We're going to break this down by ages (for "it's just old people" ageist awful ghouls") in a second, but let's orient:
On the left, big:
All-cause mortality (Vax: -34%)
On the top right:
COVID mortality (Vax: -74%)
On the bottom right:
Non-COVID mortality (Vax: -28%)
/2
In the oldest age group, 90+, we see benefit of -12.5% mortality to being vaccinated. For COVID mortality, it's -60% and for non-COVID mortality it's -4%
(likely healthy vaccine effect w/early non-covid benefit, so more likely negligable non-COVID mortality benefit)
/3
Ages 80-89: -29% mortality benefit to being vaccinated.
For COVID mortality, it's -73%
For non-COVID mortality it's -21%
("healthy vax fx" NOT likely outside of first few months, COVID death undercount is a more likely explanation as it matches covid death wave)
/4
Ages 70-79: -43% mortality benefit to being vaccinated.
For COVID mortality, it's -80%
For non-COVID mortality it's -37%
my guess as to the "blip" of increased mortality in 2022 is breakthrough serious infection in previously protected people, perhaps delayed vax rollout?
/5
Ages 60-69: -45% mortality benefit to being vaccinated.
For COVID mortality, it's -84%
For non-COVID mortality it's -39%
Just absolutely stark, likely combination of healthy vax fx + undercount COVID + med complications of covid infection.
/6
Ages 50-59: -42% mortality benefit to being vaccinated.
For COVID mortality, it's -85%
For non-COVID mortality it's -36%
Absolutely obvious and incredible mortality protection by being vaccinated.
/7
Now interestingly, below 50, we have strong evidence an inverse to "Healthy vax effect"- under 50, vaccines were first offered to the very sick.
(to illustrate: non-COVID mortality for 90+y with likely "healthy vax fx", for 40-49y with "unhealthy vax fx")
Lets press on!
/7
Ages 40-49: -27% mortality benefit to being vaccinated.
For COVID mortality, it's -81%
For non-COVID mortality it's -19%
And this is **battling** a likely "unhealthy vaccine effect" early on. Note at this young age, NO increased non-covid mortality even by may 2023!
/8
Ages 18-39: -13% mortality benefit to being vaccinated.
For COVID mortality, it's -75%
For non-COVID mortality it's -7% (with a HUGE headwind of early "unhealthy vax effect")
Unequivocally: even for the very young, COVID vaccination prevented death and did NOT cause death.
/9
Quick data note - for COVID deaths at this young age group, we are dealing with very few covid deaths in 2022 onwards. However, given the overall mortality benefit, no evidence of non-covid mortality harm, the risk-benefit analysis clearly supports vaccinating this group.
/10
So how do the misinfo ghouls lie?
1) cherrypicking extremes
We are usually dealing with 95% error, so if we chop data into 2 sexes, 9 vaccination states, & 7 age groups, we are making 126 comparisons per month. To adjust, error would have to be 99.96%!
/11
2) ignoring the very likely "unhealthy vax effect"
The ghouls LOVE to show a subset where 18-39 year olds are more likely to die:
However, as i've demonstrated, the "Healthy vaccine effect" INVERTS at age 50, meaning that in this young group, likely a very strong effect.
/12
Almost certainly, each subsequent booster, especially without mandates, will be taken by the people with the greatest health problems in the 18-39 age group.
After the winter campaign for 4th boosters in the UK, only 9% of 18-39 year olds received a vaccine!
/13
3) COVID ghouls love to lie about the data. Lies, unfortunately, can be hidden behind clever use of statistics or ignoring biases.
/14
CONCLUSION:
ONS UK data clearly shows that for both NON-COVID and COVID deaths, mortality benefit of vaccination exceeds mortality risk, and vaccination saves lives.
(note: non-covid benefit mostly derived from 2021-mid 2022, but still slightly beneficial 2022-2023)
Simple graphs to counter COVID denialism & revisionism
1) American Deaths '99-'23 (raw)
Sometimes simple graphs tell a powerful story. When we talk about the United States having > 1 million excess deaths due to the pandemic, this is the visualization of what that looks like.
