Vaccine Progress, Approval, and Deployment

[Bald Rick]

No they wouldn't catch it quickly. There would be too many people vaccinated or who'd already had it to allow for quick spread. It really isn't as easy to catch as people think.

Well that depends on what you call ‘quickly’.

The models I’ve seen show that with a progressive removal of all restrictions between early Feb and July, and a 60% vaccine efficacy for infection transmission, we’d have a big wave of infections in early autumn, peaking at about twice the level we saw in early January. The assumptions in that model are conservative - vaccine take up is better than expected, as is the rate of vaccination, and the efficacy is slightly better. We are also going to be a month later in restriction easing, and I suspect it will be gently, gently at first. That research needs updating with better assumptions - and I bet there is someone tousling his hair over that very data in No 10 right now.

The people I know in this field have been pretty consistent in their message that we will have another and final bout of infections (and, sadly, deaths) in Sept/Oct this year. The size of it depends on how many are vaccinated before we significantly ease restrictions (ie the rate of vaccination take up, the date of easing, and how they are eased), and the vaccine efficacy. Various levels to pull in the model, and the end result is how many extra tens of thousands of people end up dead by the at the end of the year.

It’s simple, but brutal. Not a decision I would want to have to take.

 

[Bantamzen]

Well that depends on what you call ‘quickly’.

The models I’ve seen show that with a progressive removal of all restrictions between early Feb and July, and a 60% vaccine efficacy for infection transmission, we’d have a big wave of infections in early autumn, peaking at about twice the level we saw in early January. The assumptions in that model are conservative - vaccine take up is better than expected, as is the rate of vaccination, and the efficacy is slightly better. We are also going to be a month later in restriction easing, and I suspect it will be gently, gently at first. That research needs updating with better assumptions - and I bet there is someone tousling his hair over that very data in No 10 right now.

The people I know in this field have been pretty consistent in their message that we will have another and final bout of infections (and, sadly, deaths) in Sept/Oct this year. The size of it depends on how many are vaccinated before we significantly ease restrictions (ie the rate of vaccination take up, the date of easing, and how they are eased), and the vaccine efficacy. Various levels to pull in the model, and the end result is how many extra tens of thousands of people end up dead by the at the end of the year.

It’s simple, but brutal. Not a decision I would want to have to take.

Infections do not necessarily equal serious illnesses or death. Its time to stop talking up the risk.

 

[Bald Rick]

I don't believe that's right, nowhere all the people in the 5-9 categories are at risk, so we don't near anywhere near 90% compliance.

The stats are clear: c90% of the deaths are in the top 4 categories, c90% of the hospitalisations are in the top 9. It’s the latter that is the prime consideration for Government, as it is this that stretches the NHS.

Death rates are going to fall rapidly in the next month as the top 4 categories successful vaccinations take effect; particularly so as second doses are being administered in quantity from next month. Hospitalisations are falling rapidly now, mostly due to lockdown. They will continue to fall as more and more people are vaccinated, but it will be a while before it is done to a sustainable level.

Infections do not necessarily equal serious illnesses or death. Its time to stop talking up the risk.

Of course not, but there is a pretty well established relationship between infection rates, hospitalisation rates and death rates! Albeit that is changing (for the better) with vaccination.

 

[Bantamzen]

The stats are clear: c90% of the deaths are in the top 4 categories, c90% of the hospitalisations are in the top 9. It’s the latter that is the prime consideration for Government, as it is this that stretches the NHS.

Death rates are going to fall rapidly in the next month as the top 4 categories successful vaccinations take effect; particularly so as second doses are being administered in quantity from next month. Hospitalisations are falling rapidly now, mostly due to lockdown. They will continue to fall as more and more people are vaccinated, but it will be a while before it is done to a sustainable level.

Just because 90% of hospitalisations come from those groups doesn't mean that all have the same risks. We know that there are various health factors that increase risk, particularly in younger people, and the NHS is now targeting them rather than trying to blanket cover all.

