Covid Data Science

Again, you are completely wrong and simply spouting vaccine conspiracy nonsense. It is very clear from the study that those "suspected but unconfirmed COVID-19 cases" are those who reported one flu-like symptom but had a NEGATIVE PCR test. A very detailed explanation of why you are wrong to want to include those in the efficacy percentage number can be found here: https://www.covid-datascience.com/post/refuting-peter-doshi-s-claims-d… It is not clear whatsoever that the study was bullshit, and the results speak for themselves. An anecdote about your Prime Minister is meaningless and occurred AFTER this initial study period when the variants had changed the landscape and these numbers were no longer applicable.

According to that article this is being driven almost entirely by hospitalizations in the 70+ population. That group is nearly 100% double vaccinated in Scotland. What you are seeing here is classic Simpson's Paradox. It is exactly the same thing we saw earlier with Israel with delta. Here was an article on that which covers how Simpson's Paradox works in that situation [1]. [1] https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v…

Simpson's Paradox applies here: once you slice by age, the data tells a different story. Even when vaccines are highly effective, the confounding factor that gets lost when you aggregate across ages is that the most vulnerable (the elderly) are also the most vaccinated. I highly recommend anybody concerned about hospitalization among the vaccinated read this post. It's about Delta hospitalizations in Israel 2021, but we should expect the same effect from Omicron and across countries: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v…

Worth noting that when you break down the data into age groups the efficacy is substantially higher than that overall number, since the vaccinated population skews older. An example of Simpson's paradox. Good explanation here: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v…

This sounds like Simpson's paradox again: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v… Anecdotally, my wife has been working covid ICU this month and to her it's not surprising. The people with breakthrough cases apparently have so many comorbidities that it's not surprising at all they'd get vaccinated with high priority but also have an immune system that doesn't do a great job learning new skills from the shots. The people for whom the shot is least effective are the ones that get it the most.

> How should I interpret for example the data from Israel? By reading the following analysis and familiarizing yourself with Simpson's paradox: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v… You're spreading misinformation. Stop it.

The 60% is also not adjusted for age, see here for good explanations: https://www.washingtonpost.com/outlook/2021/08/31/covid-israel-hospita… https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v… This is such an important point, I think it's worth including the key table here: Vaccine effectiveness vs. severe disease for younger (<50yr) = 1 - 0.3/3.9 = 91.8% Vaccine effectiveness vs. severe disease for older (>50yr) = 1- 13.6/91.9 = 85.2% But if you combine the two, you get 67.5% !!!! which is the # people have been wrongly focused on.

Expanding on a sibling comment's reference to Simpson's paradox is the following analysis of the data: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v… Basically, Israel is doing quite well even still. Yes, there's a reduction in efficacy, and yes there's breakthrough infections, but your risk of serious outcomes is still reduced by ~80%+, even in older populations _when correctly compared to unvaccinated older populations_.

That number does not mean what people think it does and is being bandied around as a result of extreme statistics illiteracy: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v… Also, what is the world is going on in that preprint[0]? I can not follow their conclusions in Table 1 model 1. Looking at Table 1 - Model 1, it looks like the 40+, early vaccinated cohort has only a 15%-30% chance of being infected as compared to the 40+, late vaccinated cohort. The paper itself says they did some adjustments for co-morbidities, but has no description of what they did. So it looks like the paper's conclusion is directly the opposite of what's being shown in the Table 1 - Model 1. [0]https://www.medrxiv.org/content/10.1101/2021.07.29.21261317v1.full.pdf

In addition to the clear example of Simpson’s paradox presented here: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v… , I offer you the following intuition: if only 4 people in Israel were unvaccinated and 2 were in the hospital with 2 vaccinated patients, would you consider the vaccine effective? It’s important to look at the right numbers.

