COVID19

[Doobs]

Given it is past Wednesday. I am going to do an update. 

I am thinking this will be the last slightly positive update for a while.

Deaths halved in the week up to 19 May, but it is very unlikely they will go lower, as the provisional figures look higher going forward.

Hospitalisations.  They did drop beloe 100 for 3 days, but have been stubbornly above 100 every day since the 17 May.

Cases.  The cases are now above 3,000 a day for the first time since the first week in April.

Vaccinations have been holding up though with a steadily increasing trend in both first and second vaccinations.

I think it is all about the hotpsots now.  You can actually see the data at hospital level; and as an example Bedfordshire is getting 2 hospitalisations a day right now versus 25 a day at the beginning of February.  Bolton is getting 5 a day vs 24 at the beginning of February.  You can see on the visualisations that the mean age of each patient is much lower now and they are more likely to be vaccinated than not (even if more than half are not vaccinated as is claimed).  So even in the hotspots we aren't going to be seeing deaths at the levels we saw earlier in the year anytime soon. 

I think it is pretty clear now that the dominant Indian variant is probably about 50% more transmissible than the Kent variant, but I don't think we know enough about how likely it is to kill us.  I have read that the first vaccine dose offers less protecion than before, but they think you get similar levels of protection after two vaccinations.  Overall deaths are unlikely to be much more than 10% of previous levels. 

So overall, there are some worrying signs, but don't think we should worry too much. 

For more into, John Burns Murdoch has done some good work on the Indian variant here https://twitter.com/jburnmurdoch/status/1397995466701344769.

 

That James Ward thread is good, so won't step too much on his shoes.

The reporting of deaths this week has been a bit questionable.  I didn't notifce anyone in the press mention the bank holiday when they reported zero deaths the other day (though just checked and it is buried in the BBC as if it wasn't that significant).  I was hoping Marky was going to pop up yesterday and say "that's because it was a bank holiday, duh!" and then I could have retired knowing my work was done.

Anyway deaths up until 7 days ago were up 20% vs the week before to about 5 a day in England, but the week before they had fallen by 50%.  So overall. whatever the concerns are about the Indian variant (or Delta?), there hasn't really been a big rise in deaths. 

I'd say much the same about hospital admissions too.  We are at levels seen as long ago as a couple of weeks previously, so I am not going to get hysterical yet.  I saw someone from Independent Sage (can't seem to get rid of them after Marky mentioned them) saying Cases, hospitalisations and deaths were all rising exponentially!!! the other day.  The exclamation marks are my interpretation of the tone of her voice when she said it.  The reality is different.

Anyway, cases are rising pretty sharply, so there is a bit of a worry there, but cases are still at a relatively low level nationally.   

My overall take on this is that most over 50s are double vaccinated already, so much of any spread is going to be amongst those who are less likely to be badly affected.  I realise there are people who can't take the vaccine and have compromised immune systems, but even then I struggle to get to a large number of excess deaths in the exit wave.  I'd consider myself about the level of an optimistic James Ward right now.  We still have 3 wweks or so until step 4, so I think we shouldn't be rushing to announce anything.

By next week, I'll have a better look at the numbers.   

[Marky147]

Hope they stay good, as I've just got a Wynn summer offer through, and got 8 nights at Encore for a monkey

[Doobs]

Hope they stay good, as I've just got a Wynn summer offer through, and got 8 nights at Encore for a monkey

ooh, seems an absolute steal.  resort fee a bazillion dollars though

[Marky147]

Hope they stay good, as I've just got a Wynn summer offer through, and got 8 nights at Encore for a monkey

ooh, seems an absolute steal.  resort fee a bazillion dollars though

About $900 all in, but I've always managed to get the RF made optional when I check-in, and then just paid for internet on the days I needed to work.

[nirvana]

This thing goes to new levels of cuntiness every day. Used to just be corona virus/covid , and a few other terms.

Now we has 'the variant first identified in India now known as Delta'. Covidiocy. Fkn fumming, can barely get thru the news without losing it.

