COVID-19

[Grumble]

You really haven't. There's never been a chart like that before. Look at both the y axis.

It would be nice to see that graph going back to last September

[shpalman]

The right (deaths) axis is set for the ~2% CFR you used to have, when you could now maybe reset it to 0.1—0.2%.

Great you've got an extra month of advance warning about what's going to happen.

[shpalman]

You really haven't. There's never been a chart like that before. Look at both the y axis.

It would be nice to see that graph going back to last September

uk-cases-and-deaths-log-20210628-long.png (54.01 KiB) Viewed 2503 times

Day 150 was the 19th of July.

[lpm]

Day 230 to 240 is the place to look at. Early October. 3rd wave cases today have risen to where the 2nd wave reached about then.

Today's yellow shows the changed gearing in cases to deaths.

[shpalman]

Day 190 to 210 is the place to look at. Early September. Deaths started rising just as deaths are rising now. What do you expect is going to happen to make them stop rising? There's still 8 million adults who haven't had a first dose* and the cases in England are definitely peaked around the 20-24 age group (the 20-24 age group on its own has about 3.6 million people in it).

If nothing gets done to stop it, covid will spread to a substantial fraction of these people. The case doubling time is around 10 days, so in a month you get a factor of 8. The 7-day average for cases in the UK is currently about 16,000 so >100,000 per day by the end of July?

Yeah yeah different gearing between cases and deaths, but deaths are going up too and it's either young unvaccinated people (at ~160,000 per day how are you going to have millions done in the next two weeks so that they'll have some antibodies by the end of July?) or it's older people for whom the vaccination didn't give enough protection.

So, a couple of hundred deaths per day in mid July. Are we ok with that? That's roughly what we had at the end of October last year when the second lockdown was announced.

* - and there's another 12 million or so who have had one dose but not the second one.

(First doses are currently running at about 200,000 per day but second doses has dropped to less than that, around 160,000 per day recently.)

[lpm]

Day 190 to 210 is the place to look at. Early September. Deaths started rising just as deaths are rising now. What do you expect is going to happen to make them stop rising?

Huh? Of course nothing will stop deaths rising while the 3rd wave of cases rise. I've always stated clearly that the "decoupling" idea is a dud and that it's more a change in gearing.

Look again at early September. Look again at early October.

7 September: 2,000 cases per day, 8 deaths per day.
15 October: 16,000 cases per day, 100 deaths per day.
Today: 16,000 cases per day, 17 deaths per day.

There's still 8 million adults who haven't had a first dose* and the cases in England are definitely peaked around the 20-24 age group (the 20-24 age group on its own has about 3.6 million people in it).

This age group simply doesn't die of Covid as a population. They are being vaccinated last week, this week and next week, and will be more than adequately protected from the 3rd wave for August onwards. The official target of 19 July will easily be beaten, by around a week.

If nothing gets done to stop it, covid will spread to a substantial fraction of these people. The case doubling time is around 10 days, so in a month you get a factor of 8. The 7-day average for cases in the UK is currently about 16,000 so >100,000 per day by the end of July?

Something is being done to stop it - vaccinations. Plus schools shutting around 21 July.

Yeah yeah different gearing between cases and deaths, but deaths are going up too and it's either young unvaccinated people (at ~160,000 per day how are you going to have millions done in the next two weeks so that they'll have some antibodies by the end of July?) or it's older people for whom the vaccination didn't give enough protection.

So, a couple of hundred deaths per day in mid July. Are we ok with that? That's roughly what we had at the end of October last year when the second lockdown was announced.

A couple of hundred deaths per day in mid July is nothing like possible. The maths doesn't add up.

We're currently tracking fairly close to my guestimate which saw a peak (and a long plateau) at 60,000 per day and around 5,000 deaths across the summer. But yes it could be higher. Call it 8x current levels. That gives 136 deaths a day by August, 8,000 in total across the two months Aug and Sept. Half of those will be those who've declined the vaccine.

That's less than the daily fatality caused by smoking. Why not oppress the rights of nicotine addicts instead of taking away my freedoms? Currently Covid deaths are lower than alcohol deaths - the risk of people gathering to get drunk and watch football tonight is about the same from the virus and from the poisonous liquid.

It's tough because we are used normal being The Before and we still haven't mentally got accustomed to the world of The After - but the government is right we have to live with the virus for the rest of our lives. Covid-19 in a vaccinated population has very similar characteristics as flu: a regular killer of the elderly, worse some months than others, occasional very bad waves.

