analysis of these figures showed that it was very unlikely that the proportion of unvaccinated could be less than 20%. This of course would make the blue and orange lines on your final chart even further apart.
It sure is! I've just been hammering away for 5 hours with several spreadsheets and discovered in excess of 100% uptake for the 65-69y and 75-79y age groups. In short we can't trust the dose data, but I'll be writing all this up for part 2, which is going to require another day or so.
When people went into walk-in vax centres, how were they accounted for? did you have to be registered with a GP? or could you just get jabbed and walk away with no jab record?
No it’s recorded, they have a database which are coded with those who took and those who remained unvaxxed. The official numbers are vague because they don’t want to inform the public of how many really are here.
A great question Dave. I saw a sobering post by a whistleblower who revealed just how sloppy data entry and record keeping was. I'm expecting jabbed folk with no record and records with no jabbed folk. Unless a centre manager can enlighten us I shall expect the worst. This really is criminally poor for such a momentous event.
John Dee you should know they harassed, bullied and badgered my friend trying to get her to vaccinate her autistic son.
What was outrageous and disgusting was how they sent a letter to her 14yr old daughter offering her a tenner to go get vaxxed, they’ve kept better track of those who haven’t taken anything.
Outrageous. I think things gotten so bad so quickly that it went over most people's heads. Effectively they went bed Sunday and woke up Monday morning to full-blown fascism; something they didn't recognise because it was writ huge across society.
https://open.substack.com/pub/usmortality/p/uk-ons-updates-english-dataset-after
Did you see this about problems with the data released yesterday?
Yep. I'm keeping work on this minimal until I can trust the numbers
The main issue with trying to determine the health of the unvaccinated is "confounding by contraindication" (CBC). This is way worse than health user biases, which is already wtf:
https://www.frontiersin.org/articles/10.3389/fpubh.2022.918743/full
(Chinese study says education, healthy body weight, healthy diet, exercise, sleep, regular physical examination, hand washing, using sanitizers, wearing masks, social distancing, decreasing social gathering activities, and abstaining from smoking and drinking were all predictors of vaccination receipt. )
The issue with CBC is not just that the sickest people fear vaccines the most and avoid them. It is also that people who are correctly able to predict their susceptibility to vaccine injury also avoid vaccines. Examples:
1) I have autoimmunity. I don't want it to get worse. I will skip vaccines.
2) My family member had a reaction. I don't want it.
3) The first dose gave me a reaction. I don't want more.
So we have:
1) Some cases of sick people going into the unvaccinated group.
2) Some cases of people - sick or not - successfully preventing their own vaccine injuries, making vaccines look safer than they would be under coercive circumstances.
3) Some cases of vaccine harm that has already occurred causing vaccine avoidance, thereby making say 1 dose look relatively safer than dose 2.
Susceptibility to vaccines injuries does run in families. For example, a couple recent cases were published where siblings (in one case twins) both got myocarditis.
I've done a lot of research on CBC in vaccine-autism studies. Adjustment is basically impossible unless you have a huge volume of literature on vaccine hesitancy to mine that lets you figure out like 13 different parameter values and that describes every single type of contraindication a person may consider. Even then you need to make like 10 assumptions and do a lower-bound estimate. I can conservatively estimate MMR-autism cohort studies are off by a factor of at least 1.3, but probably more.
Matching does nothing to fix this. Nor do "cumulative antigen exposure" studies.
Example: Children who have an older sibling with autism have a 20-fold risk of ending up with autism, relative to children who have an older sibling without autism. Such children also only have an 80% MMR vaccination rate, whereas the overall vaccination rate is 90%. Thus people at higher risk of the adverse outcome are also concentrated in control groups.
There are only a few things that can done afaik:
1) Confirm all the unaddressable biases point in the same direction (i.e. makes vaccines look better). Then if you still get signal of harm, you can be confident it is real.
2) Take some other matched studies that find decreased non-covid mortality among the vaccinated, assume it's all bias (or try to prove that assumption somehow), and then just use that as the fudge factor to correct population level data. Or do sensitivity analysis across a reasonable range of values that fudge factor may take. (One CDC study says 60-70% lower non-covid mortality fudge. Another study says 30% fudge. )
3) Or do something similar to what Ethical Skeptic has done for the US. Regress county-level vaccination rate against all-cause mortality *IN 2020* (or maybe 2019?). Since vaccines cannot time travel, any apparent benefit is the trajectory of the bias. Then see if you can interpolate the ACM of a 0% vax county and a 100% vax county. Divide the two. That is the correction factor.
That is a fabulous summary of the impossible situation we are in. I've started with crude mortality and will be showing how this gets bounced around as we try to compensate. The pitfalls of age standardisation will go through the mangle first, and then I'd like to introduce the results of a study that never got published, that uses what we may call 'health standardisation'. I'll then be asking what else we need to consider, and by then I hope folk will realise this is a bottomless pit. The aim, I guess, is not so much to evaluate the vaccine (I can't possibly do that) but 'expertism'. If folk start to snarl at the phrases 'experts say...' or 'a new study shows...' then I'll have done my job. Thank you for contributing to my wicked scheme!
