COVID-19: Vaccine and testing procedures Megathread Part 2 [Mod Warning - Post #1]

Cork2021

ACitizenErased wrote: »

Must turn on the South African news

Ah yes. Must sky plus it

latency89

caveat emptor wrote: »

No worries. Good to know ACE knows more than the nation of South Africa. Thanks for keeping us safe. These scientists and countries haven't a clue!

I don't know how the mods let them get away with it tbh

I had the pleasure of isthatso in the frozen covid origin thread, good grief lad, he wouldn't know a science book if it hit him in the head and he's going around to each thread citing studies, writing off treatments as no medical basis and backing up his educated friends here, it's a funny forum this.

I like the covid thread on merdic, you need an actual field qualification before you can post and I can tell you alot of them are fearful of these variant's, they respect mother nature and know what it can do.

In here, its ridiculous how confident they are, "it is not so much that I have confidence in scientists being right, but that I have so much in nonscientists being wrong"

Apogee

NPHET also mentioned that they are now sequencing ~10% of samples to monitor variants. They expect the sequencing rate to increase further.

ACitizenErased

Apogee wrote: »

NPHET also mentioned that they are now sequencing ~10% of samples to monitor variants. They expect the sequencing rate to increase further.

and that includes all tests of people who travel from SA and Brazil. Well on top of it to be fair to them.

Frank Bullitt

ACitizenErased wrote: »

and that includes all tests of people who travel from SA and Brazil. Well on top of it to be fair to them.

That kind of compliment will get you 50 whips at dawn.

Danzy

caveat emptor wrote: »

No worries. Good to know ACE knows more than the nation of South Africa. Thanks for keeping us safe. These scientists and countries haven't a clue!

How did a country like that produce someone like Elon Musk?

There are questions about South Africa's decision.

Turtwig

caveat emptor wrote: »

No worries. Good to know ACE knows more than the nation of South Africa. Thanks for keeping us safe. These scientists and countries haven't a clue!

How did a country like that produce someone like Elon Musk?

What would you consider the limitations of their study then? Surely you do not think it was perfect?

caveat emptor

Turtwig wrote: »

What would you consider the limitations of their study then? Surely you do not think it was perfect?

Of course sample size could be bigger. The attack rate was similar in both the seropositive (previous infection old strain) and the seronegatives even considering the confidence interval which wasn't that wide. i.e good indicator that reinfection with the new strain can occur REGULARLY.

But seriously. 26 people who were proven to have had a previous infection out of a sample size of 674 were REINFECTED.

This is separate to the other trial that showed the AZ vaccine didn't work against the new strain.

Listen people are free to believe what they want.

The world is full of people who don't want to listen to something that challenges their world view. Most people told them selves reinfections are rare and can't occur and here's some very hard data on it.

Situation has changed and the actions that countries are taking should tell you that. Not some randomer on the internet.

timsey tiger

Turtwig wrote: »

What would you consider the limitations of their study then? Surely you do not think it was perfect?

I was under the impression that the study was fine, co authored by Oxford. The problem is with tools taking a point estimate in a large CI as being something other than a very vague estimate.

caveat emptor

caveat emptor wrote: »

Of course sample size could be bigger. The attack rate was similar in both the seropositive (previous infection old strain) and the seronegatives even considering the confidence interval which wasn't that wide. i.e good indicator that infection with the new strain can occur REGULARLY with the new strain.

But seriously. 26 people who were proven to have had a previous infection out of a sample size of 674 were REINFECTED.

This is separate to the other trial that showed the AZ vaccine didn't work against the new strain.

Listen people are free to believe what they want.

The world is full of people who don't want to listen to something that challenges their world view. Most people told them selves reinfections are rare and can't occur and here's some very hard data on it.

Situation has changed and the actions that countries are taking should tell you that. Not some randomer on the internet.

“I don’t know what a confidence interval is”

https://www.statnews.com/2021/02/07/south-africa-halts-rollout-of-astrazenecas-covid-19-vaccine-after-shot-falters-against-variant/

Those data appear unreliable, however. They were given with confidence intervals, which propose a range of plausible outcomes. For the 22% number, those ranged from -50% to 60%, meaning that more data would be needed to be collected to trust the figure.

