People are not trusting anything regarding this...
1. Instituto Butantã, claims they can't publish the results of the trial because of NDA with Sinovac, and won't comment about the effectiveness.
2. Meanwhile Sinovac is being tried for corruption in China, and their defense was actually confirm they paid bribes to officials in various countries to approve their vaccines (not Coronavirus ones, the trial is about 2011 stuff), but claim that the officials (including Instituto Butantã related officials) that used extortion, and thus they are victims, not perpetrators of the crime.
In this situation, when allegedly bribery is afoot, do you really think we should blindly trust the alleged criminals numbers?
> Meanwhile Sinovac is being tried for corruption in China, and their defense was actually confirm they paid bribes to officials in various countries to approve their vaccines (not Coronavirus ones, the trial is about 2011 stuff), but claim that the officials (including Instituto Butantã related officials) that used extortion, and thus they are victims, not perpetrators of the crime.
Where can I read about this? Would like to know more about it since the National Government here in the Philippines will use the one created by Sinovac for vaccination and they're trying their best to convince people despite the public skepticism regarding the lower efficacy rates which seems to indicate what you've said.
"In 1985, Mr. Yin, then 21, was the first to isolate the hepatitis A virus, and in 2001, he founded Sinovac Biotech ... From 2002 to 2011, Mr. Yin paid $77,000 to a senior Chinese drug regulator, Yin Hongzhang, and his wife to approve his vaccines, according to a 2018 filing to the Securities and Exchange Commission, citing court documents."
Turkey claimed 90%+ but the numbers in their tests were tiny:
> The Turkish trial data for CoronaVac is based on analysis of 1,322 participants which included 29 infected people, and efficacy evaluations were made 14 days after the second dose was administered.
Plus I’m not sure comparing countries at that level is a good base line. There could be major demographic (vitamin D which has been shown to play a role has also been shown to be less prevalent in darker skin) and even the details of the trials (Turkeys itself usinf relatively small samples). Or even the quality of the healthcare system treating these people after they get sick.
From the same article:
> Experts say it’s not unusual for a vaccine to show different efficacy rates in various settings, as trial protocols, data size and population could impact results, but the way CoronaVac data has been released created some confusion.
Well, besides the numbers being hearsay, this refers to "full efficiency" which is pretty much a meaningless number in this case. It refers to the number of people who had any kind of symptom manifestation
Now do you care that instead of dying you got a runny nose? It's a useful number for statistic purposes, not practical ones
Correct, you do care about sterilizing immunity, but you need to do PCR tests on everybody, I'm not sure it was the case there. That is not what this number is referring to though.
60% is pretty good for an inactivated full virus vaccine. That's about what the trivalent seasonal influenza vaccines (using the same general technique) achieve.
>That's about what the trivalent seasonal influenza vaccines (using the same general technique) achieve.
Doesn't the seasonal flu vaccine also have the issue that they have to "predict" what strains would be active? Is that already factored into the effectiveness rate?
Yes it is already factored in. Influenza vaccine efficacy is usually calculated separately for "matched" influenza and all influenza. 60% range is for matched.
Perhaps, but the last time I checked you needed around 60% of the population to be immune in order to halt the pandemic, so I'm not sure if 60% is good enough in this case.
At the rates at which we're going, 10-20% of the population is going to be immune by other means by the time everyone who wants it has been vaccinated.
It might work the other way - the immune system may get used to a given virus and begin ignoring it.
This is the way anti-allergy therapies work - you inject patients with a progressively higher amounts of the substance the person is used to and the immune system stops responding to it.
With vaccines, there have been cases of this effect (found during testing, and not with covid vaccines so far).
The first sentence of the article ends with "citing two people who seen the results." How does such a glaring grammatical error get past Reuters editors?
Does anyone know of a good resource that explains what 'efficacy' means for a vaccine, and what can (and cannot) be extrapolated from it? The reported efficacies are (from what I understand) 1 - (n infections in experimental group / n infections in control group). But then I see people draw wildly varying conclusiona from that, most of which I suspect are wrong (e.g. the relationship between 'efficacy rate' and 'group immunity' after everyone is vaccinated). But despite looking long and hard, I can't anyone explaining in detail what such efficacy rates mean (and don't mean).
I may screw this up, but efficacy describes how effective the vaccine was in the trial vs the control group. So if each group had an equal number of people, and 100 people in the control group got sick, but only 5 people who received the vaccine group got sick, the efficacy rate would be 95%.
The hope, of course, is that the vaccine's effectiveness (how well it does at preventing covid) is close to the efficacy, but it's very hard to know for sure. That's why it's best to have the trials spread out among demographics and even locations, to better try and identify what may cause the vaccine to be more or less effective.
Very good point. I read recently that there were no hospitalizations in the AstraZeneca vaccinated group, so by this metric it's 100% efficient. I really want to hear those numbers as well, it makes a huge difference in the decision to promote a certain vaccine.
> Even after removing cases occurring within 7 days of vaccination (409 on Pfizer’s vaccine vs. 287 on placebo), which should include the majority of symptoms due to short-term vaccine reactogenicity, vaccine efficacy remains low: 29% (see footnote).
The 95% claim assumes you've had both doses and five weeks have passed since the first dose. I don't think that's a fair way to dice the data. Also, "covid-19 symptoms" (the proxy that opinion writer used) are very non-specific. Headache, sore throat, cough - there's myriad causes. That's why PCR+ is the only way to get a decent signal.