[shpalman]

Day 190 to 210 is the place to look at. Early September. Deaths started rising just as deaths are rising now. What do you expect is going to happen to make them stop rising?

Huh? Of course nothing will stop deaths rising while the 3rd wave of cases rise. I've always stated clearly that the "decoupling" idea is a dud and that it's more a change in gearing.

Look again at early September. Look again at early October.

7 September: 2,000 cases per day, 8 deaths per day.
15 October: 16,000 cases per day, 100 deaths per day.
Today: 16,000 cases per day, 17 deaths per day.

We had 17 deaths per day on the 16th of September and 100 deaths per day on the 16th of October (7-day averages for England only) so that took a month.

There's still 8 million adults who haven't had a first dose* and the cases in England are definitely peaked around the 20-24 age group (the 20-24 age group on its own has about 3.6 million people in it).

This age group simply doesn't die of Covid as a population. They are being vaccinated last week, this week and next week, and will be more than adequately protected from the 3rd wave for August onwards. The official target of 19 July will easily be beaten, by around a week.

If "this age group simply doesn't die of Covid as a population" then it's not members of this age group showing up in the increasing death figures now, and vaccinating them won't do anything direct to stop deaths from rising, if they're actually rising in older double-vaccinated people who are as protected as they're ever going to get (and it turns out that for some of them it's still not enough).

(ETA vaccinating young people will of course help reduce how much of it spreads to older people, but then so would some NPIs to reduce the overall case numbers instead of letting them double every 10 days.)

uk-vax-projection.png (25.06 KiB) Viewed 2388 times

At current rates, the UK won't have first-dosed all adults before the end of September.

If nothing gets done to stop it, covid will spread to a substantial fraction of these people. The case doubling time is around 10 days, so in a month you get a factor of 8. The 7-day average for cases in the UK is currently about 16,000 so >100,000 per day by the end of July?

Something is being done to stop it - vaccinations. Plus schools shutting around 21 July.

A month is a long time in exponential growth.

Yeah yeah different gearing between cases and deaths, but deaths are going up too and it's either young unvaccinated people (at ~160,000 per day how are you going to have millions done in the next two weeks so that they'll have some antibodies by the end of July?) or it's older people for whom the vaccination didn't give enough protection.

So, a couple of hundred deaths per day in mid July. Are we ok with that? That's roughly what we had at the end of October last year when the second lockdown was announced.

A couple of hundred deaths per day in mid July is nothing like possible. The maths doesn't add up.

Well if we go by what happened in September/October last year, you'd expect 100 deaths per day by the end of July. But in mid September some restrictions were (re)introduced.

We're currently tracking fairly close to my guestimate which saw a peak (and a long plateau) at 60,000 per day and around 5,000 deaths across the summer. But yes it could be higher. Call it 8x current levels. That gives 136 deaths a day by August, 8,000 in total across the two months Aug and Sept. Half of those will be those who've declined the vaccine.

That's less than the daily fatality caused by smoking. Why not oppress the rights of nicotine addicts instead of taking away my freedoms? Currently Covid deaths are lower than alcohol deaths - the risk of people gathering to get drunk and watch football tonight is about the same from the virus and from the poisonous liquid.

It's tough because we are used normal being The Before and we still haven't mentally got accustomed to the world of The After - but the government is right we have to live with the virus for the rest of our lives. Covid-19 in a vaccinated population has very similar characteristics as flu: a regular killer of the elderly, worse some months than others, occasional very bad waves.

This is all fine, but the problem with an "acceptable" level of deaths from covid is that a couple of weeks later it tends to already be an unacceptable level and no matter what you do at that point you've still got another couple of weeks of it getting worse before it gets better.

Alcohol and smoking don't spread exponentially through a population.

[lpm]

We had 17 deaths per day on the 16th of September and 100 deaths per day on the 16th of October (7-day averages for England only) so that took a month.

You keep implying that the 3rd wave will follow the 2nd wave trajectory. This is wrong wrong wrong. All gearings have changed - something that took a month before takes two months now etc etc.

If "this age group simply doesn't die of Covid as a population" then it's not members of this age group showing up in the increasing death figures now, and vaccinating them won't do anything direct to stop deaths from rising, if they're actually rising in older double-vaccinated people who are as protected as they're ever going to get (and it turns out that for some of them it's still not enough).

Vaccination reduces transmission, which slows the spread from young to the old, which prevents deaths of the imperfectly-vaccinated elderly. Covid depends on super-spreader events - which currently happen among the young and at school. Vaccination holds back super-spreading.

