Loading Notifications...

What happens after the English lockdown

Please Register as a New User in order to reply to this topic.

Couldn’t resist starting another Covid thread... What do people think the English rules will be from 2 December?

First of all, there’s no guarantee the lockdown won’t be extended for another week or two. However given the restlessness of some Tory backbenchers and indeed Cabinet ministers, I suspect there will be a relaxation from the 2nd. The idea is a tiered approach but given the nationwide NHS capacity issues, I suspect most areas would end up in the same tier anyway.

My prediction is they will open non essential retail as that’s good for the economy, particularly in the run up to Xmas. Pubs and restaurants to open to single household groups only or perhaps 2-3 people from different households. Walls and gyms might be allowed to reopen but I think that’s 50/50. Most other rules to stay the same. Then a rule of 6 type relaxation for private gatherings for the last two weeks of December to ‘save’ Xmas and NY. Realistically many (most?) people will do it anyway, so no point criminalising / annoying them.

Of course any significant relaxation, particularly around private gatherings and the hospitality sector, will drive infections again. I imagine R will drop slightly below 1 by the 2nd but we’ll still be running at around 50,000 cases a day, which is a lot and that’s before R inevitably starts creeping up again. That’s a very different position compared to   the summer, when R wasn’t that much lower than 1 but daily cases were in the very low 1,000s. This time it will get to breaking point a lot quicker. I expect another lockdown in January to deal with the impact of Xmas. It would probably be needed right after NY but BoJo will probably f*ck around again until it’s too late, so we’ll get another 4 weeks from mid or late January. Another relaxation after that, similar to December. We might need another lockdown in March but, if things go well with the vaccine rollout to vulnerable groups and with a fair wind, we might just about get through to the Easter school / Uni break without locking down again.

I’m hopeful that things will get gradually better after Easter as more vulnerable people get vaccinated and as we move out of the peak season for respiratory diseases. However I don’t anticipate a wholesale lifting of restrictions until the summer at the earliest. There will probably be a gradual relaxation but things like going to the pub or working in an office without SD, as well as theatres, gigs, spectator sports and night clubs will just have to wait... I think we’ll still be in a world of face masks and SD for most of next year. A lot depends on how effective and long lasting the various vaccines will be. I’ve been wondering whether it will actually be feasible to vaccinate the non-vulnerable population even by the end of next year. Time will tell...

Report
 veteye 14 Nov 2020
In reply to Misha:

What happens is that we watch all the bickering and in-fighting across all parties, and all four nations.

Report
In reply to veteye:

Ah yes, very good point. 

Report
 wintertree 14 Nov 2020
In reply to Misha:

Cases fall more slowly in lockdown than they grow before lockdown.  I've made a plot of the average number of cases in the last week vs the rate of change in that number - black is growing, blue is shrinking.  A lot of former T2/T3 regions now have falling case numbers which gives me hope that T2/T3 made a difference, and that lockdown will soon start making that same difference elsewhere.  The period for this data ended on Nov 9th, and ZOE - which leads this data by perhaps 7 days - is suggesting more areas are tipping over in to decay.  So hopefully in a week much of this plot will be blue.  But most places on this plot will need over 40 days to get to low prevalence unless lockdown-lite further improves the decay rate beyond T2/T3; I'm skeptical.

So I think we're going to have high prevalence when we get to Dec 2nd, and a release of lockdown combined with Christmas rush behaviour would just lead to another rapid rise in case numbers and another panicked lockdown to protect healthcare.  I suspect lockdown is going to have to be held, or released into T3+ for most locations, until shortly before Christmas.   

When lockdown was announced, the course we were on had us exceeding nationwide "surge" capacity in healthcare in a couple of weeks.  Since then the effects of T2/T3 restrictions (any perhaps school half-terms) have fed through to levelling cases of at the national level and seeing hospitalisations slow down.  The lockdown-lite should see that trend continue, but hospital occupancy is worryingly close to the red line nationally speaking, regionally it's worse in some places.  What happens next depends massively on the flu season wild card; if it's mild that's great, if there're a lot of influence hospitalisations we've got a real problem that it's too late to fix, other than by not ending lockdown until the flu season is passed.  

Another wild card is the return of undergraduates to their family homes.  The false negative rates of the lateral flow tests makes me think they're not a magic bullet; if prevalence in the undergraduate population isn't held meticulously low throughout lockdown, this could drive another big round of infections - but this time in older, more vulnerable people.  I can see why the government are so keen to send them home the moment lockdown is set to end.

