FDA Releases More Documentation on Pfizer Vaccine

[sesquiterpene]

The FDA released briefing documents for it's Dec. 10 meeting to discuss EUA for the Pfizer Covid19 vaccine, 55 pages for those wanting to look in detail at the ongoing study.
VRBPAC December 10, 2020 Meeting Announcement

Here's the money shot, vaccinated in blue, placebo in red. Note that the lines begin to diverge ~14 days. The booster shot is at 21 days, indicating some efficacy with only one shot. The primary efficacy endpoint starts counting cases on day 28.

 

[Hans Blaster]

I'm not familiar with the analyses used here, but it looks like for the first 10-12 days after injection the subjects in both groups get infected at the same rate (no effectiveness) and then around that time it "kicks in" and blocks most infections. (After day 14 the placebo group gets about 21 "units" of infection and the vaccinated group about 1, thus the effectiveness of ~95%.) I don't see error bars or a likelihood region, so the certainty level is hard to assess.

It's too bad they didn't do a third group that didn't get the booster, perhaps it was needed in the clinical setting. (I'm assuming there was some sort of antibody activity drop that they thought the needed to avoid to provide protection.)

 

[sesquiterpene]

(After day 14 the placebo group gets about 21 "units" of infection and the vaccinated group about 1, thus the effectiveness of ~95%.) I don't see error bars or a likelihood region, so the certainty level is hard to assess.

The total efficacy reported is 94.6%, with a 95% confidence interval of 89.6 to 97.6 The study is continuing, so that is just a snapshot of the efficacy at one time point.

 

[sesquiterpene]

It's too bad they didn't do a third group that didn't get the booster, perhaps it was needed in the clinical setting. (I'm assuming there was some sort of antibody activity drop that they thought the needed to avoid to provide protection.)

It's easy to read a little too much into the early data points. The infection rate was probably lower when they first started the study. The late data points are also skewed - not everybody in the trial had reached day 113. I believe the FDA told them they had to wait until at least 50% of the study had reached two months after the booster, so ~ day 90 on the graph only represents half of the total participants.

The timing of the Moderna study is different - they wait 28 days for the booster. They will have a slightly better idea of the one-shot efficacy.

Neither company had a control group taking only the one shot. I think part of it was designing a protocol with the largest dosages of vaccine they thought would be well tolerated, and with the greatest chance of good efficacy.

 

[Hans Blaster]

It's easy to read a little too much into the early data points. The infection rate was probably lower when they first started the study. The late data points are also skewed - not everybody in the trial had reached day 113. I believe the FDA told them they had to wait until at least 50% of the study had reached two months after the booster, so ~ day 90 on the graph only represents half of the total participants.

The timing of the Moderna study is different - they wait 28 days for the booster. They will have a slightly better idea of the one-shot efficacy.

Neither company had a control group taking only the one shot. I think part of it was designing a protocol with the largest dosages of vaccine they thought would be well tolerated, and with the greatest chance of good efficacy.

Yeah, that's a good point. It wasn't 20,000 people in each group getting shots on the same day.

 

[Kenny'sID]

There are probaly a few of us who get nothing out of the graph, so this is good news? Could we get a simple explanation why please?

 

[sesquiterpene]

There are probaly a few of us who get nothing out of the graph, so this is good news? Could we get a simple explanation why please?

The graph is showing the number of people who got Covid over time since vaccination. As expected from the preliminary reports, a lot more people who got the placebo got sick. What is interesting is that it looks like just one shot (instead of the two they used) helped to prevent at least some disease. That would mean we could vaccinate twice as many people. Unfortunately, the data looks somewhat weak as to exactly how strong that protection will be, and it would be pretty nervy for the FDA to suggest people get only one shot. It will take some very careful monitoring once the vaccination campaign gets underway to make that call in time to make a large difference. Let's hope they can do it.

 

[Kenny'sID]

The graph is showing the number of people who got Covid over time since vaccination. As expected from the preliminary reports, a lot more people who got the placebo got sick. What is interesting is that it looks like just one shot (instead of the two they used) helped to prevent at least some disease. That would mean we could vaccinate twice as many people. Unfortunately, the data looks somewhat weak as to exactly how strong that protection will be, and it would be pretty nervy for the FDA to suggest people get only one shot. It will take some very careful monitoring once the vaccination campaign gets underway to make that call in time to make a large difference. Let's hope they can do it.

Thank you...very helpful.

 

[sesquiterpene]

View attachment 290222

For comparison, here's a similar figure for AstraZeneca/Oxford vaccine:

However, they don't show the data starting at day 1. Claimed efficacy was 70·4% (95% CI 54·8 to 80·6)
https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)32661-1/fulltext

 

[RestoreTheJoy]

The FDA released briefing documents for it's Dec. 10 meeting to discuss EUA for the Pfizer Covid19 vaccine, 55 pages for those wanting to look in detail at the ongoing study.
VRBPAC December 10, 2020 Meeting Announcement

Here's the money shot, vaccinated in blue, placebo in red. Note that the lines begin to diverge ~14 days. The booster shot is at 21 days, indicating some efficacy with only one shot. The primary efficacy endpoint starts counting cases on day 28.
View attachment 290222

Well, this is important information IF the subjects are similarly situated (and I don't know if they are or not from this). That matters.

If the group getting placebos are also primarily first responders/people out and involved in the community like delivery, food service, grocery, uber, etc, those engaged with lots of people, while those getting the actual vaccine are people who mostly stay at home and are low risk anyway, then the results aren't accurate.

 

[sesquiterpene]

Well, this is important information IF the subjects are similarly situated (and I don't know if they are or not from this). That matters.

If the group getting placebos are also primarily first responders/people out and involved in the community like delivery, food service, grocery, uber, etc, those engaged with lots of people, while those getting the actual vaccine are people who mostly stay at home and are low risk anyway, then the results aren't accurate.

That's why randomized, placebo-controlled studies are so important. The randomization of the study participants means that equal numbers of types of people end up in the placebo vs vaccine groups - even "types" that you didn't think were important are controlled.

 

[RestoreTheJoy]

That's why randomized, placebo-controlled studies are so important. The randomization of the study participants means that equal numbers of types of people end up in the placebo vs vaccine groups - even "types" that you didn't think were important are controlled.

If it is actually random. Even random picks can sometimes be very weighted.

 

[sesquiterpene]

If it is actually random. Even random picks can sometimes be very weighted.

All right, so you are clueless as to how clinical trials are actually conducted. Why double down on this nonsense? Try looking at all the data at the link I provided in the OP. Is there any category that doesn't look randomized to you?

 

[RestoreTheJoy]

All right, so you are clueless as to how clinical trials are actually conducted. Why double down on this nonsense? Try looking at all the data at the link I provided in the OP. Is there any category that doesn't look randomized to you?

So you cannot recognize the logic. That's a shame.

Since this appears to be aimed at the American people very shortly, we shall find out real results in a few years. Not everyone will react immediately, as did the two who had anaphylactic responses right away.