At current rates, the UK won't have first-dosed all adults before the end of September.

No, obviously, because that isn't the target. The UK will have offered vaccines to all adults by about 12 July. After this point it's all about chasing stragglers, can't-be-bothereds and nutters. We don't expect to beat the 90% level - which is a huge achievement, far higher take up than anyone ever expected.

Your maths is awry if you think 1st doses of all adults won't be complete by mid July for everyone coming forward.

Well if we go by what happened in September/October last year, you'd expect 100 deaths per day by the end of July. But in mid September some restrictions were (re)introduced.

WE CAN'T GO BY WHAT HAPPENED LAST YEAR YOU CAN'T JUST COPY A CHART FROM AN UNVACCINATED COUNTRY AND EXPECT IT TO REPEAT IN A VACCINATED COUNTRY.

This is all fine, but the problem with an "acceptable" level of deaths from covid is that a couple of weeks later it tends to already be an unacceptable level and no matter what you do at that point you've still got another couple of weeks of it getting worse before it gets better.

Alcohol and smoking don't spread exponentially through a population.

The same applies to flu, so let's apply the flu approach to Covid. Let's return to acknowledging that death from an infectious respiratory disease is an acceptable outcome for the very old and frail.

Alcohol and smoking don't spread exponentially but neither would they decline exponentially from universal vaccination. Like Covid fatalities among the unvaccinated, alcohol and smoking deaths are biased to the disadvantaged, poorly educated and reckless. These are far bigger health problems for the next decade than Covid in terms of burden on the NHS and early deaths. If we're protecting our vulnerable population by locking up international travellers in quarantine hotels/prisons, why can't we protect our vulnerable population by locking up cigarette manufacturers and sellers?

[shpalman]

We had 17 deaths per day on the 16th of September and 100 deaths per day on the 16th of October (7-day averages for England only) so that took a month.

You keep implying that the 3rd wave will follow the 2nd wave trajectory. This is wrong wrong wrong. All gearings have changed - something that took a month before takes two months now etc etc.

I'm just looking at the slopes on the (semi-logarithmic) graphs. The slope now corresponds to a 10-day doubling time. If vaccination is going to work as you say it will, we'll see the new cases slope get flatter.

Similarly, deaths are going up at the same doubling time, it's just that there's a factor of 10 lower CFR. If deaths stop going up even as cases keep doubling, then vaccination is working to stop deaths. But at the moment, deaths are going up in a way which follows cases with the usual lag.

[shpalman]

great news today's covids aren't as bad as yesterday's!

Still nearly double what you had at the beginning of last week though.

[Herainestold]

Why can't we have a government like Australia where they are locking down over a hundred cases, protecting the population, while they ramp up vaccination s?
In August we will have between 100 and 1000 deaths per day, even the lower number is unacceptable . It affects mainly the older and immuno compromised, working class, BAME and POC people. How is right to sacrifice them so the middle class can go the pub?
Even in America, in LA they are re-introducing masking for vaccinated people, they can see the writing on the wall.

Delta changes everything!

[Herainestold]

LA mask mandate

The Los Angeles County Department of Public Health backtracked on its Covid guidelines Monday, “strongly” recommending masking in all indoor public places, due to the rapid spread of the Delta variant.

The County is advising that all residents mask up indoors, regardless of Covid-19 vaccination status. Public Health’s “recommendation” is currently just that, rather than an official health order.

Today’s recommendation in L.A. mirrors one made last weekend by the World Health Organization. It also follows the recent reinstatement of indoor mask requirements in Israel, due to the spread of the Delta variant.

https://deadline.com/2021/06/la-county- ... 234783132/

[shpalman]

At current rates, the UK won't have first-dosed all adults before the end of September.

No, obviously, because that isn't the target. The UK will have offered vaccines to all adults by about 12 July. After this point it's all about chasing stragglers, can't-be-bothereds and nutters. We don't expect to beat the 90% level - which is a huge achievement, far higher take up than anyone ever expected.

Your maths is awry if you think 1st doses of all adults won't be complete by mid July for everyone coming forward.

Well it doesn't matter why they're not vaccinated, for the purposes of figuring out how many won't be vaccinated at that point. At current rates by 12th July there will still be 12 million adults who haven't even received a first dose. The covids don't care about whether you've been offered the vaccine.

[lpm]

Your numbers are wrong. We're already past that. It'll be about 5 million.