Sure. I'm not going to do a full lit review for you. You can search PubMed and do that on your own if you like. But here are a few good ones. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8556631/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8553028/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8513403/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8162901/ https://pubmed.ncbi.nlm.nih.gov/35901091/ https://pubmed.ncbi.nlm.nih.gov/35891173/ https://pubmed.ncbi.nlm.nih.gov/35891151/ https://pubmed.ncbi.nlm.nih.gov/35855494/ https://pubmed.ncbi.nlm.nih.gov/35821984/ https://www.covid-datascience.com/ also has some helpful discussions of this stuff if you don't want to have to dig through a bunch of studies.

It's just a preprint and a literature review? Criticism: https://www.covid-datascience.com/post/does-mccullough-s-paper-really-…

It is a joke as the author states in a more recent article[0] [0] https://www.covid-datascience.com/post/what-do-uk-data-say-about-impac…

This is meaningless without context. What percentage of the population is vaccinated? How old are the patients? What percentage of the hospitalized are extremely old/frail/comprosmised? Remember: if 100% of your population is vaccinated, then 100% of your hospitalizations and deaths will be in vaccinated people. https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v…

The first study is clearly underpowered to say anything about the risk after more than around 150 days. This can be seen in figures 2 and S1, where the confidence intervals get larger than the effect size at around this point. Combining multiple age groups together also leads to a large risk of getting incorrect results due to Simpson's paradox [1]. But with so little data, the authors probably couldn't stratify for age groups without the confidence intervals exploding. A much larger study from the US [2] shows no measurable decline in the protection against hospitalization due to COVID over more than 5 months after vaccination. [1] https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v… [2] https://doi.org/10.1016/S0140-6736(21)02183-8

You picked a source that not age-stratified and also looks at a timeframe when many were not vaccinated in the first place. I can give you a source for deaths in the past few weeks in Israel, but only stratified between over/under 60: https://www.covid-datascience.com/post/what-do-new-israeli-data-say-ab… The picture is similar: Efficacy of around 70% (two doses) and 95% (fresh booster) against death. This is far from "one in a million". If you age-stratified this data further, you would likely get better efficacy though.

There is a lot of shoddy methodology and data interpretation in Kirsch's testimony. https://www.covid-datascience.com/post/do-pfizer-vaccines-kill-2-peopl…

Vaccine efficacy on the Delta variant is still a debatable topic, and it's dangerous to make any conclusion from a single reference limited to a very specific, narrow subset of sample. You probably remember Israel's report (which is strongly motivated to promote booster shots) indicated that vaccine's efficacy was pretty low, but it turns out to be a textbook case of Simpson's paradox[1] from extremely biased vaccination rate. From what I saw from KDCA's daily reports (which performs extensive contact tracing with a pretty high coverage, not even remotely comparable to the US), it's more of 70% protection against Delta. Though it's possible that its efficacy is waning over time, which may explain the gap between Korea and Israel. [1] https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v…

I think your analysis doesn't account for the fact that the majority of double vaccinated people are also at far higher risk of covid to begin with. There was an interesting article about this (Simpson's Paradox, as it's known in the world of data analysis) recently: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v… -- the overall trend is reversed when the data are grouped properly. The preponderance of evidence shows that the covid vaccines are overwhelmingly effective and safe for the vast majority of people. Your alarmism and calling Government vaccination campaigns propaganda is contributing to the spread of an incurable but now preventable disease that is killing thousands of people daily and you should feel bad about that. You are a menace to society.

The evidence definitely points the other way, see [0]. The vaccines dramatically reduce the rate of severe infection and hospitalization. The vaccines do not prevent infection, they prime your immune system to fight the infection. That is why the endpoint for the clinical trials is not infection, which would have to be determined by testing, but rather symptomatic infection. [0] https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v…

The vaccines (of all types) aren't somehow an alternate to our immune systems, they're a way to boost our immune systems' response. And yes, they do make the immune system significantly more effective in responding to COVID infection. Go read this: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v…

