8 ms·

Also note efficacy can be misleading. Pfizer was 100% effective at preventing death or serious illness in their trials, which is what's really important. It was

Also note efficacy can be misleading. Pfizer was 100% effective at preventing death or serious illness in their trials, which is what's really important. It was "only" 95% effective against getting the disease at all, which isn't particularly important. i.e. even a vaccine that has 0% efficacy (doesn't prevent mild cases at all) and yet is ~100% at preventing severe cases would be on par with Pfizer's, whereas one that does the converse (prevents all mild cases but no severe ones) would be drastically worse. Not saying these extremes are likely, but we need numbers on the former rather than the latter to say whether another vaccine truly 'rivals' that of Pfizer. P.S. (rant): Why don't we just define a single word for the metric for what really matters and just use that?? It boggles my mind.

cm2187 5y ago
I believe it is 95% effective at getting the disease with symptoms. I understand they never tested people who had no symptoms in their trials.
- dataflow 5y ago
  Oh interesting. Would you have a link on this?
  - cm2187 5y ago
    That's what I have seen mentioned repeatedly on HN. Some of the comments link to articles and studies, but not being a doctor myself, I have not verified them https://hn.algolia.com/?dateRange=pastYear&page=0&prefix=false&query=pfizer%20symptoms&sort=byPopularity&type=comment https://hn.algolia.com/?dateRange=pastYear&page=0&prefix=fal...
    - dataflow 5y ago
      Ok thanks. Btw, I think my comment is actually correct with that definition, given how having the disease (COVID-19) implies having symptoms. IIRC that's precisely when they say you have SARS-CoV-2 (the virus, which you'd be an asymptomatic carrier of), but not COVID-19 (the disease).
smiley1437 5y ago
Marketing
kube-system 5y ago
Another example: Sinovac's Vaccine is 100% effective (at preventing death in Brazil) and 50.7% effective (at preventing symptoms in Brazil)
- marcodiego 5y ago
  Coronavac efficacy in Brazil has been way higher than 90% at preventing death, right: https://www.bbc.com/portuguese/geral-55639265 https://www.bbc.com/portuguese/geral-55639265 ?
  - kube-system 5y ago
    Yep, fixed. Copied the wrong number.
sgc 5y ago
Preventing infection is very important in reducing spread of the disease to others. Vaccines in the end work best when we reach herd immunity, and that occurs by preventing most infection, and thus the possibility of mutation.
- dataflow 5y ago
  > Preventing infection I agree! I didn't say anything about the importance of preventing infection.
  - sgc 5y ago
    Getting the disease is being infected. It is important that people not get it in the first place, so they can't transmit it to others.
    - dataflow 5y ago
      Not quite. As far as the terminology goes, I don't think they're the same, since to me having disease implies feeling ill—note "disease" = "not at ease" = "COVID-19"—whereas infection implies merely carrying the SARS-CoV-2 virus in your body regardless of any visible effects. (Having SARS-CoV-2 does not imply you have COVID-19!) But in any case, let me clarify this. What I agree with (and in my initial comment I said nothing about) is that it matters if people can spread the virus to others. What I don't with is that it matters if you're "infected" if you can't spread it to others (and don't have severe symptoms). The important thing here is that, from what I recall Dr. Fauci said a month or so ago, there have been promising indications that vaccinated people who get infected with the virus are both significantly less likely to get the disease (COVID-19) and to transmit the virus (SARS-CoV-2) to others (independently of the disease, COVID-19). You can read about this on the CDC website [1]. (And my assumption was that was a big factor in why they lifted the mask mandate... because fewer people are capable of spreading it even if they get the virus.) So no, it doesn't matter if you merely get infected. What really matters is how many people can become severely sick as a result of your infection, whether that's you or others. Which is important because these vaccines seem to limit transmission itself. [1] https://www.cdc.gov/coronavirus/2019-ncov/science/science-briefs/fully-vaccinated-people.html https://www.cdc.gov/coronavirus/2019-ncov/science/science-br...
mfsch 5y ago