 

[Philip]

The stats are clear: c90% of the deaths are in the top 4 categories, c90% of the hospitalisations are in the top 9. It’s the latter that is the prime consideration for Government, as it is this that stretches the NHS.

Death rates are going to fall rapidly in the next month as the top 4 categories successful vaccinations take effect; particularly so as second doses are being administered in quantity from next month. Hospitalisations are falling rapidly now, mostly due to lockdown. They will continue to fall as more and more people are vaccinated, but it will be a while before it is done to a sustainable level.



Of course not, but there is a pretty well established relationship between infection rates, hospitalisation rates and death rates! Albeit that is changing (for the better) with vaccination.

By July the whole adult population should have all been offered the first vaccination, along with a fair proportion with natural immunity. I don't see why it will be any worse than the flu season after this? There is also a chance too that this virus isn't as severe as flu...lack of immunity could be the reason for the higher IFR.

 

[yorksrob]

Because the unconstrained R rate is relatively high, ie this virus is relatively easy to ‘catch’.

Strictly speaking it doesn’t need high vaccine compliance, but the only other option is to let the unvaccinated catch it. If 80% of the adult population take the vaccine, and it is 80% effective, that still leaves over a third of the adult population unprotected, and with all restrictions removed most of them would, quickly. That is enough to cause the NHS issues, I imagine.

As it happens, over 10%* of the population have had the virus, and nearly 30% have had the first shot of the vaccine. Whilst there is some overlap it’s reasonable to sssume that over a quarter of the population now has some protection, including nearly all the most vulnerable.

*officially it’s about 6%, but it’s reasonable to assume many people have had it without realising, or simply didn’t ge5 a test.

I've never been convinced by the mathematics of this argument. In that scenario, over a third of the population are not at risk, because we know that even without vaccines, only a comparatively small proportion of the population are likely to be hospitalised and that will be as true of the 20% who aren't vaccinated.

 

[hwl]

Because the unconstrained R rate is relatively high, ie this virus is relatively easy to ‘catch’.

Strictly speaking it doesn’t need high vaccine compliance, but the only other option is to let the unvaccinated catch it. If 80% of the adult population take the vaccine, and it is 80% effective, that still leaves over a third of the adult population unprotected, and with all restrictions removed most of them would, quickly. That is enough to cause the NHS issues, I imagine.

As it happens, over 10%* of the population have had the virus, and nearly 30% have had the first shot of the vaccine. Whilst there is some overlap it’s reasonable to assume that over a quarter of the population now has some protection, including nearly all the most vulnerable.

*officially it’s about 6%, but it’s reasonable to assume many people have had it without realising, or simply didn’t ge5 a test.

For most of the last 9 months the ONS surveillance studies have indicated that the testing programmes (NHS staff or pillar 2 etc) have been capturing between 45-55% of cases, hence doubling reported case numbers is a good rough estimate. The exception being in September when not enough testing capacity was available and more recently with the least well off (highest prevalence of cases) becoming more and more unlikely to get tested thus the case rate under recording a bit more than normal.

Antibody testing puts numbers higher still but have seem much recent analysis on this recently.

 

[yorksrob]

The stats are clear: c90% of the deaths are in the top 4 categories, c90% of the hospitalisations are in the top 9. It’s the latter that is the prime consideration for Government, as it is this that stretches the NHS.

Death rates are going to fall rapidly in the next month as the top 4 categories successful vaccinations take effect; particularly so as second doses are being administered in quantity from next month. Hospitalisations are falling rapidly now, mostly due to lockdown. They will continue to fall as more and more people are vaccinated, but it will be a while before it is done to a sustainable level.



Of course not, but there is a pretty well established relationship between infection rates, hospitalisation rates and death rates! Albeit that is changing (for the better) with vaccination.

If you think back to last year, the virus remained at very low levels throughout the summer, then rose from August. This seems inevitable with the cooler weather.

I really don't see what's to be gained by locking everyone up during the one time of the year when the weather is on our side.