It's almost a myth that Israel is highly vaccinated, and the vaccine is far more effective in Israel than first meets the eye. A significant % of Israel's entire population is currently too young to be eligible for vaccination. There are two commonly reported %s: percent of population and percent of eligible population. That distinction matters. https://en.wikipedia.org/wiki/Demographics_of_Israel#Structure_of_the_… https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v…

Okay, I looked, and found this. https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v…

Pfizer is performing well under delta: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v…

Things that matter a lot these days: https://en.wikipedia.org/wiki/Base_rate_fallacy https://en.wikipedia.org/wiki/Simpson%27s_paradox Someone that's done some analysis using the above mental models: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v…

It's a statistical effect called Simpson's Paradox. The 60% is nothing to be worried about. https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v… https://twitter.com/jsm2334/status/1427465566210695178

https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v… argues that people are misinterpreting the Israeli data due to Simpson's paradox: In the case of vaccine efficacy vs. severe disease, it is the fact that both vaccination status and risk of severe disease are systematically higher in the older age group that makes overall efficacy numbers if estimated without stratifying by age misleading, producing a paradoxical result that the overall efficacy (67.5%) is much lower than the efficacy for either of the age groups (91.8% and 85.2%).

Apparently not at the same rate. In the UK, ages 1-59 year olds are dying at almost twice the rate if vaccinated than not. The forth chart from the top "ages 10-59" https://www.covid-datascience.com/post/what-do-uk-data-say-about-real-… This site is very pro vaccine, but even here they have to do a lot of hand waving to try and explain it away. And Suddenly everything is fuzzy, they can't quite put the two together. Which is almost funny considering these same people were touting the effectiveness of the vaccine on cherry picked statistical data with utmost certainty. APPENDIX 1. LIST OF ADVERSE EVENTS OF SPECIAL INTEREST from the pfizers own study. http://supremelaw.org/authors/pfizer/1290.adverse.events.htm Here's another one, young person, story from today. https://www.cbc.ca/news/canada/kitchener-waterloo/ayr-centennials-hock… [Edit] Corrected the link https://www.covid-datascience.com/post/what-do-uk-data-say-about-real-…

There is not strong evidence supporting a 3rd booster dose in immunocompetent people. We have partly seen a misinterpretation of the Israeli data [1] and specifically the single low effectiveness value that appears to be the result of Simpson's Paradox [2]. Dr. Fauci reported this single value decline from the high nineties to the high seventies on Andy Slavitt's podcast [3]. Pfizer's CEO has promoted this same data, Israel's 3rd dose campaign is well under way, and Fauci's public statements are a strong indicator that the U.S. will follow the same path. British Columbia's Dr. Bonnie Henry indicated yesterday [4] that the Canadian numbers do not yet support a general 3rd dose booster but, like the UK, a longer 2nd dose interval due to a First Doses First (FDF) strategy may be a factor. The 3rd dose is safe and will probably provide a small benefit to the recipient as Shane Crotty described in TWiV 802 [5]. The downside is that these doses are in short supply globally where they could make a significant difference. [1] https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v… [2] https://en.wikipedia.org/wiki/Simpson%27s_paradox [3] https://lemonadamedia.com/podcast/dr-fauci-answers-your-biggest-questi… [4] (YouTube ~3.5min @36m45s) https://youtu.be/93Rnjmr7iCk?t=36m45s [5] https://www.microbe.tv/twiv/twiv-802/

This may be explained by Simpson's Paradox. Something similar happened in Israel, if you look at the raw case numbers for the current wave, it looks like the vaccine only has a ~67% efficiency: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-v… The UK and Israel both have very high vaccination rates of older people, and lower fully vaccinated rates of younger people. Someone who is 75 most likely has a much weaker immune system than someone who is 25, to the point where someone who is 75 and fully vaccinated may have a much harder time fighting the virus than someone who is 25 and unvaccinated. The data from Israel if broken down by under and over 50s, shows the vaccine efficiency is 92% for under 50s and 85% for over 50s. Israel mainly used Pfizer though, where as the UK mainly AstraZeneca, so there could also be differences there.