The main problem is that you need a lot more data to make confident statements about severe cases than just about symptomatic cases. That 100% number of the Pfizer trial had a huge uncertainty interval. So even though severe cases are what you care most about, you usually don’t want to wait for a study that gives you good data on those and measure (and compare) symptomatic cases instead. Zeynep Tufekci discussed this issue of study endpoints in one of her recent articles: https://www.theinsight.org/p/vaccine-efficacy-statistical-power https://www.theinsight.org/p/vaccine-efficacy-statistical-po...
- dataflow 5y ago
  But couldn't they just tell people about the uncertainty? (like "90-98% effective"?) Or underestimate the efficacy (in some objective way, like "at least 90% effective"). Or something else, as long as they're ultimately measuring what matters and what people actually care about. It's not like that 100% with their sample sizes would realistically ever suddenly turn into 50%, would it? It seems just inherently wrong to measure the wrong thing and convey it with certainty than to measure the right thing and convey it with the associated uncertainty, does it not?
  - mfsch 5y ago
    They actually had a table on the efficacy for severe cases in the supplemental material of the original Pfizer study (table S5 in [1]), but all it shows is that the study doesn’t really give any information on this. The 95% confidence interval given for severe cases after the first vaccine is 20.1–99.7%. From what I understand, it’s really a trade-off between waiting for proper data about the efficacy you’re most interested in, or starting to vaccinate people as soon as you have efficacy data for an endpoint that’s still meaningful but easier to reach. Having much fewer symptomatic cases is already a pretty good result, and I think it was expected that the protection would also extend to severe cases. [1]: https://www.nejm.org/doi/suppl/10.1056/NEJMoa2034577/suppl_file/nejmoa2034577_appendix.pdf https://www.nejm.org/doi/suppl/10.1056/NEJMoa2034577/suppl_f...
chiefalchemist 5y ago
How much data (read: jabs) is necessary to evaluate the ability to prevent hospitalizations when even without the jab only a small percentage of cases lead to hospitalization? How much care is given to ensure the jab group's demographics reflect the "control" (i.e., the broader population). For example, ppl getting chemo aren't given the jab, are they? Yet they are counted if they get the virus, end up in hospital, etc. The point being, extrene Covid preys on the (immune) weak. Yet the jabs aren't tested on ppl who profile as such. How is that accounted for in evaluating a vaccine?
- dataflow 5y ago
  I'm a layman so I don't know if I can really give a satisfying answer to your question. All I know is what I see just doesn't make sense to me. Even in April 2020 we had rough but useful numbers on the hospitalization rate. Just some quick Googling turns up e.g. [1] which says that around 20% of those who test positive end up hospitalized. Now the Pfizer vaccine by itself had like 40k participants, with 9 severe (out of 162 confirmed cases) of COVID-19 in the placebo group, and 1 severe case (out of 8 confirmed cases) in the vaccine group. [2] [3] So right off the bat it looked pretty promising, and IMHO they could've just cited these numbers and let people judge for themselves... it's not like people don't have the intuition to realize that 1 patient couldn't have been 2-3 instead. But even if you think that wasn't enough data to be convincing, at least now we have mass vaccination rates they could be citing. Like literally ~40k people have been hospitalized with COVID-19 in just the last week and nearly all COVID deaths are among the unvaccinated now, and it's not like the vaccine just came out last week. An Associated Press analysis of available government data from May shows that “breakthrough” infections in fully vaccinated people accounted for fewer than 1,200 of more than 853,000 COVID-19 hospitalizations. That’s about 0.1%. [4] How is that still not enough data to switch metrics at this point? Why do they still think efficacy rates are more relevant for people instead? [1] https://www.tfah.org/wp-content/uploads/2020/04/COVIDunderlyingconditions040320.pdf#page=2 https://www.tfah.org/wp-content/uploads/2020/04/COVIDunderly... [2] https://www.pfizer.com/news/press-release/press-release-detail/pfizer-and-biontech-conclude-phase-3-study-covid-19-vaccine https://www.pfizer.com/news/press-release/press-release-deta... [3] I don't know why this reporting differs from the 100% I see claimed elsewhere; 1 severe case doesn't sound like 100% effective at preventing severe cases? I don't know if they're discussing different trial phases or have different thresholds for "severe" or something else... but regardless, it's irrelevant to my point. [4] https://apnews.com/article/coronavirus-pandemic-health-941fcf43d9731c76c16e7354f5d5e187 https://apnews.com/article/coronavirus-pandemic-health-941fc...