 

[hwl]

Well that depends on what you call ‘quickly’.

The models I’ve seen show that with a progressive removal of all restrictions between early Feb and July, and a 60% vaccine efficacy for infection transmission, we’d have a big wave of infections in early autumn, peaking at about twice the level we saw in early January. The assumptions in that model are conservative - vaccine take up is better than expected, as is the rate of vaccination, and the efficacy is slightly better. We are also going to be a month later in restriction easing, and I suspect it will be gently, gently at first. That research needs updating with better assumptions - and I bet there is someone tousling his hair over that very data in No 10 right now.

The people I know in this field have been pretty consistent in their message that we will have another and final bout of infections (and, sadly, deaths) in Sept/Oct this year. The size of it depends on how many are vaccinated before we significantly ease restrictions (ie the rate of vaccination take up, the date of easing, and how they are eased), and the vaccine efficacy. Various levels to pull in the model, and the end result is how many extra tens of thousands of people end up dead by the at the end of the year.

It’s simple, but brutal. Not a decision I would want to have to take.

Completely agreed on the analysis I've been seeing too.

While effectiveness in reducing transmission is better than expected with the older strains (e.g. Ox/AZ at 67%) it will be lower with the new strains especially those with 484K mutations which make it far harder for antibodies to "stick" resulting in a longer period where a person is infectious. Hence whether removing too much of the conservativeness is wise at this stage worth leaving in for a while.

 

[yorksrob]

Well, it all goes to show that the vaccine is the only game in town. Once that has been exhausted, they should open everything up as there are no further rainbows on the horizon.

 

[hwl]

The stats are clear: c90% of the deaths are in the top 4 categories, c90% of the hospitalisations are in the top 9. It’s the latter that is the prime consideration for Government, as it is this that stretches the NHS.

And 75-80% of ICU admissions in groups 1-9 which is the real NHS stretcher. The other 20-25% (20% being healthy pre Covid) being a problem that won't be addressed for vaccination for months to come.

Death rates are going to fall rapidly in the next month as the top 4 categories successful vaccinations take effect; particularly so as second doses are being administered in quantity from next month. Hospitalisations are falling rapidly now, mostly due to lockdown. They will continue to fall as more and more people are vaccinated, but it will be a while before it is done to a sustainable level.

Of course not, but there is a pretty well established relationship between infection rates, hospitalisation rates and death rates! Albeit that is changing (for the better) with vaccination.

This FT analysis is noticeably better than other public ones on vaccine roll out through age groups. Not the time taken to start on the under 50s.

The government population estimates are also a bit low compared to real people numbers hence in some locations more than 100% of older age groups have already been vaccinated (an achievement with some people declining to be vaccinated!)

https://twitter.com/jburnmurdoch/status/1361959977116700675/photo/1

 

[Bald Rick]

I've never been convinced by the mathematics of this argument. In that scenario, over a third of the population are not at risk, because we know that even without vaccines, only a comparatively small proportion of the population are likely to be hospitalised and that will be as true of the 20% who aren't vaccinated.

I don’t follow your maths, sorry.

The trouble is, without any NPIs those that aren’t protected will catch it. And even if it is only 20% of the population, and only 2%* of that 20% end up in hospital, that’s still 250,000 people in hospital in a relatively short time. Compared to 19,000 now. So, government needs to get as many people protected as possible, whether through vaccination, or some form of restrictions. One course of action could be to require those at much higher risk to shield for longer until all double vaccinated and it has taken effect - as was started this week.

*the hospitalisation rate to date has been 10.5% of recorded infections. If we assume that there are twice as many infections as recorded, then it’s reasonable to assume the real hospitalisation rate is 5% of those infected, rather than the 2% I have suggested.


I want to be clear that I am not arguing for or against any particular course of action. I want lockdown to end just as soon as the next person (and the next person is Mrs BR who is desperate for it to end!). I’m just trying to explain the challenge that Government has on its hands.

 

[yorksrob]

I don’t follow your maths, sorry.