Like you, I don't see the vaccine putting a stop to cycles of rising cases and lockdowns until next summer.  

I'm done trying to predict what the government will do.  Perhaps they'll get an advisor who understands what is going on, and start listening to them.  We can but hope.


Report
In reply to wintertree:

Thank you for the detail, that all makes sense. Winter will be hard - but it was always predicted to be. More people seem to be realising this now...

Report
In reply to wintertree:

> Cases fall more slowly in lockdown than they grow before lockdown

Sadly, that seems to be mostly true. But, as we've discussed earlier, there are examples of falls being as rapid as the rise; Belfast for instance (although that now seems to have stalled). So it does look like it is possible to bring numbers down rapidly (assuming there isn't some reporting or other anomaly going on in Belfast).

We need to find out what they're doing in the areas where numbers are falling rapidly.

Given that I've heard figures of 11-14% for compliance with self-isolation after track and trace, I'm not surprised that numbers either aren't falling, or are only falling slowly...

Report
 jkarran 14 Nov 2020
In reply to Misha:

My guess is this 'lockdown' isn't going to be that effective: restrictions are lax, schools and most workplaces open, covid is everywhere, adherence is patchy, community spirit lost and enforcement absent. Sending the unis home (despite testing) is likely to trigger a pre and into Christmas infection wave that given family extended gathering habits over the holidays could be quite lethal.

So, we'll know it's proving quite ineffective soon (if I'm right) and I can't be the only one seeing the potential interaction between students and xmas. Also I doubt there will be reactive tightening of restrictions this week as the data comes in.

Early December we have the desperate need to get the high streets open for their most profitable time of year weighed against the need to not have overcrowded suggesting a policy of reopening shops as early as possible, perhaps with encouragement to shop early.

All the while prevalence starts high and grows fast again as the economy reopens. Factor in the likely spike because of unis/xmas. Something needs to be done to limit growth after the end of 'lockdown' to push the unavoidable problem well into new year.

If it was my job to guess at this with limited data and understanding: I'd probably keep the pubs and eateries closed another week, xmas parties banned anyway so no huge additional loss, I'd probably keep the inter-household and group size restrictions in place well into mid-late December and I'd expect to be back in 'lockdown' early Feb if forced, week one or two of Jan if taking planned action to complete the work of this November 'lockdown'.

All that is complicated by the 'VV day' feeling and resulting individual relaxation in adherence there will be as (if) the vaccine rolls out in that period. The 'it's all over' will be hard to counter because doing so will be unpopular, it doesn't boost the government in the polls, it doesn't sell copy. Also there is whatever antisocial agitation we have to deal with in this period, I don't know how effective this is in reality.

The period between first vaccinations and hitting meaningful herd immunity (not from this vaccine if only 40M vials available) will be the biggest challenge of the pandemic yet, assuming this is understood in Whitehall which it should be, it's going to take exceptional leadership to answer the frantic calls (press, public and backbench and saboteur) for early and complete relaxation of restrictions then assuming sense prevails to maintain public adherence to any remaining restrictions. With spring arriving restrictions will by then both feel very old and quite unnecessary since 20M vaccines administered wisely should have seen the death rate drop off a cliff (in the presence of then proportionate restrictions). 20M vaccinated (even done well) won't however be sufficient to prevent a very deadly third wave in spring if we behave like it's all over.

That's how I see it and what I think probably should happen. I genuinely have no idea what will happen and if brexit triggers another wave of panic buying we have a whole new factor to consider.

jk

Post edited at 23:55
Report
In reply to Misha:

Thought occurred to me tonight if schools will close a week early before Xmas to give a bit more space before family gatherings?

Post edited at 00:30
Report
 profitofdoom 15 Nov 2020

In reply:

Thanks for posting, everyone. Very helpful stuff and I'm trying (with my little brain) to figure it all out

I wish people would stop talking about a "lockdown" this month, though. In no possible way, shape or form do we have a lockdown now - a lockdown is when no-one goes out for a set period

Report
In reply to profitofdoom:

I agree, the first lockdown was lockdown lite, this one is just none essential shops and pubs are shut. No one is actually locked down. 

Report
In reply to wintertree:

With regard to flu season, if people are following the covid rules they shouldn't be getting the flu. 

Report
In reply to Misha:

Folk will rush out and shop, drink and party, then numbers will rise again. 