Sandwich boards looking good

[Marky147]

Amazing how they make themselves look more stupid every day, too.

Shapps goes on talking about a Nepal variant, which has already been discounted by the WHO as nothing.

Just the mugs in the mail reporting it, after some dimwit on Reddit probably emailed them with a scoop

I know the WHO are a ballache, and obviously a shitshow, but we make them look like an organisation run with the efficiency of Amazon.

[nirvana]

Amazing how they make themselves look more stupid every day, too.

Shapps goes on talking about a Nepal variant, which has already been discounted by the WHO as nothing.

Just the mugs in the mail reporting it, after some dimwit on Reddit probably emailed them with a scoop

I know the WHO are a ballache, and obviously a shitshow, but we make them look like an organisation run with the efficiency of Amazon.

Definitely one of those things where it makes most sense to know just enough to keep yourself safe and be lucky enough to have a job where you can do that. And then, somehow, find the ability to switch off from all of it

[Doobs]

not good news (nirvana, look away pls)

We've had the new PHE data that suggests that if you get the Delta variant you're 2 1/2 x more likely to be hospitalised (on top of known increased transmissibility and vaccine escape after 1 dose)

https://twitter.com/john_actuary/status/1400547650034937856

Note the big confidence interval as this is based on low numbers, so 2.5x may well be an estimate that is too high (though may also be too small).  

[nirvana]

not good news (nirvana, look away pls)

We've had the new PHE data that suggests that if you get the Delta variant you're 2 1/2 x more likely to be hospitalised (on top of known increased transmissibility and vaccine escape after 1 dose)

https://twitter.com/john_actuary/status/1400547650034937856

Note the big confidence interval as this is based on low numbers, so 2.5x may well be an estimate that is too high (though may also be too small).  

Haha, I can cope, I'm just squarely in the so what category in the sense that I think we need a decent period of no restrictions to see what the real ongoing situation of living with the virus will turn out to be. Hopefully people can find whatever that situation is to be tolerable because the alternative of permanent (albeit fluctuating) controls and restrictions is pretty unthinkable.

Anyway, can you interpret that a bit more.

Delta resulting in 2.5 x covoriginal hospitalisations today would still be a massively lower number of hospitalisations in absolute terms than pre vax days ?

If that's so, then the headline could be , even with delta, you are still 30x (or whatever the number is)  less likely to end up seriously ill if you are fully vaccinated than you would have been pre vax. Not sure I see the point of the 2.5x number other than to cause alarm and serve as a backdrop to maintaining increasingly harmful restrictions.

Also, would have a friendly fiver with anyone that the 2.5 x proves to be an overestimate.  Seen this CYA estimating happen enough now to know that this is pretty near a banker

[Doobs]

not good news (nirvana, look away pls)

We've had the new PHE data that suggests that if you get the Delta variant you're 2 1/2 x more likely to be hospitalised (on top of known increased transmissibility and vaccine escape after 1 dose)

https://twitter.com/john_actuary/status/1400547650034937856

Note the big confidence interval as this is based on low numbers, so 2.5x may well be an estimate that is too high (though may also be too small).  

Haha, I can cope, I'm just squarely in the so what category in the sense that I think we need a decent period of no restrictions to see what the real ongoing situation of living with the virus will turn out to be. Hopefully people can find whatever that situation is to be tolerable because the alternative of permanent (albeit fluctuating) controls and restrictions is pretty unthinkable.

Anyway, can you interpret that a bit more.

Delta resulting in 2.5 x covoriginal hospitalisations today would still be a massively lower number of hospitalisations in absolute terms than pre vax days ?

If that's so, then the headline could be , even with delta, you are still 30x (or whatever the number is)  less likely to end up seriously ill if you are fully vaccinated than you would have been pre vax. Not sure I see the point of the 2.5x number other than to cause alarm and serve as a backdrop to maintaining increasingly harmful restrictions.

