6 ms·

So that's what 0.0000095%? I'll take 5 nines success chance.

jozvolskyef 5y ago
From what David Spiegelhalter said earlier it's about 1 in 250k for the under-30s. That's 0.0004%. The problem is that if you control for the right parameters the expected outcome is worse, possibly by multiple orders of magnitude, than that of being infected with covid.
- spoonjim 5y ago
  What do you mean by that? Control in what way?
  - ggm 5y ago
    It's an odd way of saying if you pre screen you can say which people probably shouldn't get this vaccine.
- data_acquired 5y ago
  This pre-print (edit : now published apparently!) argued that the likelihood of clots from Covid was still higher than AZ vaccines https://www.medrxiv.org/content/10.1101/2021.03.19.21253980v1.full https://www.medrxiv.org/content/10.1101/2021.03.19.21253980v.... Fauci agreed with their conclusions (https://www.cbsnews.com/news/covid-vaccine-blood-clots-risk-disease/ https://www.cbsnews.com/news/covid-vaccine-blood-clots-risk-...) but disagreed with the aspects of this study that looked at mRNA vaccines and blood clots (but hasn't given reasons why he disagrees with this aspect).
  - jozvolskyef 5y ago
    You seem to have linked a different article.
    - data_acquired 5y ago
      My bad. Have fixed the link. Thanks!
    - data_acquired 5y ago
      "While the AstraZeneca vaccine has been the focus, other vaccines have also been reporting adverse events. Searching the EMA EV database for Moderna vaccine reports for deep vein thrombosis and pulmonary embolism identified 29 DVTs, of which 12 also had pulmonary embolism; no reported deaths or fatal events [16]. Similarly, searching for the Pfizer vaccine reports, 13 DVTs were found in the same study period, from 17 February to 12 March, as this paper reporting on the AstraZeneca vaccine adverse events. With the Pfizer vaccine, 11 people also developed pulmonary embolism and two died [16]." Seems to be the case that Pfizer and Moderna didn't get in as much trouble with clot risk I guess?
      - jozvolskyef 5y ago
        Not necessarily, we can't draw any conclusions from these numbers without knowing how many vaccines were administered and to whom.
        data_acquired 5y ago
        Agreed, but I doubt that the difference in # of AZ and Moderna/Pfizer doses administered is off by too many orders of magnitude [edit : on second thought, could be off by a factor of 10]. The central issue in my mind is the difference in populations since Americans have predominantly received Pfizer/Moderna while Brits have predominantly received AZ. Its possible that Brits have an intrinsically higher clot risk with the AZ preparation than Americans do with the Pfizer/Moderna vaccines (although, the J&J clots in the US certainly questions this idea too).
        jozvolskyef 5y ago
        I edited my comment before noticing you already replied. I agree.
- outworlder 5y ago
  > The problem is that if you control for the right parameters the expected outcome is worse, possibly by multiple orders of magnitude, than that of being infected with covid. How come? COVID infections have a high probability of causing blood clots.
nonbirithm 5y ago
I knew someone with a preexisting blood condition who got clotting after taking the vaccine. So for those people the chance is way higher. But the risks for clotting were not as well known at the time as they are now, and she didn't do her own research.
- throwawayboise 5y ago
  People "doing their own research" about COVID vaccines get ridiculed as science-denying nutjobs. Has something changed?
  - dkarras 5y ago
    nope, they are still nutjobs.
    - data_acquired 5y ago
      "Nutjobs" is still a bit harsh, no? :) I do know a fair number of people who fall into the first category (but unfortunately, know far more in the second category).
      - dkarras 5y ago
        I'd agree but I have never seen someone that said "I did my own research" in this context who is not a nutjob. "I did my own research" in this context deeply implies that there is a conspiracy going on. This disease has been the primary focus of the whole world for the last 1.5 years, all data is publicly available, and is publicly interpreted by countless independent top experts in the field frequently. If you "did your own research" and believe things to be opposite of countless independent experts are saying, you probably believe in a conspiracy around all this - that is nutjob territory to me. If you are researching a very niche, understudied subject, sure. But in contexts like this, you probably won't hear the words "I did my own research" from people who know what they are saying.
