Vaccine Effectiveness in Ireland

godzilla1989

https://www.boards.ie/discussion/comment/117802422#Comment_117802422

What's point continuing the trial if the control group is gone?

timmyntc

https://www.boards.ie/discussion/comment/117802422#Comment_117802422

"To collect efficacy and safety data"

You cannot measure efficacy or safety data without an adequate control group to compare against. If in 1 years time, theres 100 heart attacks observed - we cannot know if that is above the baseline or not if all the participants have been vaccinated.

The reality is there is no long term safety data available, and there will not be any because of the way this trial has been conducted. The push to get 100% vaccination coverage in a population wide scale also, means we cannot get accurate safety data either.

Unicorn Milk Latte

Percentage calculations are completely meaningless.

If you have 0% of the population vaccinated, like a year ago, and 500 people in hospitals, 100% are unvaccinated.

If you have 100% of the population vaccinated, and 5 people are in hospitals, 100% of people in hospitals are vaccinated.

Here are two studies/papers out of Israel (I already posted those in the Delta Variant thread..) that look at the actual numbers.

https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-vs-severe-disease-be-strong-when-60-of-hospitalized-are-vaccinated

if you have breakthrough infections, any factors that indicate a weakened immune system - being immunosuppressed, pre-existing conditions, and age - need to be considered for a realistic assessment.

https://www.clinicalmicrobiologyandinfection.com/article/S1198-743X%2821%2900367-0/fulltext

The results indicate that vaccines do exactly what they are supposed to do - reduce possibility of death and hospitalisation significantly - and that when you look at hospitalisation and death from break through infection, only 4% have no pre-existing conditions, and 40% are immunocompromised.

astrofool

https://www.boards.ie/discussion/comment/117802440#Comment_117802440

Because you don't need a control group to continue measuring efficacy (against various measures, e.g. antibody count, #Infected by age etc.).

https://www.boards.ie/discussion/comment/117802455#Comment_117802455

That's just not how medical trials are run, you can continue disagreeing, but it won't change what happens in the real world with medical trials.

The smaller the number of participants, the longer the trial takes to run to get to an agreed confidence interval, having a large amount of participants ensures a large amount of events can occur in a shorter space of time to reach the expected confidence interval that the medicine works or not. It's why the vaccines had such large numbers in the trials and why people need to think in terms of number of events and not linear time.

Pfizer are continuing to collect efficacy and safety data despite the control group now being vaccinated, are they just doing this for laughs?

lee_baby_simms

Is the concept of waning immunity from vaccination limited to pzifer or do we have similar observations with AZ or Janssen for example?

astrofool

https://www.boards.ie/discussion/comment/117802652#Comment_117802652

The antibody count drop is similar for AZ, Janssen is probably too new to have the full set of data yet. The antibody count drop is expected and the numbers seem to hold up quite well, but is an issue for those with underlying conditions or with a poorer immunity. We'll see if it becomes an issue for gen. pop. (but do expect to see lots of scary headlines extrapolating on the data).

timmyntc

https://www.boards.ie/discussion/comment/117802630#Comment_117802630

efficacy and safety figures are based on comparison with a control group. That is how they are defined.

Typical Phase 3 trials are several years long, to monitor efficacy over prolonged periods of time (antibody waning would be caught in these studies), as well as to monitor for any long-term safety concerns.

It is not standard practice to vaccinate your control group while a study is ongoing. This does not happen for any other vaccine trials in the past.

Confidence intervals are all well and good - but you cant estimate confidence interval for something that has not yet been observed - i.e. long term efficacy or side effects. And you cannot observe them if you do not have some control group to compare against. This is how scientific trials are conducted.

https://www.boards.ie/discussion/comment/117801150#Comment_117801150

All you have to do is point out where you think their analysis is wrong, as others have done, instead of posting a snotty reply. That just says you don't understand anything and are just trying to pretend to be smart.

astrofool

https://www.boards.ie/discussion/comment/117802776#Comment_117802776

That is not how medical trials are conducted, as you have already said, ethically, once you know you are saving or improving lives, then the control group also gets access to the treatment.