You do know that the CI on that study is so wide that there is a substantial chance that if you take those results at face value AZ make it 50% worse. In other words the study is extremely limited

latency89

[Deleted User] wrote: »

“I don’t know what a confidence interval is”

https://www.statnews.com/2021/02/07/south-africa-halts-rollout-of-astrazenecas-covid-19-vaccine-after-shot-falters-against-variant/



You do know that the CI on that study is so wide that there is a substantial chance that if you take those results at face value AZ make it 50% worse. In other words the study is extremely limited

Why did South Africa stop using it then?

latency89

latency89 wrote: »

Why did South Africa stop using it then?

Because they decided they wanted more data

latency89

[Deleted User] wrote: »

Because they decided they wanted more data

Why if it went through trials?

In that country infact by the manufacturer

pconn062

ACitizenErased wrote: »

NPHET believe we're starting to see an effect in hospitals from HCW vaccination

Until they start linking that type of information into what it means regarding opening up the country, it's good news but rather pointless on it's own.

latency89

latency89 wrote: »

Why if it went through trials?

In that country infact by the manufacturer

-50 to +60%. That’s why they want more data.

Melanchthon

[Deleted User] wrote: »

“I don’t know what a confidence interval is”

https://www.statnews.com/2021/02/07/south-africa-halts-rollout-of-astrazenecas-covid-19-vaccine-after-shot-falters-against-variant/



You do know that the CI on that study is so wide that there is a substantial chance that if you take those results at face value AZ make it 50% worse. In other words the study is extremely limited

Haven't read the study but what's the confidence interval on re-infection with the South African strain?

latency89

[Deleted User] wrote: »

-50 to +60%. That’s why they want more data.

So that data is good enough for EMA but not South Africa?

Why so?

latency89

latency89 wrote: »

So that data is good enough for EMA but not South Africa?

Why so?

Two completely separate things.

hmmm

latency89 wrote: »

So that data is good enough for EMA but not South Africa?

Why so?

The EMA aren't going to make a major decision like rejecting a vaccine off of a single study of 1500 30-year olds. I don't know what the South Africans are thinking.

Melanchthon

Melanchthon wrote: »

Haven't read the study but what's the confidence interval on re-infection with the South African strain?

Based on 42 cases of Covid there is 95% confidence that the AZ vaccine, when faced with the SA variant, will result in somewhere between 60% less and 50% more cases that in an unvaccinated person. That is why so many here claim the study is worthless. It doesn’t tell us anything other than more data is required. The number of subjects enrolled in such a trial is not important. It’s the number of cases that result in the overall population in the study, placebo and vaccine . That way you can compare the two proportions. When there are only 42 cases you get a very wide CI. The easiest way to conduct such a study to determine efficacy is not to enroll thousands of subjects, it’s to deliberately expose a couple of hundred. There are ethical problems around that however.

caveat emptor

Deleted User wrote: »

“I don’t know what a confidence interval is”

https://www.statnews.com/2021/02/07/south-africa-halts-rollout-of-astrazenecas-covid-19-vaccine-after-shot-falters-against-variant/



You do know that the CI on that study is so wide that there is a substantial chance that if you take those results at face value AZ make it 50% worse. In other words the study is extremely limited

I get what you are saying but pre October 31st i.e pre variant there was a clear effect / difference albeit with wide confidence interval also. Why would S.A stop using it though. I get trying to say a point estimate is pointless with such wide confidence interval but it's strange. Have you seen the other evidence with the neutralising assays. There's quite a bit of circumstantial evidence now. I stress circumstantial but none of the results of the trials had super tight confidence intervals.





In the original trial with the different dosing there were two different dosing regimes. The Low dose / standard dose was actually better in the trial. Why was that discounted when it had a better efficacy and similar confidence interval?

https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)32661-1/fulltext

Hmmzis

Didn't SA decide to roll out AZ anyway after the first panic reaction wore off, and just keep a close eye on hospitalization rates?

hmmm

caveat emptor wrote: »

I get what you are saying but pre October 31st i.e pre variant there was a clear effect / difference albeit with wide confidence interval also. Why would S.A stop using it though. I get trying to say a point estimate is pointless with such wide confidence interval but it's strange. Have you seen the other evidence with the neutralising assays. There's quite a bit of circumstantial evidence now. I stress circumstantial but none of the results of the trials had super tight confidence intervals.