  - thelean12 5y ago
ku-man 5y ago
But you can have 0. Just avoid the AstraZeneca one.
kasperni 5y ago
According to the numbers that made Denmark stop the AstraZeneca rollout. They saw about one in 40,000 people was effected by clots in some way [1]. [1] https://www.bbc.co.uk/news/world-europe-56744474 https://www.bbc.co.uk/news/world-europe-56744474
Jabbles 5y ago
Not sure where your denominator came from, but the article says 22 million first doses have been given: 41/22e6 = 0.0000019 = 0.00019% So yeah, easily 5 nines, nearly 6. For this particular metric (i.e. death).
- argvargc 5y ago
  GP is doing cases of blood clots rather than deaths. (209/22e6) Regardless, you're both doing it wrong (as mentioned in sibling). To find the correct denominator of persons-fully-vaccinated (second doses), visit the regulators YellowCard report for the vaccine adverse reactions. It's a sobering read. https://www.gov.uk/government/publications/coronavirus-covid-19-vaccine-adverse-reactions/coronavirus-vaccine-summary-of-yellow-card-reporting https://www.gov.uk/government/publications/coronavirus-covid...
  - Jabbles 5y ago
    No, the denominator should not be fully-vaccinated people - how could that possibly account for cases of blood clots after the first dose? It's not sobering, it's scientific or clinical. It's what doctors do on a daily basis even without a pandemic.
    - argvargc 5y ago
      Nope. If the shots happened to take 8 doses to be completely delivered, by tracking shots instead of complete treatments as your denominator, you've just artificially deflated any concern level by 8 times. 100 deaths / 8,000,000 doses = 0.0000125 (in 80,000 people, one is killed - not too worrying, or so it would seem) 100 reactions / 1,000,000 treatments = 0.0001 (in 10,000 people, one is killed - uh oh) The issue you are tracking is how many humans are impacted by the treatment. In the case of COVID-19 treatments (which are two-dose in the UK) - if you choose to track shots versus reports, you lose accuracy by a factor of 2, and this accuracy is skewed against safety. By tracking treatments, you get slower feedback, but it is accurate and supports the correct interpretation of safety levels. Tracking shots may be partially useful as an early warning signal. Other than that though, it's not just not useful, it's completely wrong. The "sobering" references the massive disconnect between what is typically reported by media, and what is reported in the system by doctors and nurses on behalf of, or in limited cases directly by, harmed people. If you read the entire report, and bear in mind these are under-reporting figures, you may - if you're lucky, perhaps - experience a moment of sobriety when it comes to the actual experienced level of safety of what's being assured as "safe".
argvargc 5y ago
Remember to divide by fully-vaccinated persons, not shots administered. It's not the shots that experience the blood clotting. Also, bear in mind that adverse event reporting has been clocked (in pretty much every study) at under-reporting adverse reactions anywhere from a handful of times to hundreds of times (even for serious or fatal reactions).[0] Finally, after you've accounted for those trifling caveats, you've now arrived at an approximation of the odds for a princely one adverse effect, that happened to be apparently causal enough to be readily discovered, and also accurately reported enough to be identified almost immediately. There could be many others of course, complicated by genetics and unforeseen reactions with other medications or conditions, along with that not insignificant variable of time. (Ordinarily, for new substances injected intravenously, a generous several years is afforded to that one.) You also don't know if there is a large demographic out there who share both a reluctance to sign up for medical trials, along with a propensity to exhibit an as-yet undetected adverse reaction. (Quite a reasonable proposition, if you think about it. Don't we always find things in the last place we look?) And all this for a disease most people below average life-expectancy have little risk of dying from, even if they catch it and undergo no treatment, and for which safe, high-efficacy preventative measures and treatments already exist.[1] [0] keywords: pharmacovigilance + under-reporting [1] https://vitamind4all.org https://vitamind4all.org