It's a game of politics open up on  2 Dec and numbers will be climbing before Xmas and they'll be under pressure to cancel Santa. Extend the current restrictions by a week or two and the numbers won't be climbing until santas been. 

What they should be making sure is that folks know all the big figure new years eve events are totally  off the agenda. 

Report
 Si dH 15 Nov 2020
In reply to Dax H:

> With regard to flu season, if people are following the covid rules they shouldn't be getting the flu. 

My understanding is that kids at school are known to be one of the biggest vectors of flu so flu season would still be expected to happen. Of course, if kids aren't seeing their grandparents then not so many vulnerable people should get it.

They have extended the flu vaccine programme this year so hopefully that will help.

Post edited at 08:21
Report
 Si dH 15 Nov 2020
In reply to Misha:

From reviewing the trend of data over the last few weeks I'm confident infections will be dropping fairly fast by the end of lockdown but at that point the actual prevalence may be quite high, certainly in the data .

Infection rates in Wales are now shown to have dropped quickly across the country, especially in high prevalence areas, but it didn't become really clear in the data until the last couple of days, which was 2.5 weeks after their firebreak started. All LAs in South and North East Wales, where rates were high and there are denser populations, are now (in the data) dropping by between 15 and 50% week on week.

Evidence is now clear that Tier 3 was enough to hold down infections and , in the vast majority of places, turn them over a bit (even in GM now) but not enough to make them reduce quickly. Bits of Liverpool have now levelled off rather than continuing to fall as they have been over the last month. It's also clear that Tier 2 hasn't had enough effect to stop rises and Tier 1 has allowed for very fast rises, ie many doubling times of a week or even less on a local authority basis.

I think if we are to allow people to get together in groups of six over Christmas without catastrophic effect and another lockdown in short order, we really need to keep at Tier 3 through December in most places or everywhere, then just drop to a Tier 1 or equivalent for a few days at most. There is certainly no justification in the data for relaxing things more than that. Of course it may still happen.

I think if we stuck to what I suggest then went back to Tier 3 in January and if the testing programmes were working more effectively (not necessarily to break the chain, but just to spot increases reliably and quickly in a population to inform govt action) then we could proceed to gradually relax from a Tier 3 position while closely monitoring local infection rates (reintroducing Tier 3 after a relaxation if necessary), and probably get through without a lockdown to the point that either vulnerable people are vaccinated or the weather reduces the size of the problem a bit again. At that point we will relax back as we did this last summer, whether appropriate or not.

Edit to add, on the point of what happens after vulnerable people are vaccinated - Van Tam even said at the last briefing that he would hope at that point restrictions could be relaxed. I really don't think Govt will see any justification for significant restrictions to prevent the rest of us getting it if the death rate is low.

Post edited at 08:27
Report
 Si dH 15 Nov 2020
In reply to wintertree:

On mass testing, there is now data being produced about the results of a full week's worth of lateral flow testing in Liverpool. They have also been doing asymptomatic PCR testing in parallel but haven't released any of the results from it yet. As long as they do so in a clear way then it should be straightforward to review the comparative positivity rates for a large asymptomatic population from the two types of test, which would say a lot about how effective the lateral flow tests are.

Of course if the results are bad (far higher positivity in PCR tests) then they might not release them.

As of Friday night they had tested over 50000 with lateral flows and about 40000 with PCRs.

Report
 kaiser 15 Nov 2020
In reply to Misha:

A lot will depend on Carrie now she's in charge.

My guess is that she'll be much keener on lockdown than Dom was.

Report
 wintertree 15 Nov 2020
In reply to Dax H:

> With regard to flu season, if people are following the covid rules they shouldn't be getting the flu. 

In reply to Si dH:

> My understanding is that kids at school are known to be one of the biggest vectors of flu so flu season would still be expected to happen.

If you look at Figure 12 in the latest PHE surveillance report for this year [1] Rhinovirus positivity rose especially in school aged children.

I think this normally comes before the Influenze rise, although I can't now find the figure I got that from.  Last winter the influenza rise started around now,  peaking around the end of December - figure 28 in [2].

Looking at figure 28 in [2] and figure 11 in [1],  we might expect (if this was a normal year) to  be in the foothills of the influenza cases right now, and they look absent from this year's data.  But it's small numbers, and there are many competing explanations for the difference.  If there is still a much lower than typical rising rate of Influenza in a few weeks time, that's really positive.   The typical spread of rhinovirus in junior school aged children suggests the control measure's aren't working well against traditional respiratory illnesses so keeping kids and grandparents separated is probably wise for now, as is making full use of the flu vaccine for both.