The trouble is, without any NPIs those that aren’t protected will catch it. And even if it is only 20% of the population, and only 2%* of that 20% end up in hospital, that’s still 250,000 people in hospital in a relatively short time. Compared to 19,000 now. So, government needs to get as many people protected as possible, whether through vaccination, or some form of restrictions. One course of action could be to require those at much higher risk to shield for longer until all double vaccinated and it has taken effect - as was started this week.

*the hospitalisation rate to date has been 10.5% of recorded infections. If we assume that there are twice as many infections as recorded, then it’s reasonable to assume the real hospitalisation rate is 5% of those infected, rather than the 2% I have suggested.


I want to be clear that I am not arguing for or against any particular course of action. I want lockdown to end just as soon as the next person (and the next person is Mrs BR who is desperate for it to end!). I’m just trying to explain the challenge that Government has on its hands.

We're talking about twenty percent of people who have the vaccine and who aren't protected against catching the virus. That doesn't mean to say that there won't be mitigation against some of the worst effects.

Then there's the twenty percent who don't have the vaccine. The same proportion of these will be likely to be hospitalised as that of the whole pre-vaccine population.

Hospitalisation rate currently 10%.

A third of that is 3.3%, but a half of that (roughly 1.7) will be at 10% so that's under 0.2 of a percent added to the 1.7 is about 1.9 % of those infected hospitalised. Even then, what's the likelihood of that 1.9% all being infected at the same time ? Even last March before interventions, there was never a time when everyone in the whole country was infected. This will be even less likely with lots of vaccinated people wandering around.

 

[Yew]

I think the ability to board a plane going outside the UK (well, more precisely, alight at the other end) as countries impose vaccination requirements at their borders may focus a lot of minds. And if we want herd immunity, that is going to need very high levels of compliance, from all groups.

That is by no means a certainty, and the Council of Europe are passing resolutions against vaccine status discrimination

As it happens, over 10%* of the population have had the virus, and nearly 30% have had the first shot of the vaccine. Whilst there is some overlap it’s reasonable to sssume that over a quarter of the population now has some protection, including nearly all the most vulnerable.

*officially it’s about 6%, but it’s reasonable to assume many people have had it without realising, or simply didn’t ge5 a test.

Most studies and surveys suggest well over 20%.

Well that depends on what you call ‘quickly’.

The models I’ve seen show that with a progressive removal of all restrictions between early Feb and July, and a 60% vaccine efficacy for infection transmission, we’d have a big wave of infections in early autumn, peaking at about twice the level we saw in early January. The assumptions in that model are conservative - vaccine take up is better than expected, as is the rate of vaccination, and the efficacy is slightly better. We are also going to be a month later in restriction easing, and I suspect it will be gently, gently at first. That research needs updating with better assumptions - and I bet there is someone tousling his hair over that very data in No 10 right now.

The people I know in this field have been pretty consistent in their message that we will have another and final bout of infections (and, sadly, deaths) in Sept/Oct this year. The size of it depends on how many are vaccinated before we significantly ease restrictions (ie the rate of vaccination take up, the date of easing, and how they are eased), and the vaccine efficacy. Various levels to pull in the model, and the end result is how many extra tens of thousands of people end up dead by the at the end of the year.

It’s simple, but brutal. Not a decision I would want to have to take.

Who is going to be infected, exactly? All of those people for whom this is just a bit of a cough for a few days?

 

[Crossover]

I'd rather they spend the money on duplicate letters than start missing people en masse.
The cost of the vaccination programme is macroeconomically irrelevant.

Maybe, but it isn’t beyond possibility to get it right first time. Though it is likely to be just the tip of the inefficiency iceberg

 

[Yew]

*the hospitalisation rate to date has been 10.5% of recorded infections. If we assume that there are twice as many infections as recorded, then it’s reasonable to assume the real hospitalisation rate is 5% of those infected, rather than the 2% I have suggested.