From what David Spiegelhalter said earlier it's about 1 in 250k for the under-30s. That's 0.0004%. The problem is that if you control for the right parameters the expected outcome is worse, possibly by multiple orders of magnitude, than that of being infected with covid.

spoonjim 5y ago
What do you mean by that? Control in what way?
- ggm 5y ago
  It's an odd way of saying if you pre screen you can say which people probably shouldn't get this vaccine.
data_acquired 5y ago
This pre-print (edit : now published apparently!) argued that the likelihood of clots from Covid was still higher than AZ vaccines https://www.medrxiv.org/content/10.1101/2021.03.19.21253980v1.full https://www.medrxiv.org/content/10.1101/2021.03.19.21253980v.... Fauci agreed with their conclusions (https://www.cbsnews.com/news/covid-vaccine-blood-clots-risk-disease/ https://www.cbsnews.com/news/covid-vaccine-blood-clots-risk-...) but disagreed with the aspects of this study that looked at mRNA vaccines and blood clots (but hasn't given reasons why he disagrees with this aspect).
- jozvolskyef 5y ago
  You seem to have linked a different article.
  - data_acquired 5y ago
    My bad. Have fixed the link. Thanks!
  - data_acquired 5y ago
    "While the AstraZeneca vaccine has been the focus, other vaccines have also been reporting adverse events. Searching the EMA EV database for Moderna vaccine reports for deep vein thrombosis and pulmonary embolism identified 29 DVTs, of which 12 also had pulmonary embolism; no reported deaths or fatal events [16]. Similarly, searching for the Pfizer vaccine reports, 13 DVTs were found in the same study period, from 17 February to 12 March, as this paper reporting on the AstraZeneca vaccine adverse events. With the Pfizer vaccine, 11 people also developed pulmonary embolism and two died [16]." Seems to be the case that Pfizer and Moderna didn't get in as much trouble with clot risk I guess?
    - jozvolskyef 5y ago
      Not necessarily, we can't draw any conclusions from these numbers without knowing how many vaccines were administered and to whom.
      - data_acquired 5y ago
        Agreed, but I doubt that the difference in # of AZ and Moderna/Pfizer doses administered is off by too many orders of magnitude [edit : on second thought, could be off by a factor of 10]. The central issue in my mind is the difference in populations since Americans have predominantly received Pfizer/Moderna while Brits have predominantly received AZ. Its possible that Brits have an intrinsically higher clot risk with the AZ preparation than Americans do with the Pfizer/Moderna vaccines (although, the J&J clots in the US certainly questions this idea too).
        jozvolskyef 5y ago
        I edited my comment before noticing you already replied. I agree.
outworlder 5y ago
> The problem is that if you control for the right parameters the expected outcome is worse, possibly by multiple orders of magnitude, than that of being infected with covid. How come? COVID infections have a high probability of causing blood clots.

nonbirithm 5y ago

I knew someone with a preexisting blood condition who got clotting after taking the vaccine. So for those people the chance is way higher. But the risks for clotting were not as well known at the time as they are now, and she didn't do her own research.

throwawayboise 5y ago
People "doing their own research" about COVID vaccines get ridiculed as science-denying nutjobs. Has something changed?
- dkarras 5y ago
  nope, they are still nutjobs.
  - data_acquired 5y ago
    "Nutjobs" is still a bit harsh, no? :) I do know a fair number of people who fall into the first category (but unfortunately, know far more in the second category).
    - dkarras 5y ago
      I'd agree but I have never seen someone that said "I did my own research" in this context who is not a nutjob. "I did my own research" in this context deeply implies that there is a conspiracy going on. This disease has been the primary focus of the whole world for the last 1.5 years, all data is publicly available, and is publicly interpreted by countless independent top experts in the field frequently. If you "did your own research" and believe things to be opposite of countless independent experts are saying, you probably believe in a conspiracy around all this - that is nutjob territory to me. If you are researching a very niche, understudied subject, sure. But in contexts like this, you probably won't hear the words "I did my own research" from people who know what they are saying.
- thelean12 5y ago

ku-man 5y ago

But you can have 0. Just avoid the AstraZeneca one.