David, would I be allowed to paste your text into my next article along with my response? I thought I could take folk through a few things when it comes to understanding mortality figures but it seems they don't read what I've written or stop to think - they snatch what they want and run off with it. If this is a sign of the times I may well stop all work on the ONS deaths by vaccination status datafile and produce something less volatile.... or perhaps put this behind a paywall.
It would be my honor.
Addendum:
"Factoid: Imminent death is a predictor of flu vaccine avoidance in the elderly. No clue if that still exists, or even reverses with covid vaccines."
Thank you kindly! I thought it prudent to hit readers right up front with how tricky this all is before I start producing slides and tables. Also, I appear to have discovered a major discrepancy in ONS' revised file leading to 80,067 missing deaths. I'm not keen on wasting time on bad data so shall probably let the dust settle and continue with part 4 of my further considerations series. Sometimes plain buttered toast is as good as one covered in chilli beans.
One thing that might still shine through with this data is vaccinated death spikes around vaccination waves. They supposedly fixed the vaccine misclassification issue in this data. And merging unvaccinated population count time series from some other data source can patch the denominator issue. Basically, what Fenton tried to due in 2021, but this time in may work. Maybe he's already doing it.
Nice one! I've been putting in a fair few hours behind the scenes and am sitting on tables of vaccinated and unvaccinated populations by age band and month. This was tedious work because you have to dredge-up all the monthly archived files on dose distribution then reconcile coding frame changes. It's done now, thank goodness, and whilst I was at it I produced monthly population estimates for England by age band. That required a bit of modelling and was far more enjoyable! Perhaps I'll pen an article on this work because the graphs are quite pretty.
Brilliant. However there is a problem with 3. I'll copy paste this bit from another post I made in this thread:
These are simple regressions across 52 jurisdictions, 1 per week:
https://substack.pervaers.com/USA_Misc/temp_final_rate_0p_weekly_excess_deaths.png
And for reference, US mortality, COVID mortality and vaccinations:
https://substack.pervaers.com/USA_Misc/allcause_covid_mortality_vaccinations.png
The states which later vaccinated the most were clearly worst off in Q2'2020. I think this is very remarkable, because during every COVID wave after that, the most vaccinated states are better off than all others.
Here's ET's graph:
https://twitter.com/EthicalSkeptic/status/1550995092198117376/photo/1
I think maybe I've used some misleading terms. I'm wondering if something like this can be extrapolated to guess at the difference between a would-be 100% vaxxed population and a would-be 0% vaxxed population in 2020 or 2019
This still doesn't really illuminate what happened I think.
I prefer plotting correlations and regression slopes over time. I know these charts aren't pretty to look at, but when it comes to understanding what's going on, I think they are worth their weight in gold. Oh wait, they don't have a weight.
https://usa-correlations.pervaers.com/
This is for personal use, non-optimized, needs to buffer lots of data before the page loads.
The first chart that pops up is interesting. You can see the correlation go from -0.8 to 0.7 within 3 months.
This is due to the nature of sigmoid curves and the differences in vaccination speed (approximated by weekly doses/final doses) between states.
Those states that vaccinate fastest, also slow down soonest, leaving their population with better protection. It's an unblinded prisoner's dilemma.
His chart looks really pretty though, I love it.
Before we beat ourselves up over what might appear to be a monstrous cock-up, let us also remember what brought us here in the first place.
I think we all started out as concerned citizens, naturally worried about the news reports that a deadly new virus from China (yes, China, and Bats)! was about to slay us all.
I'm not just a bat-lover (I was a front-line international cave explorer in exotic places like Papua New Guinea and Japan long before I trained in Public Health Management) , I also have a good nose for bat-shit.
The problem we faced in early 2020 was one of exceptionalism and media sensation: we were at the start of a storm of bat-shit.
Now, reflecting back, we may be at the start of a different storm: one of recrimination as the tide turns
Yes, personally I wish to see those b'stards exposed and tortured for what they did, but I also have to admit that every day, in every way, the world is killing people, and that we need to step back a bit: if we care about humanity then we also must care about the many millions of ghastly deaths and injustices perpetrated globally due to the boring old horsemen of the apocalypse that Covid has somewhat obscured: famines in North Korea, Africa and the Middle East, War in Syria, Congo and Ukraine, mass forced migration from the poor world towards the developed world, but mainly the endless drive of the rich countries to steal the wealth of the poorest, through bribery and a fountain pen.
So: lets avoid exceptionalism: the world is indeed becoming a stinking rotten midden, but Covid is just a symptom of that sickness, and not the cause.
Quite. The COVID circus has eclipsed a great deal, and not just in healthcare. It has usefully shown just how corporately captured our legacy and social media, regulators, experts and governments are.