However we've vaccinated the most vulnerable groups, so we can expect that to drop significantly over the next two weeks.

 

[Bald Rick]

Most studies and surveys suggest well over 20%.

The latest ONS study records that 18.5% of the population of England had COVID antibodies (data point 1 Feb). This accounts for those who have had the virus and those who have been vaccinated; the number nearly doubled since December which rather proves that point.

We are three weeks further on now, so that number will be nearer 30%, of which 2/3rds approx will have been due to vaccination.

 

[Bantamzen]

The latest ONS study records that 18.5% of the population of England had COVID antibodies (data point 1 Feb). This accounts for those who have had the virus and those who have been vaccinated; the number nearly doubled since December which rather proves that point.

We are three weeks further on now, so that number will be nearer 30%, of which 2/3rds approx will have been due to vaccination.

Antibodies are not the only indicator of immunity.

 

[Bald Rick]

However we've vaccinated the most vulnerable groups, so we can expect that to drop significantly over the next two weeks.

No, we’ve vaccinated the very most vulnerable groups at risk of being killed by the virus, and it will take another 3 - 4 weeks for the full effect of that to show. (1-2 weeks to catch it, 2 weeks before hospital admission, 1-2 weeks before dying). And of course better protection only comes with the second dose, and that won’t be achieved for this group for another 6 weeks or so. But we are seeing the start of the heavy reduction trend now.

This group only represents half those most at risk of being hospitalised. We have only just started vaccinating the other half. Hospitalisations will continue to fall as a result of lockdown and the vaccination of the first half of the population at high risk of hospitalisation. But the reduction trend will slow down about 4 weeks after any significant easing in restrictions across the whole population.

Antibodies are not the only indicator of immunity.

True, but it is the one that’s being measured.

 

[hwl]

That is by no means a certainty, and the Council of Europe are passing resolutions against vaccine status discrimination

Which are unenforceable

Most studies and surveys suggest well over 20%.

The levels in Manaus are over 80% but that didn't stop high case rates with newer strains...

 

[Bantamzen]

True, but it is the one that’s being measured.

That doesn't make it accurate.

 

[Tezza1978]

Which are unenforceable

The levels in Manaus are over 80% but that didn't stop high case rates with newer strains...

The Manaus data is very uncertain - I've read several articles that state the estimates of antibodies there were vastly overstated and were more like 30% or just above from wave 1 - meaning that there were still plenty of people to fuel their 2nd wave -many of whom live in multigenerational households in poor conditions - hence the serious problems they had

 

[Bald Rick]

That doesn't make it accurate.

“All models are wrong, but some are at least useful”

 

[Philip]

No, we’ve vaccinated the very most vulnerable groups at risk of being killed by the virus, and it will take another 3 - 4 weeks for the full effect of that to show. (1-2 weeks to catch it, 2 weeks before hospital admission, 1-2 weeks before dying). And of course better protection only comes with the second dose, and that won’t be achieved for this group for another 6 weeks or so. But we are seeing the start of the heavy reduction trend now.

This group only represents half those most at risk of being hospitalised. We have only just started vaccinating the other half. Hospitalisations will continue to fall as a result of lockdown and the vaccination of the first half of the population at high risk of hospitalisation. But the reduction trend will slow down about 4 weeks after any significant easing in restrictions across the whole population.



True, but it is the one that’s being measured.

In terms of the two doses of the vaccine, it is the first dose which provides the vast bulk of protection; the second dose is little more than a booster/topping up job. Whilst the second dose is important as a top up when the time comes to have it, I don't think people are at significantly more risk of severe disease after the first dose, compared to after having both doses.

Hospital admissions and numbers will increase when the majority of the hospitality and social contact restrictions are lifted, but this will be from low numbers and at only a moderate rate by the time this is expected to happen; sometime during April or early May.