Is "ongoing phase 3 trials" a talking point you use?

Mr.StRiPe

https://www.boards.ie/discussion/comment/117800856#Comment_117800856

This is a very good point i.e. we know the vaccine breakthrough cases were admitted to hospital but not when they were admitted as the HSE haven't given this breakdown and dont consistentantly give us the data on how many are discharged from hospital each week as far as I can see. However, the average hospital stay seems to be holding at 6.5 days.

All we know from the data the HSE have released so far is on the 12th of August 49% of peope in hospital were breakthrough cases and on the 19th of August 45% of people in hospital were breakthough cases so the admissions each week on average will reflect this with an average turnover of 6.5 days but I agree we don't know the exact number.

Are you happy if I go with the rolling average % of people with breakthough cases in Ireland until the HSE gives us this data? i.e. 47% (49+45)/2

If not, what would you suggest is the most accurate way to reflect the admission % based on the data the HSE are giving us.

If we can agree on the most accurate way to reflect the admission % are you happy the spreadsheet accurately reflects the HSE vaccine protection model presented by professor Nolan?

Mr.StRiPe

https://www.boards.ie/discussion/comment/117801150#Comment_117801150

All I saw was some random person online that seems to have no expertise whatsoever and less grasp of how figures work trying to show that he/she is an" expert" (YET ANOTHER EFFIN ONLINE EXPERT)

All I did was input HSE data into a HSE model on vaccine protection so all of the figures are from the HSE

It is the HSE's expertise that created the model which was presented by professor Nolan on the 12th of August do you trust their expertise and figures?

For me, a quick glace showed that you really have not got a clue

I'd appreciate if you could highlight what part of my post you had an issue with.

As I've said previously I have attached the spreadsheet that represents the model professor Nolan presented for anyone to pull apart as I want to make sure it is accurate.

Mr.StRiPe

https://www.boards.ie/discussion/comment/117802493#Comment_117802493

Percentage calculations are completely meaningless.

If you have 0% of the population vaccinated, like a year ago, and 500 people in hospitals, 100% are unvaccinated.

If you have 100% of the population vaccinated, and 5 people are in hospitals, 100% of people in hospitals are vaccinated.

When you have 0% of the population vaccinated the % of hospital admissions for covid is the baseline you measure the effectiveness of the vaccine against.

If you look at the HSE vaccine protection model their assumption on risk of hospitalisation in an unvaccinated case is 5% this means if the vaccine is 0% effective you would still have 5% risk of hospitalisation if the vaccine is 100% you would have 0% risk of hospitalisation i.e. the closer the rate gets to 0% afther vaccination the greater the effectiveness of the vaccine.

This is my understanding anyway.

Unicorn Milk Latte

https://www.boards.ie/discussion/comment/117802652#Comment_117802652

I think it's probably a misconception to see the reason of possible waning immunity in the vaccine, rather than the virus itself.

When you look at research of coronaviruses pre-Covid, like coronaviruses that cause common cold, you will find that when a person got naturally infected and therefore produced antibodies, these antibodies gradually disappear over a time span of around 5 years. Unlike some other viruses, where antibodies produced after an infection remain in the human body indefinitely.

In other words, if there is waning immunity with SARS-COV-2, it is most likely identical between vaccinated people, and people who got infected with SARS-COV-2 and produced antibodies that way.

Unicorn Milk Latte

https://www.boards.ie/discussion/comment/117803237#Comment_117803237

But I'm not talking about the 'risk of hospitalisation' percentage of a vaccine - in that case, having percentage numbers makes sense, of course.

I'm talking about drawing conclusions about the efficacy of vaccines from 'x% of hospitalised are vaccinated', because that is misleading when the number of hospitalised overall is small, while the number of overall vaccinated in the population is high.

