1/ I actually don't like working with VAERS. One would need a data-prep tool, which CDC has, but me not (out of scope to develop). Still, I took a look.
The issues with lot IDs and VEARS deaths/AE are:
-Age confounded (average administered dates).
-Different batch sizes.
2/ We immediately see the high age of most deaths and related batches. We do have confounding to resolve. Deaths, both C19 and vaxx, is an elderly business and includes a lot of comorbidities in both cases. The young don't die in relevant numbers from Covid not vaxx.
3/ One thing that looks strange are the low death batches / weeks in many cases for the very old groups. It should follow an exponential life table to my view. There should be no points in the red circled area (high age, low deaths).
4/ Those lots here look strange. Almost like if they were placebo. But even then, we should expect to see the age confounding (so died "with" vaxx). But maybe the lot size was small. Without batch sizes...very difficult to say much.
5/ Filtering to the first 30 weeks, we get a quite decent exponential trend, like expected in the case of age confounding (died “with” or died from vaxx due to higher risk when old and fragile).
6/ The death age relation follows an exponential trend. It would be clearer if normalizing by doses per age bin. We also see Moderna having same count similar as Biontech, despite a 20% lower number of doses.
This is likely a true signal: the 3x higher dose in Moderna.
7/ To do this better we need to merge in:
Doses by lot, manufacturer and age (lso sex).
Categories: pre-existing conditions.
Population sizes.
As this is not straightforward, I dislike to work with VAERS. It's still much better than having nothing (like in EMA, PEI, Lareb).😌
• • •
Missing some Tweet in this thread? You can try to
force a refresh
1/ After checking, that mortality.org is actually showing the age standardized mortality rates under "death rate, total", we can work with their all age value.
Germany: disaster. The lockdowns finally worked. The other way around. Record excess. And continuously over the weeks. This is what politicians and hysteria did. Not the virus.
Austria lockdown master: disaster. The lockdowns finally worked. The other way around. Record excess. And continuously over the weeks. This is what politicians and hysteria did. Not the virus.
@MntyP1@knigotnik@systemanalysen They can't explain it. But it's quite easy. The excess is dominated by 85+. Those are fragile people, and having had 2 years lockdowns and panic is not healthy. In Sweden this group is doing fine, as living a social life as mammals are supposed to live. See here.
@MntyP1@knigotnik@systemanalysen Both SWE and NL have vaxx levels of 95% in this group. It's irrelevant for the excess. No vaxx can make a damage at this level like NL. It's the lockdown, fear and isolation. Prisoners don't live long. If we continue, we can probably reduce life expectancy by 10 years.
@MntyP1@knigotnik@systemanalysen This excess is also in the 75-84 cohort in NL. Also 93% vaxxed in both countries. Nothing in SWE. It's the lockdown. Isolation and fear is unhealthy for social mammal life. Destroys resilience. I bet it's all cardiovascular and respiratory. @rubenivangaalen@KoudijsHenk
1/ Netherlands ASMR mortality update - now up to week 47 (end November).
Left: perception and panic (unadjusted crude mortality)
Right: Age Standardized Mortality (reality)
2/ ASMR By age group and sex. Problem or not?
All of the 2021 apparent excess in tweet 1 on the left is the illusion of demographics and mortality in the 75+ age population which is getting bigger and bigger every year.
Standardized, the statistical illusion disappears.
2/ All of the Balkan countries show the spike in season 2020, so July 2020 - June 2021.
Interesting.
We may need to expand on the hypothesis laid out by @hmatejx to understand this as Croatia really isn't known for having done hard lockdowns like the Slovenian neighbours.
1/ The main stream media @TheEconomist and also @OurWorldInData continue their "Our Age Confounded Useless Data" game combined with excess metrics which mislead.
2/ I don't have 26M followers @TheEconomist, but I do know that using Age Standardized Mortality Rates is the only meaningful metric unless you want to show that Nigeria with a low excess is better than living in Italy (life expectancy 82 years).
But here an example: Denmark
3/ The @TheEconomist is here suggesting that Denmark with an 46 week ASMR 2021 of 8.48 is better than Norway at 6.66.
30% higher ASMR is a bad.
Having a lot of deaths at a stable baseline is better than having variance at ultra low baselines?
The OWID website is one of the worst. After almost a 2 years, there still aren't any age de-confounded ASMR, and the coverage time series only include misleading aggregated figures. OWID delivers one Simpson's paradox after another.
2/ Here some input for @redouad. OWID likes to fear porn with unadjusted death figures. Even better with cumulated "total deaths" (sounds scary right?) over several years.