[1] https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/934810/Weekly_Flu_and_COVID-19_report_w46_FINAL.PDF

[2] https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/895233/Surveillance_Influenza_and_other_respiratory_viruses_in_the_UK_2019_to_2020_FINAL.pdf

Report
 profitofdoom 15 Nov 2020
In reply to Dax H:

> ................if people are following the covid rules they shouldn't be getting the flu. 

But flu can be brought into the house by kids coming from school. and/or by people coming from work or coming from anywhere else?

Report
 elsewhere 15 Nov 2020
In reply to Misha:

Cowardice, prevarication, cronyism, new promises to distract from failures to deliver on earlier promises. More of the same.

Post edited at 11:08
Report
In reply to Bobling:

> Thought occurred to me tonight if schools will close a week early before Xmas to give a bit more space before family gatherings?

Given how teenagers are spreading it around like there's no tomorrow, I would see great benefit in say a 4 week Christmas holiday for secondary education.  It might well bring cases down more than other measures.

Report
In reply to profitofdoom:

> But flu can be brought into the house by kids coming from school. and/or by people coming from work or coming from anywhere else?

It can, but it's still low this year so far because the COVID measures also stop other infectious diseases from spreading as widely.

Report
 wintertree 15 Nov 2020
In reply to Si dH:

You don't happen to have produced a nice list of dates and Tier levels for each UTLA as part of the detailed pouring through regions?  I ask as I'd like to colour a plot...

I changed my UTLA plot a bit so it now show's the rolling average case number (x) and the rate of change as a day-on-day percentage (y).  

The second plot adds trails showing the location of UTLAs over the last 21 days, with them fading from invisible at 21 days ago to dark lines by the present day (Nov 9th, including the lag effects of the 7-day rolling average PHE needlessly apply to the data before releasing it).  I've also put a polynomial filter on them as they're quite raggedy.  If you had something I could easily digest, I could colour the trails by the tier and lockdown level at the time...  This is a bit of a crap plot but if you stare at it it's quite interesting - everywhere has a decreasing day-on-day rate but what's curious is that for places where the rate has gone negative and cases are decreasing, the day-on-day rate then decreases in magnitude towards 0% before cases get to 0.  This is contra to a simple exponential growth model and I've not finished thinking through the implications of what it tells us about what's going on.  It could be that it's an effect suddenly fragmenting a transmission network by removing some links, with some fast mechanics gone another other slower ones still in play.  The good news is that everywhere looks to be moving in the right direction.  

Edit:  I'm wondering if the reduction of % rates following reduction in numbers is actually an artefact of the way the 7-day moving average is done in the UTLA data.  I'll have to look at this same period using the data released in a week's time and see if it changes.

I know the second plot looks like a dog's dinner!  

Post edited at 11:43

Report
 wintertree 15 Nov 2020
In reply to jkarran:

> With spring arriving restrictions will by then both feel very old and quite unnecessary since 20M vaccines administered wisely should have seen the death rate drop off a cliff (in the presence of then proportionate restrictions).

Depending on how cynical you are, the total absence of an age-based breakdown of efficacy in the Pfizer press release suggests they've got nothing to shout about when it comes to protecting the over 85s (either no data, or not compelling data), where much of the deaths are coming from and where the immune systems a vaccine relies on are low functioning.  

Another possibility is that the vaccine reduces the severity of infection in the oldest people and drops their death rate but doesn't stop things reaching the hospitalisation stage, so paradoxically vaccination of the most elderly first could raise total hospital occupancy in the same way some have suggested that the improved treatments have; if someone dies they stop taking up a hospital bed, but if they come back from the brink it takes some time to get them well enough to be discharged.

Until the Phase 3 results are published I think it's really hard to predict what's going to happen - and the optimums vaccination strategy is going to depend strongly on those results.

Report
In reply to wintertree:

> Cases fall more slowly in lockdown than they grow before lockdown.