Pete_Cavan

OP, the figures you keep quoting are not what you claim them to be. The figure of 49% which you say is Covid patients in hospital fully vaccinated clearly includes partially vaccinated people. Some of those who are fully vaccinated could also be within two weeks of their second dose so outside the parameters set by Prof. Nolan.

The figure of 45% Covid patients in hospital fully vaccinated also includes those who caught Covid within two weeks of when they received their second dose.

Neither of those figures are what Prof. Nolan's spreadsheet was intended to measure so it should be no surprise that he didn't come up with either. I understand that we don't have the exact % of Covid patients in hospital having received their second dose more than two weeks previously, but that doesn't mean you can just pick a different number and throw it in instead.

Mr.StRiPe

https://www.boards.ie/discussion/comment/117803403#Comment_117803403

Some of those who are fully vaccinated could also be within two weeks of their second dose so outside the parameters set by Prof. Nolan.

The HSE define fully vaccinated as follows:

I believe it is clear when the HSE refer to fully vaccinated in their data they are referring to same definition as professor Nolan which does not include people who haven't reached the recommented time period after getting their second shot.

The figure of 49% which you say is Covid patients in hospital fully vaccinated clearly includes partially vaccinated people.

You appear to agree with the 45% figure as Paul Reid explicity says it on August 19th? he also goes on to say 52% not vaccinated and 3% no status he does not mention any figure for partially vaccinated but as the other figures add up to 100% we know there were none.

On both the 12th and the 19th he breaks the ICU data into fully and partially vaccinated and we know there were no partially vaccinated reported in the hospitilisation breakthroughs on the 19th which was the following week. Do you not think he would have reported the partially vaccinated figure on the 12th if he had it? Not reporting the partially vaccinated figure, if it was available, would be very misleading to the public so it is hard to believe he would choose to do this. Surely you can't believe Paul Reid choose to omit it? and if he omitted it in error I'm sure they would have corrected the record once they realised.

Even if you belive he did deliberately omit it hopefully the clarification on the HSE's definition of fully vaccinated addresses your concerns that the figures I quoted do not include those who caught Covid within two weeks of when they received their second dose? If so, then you shouldn't have any issue with the examples when the spreadsheet is set to achieve the 45% breakthough.

yoke

I think it’s important to note that mr.stripe is not actually saying that the vaccines are useless, I think he’s more making the point that the manufacturers claims (like most manufacturers claims) are overly optimistic, based on the data available.

The problem with relying purely on overly-optimistic manufacturer’s claims is that it can lead to policy errors down the line, ie. if the vaccine is 93% effective at preventing something, then you might implement different policies than if it was only 80% effective.

My initial reaction was also “his figures are probably fudged”, but actually on closer inspection it looks like international consensus may be moving towards reviewing exactly how efficient the vaccines are, as it seems the manufacturer claims may have been overly optimistic (which is usually the case with all manufacturers claims, to be fair).

eg: https://www.yalemedicine.org/news/covid-19-vaccine-comparison

Lumen

@yoke no, Mr.StRiPe is claiming that Irish data fed into his spreadsheet model shows that the vaccines are significantly worse at preventing serious disease than has been demonstrated in Israel.

The article you linked to says

Another study, not yet peer-reviewed, provided more new data that brought the efficacy number down to 84% after 6 months, although efficacy against severe disease was 97%.

Pete_Cavan

https://www.boards.ie/discussion/comment/117804395#Comment_117804395

I don't think "overly-optimistic manufacturer’s claims" is an accurate or fair way of putting it. They gave the data they got from their trials and that was reviewed by the regulators. It is expected that real world experience can differ from trial results.

Also the earlier efficacy results which we keep going back to were a snapshot at that moment in time. Efficacy can change over time, that seems to be an established fact now with these Covid vaccines. I don't think anybody is "relying purely" on manufacturer’s claims, there are lots of studies ongoing at the minute.