kasperni 5y ago

According to the numbers that made Denmark stop the AstraZeneca rollout. They saw about one in 40,000 people was effected by clots in some way [1]. [1] https://www.bbc.co.uk/news/world-europe-56744474 https://www.bbc.co.uk/news/world-europe-56744474

Jabbles 5y ago

Not sure where your denominator came from, but the article says 22 million first doses have been given: 41/22e6 = 0.0000019 = 0.00019% So yeah, easily 5 nines, nearly 6. For this particular metric (i.e. death).

argvargc 5y ago
GP is doing cases of blood clots rather than deaths. (209/22e6) Regardless, you're both doing it wrong (as mentioned in sibling). To find the correct denominator of persons-fully-vaccinated (second doses), visit the regulators YellowCard report for the vaccine adverse reactions. It's a sobering read. https://www.gov.uk/government/publications/coronavirus-covid-19-vaccine-adverse-reactions/coronavirus-vaccine-summary-of-yellow-card-reporting https://www.gov.uk/government/publications/coronavirus-covid...
- Jabbles 5y ago
  No, the denominator should not be fully-vaccinated people - how could that possibly account for cases of blood clots after the first dose? It's not sobering, it's scientific or clinical. It's what doctors do on a daily basis even without a pandemic.
  - argvargc 5y ago
    Nope. If the shots happened to take 8 doses to be completely delivered, by tracking shots instead of complete treatments as your denominator, you've just artificially deflated any concern level by 8 times. 100 deaths / 8,000,000 doses = 0.0000125 (in 80,000 people, one is killed - not too worrying, or so it would seem) 100 reactions / 1,000,000 treatments = 0.0001 (in 10,000 people, one is killed - uh oh) The issue you are tracking is how many humans are impacted by the treatment. In the case of COVID-19 treatments (which are two-dose in the UK) - if you choose to track shots versus reports, you lose accuracy by a factor of 2, and this accuracy is skewed against safety. By tracking treatments, you get slower feedback, but it is accurate and supports the correct interpretation of safety levels. Tracking shots may be partially useful as an early warning signal. Other than that though, it's not just not useful, it's completely wrong. The "sobering" references the massive disconnect between what is typically reported by media, and what is reported in the system by doctors and nurses on behalf of, or in limited cases directly by, harmed people. If you read the entire report, and bear in mind these are under-reporting figures, you may - if you're lucky, perhaps - experience a moment of sobriety when it comes to the actual experienced level of safety of what's being assured as "safe".

argvargc 5y ago

Remember to divide by fully-vaccinated persons, not shots administered. It's not the shots that experience the blood clotting. Also, bear in mind that adverse event reporting has been clocked (in pretty much every study) at under-reporting adverse reactions anywhere from a handful of times to hundreds of times (even for serious or fatal reactions).[0] Finally, after you've accounted for those trifling caveats, you've now arrived at an approximation of the odds for a princely one adverse effect, that happened to be apparently causal enough to be readily discovered, and also accurately reported enough to be identified almost immediately. There could be many others of course, complicated by genetics and unforeseen reactions with other medications or conditions, along with that not insignificant variable of time. (Ordinarily, for new substances injected intravenously, a generous several years is afforded to that one.) You also don't know if there is a large demographic out there who share both a reluctance to sign up for medical trials, along with a propensity to exhibit an as-yet undetected adverse reaction. (Quite a reasonable proposition, if you think about it. Don't we always find things in the last place we look?) And all this for a disease most people below average life-expectancy have little risk of dying from, even if they catch it and undergo no treatment, and for which safe, high-efficacy preventative measures and treatments already exist.[1] [0] keywords: pharmacovigilance + under-reporting [1] https://vitamind4all.org https://vitamind4all.org