 

[Yew]

https://pace.coe.int/en/files/29004/html

7.3.1 ensure that citizens are informed that the vaccination is NOT mandatory and that no one is politically, socially, or otherwise pressured to get themselves vaccinated, if they do not wish to do so themselves;

7.3.2 ensure that no one is discriminated against for not having been vaccinated, due to possible health risks or not wanting to be vaccinated;

For reference, here is the council of Europes resolution

“All models are wrong, but some are at least useful”

Apart from those made by Neil Fergurson?

 

[HSTEd]

As Bald Rick points out, if we release restrictions now there will be another wave.

The question you have to ask is how destructive it would be, and whether that constitutes acceptable losses.

I lean towards it being acceptable within a matter of weeks as it seems unlikely that healthcare capacity would be grossly overmatched.
It depends on the transmission impacts of the vaccine.

EDIT:

Given what we know about vaccine capability against hospitalisation/serious illness (near complete), and given the very high takeup amongst high risk groups (very high for the most part) I don't think we have too much to worry about.

Most of the vaccine refuseniks would rush to get a vaccination if another wave was building I think.

 

[liam456]

I'm the last person to support lockdowns but Bald Rick isn't usually wrong.....

Well that depends on what you call ‘quickly’.

The models I’ve seen show that with a progressive removal of all restrictions between early Feb and July, and a 60% vaccine efficacy for infection transmission, we’d have a big wave of infections in early autumn, peaking at about twice the level we saw in early January. The assumptions in that model are conservative - vaccine take up is better than expected, as is the rate of vaccination, and the efficacy is slightly better. We are also going to be a month later in restriction easing, and I suspect it will be gently, gently at first. That research needs updating with better assumptions - and I bet there is someone tousling his hair over that very data in No 10 right now.

The people I know in this field have been pretty consistent in their message that we will have another and final bout of infections (and, sadly, deaths) in Sept/Oct this year. The size of it depends on how many are vaccinated before we significantly ease restrictions (ie the rate of vaccination take up, the date of easing, and how they are eased), and the vaccine efficacy. Various levels to pull in the model, and the end result is how many extra tens of thousands of people end up dead by the at the end of the year.

It’s simple, but brutal. Not a decision I would want to have to take.

A few weeks ago I was in a really bad place but I've began to mellow, and will exercise a little more patience. If only the government could communicate as effectively as the above. I'm reminded of clips of Merkel at the start of the pandemic explaining R and exponential spread in a manner which made her look like she really knew what she was talking about. As for out cabinet........

 

[Bantamzen]

“All models are wrong, but some are at least useful”

Only if the data is anywhere complete.

As Bald Rick points out, if we release restrictions now there will be another wave.

The question you have to ask is how destructive it would be, and whether that constitutes acceptable losses.

I lean towards it being acceptable within a matter of weeks as it seems unlikely that healthcare capacity would be grossly overmatched.
It depends on the transmission impacts of the vaccine.

With the vaccines the transmission rate is not the primary concern.

 

[HSTEd]

With the vaccines the transmission rate is not the primary concern.

Unfortunately without total vaccination, transmission rate is still a significant figure.

 

[brad465]

As Bald Rick points out, if we release restrictions now there will be another wave.

The question you have to ask is how destructive it would be, and whether that constitutes acceptable losses.

I lean towards it being acceptable within a matter of weeks as it seems unlikely that healthcare capacity would be grossly overmatched.
It depends on the transmission impacts of the vaccine.

EDIT:

Given what we know about vaccine capability against hospitalisation/serious illness (near complete), and given the very high takeup amongst high risk groups (very high for the most part) I don't think we have too much to worry about.

Most of the vaccine refuseniks would rush to get a vaccination if another wave was building I think.

At the moment infections could well go up, but as the seasons get better we should see less risk of infections going up, in the same way hospitality reopening in July last year didn't cause any problems (the problem coming 2 months later when schools reopened and the seasons got against us). As you allude to, another wave emerging should not be a concern if all those most at risk are vaccinated, and should this summer be a low transmission one just by virtue of it being unfavourable for spread, vaccinating the younger population throughout it will help mitigate any further rise in autumn anyway.