cm2187 5y ago

I believe it is 95% effective at getting the disease with symptoms. I understand they never tested people who had no symptoms in their trials.

dataflow 5y ago
Oh interesting. Would you have a link on this?
- cm2187 5y ago
  That's what I have seen mentioned repeatedly on HN. Some of the comments link to articles and studies, but not being a doctor myself, I have not verified them https://hn.algolia.com/?dateRange=pastYear&page=0&prefix=false&query=pfizer%20symptoms&sort=byPopularity&type=comment https://hn.algolia.com/?dateRange=pastYear&page=0&prefix=fal...
  - dataflow 5y ago
    Ok thanks. Btw, I think my comment is actually correct with that definition, given how having the disease (COVID-19) implies having symptoms. IIRC that's precisely when they say you have SARS-CoV-2 (the virus, which you'd be an asymptomatic carrier of), but not COVID-19 (the disease).

smiley1437 5y ago

Marketing

kube-system 5y ago

Another example: Sinovac's Vaccine is 100% effective (at preventing death in Brazil) and 50.7% effective (at preventing symptoms in Brazil)

marcodiego 5y ago
Coronavac efficacy in Brazil has been way higher than 90% at preventing death, right: https://www.bbc.com/portuguese/geral-55639265 https://www.bbc.com/portuguese/geral-55639265 ?
- kube-system 5y ago
  Yep, fixed. Copied the wrong number.

sgc 5y ago

Preventing infection is very important in reducing spread of the disease to others. Vaccines in the end work best when we reach herd immunity, and that occurs by preventing most infection, and thus the possibility of mutation.

dataflow 5y ago
> Preventing infection I agree! I didn't say anything about the importance of preventing infection.
- sgc 5y ago
  Getting the disease is being infected. It is important that people not get it in the first place, so they can't transmit it to others.
  - dataflow 5y ago
    Not quite. As far as the terminology goes, I don't think they're the same, since to me having disease implies feeling ill—note "disease" = "not at ease" = "COVID-19"—whereas infection implies merely carrying the SARS-CoV-2 virus in your body regardless of any visible effects. (Having SARS-CoV-2 does not imply you have COVID-19!) But in any case, let me clarify this. What I agree with (and in my initial comment I said nothing about) is that it matters if people can spread the virus to others. What I don't with is that it matters if you're "infected" if you can't spread it to others (and don't have severe symptoms). The important thing here is that, from what I recall Dr. Fauci said a month or so ago, there have been promising indications that vaccinated people who get infected with the virus are both significantly less likely to get the disease (COVID-19) and to transmit the virus (SARS-CoV-2) to others (independently of the disease, COVID-19). You can read about this on the CDC website [1]. (And my assumption was that was a big factor in why they lifted the mask mandate... because fewer people are capable of spreading it even if they get the virus.) So no, it doesn't matter if you merely get infected. What really matters is how many people can become severely sick as a result of your infection, whether that's you or others. Which is important because these vaccines seem to limit transmission itself. [1] https://www.cdc.gov/coronavirus/2019-ncov/science/science-briefs/fully-vaccinated-people.html https://www.cdc.gov/coronavirus/2019-ncov/science/science-br...

mfsch 5y ago

The main problem is that you need a lot more data to make confident statements about severe cases than just about symptomatic cases. That 100% number of the Pfizer trial had a huge uncertainty interval. So even though severe cases are what you care most about, you usually don’t want to wait for a study that gives you good data on those and measure (and compare) symptomatic cases instead. Zeynep Tufekci discussed this issue of study endpoints in one of her recent articles: https://www.theinsight.org/p/vaccine-efficacy-statistical-power https://www.theinsight.org/p/vaccine-efficacy-statistical-po...