2) Age-Standardized Mortality, 1999-2023
The gold standard in mortality is "age-standardized mortality", which ensures that we are controlling for population and age shifts in the US.
Again, COVID a stark shift unlike any "flu" season previous.
(model fit 2014+ is plotted)
3) Excess Mortality 2014-2023
Using a 5-year seasonal model, we can calculate excess mortality. The pandemic DWARFS previous seasonal/flu shifts. >1,000,000 Excess American Deaths during the pandemic, and in December 2020 (before vaccines), 40% higher-than-expected mortality!
Lets make schools safer! #WorldSuicidePreventionDay
1) School can be exciting, stressful, & terrifying. Listen to kids if are distressed!
2) Suicide rates, and suicidal visits to the ER are directly correlated to school days (+50-+100%)
/1
3) When we listen to kids, they tell us they feel the pressure! Surveys of kids reporting severe stress caused by school burdens by the American Psychological Association range from 40% to 80%!
A 2019 pew survey shared some of the top pressures.
/2
4) Significant academic stress is strongly correlated with a number of negative outcomes, like poor sleep quality, substance use, poor academic performance, mental health diagnoses.
Thank you so much to @mehdirhasan for this incredible deep dive into the circumstances that we faced in 2020, the decisions that we made with respect to schools, and the impact that covid has had on children both the virus itself as well as our response to it.
I worked closely with production team as they asked me about the numbers that I've posted and they worked so hard to get it right.
Hours of discussions and data analysis went into informing the @MehdiHasanShow team as best I could.
/2
Example: in learning loss data we have so much variation by district and state that rather than making a bold claim we simply stated what is undeniably true: there was a weak relationship between learning loss and school modality, & tying it directly to school closures hard.
/3
VAERS database deaths:
Trumped by political partisanship
A Twitter Mini-Paper
VAERS is a database storing "side effect" reports (note: VAERS can never show cause, only potential link). Given the subjective nature of entry, I wonder, does political bias affect VAERS?
/1
Why do I say that VAERS is subjective in it's entry? For a number of reasons:
a) entry into VAERS is requested but voluntary when someone suspects a vaccine-induced event
b) no standard exists for VAERS entries
c) can be a professional or layperson
d) events are NOT verified
/2
Each of these could be subject to political bias - right wing politicians have promoted "your choice about vaccination" principles and anti-vaccination ideology and left-wing politicians have "you should/must get vaccinated" principles and pro-vaccination ideology.
/3
Imagine a database, BAERS: Breakkie Adverse Eating Reaction System. Whenever you feel like it, you can log when you notice someone gets injured, sick, or dies after eating breakfast.
Heart attack after toast? BAERS.
Stomach ache after cereal? BAERS.
That's what VAERS is.
/1
Millions of people have breakfast daily, so we would expect events ranging from indigestion, stomach pain (most common) to heart attacks & even to deaths.
Some group of anti-breakfasters would claim at some point that breakfast is causing those.
Thats who antivaxxers are.
/2
Every now & then, the anti-breakfasters would share some person who hates breakfast, Deli Bigfeast, or KFCjr, who would take snippets of BAERS & show that BREAKFAST CAUSES MYOCARDITIS due to the # of BAERS reports on it.
That's who the grifters are, and its very lucrative!
/3
This study that's going around is a fantastic example of an awful meta-analysis that adds no new data but makes a surprising and unfounded claim.
/1
There are technical reasons that this study has no superiority for confounding vs other studies, despite the authors claims, but I'm not going to get into that. I'm going to get into the logic.
/2
There are 3 studies included, one from Italy, one from England, and one from Florida (the infamous and extremely biased Ladapo study).
Due to the metaanalytic method, ladapos study was weighted 42%, Italian 40%, and England 17%.
/3