I get what you're saying, but in South Africa the number of hospitalisations and cases is collapsing.

If they were shooting upwards you'd be thinking "oh-oh", particularly as some claim the south-african strain is reinfecting previously infected people.

The evidence does not match the doomsday predictions.

latency89

hmmm wrote: »

The EMA aren't going to make a major decision like rejecting a vaccine off of a single study of 1500 30-year olds. I don't know what the South Africans are thinking.

AZ vaccine trial we are talking about here hmmm

It took place in South Africa, UK,Brazil, India, Japan with 30,000+ people

One EMA approved based on its data, which wasn't good enough for South Afica according to one poster

Your thinking of the university fied trial in South Africa with that 1500 30-year olds, it was 2000 btw

latency89

hmmm wrote: »

I get what you're saying, but in South Africa the number of hospitalisations and cases is collapsing.

If they were shooting upwards you'd be thinking "oh-oh", particularly as some claim the south-african strain is reinfecting previously infected people.

The evidence does not match the doomsday predictions.

Its summer there, virus is somewhat seasonal based on data from Europe

caveat emptor

caveat emptor wrote: »

I get what you are saying but pre October 31st i.e pre variant there was a clear effect / difference albeit with wide confidence interval also. Why would S.A stop using it though. I get trying to say a point estimate is pointless with such wide confidence interval but it's strange. Have you seen the other evidence with the neutralising assays. There's quite a bit of circumstantial evidence now. I stress circumstantial but none of the results of the trials had super tight confidence intervals.





In the original trial with the different dosing there were two different dosing regimes. The Low dose / standard dose was actually better in the trial. Why was that discounted when it had a better efficacy and similar confidence interval?

https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)32661-1/fulltext

On your LD SD question. Because the populations were different

Hmmzis

caveat emptor wrote: »

I get what you are saying but pre October 31st i.e pre variant there was a clear effect / difference albeit with wide confidence interval also. Why would S.A stop using it though. I get trying to say a point estimate is pointless with such wide confidence interval but it's strange. Have you seen the other evidence with the neutralising assays. There's quite a bit of circumstantial evidence now. I stress circumstantial but none of the results of the trials had super tight confidence intervals.





In the original trial with the different dosing there were two different dosing regimes. The Low dose / standard dose was actually better in the trial. Why was that discounted when it had a better efficacy and better confidence interval?

https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)32661-1/fulltext

LD/SD was confounded by the dosing interval being 12+ weeks. The more you look into the data presented in The Lancet the less clear the picture becomes.

Btw. Regarding the seropositive attack rate, the same thing was noticed in the Pfizer/BNT phase 3 trial, but it's not reflected in wider population based studies. It's in the FDA submission document if you're interested in more details.

latency89

Hmmzis wrote: »

Didn't SA decide to roll out AZ anyway after the first panic reaction wore off, and just keep a close eye on hospitalization rates?

Selling it apparently

https://businesstech.co.za/news/government/467045/south-africa-considers-selling-unused-astrazeneca-vaccines-to-other-countries/amp/

Starting to vaccinate with J&J next week, not even approved yet for use

They don't need all the data on that one, weird.

caveat emptor

hmmm wrote: »

I get what you're saying, but in South Africa the number of hospitalisations and cases is collapsing.

If they were shooting upwards you'd be thinking "oh-oh", particularly as some claim the south-african strain is reinfecting previously infected people.

The evidence does not match the doomsday predictions.

No doomsday intended. This is science. Leave your positive / negative biases at the door science. It's about understanding and it's hard.


This is a fact. In the novovax trial 26 / 674 who were seropositive for a previous infection also ended up getting reinfected.

versus

58 who tested positive out of 1494 who were seronegative.

That's gives a point estimate attack rate of 3.9% in both groups but more importantly a confidence interval of [3.0-5.0] vs [2.5-5.6]

i.e nothing separating them. They are not insignificant sample sizes.

Please discus. It's kinda important.

edit: Saw reply above thanks Hmmzis.

Head well and truly wrecked on this one. Going to bed soon. Not doom mongering just want to understand.

latency89

latency89 wrote: »

Its summer there, virus is somewhat seasonal based on data from Europe

Started rising in November in South Africa, surged December, peaked early January. Would that not be summer?