Also, would have a friendly fiver with anyone that the 2.5 x proves to be an overestimate.  Seen this CYA estimating happen enough now to know that this is pretty near a banker

The point of the 2.5x number isn't to cause alarm, it is just the result given by the data.  The confidence interval shows how accurate it is likely to be.   My instinct is 2.5 is more likely to be too high than too low, but it isn't good news either way, and calling the unders a banker is just wrong.  For instance, I was pretty hopeful that the 50,% extra transmissability estimate would be an overestimate, but think it is more likely to have been an underestimate right now.  I was pretty hopeful we'd get lower fatality rates in wave 2, and they weren't that different and so on.

My view is that it is a really good thing we have got so many vulnerable people vaccinated right now in this country.  That means that even if hospitalisations do take off, deaths should be in much lower proportions than before.

I am also pretty confident that John Roberts has a better understanding of this than most, so it is quite wrong to link him with some bad estimates that you may have seen elsewhere.

As an aside, I got my second jab early a couple of days ago.  Pretty glad in the circumstances, and side effects so far were much better than last time (both AZ).

[Marky147]

Why I don't get anything from the news, and just rely on people like Stuart McDonald, James Ward, John Roberts etc.

Then you just get stuff formulated from data, and no skews to fit any agenda.

Think they have to throw a massive lump of money at hospitality if they don't want to just yolo and go with it in a few weeks, because they're not going to be able to cope.

Given how a lot of people are markedly more sensible now, I'd like to think things wouldn't get as bad as the worst estimate, or hopefully even the moderately bad.

I'm probably putting far too much faith in the general public, though

[RED-DOG]

A massively low number?

[nirvana]

A massively low number?

Lower, keep up.

Doobs, 2.5x is a good faith estimate from smart people, accept that but this is betting so ill still bet the unders.

I guess my thinking is more about the presentation of that info by press and idiot ministers. I think this can be presented as something that doesnt change the overall positive narrative around vaccines and opening up, not like something that is fuel to the nutty end of progressives who would like to secure all borders indefinitely.

Cant see why we cant have a simple scorecard showing probability of death in each age group, pre covid, with covoriginal unchecked, with vax and then show the delta of any new variant to those numbers.

[Doobs]

A massively low number?

Lower, keep up.

Doobs, 2.5x is a good faith estimate from smart people, accept that but this is betting so ill still bet the unders.

I guess my thinking is more about the presentation of that info by press and idiot ministers. I think this can be presented as something that doesnt change the overall positive narrative around vaccines and opening up, not like something that is fuel to the nutty end of progressives who would like to secure all borders indefinitely.

Cant see why we cant have a simple scorecard showing probability of death in each age group, pre covid, with covoriginal unchecked, with vax and then show the delta of any new variant to those numbers.

I could probably put together a scorecard thing with no vaccination, 1 dose and 2 dose plus the deltas; I think it is a good idea.   I am doing family stuff this week though, so might be pushed for time.

[nirvana]

A massively low number?

Lower, keep up.

Doobs, 2.5x is a good faith estimate from smart people, accept that but this is betting so ill still bet the unders.

I guess my thinking is more about the presentation of that info by press and idiot ministers. I think this can be presented as something that doesnt change the overall positive narrative around vaccines and opening up, not like something that is fuel to the nutty end of progressives who would like to secure all borders indefinitely.

Cant see why we cant have a simple scorecard showing probability of death in each age group, pre covid, with covoriginal unchecked, with vax and then show the delta of any new variant to those numbers.

I could probably put together a scorecard thing with no vaccination, 1 dose and 2 dose plus the deltas; I think it is a good idea.   I am doing family stuff this week though, so might be pushed for time.

That would be great but wasnt trying to get you to do it

Would just like to see some better factual info from Govt that makes it easy for people to understand relative risk and start to think for themselves in terms of adapting their lifestyles. Too often it feels like data is delivered / presented to help achieve unstated aims and it feels like we need a much more grown up delivery to manage this situation sustainably over the next few years.