Going back to the OPs calculations, there are lots of flaws in it. It assumes one set rate for vaccine effectiveness, but in reality there are likely several rates. These range from the high manufacturer figure for those recently fully vaccinated to lower rates for people vaccinated earlier in the year. Then you have to factor in the demographic and general health differences between those vaccinated recently and those vaccinated 6+ months ago (i.e. that the most likely people to be hospitalisated currently have the lowest protection of the people fully vaccinated).

Obviously these issues are built into the spreadsheet which the OP didn't create. However, I don't think Prof. Nolan ever intended the spreadsheet be used for the purposes the OP is using it for (i.e. calculating actual vaccine effectiveness from a few broad figures). I think Prof. Nolan was just giving a high level example of how vaccines can prevent hospitalisations based on some assumptions. I'msure he would be the first to acknowledge that the assumptions don't account for several factors.

Cuddlesworth

https://www.boards.ie/discussion/comment/117804395#Comment_117804395

He is doing two things.

He is trying to extrapolate a result from the data that has been provided which validates his viewpoint.

In which is a attempt to insinuate that the given figures are wrong and therefore, create a sense of distrust in the the parties involved, government, health officials, companies that manufactured the vaccines etc etc.

The other poster is doing the same thing, the trials are still ongoing, there is no control group, it/they can't be trusted. Please start distrusting things so that my own viewpoint can be validated, while ignoring the clear and obvious fact that the vaccines clearly are working globally.

https://www.boards.ie/discussion/comment/117803142#Comment_117803142

And this word vomit is a good example of this. He is just repeating the same things over and over again to try get back to his biased conclusion that the vaccines can't be working as well as they say it is.

The figures are in, vaccines are working. Multiple organisations across the world have provided this independently of each other. And if the figures show booster shots help with that(and it appears that they do), then boosters will become available.

Mr.StRiPe

https://www.boards.ie/discussion/comment/117804858#Comment_117804858

He is trying to extrapolate a result from the data that has been provided which validates his viewpoint.

What caught my attention was when the first vaccine breakthough data released by the HSE on ICU figures

on August 12th they were:

• 24% Of those in ICU were full vaccinated
• 18% of those in ICU were partially vaccinated

and on August 19th they were:

• 23% Of those in ICU were full vaccinated
• 15% of those in ICU were partially vaccinated

This struck me as very high as there is no doubt these are classed as serious illness whatever about ending up in hospital especially if the quoted vaccine efficacy rates are 93%

Then I saw professor Nolans vaccine protection model where the point he is making is sound (in my view) but I noted the Vaccine effectiviness in preventing symtomatic infection assumption for the model was 80% which I thought might be a bit on the high side based on the data out of Israel which put this at an average of 39% as they found it started to wane down to 16% over a period of 6 momths. We are also at roughly 6 months from when the most vunerable in our society were vaccinated.

However, the data out of Isreal still reported the more important effectiviness measure vaccine effectiveness in preventing sever disease requiring hospitalisation was still very high at 91%. I found this very reassuring as a fully vaccinated person myself.

My initial thought was to see if putting the 39% Vaccine effectiviness in preventing symtomatic infection assumption into the HSE model might help to explain the very high (in my opinion) breakthough numbers in ICU.

However, when I went to do this I noted that the model returned an expected hospital breakthrough admission rate of 14% yet the data released by the HSE was a 49% breakthough rate in hospital which is not an insignifficant difference.

I can tell from your responses that you have actually read my post and understand this point hence you raised a valid point which I would like to flesh out with you as I have no interest in not being as accurate as possible with the data we have.

And this word vomit is a good example of this. He is just repeating the same things over and over again to try get back to his biased conclusion that the vaccines can't be working as well as they say it is.

The reason I am repeating the same things over and over again is people are posting comments when they appear not to have read and/or understand the point of my post which is based on a very straightforward model released by the HSE where the only input parameter we are discussing is the breakthough % in hospital that the HSE also released. So as there is only one variable this is why it is being repeat over and over again.