Why expect 100%? Nobody else does. But the UK will achieve 90%. And most crucially 96% of over 50s. That's phenomenal.

[shpalman]

Your numbers are wrong. We're already past that. It'll be about 5 million.

Why expect 100%? Nobody else does. But the UK will achieve 90%. And most crucially 96% of over 50s. That's phenomenal.

ETA wait a minute I'm getting confused between England and UK numbers (i.e. consider the England vaccination numbers out of the total number of adults in the UK) so yes you're right.

https://coronavirus.data.gov.uk/details ... me=England

England has currently given first doses to 37.4 million, and second doses to 27.5 million. This is described as being 84.5% and 62.2% of those aged 18 and over. They consider 44 million adults.

This should be correct:

England-vax-projection.png (17.25 KiB) Viewed 2317 times

(although of course second doses will go faster once there aren't any first doses to give anymore)

Even if you reach 90% you've still got 5 million susceptible unvaccinated hosts for the virus so 100,000 day new infections should be easy.

Alternatively, great, more people are vaccinated than I though. So it's had all the effect its ever going to have on the CFR already and you'll just need young people to get vaccinated so that the cases stop going up.

[KAJ]

Even if you reach 90% you've still got 5 million susceptible unvaccinated hosts for the virus so 100,000 day new infections should be easy.

Alternatively, great, more people are vaccinated than I though. So it's had all the effect its ever going to have on the CFR already and you'll just need young people to get vaccinated so that the cases stop going up.

Wouldn't such a large proportion vaccinated reduce the R? Maybe below 1?

[shpalman]

Even if you reach 90% you've still got 5 million susceptible unvaccinated hosts for the virus so 100,000 day new infections should be easy.

Alternatively, great, more people are vaccinated than I though. So it's had all the effect its ever going to have on the CFR already and you'll just need young people to get vaccinated so that the cases stop going up.

Wouldn't such a large proportion vaccinated reduce the R? Maybe below 1?

You've got 84.5% vaccinated already (with one dose) and the R_t is obviously still more than 1.

[lpm]

That's adults, the weight of unvaccinated children holds us back. We need to get 85% of total population including immunity from previously infection.

[shpalman]

That's adults, the weight of unvaccinated children holds us back. We need to get 85% of total population including immunity from previously infection.

You would have 85% of the whole population with at least one dose by the the 24th of August according to my previous projection based on the current rates, if you kept going, but only 79% of the population are adults, so unless you change your rollout rules* you'd stop before then.

What's your estimate for how many people aren't vaccinated yet but have immunity from having been infected? You could presumably get something by taking the infection rates in the younger age bands and integrating them but especially in the first wave it would have missed a lot, and possibly in children it misses most of them, so I personally am not going to bother trying. Every million you don't have to do means you arrive at your goalpost 5 days earlier, and if they're the youngest millions then they'd be the last ones you'd get around to, and especially with infections peaked in the 20-24 age range, it may well be the case that they'd get naturally infected before they got vaccinated.

* - Italy now is now vaccinating everyone 12 or over, for example.

[lpm]

Yep.

My little model assumed the vacs continued down to 16-18 and 12-16. Otherwise the Sept school return will be a.problem.

But there will still be voluntary distancing and masks, which should knock a chunk off R.

[Grumble]

We really do need vaccination for 12 and over.

[Herainestold]

We really do need vaccination for 12 and over.

We need lockdown

[Herainestold]

In the face of Delta, with a vaccine efficacy of 90% you need population coverage of 97%.
Or maybe it was the other way around. In either case it is not possible. Not without vaccinating children.

Without enough vaccination, the only way to get immunity is through infection, and Delta is very good at that, it won't take long to confer immunity to the rest of the unvaccinated population.

I just hope the morgues are ready for the onslaught.

[jdc]

In the face of Delta, with a vaccine efficacy of 90% you need population coverage of 97%.
Or maybe it was the other way around. In either case it is not possible. Not without vaccinating children.

Without enough vaccination, the only way to get immunity is through infection, and Delta is very good at that, it won't take long to confer immunity to the rest of the unvaccinated population.

I just hope the morgues are ready for the onslaught.

Sounds about right - r0 of 7, 86% coverage needed for herd immunity with 100% effective vaccine. VE of .9 would make it 95.55% needed if me maths is right.

Do we have reliable figures on vaccine efficacy against transmission of delta?

[Herainestold]

Do we have reliable figures on vaccine efficacy against transmission of delta?

Not that I have seen. People are speculating but I havent seen anything reliable