dataflow 5y ago
But couldn't they just tell people about the uncertainty? (like "90-98% effective"?) Or underestimate the efficacy (in some objective way, like "at least 90% effective"). Or something else, as long as they're ultimately measuring what matters and what people actually care about. It's not like that 100% with their sample sizes would realistically ever suddenly turn into 50%, would it? It seems just inherently wrong to measure the wrong thing and convey it with certainty than to measure the right thing and convey it with the associated uncertainty, does it not?
- mfsch 5y ago
  They actually had a table on the efficacy for severe cases in the supplemental material of the original Pfizer study (table S5 in [1]), but all it shows is that the study doesn’t really give any information on this. The 95% confidence interval given for severe cases after the first vaccine is 20.1–99.7%. From what I understand, it’s really a trade-off between waiting for proper data about the efficacy you’re most interested in, or starting to vaccinate people as soon as you have efficacy data for an endpoint that’s still meaningful but easier to reach. Having much fewer symptomatic cases is already a pretty good result, and I think it was expected that the protection would also extend to severe cases. [1]: https://www.nejm.org/doi/suppl/10.1056/NEJMoa2034577/suppl_file/nejmoa2034577_appendix.pdf https://www.nejm.org/doi/suppl/10.1056/NEJMoa2034577/suppl_f...

chiefalchemist 5y ago

How much data (read: jabs) is necessary to evaluate the ability to prevent hospitalizations when even without the jab only a small percentage of cases lead to hospitalization? How much care is given to ensure the jab group's demographics reflect the "control" (i.e., the broader population). For example, ppl getting chemo aren't given the jab, are they? Yet they are counted if they get the virus, end up in hospital, etc. The point being, extrene Covid preys on the (immune) weak. Yet the jabs aren't tested on ppl who profile as such. How is that accounted for in evaluating a vaccine?

dataflow 5y ago
I'm a layman so I don't know if I can really give a satisfying answer to your question. All I know is what I see just doesn't make sense to me. Even in April 2020 we had rough but useful numbers on the hospitalization rate. Just some quick Googling turns up e.g. [1] which says that around 20% of those who test positive end up hospitalized. Now the Pfizer vaccine by itself had like 40k participants, with 9 severe (out of 162 confirmed cases) of COVID-19 in the placebo group, and 1 severe case (out of 8 confirmed cases) in the vaccine group. [2] [3] So right off the bat it looked pretty promising, and IMHO they could've just cited these numbers and let people judge for themselves... it's not like people don't have the intuition to realize that 1 patient couldn't have been 2-3 instead. But even if you think that wasn't enough data to be convincing, at least now we have mass vaccination rates they could be citing. Like literally ~40k people have been hospitalized with COVID-19 in just the last week and nearly all COVID deaths are among the unvaccinated now, and it's not like the vaccine just came out last week. An Associated Press analysis of available government data from May shows that “breakthrough” infections in fully vaccinated people accounted for fewer than 1,200 of more than 853,000 COVID-19 hospitalizations. That’s about 0.1%. [4] How is that still not enough data to switch metrics at this point? Why do they still think efficacy rates are more relevant for people instead? [1] https://www.tfah.org/wp-content/uploads/2020/04/COVIDunderlyingconditions040320.pdf#page=2 https://www.tfah.org/wp-content/uploads/2020/04/COVIDunderly... [2] https://www.pfizer.com/news/press-release/press-release-detail/pfizer-and-biontech-conclude-phase-3-study-covid-19-vaccine https://www.pfizer.com/news/press-release/press-release-deta... [3] I don't know why this reporting differs from the 100% I see claimed elsewhere; 1 severe case doesn't sound like 100% effective at preventing severe cases? I don't know if they're discussing different trial phases or have different thresholds for "severe" or something else... but regardless, it's irrelevant to my point. [4] https://apnews.com/article/coronavirus-pandemic-health-941fcf43d9731c76c16e7354f5d5e187 https://apnews.com/article/coronavirus-pandemic-health-941fc...