What I find strange is you understand what I am trying to do as you have actually read the post however, it appears your own bias is preventing you from having a reasonable discussion with me on how to make the data as accurate as possible.

In which is a attempt to insinuate that the given figures are wrong and therefore, create a sense of distrust in the the parties involved, government, health officials, companies that manufactured the vaccines etc etc.

This is your bias and maybe your right if the real Irish data shows the efficacy is signifficantly lower in Ireland less people will choose to get vaccinated or take the booster dose or whatever but maybe you are wrong too. I believe everyone should have the right to make their own minds up on how best to protect themselves for example if the efficacy rates are signifficantly lower some people might choose not to engage in certain social interactions despite the restrictions being lifted.

The figures are in, vaccines are working. Multiple organisations across the world have provided this independently of each other. And if the figures show booster shots help with that(and it appears that they do), then boosters will become available.

I agree and have never said otherwise hence why I am fully vaccinated myself.

This issue is if the efficacy % is signifficantly lower it might encourage action such as trying to establish the antibody responses of people in nursing homes to get an indication of how big an issue this might be and inform a strategy to best protect them before its to late. Also even offering the option to everyone giving blood, for what ever reason, if they would like their antibody levels checked for free would be great data and be a much more scientific approach to try and validate the efficacy % which is in everyone interest.

We also have the situation where the group at the highest risk from getting infected my parents and either your parents and/or your grandparents, depending on your age, who have been couped up for the last year are getting on planes to travel within the EU where the majority of people on the plane (the vaccinated) can board based on a vaccine cert and no requirement of showing a recent negative test result.

I believe if people in this high risk group in terms of age thought the efficacy rate might be 39% and not 80-90% they might choose not to board the plane and if it turns out 6 months on the efficacy could have waned to 16% I believe even less of them would board the plane.

This is why I think your bias is flawed and arrogant as it not your call what people choose to do to protect themselves based on the most accurate data available.

schmoo2k

https://www.boards.ie/discussion/comment/117805715#Comment_117805715

Israel had hit a vaccine acceptance wall at about 60% though, it wasn't until Delta showed up in Israel that the uptake started to increase again:

Confirmed Cases v Deaths - Aug 24, 2021 / Gordon Smith / Observable (observablehq.com)

Mr.StRiPe

https://www.boards.ie/discussion/comment/117804395#Comment_117804395

I think it’s important to note that mr.stripe is not actually saying that the vaccines are useless

Thanks, this is correct and it appears that it is peoples perception that I am saying this that is preventing them from actually reading the post.

My initial reaction was also “his figures are probably fudged”, but actually on closer inspection it looks like international consensus may be moving towards reviewing exactly how efficient the vaccines are

I believe if most people get passed their perception that I must be anti-vax or whatever they believe I am and read the post they will understand it and then we can have a real constructive discussion.

What I didn't expect is when they do read and understand the post which might indicate the data results in something their own bias doesn't like they start to attack me personally rather than show where I have misrepresented either the HSE data and/or the HSE model.

Mr.StRiPe

https://www.boards.ie/discussion/comment/117805881#Comment_117805881

I don't follow what this has to do with the irish data and irish model released on August 12th which is clearly based on Delta data as I believe 99% of cases are now Delta and have been for some while.

Cuddlesworth

https://www.boards.ie/discussion/comment/117805715#Comment_117805715

Word vomit again. You have been told why your figure of 49% is wrong, yet you continually keep referencing it in a attempt to validate it. Its not that difficult, there is a difference between the number of people admitted and the number of people in care.

It's a very reasonable assumption that those in their later years would need significantly longer stays and would also at higher risk of hospital admittance, even with vaccination. Which would skew the numbers of those "in" hospital care towards a higher vaccination rate, since they have a significantly higher vaccination rate.