This seems like a crackpot analysis, not that I'm an expert.
Oh, and finally, relative risk reduction (that number you're quoting) is a fundamentally different property than efficacy. It's the difference between the placebo and vaccine groups in catching covid. Relative risk reduction would only equal efficacy if 100% of the placebo group caught it.
Not everything you disagree with is a crackpot analysis.
>The BMJ is a weekly peer-reviewed medical trade journal, published by the trade union the British Medical Association (BMA). The BMJ has editorial freedom from the BMA. It is one of the world's oldest general medical journals. [1]
>Peter Doshi is an associate editor at said journal. He completed a fellowship in comparative effectiveness research at Johns Hopkins and received his Ph.D. in history, anthropology, and science, technology and society from the Massachusetts Institute of Technology. [2]
I could not quote the entire article, but as we all know, failing to quote any of it usually ends up with a majority not being aware of any content within the article.
The key finding is that further investigation is required into the substantial portion of people with COVID-like symptoms that occured in both the experimental and placebo groups, which has not been adequately analyzed.
The entire article addressed VE (measured by calculating the risk of disease among vaccinated and unvaccinated persons and determining the percentage reduction in risk of disease among vaccinated persons relative to unvaccinated persons) and showed the raw numbers which allows anyone else to reach the same calculation result.
You compared 95% (the Pfizer efficacy number) to 29% (the relative risk reduction number calculated by that article.) That is incorrect. You should acknowledge that.
I originally said it looks closer to 30% than 95% based on the limited data we have available, which anyone can do using the data in the article (and can even take a step further verify that these numbers came from the FDA's report).
Vaccine Effectiveness (as defined by the CDC [1]) is calculated by:
(Risk among unvaccinated group − risk among vaccinated group) / Risk among unvaccinated group
Vaccine participants: 18198
Placebo: 18325
Confirmed COVID on vaccine: 8
Confirmed COVID on placebo: 162
Suspected COVID on vaccine (excl. Within 7 days of dose): 1185
Suspected COVID on placebo (excl. Within 7 days of dose): 1407
It didn't show that. It showed 62%. 70% is a pooled result when you put the (too few) people into the calculation who got 90% efficacy, which scientists can't explain as that testing regimen was completely fucked up by Jenner and AZN (different dosage, different timing, different age group <55yr only). So it's hard to tell if it's really 90% (due to sample size) or WHY it might be more effective than the 62% number.
The approvals so far for the AZN/Oxford vaccine have been for the full-full dose regimen, which has shown to be 62% effective only.
Yes, actually. That’s right at the inflection point where each added percent effectiveness is maximally beneficial to herd immunity. E.g. the difference from 85% to 95% is much less important than 60% -> 70%.
If R0 is 2.5, then you need to vaccinate (1 - 1/2.5) / .95 of the population = 63% to reach herd immunity. For a 60% effective vaccine, you get (1 - 1/2.5) / .6 = 100% of the population.
Yes, the slope is slightly different but the rolloff is exceptionally gentle. There is a huge difference in practicality between even an 85% and 95% effective vaccine (70% of the population vs. 63% of the population).
Not to mention that none of what we've measured is differences in transmissibility; it's quite possible that a 70% effective-against-symptomatic-illness vaccine only mitigates transmission rates by 35%, and thus can't reach herd immunity at any level of vaccination.
Not impossible, just can't be done by vaccination alone. The remaining required percentage of the population would have to get immunity from getting sick.
This is what I was getting at in the grandparent post. 50% -> 60% and 85% -> 95% might both seem like they protect 10% more of the population, but the qualitative effect on the spread of the pandemic and be quite different. 60% might be enough to get herd immunity (with masks and social distancing), while 50% is not. Both 85% and 95% are definitely enough to get herd immunity, so to some extent that added protection matters less.
It's still not any kind of "inflection point". The difference between 0 and 5% is an even sharper slope-- from infinity to 1200% of the population ;)
Also-- all we know about is effects on symptomatic illness and we can't be sure there's any large transmission benefit (and if there is, it's probably smaller than the efficacy against symptomatic illness).
If the virus is still circulating in volume, the difference between a 95% effective vaccine and a 70% vaccine is a further 6x reduction in risk for the individuals that have it. Every little bit helps, but a really high efficacy is very nice to have.
Thought experiment: how much of the population do you think you need to immunize with a 0.1% effective vaccine to stop the virus from circulating?
I find it's helpful, when I see something counterintuitive to me in math, to consider the extreme cases. Obviously a 0.1% effective vaccine won't be able to attain herd immunity. Within a set of simplified assumptions, you need to stop 60% of the transmission to get fewer infected each generation from a R0=2.5 disease, and vaccines with less than 60% efficacy against transmission cannot stop 60% of the transmission themselves (though they certainly would still help).
Is this for the 1st shot only before the booster is administered as if recall correctly for the other vaccines it was around 75-80% for the first shot and went to 95%+ after the 2nd shot. Even if it reaches 75-80% after the 2nd shot it is actually good enough and that was the figure most of the medical professionals were hoping and expecting before the 90-95% effective vaccines came out.
Less than reputable brazilian news have come up with wildly different numbers, from 40% to 63.75% to 80%. This is incredibly unreliable.
How about not writing articles about rumors from Brazilian Yahoo News?
Least of all, reports that have been dismissed by the actual institute doing the trials.