Also the HSE model is not a basis for current figures but a projection for future results. You are quite literally reading current figures, trying to change the results to suit your argument, all of which is based on a projected model which factors in the known current figures. If they state 80%, its a good starting point since they have access to the real world data.

I also doubt your actually vaccinated, seems like your playing the "I'm on your side but..." card. You bring it up to often to justify your "side" of the argument.

Mr.StRiPe

https://www.boards.ie/discussion/comment/117806334#Comment_117806334

I also doubt your actually vaccinated, seems like your playing the "I'm on your side but..." card. You bring it up to often to justify your "side" of the argument.

With 80%+ of the population vaccinated means you would be 80%+ wrong to doubt me with no other information to go on.

It doesn't make any difference to me if you are vaccinated or not as it has nothing to do with my OP!

Also the HSE model is not a basis for current figures but a projection for future results. You are quite literally reading current figures, trying to change the results to suit your argument, all of which is based on a projected model which factors in the known current figures. If they state 80%, its a good starting point since they have access to the real world data.

Wow, it seems I overestimated your understanding of the model so lets take it back a bit.

Do you agree the virus protection model presented by professor Nolan is a mathematical model that outputs the expected infection and admission % for both vaccinated and unvaccinated to demonstrate the benifits of vaccination when you enter the following inputs?

1. Background force of infection in the unvaccinated population
2. Vaccine effectiviness in preventing symtomatic infection
3. Vaccine effectiveness in preventing sever disease requiring hospitalisation
4. Risk of hospitalisation in an unvaccinated
5. % of the population vaccinated

Lumen

@Mr.StRiPe I do understand your point. It doesn't actually require a spreadsheet, just a comparison of the stats modelled in the tweet to the stats of actual split between vaxxed and unvaxxed.

I think, though, you've drawn the wrong conclusion.

Rather than the vaccine effectiveness being lower than in the model, the vaccine coverage is higher than in the model, amongst the most vulnerable who are most at risk of hospitalisation.

The higher the vaccine coverage, the greater proportion of hospitalisations will be vaccinated. In the over 60s we have well over 90% vaccination coverage.

As I wrote in my first post

Your model does not account for age related hospitalisation risk and vaccine coverage.

Cuddlesworth

https://www.boards.ie/discussion/comment/117807568#Comment_117807568

Its a spreadsheet with a small number of fields, its not hard to understand.

Here is the process to create a model, you take some known data and you use it to make a assumption of what might happen. The model is saying, over a 7 day period for 1million people, with a specific infection rate, you get the following hospitalisation rate for vaccinated and unvaccinated. There are other factors the model does not reference, its a very simple example. It doesn't account for increasing infection rates, age distribution for those infected, overall hospitalisation requirements or vaccine efficacy drops, or really any long term timeline in general or complicated factors. Its a really basic example of a "model" and I assume it was given out as such.

The tweets you reference say the known data is relatively in line with the projected model and that the vaccines are working. This is from somebody who I assume has access to the actual models and real world data. Again, in most country's across the world, they are seeing similar results.

None of your other fields matter, because they are not part of the example model given. They are external factors, with unknown data fields, not referenced in the example.. None of your conclusions are right, starting from the fact that have taken a completely different dataset and put it into the "example model" given, expecting the same result and acting surprised when you get something else.

And when this is pointed out and acknowledged, you continue down the same path attempting to justify and validate the flawed result.

Furze99

https://www.boards.ie/discussion/comment/117802455#Comment_117802455

Which could be convenient ultimately for both governments and pharma companies.

Mr.StRiPe

https://www.boards.ie/discussion/comment/117807818#Comment_117807818

Its a spreadsheet with a small number of fields, its not hard to understand.

None of your other fields matter, because they are not part of the example model given.

Ok so I take it you agree with my first question and I'll stick to the 5 input fields you agree with to keep it simple.

My next question is do you agree if you change any of the 5 input fields the model outputs the expected result for the specific combination of 5 inputs being inputted like what Nolan did when he changed the % vaccinated input from 70% to 90%?