That's because UK lockdown was lockdown lite.  This will be the problem on December 2nd. Everyone will be wanting / expecting lockdown to end, however the live case count is unlikely to have reduced enough to give sufficient breathing space - it will be like the USA in the summer, a brief flatlining of case count increase, then it'll take off again.  In a nutshell, unless they've used the 2 weeks to improve testing or T&T it will have been a waste of time (again).  As evidenced by Lemming and Stichtplate, the current level of prevalence is already at the limit, so they can't unlock very far over christmas from only a slight drop in prevalence or they'll crash the NHS in the new year.  Office Xmas Do's simply cannot happen this year. Small family gatherings will almost certainly be allowed in order to prevent 'xmas being cancelled' and mass anti-government feelings even if they're a daft idea.  Having been busy for a couple of days I looked at my stats and the live case count last night and rate of increase has increased again - what your plots show is what I posted about a couple of days ago as you don't look at the latest data (and rightly so for the things you look at).  It would appear that it takes bang on 2 weeks for the effect of restrictions to start showing in the data (see the graph below).

Anyhow, Thomas Pueyo put out some new musings which will be of interest:-https://tomaspueyo.medium.com/coronavirus-the-swiss-cheese-strategy-d6332b5939de

On the subject of travel testing and quarantine, there's hard evidence that the 'test on arrival' desired by Heathrow is not effective at protecting a population. Jersey have been doing this since early July, and since September it's allowed infective people to enter the community and they now have community seeding. Something like a 7 day (or more) quarantine + test seems to be the way forward.

Post edited at 12:28

Report
 RobAJones 15 Nov 2020
In reply to Neil Williams:

> Given how teenagers are spreading it around like there's no tomorrow

bur some people ,who just read the headline, rather than the full report will say that SAGE are saying that the children are less likely to spread it when the schools are open.

“There is no current direct evidence that transmission within schools plays a significant contributory role in driving increased rates of infection among children

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/935125/tfc-covid-19-children-transmission-s0860-041120.pdf

Report
In reply to Bobling:

> Thought occurred to me tonight if schools will close a week early before Xmas to give a bit more space before family gatherings?

If that happens and I have two weeks' holiday after a long term locked down just so that people can behave covid-recklessly over Christmas I shall be extremely unhappy. 

Report
In reply to Robert Durran:

> If that happens and I have two weeks' holiday after a long term locked down just so that people can behave covid-recklessly over Christmas I shall be extremely unhappy.

You'd best steel yourself, as that is what is going to happen, regardless of what Bozza or anyone else says.

Post edited at 12:47
Report
In reply to Neil Williams:

> You'd best steel yourself, as that is what is going to happen, regardless of what Bozza or anyone else says.

A pre-Christmas lockdown or the irresponsible behaviour at Christmas (which I agree is going to happen whatever)? 

Report
 profitofdoom 15 Nov 2020
In reply to Neil Williams:

> It can, but it's still low this year so far because the COVID measures also stop other infectious diseases from spreading as widely.

OK, I see, I missed that. Thanks a lot for your reply

Report
 didntcomelast 15 Nov 2020
In reply to Misha:

I’m curious as to how many people in the U.K. have actually had Covid. If you are to add the estimates and actuals in respect of number infected it would appear we are into the millions of U.K. residents have already had and either recovered or died from Covid.  

Going by the latest estimates of between 20000 & 50000 infections a day over say a period of say 200 days, that’s a fair few people. 

Am I over simplifying this or has it gone through a fair chunk of the population already? 

Report
 wintertree 15 Nov 2020
In reply to RobAJones:

I (still) haven’t read the full report that you also linked a few days ago...

> “There is no current direct evidence that transmission within schools plays a significant contributory role in driving increased rates of infection among children“

There is very little direct evidence about any of the transmission events driving spread in the UK.  It’s almost all reasoned supposition.  So in isolation I don’t draw much from the quote you posted from the report.  I’ll try and read it all when I have time...

Report
 RobAJones 15 Nov 2020
In reply to didntcomelast:

The problem with your logic is that for most of those 200 days there were far fewer people being infected.

1.3 million have tested positive, but lots of people will have had it without being tested.

50000-60000 have died, most estimates put the infection fatality rate for UK between 0.5 and 1.2. So a ball park figure would be between 5 and 12 million infections so far. There will surveys that have randomly tested people , so more accurate than my fag packet calcs.

Report
 didntcomelast 15 Nov 2020
In reply to RobAJones:

Yes realise without going into all the various data bases my calculations were blunt but it still shows that a huge number of people have already had Covid and if you count the folk from the first round who were not tested but had some of the now known symptoms or even no symptoms, a fair number have been infected at some point.  

Report
 RobAJones 15 Nov 2020
In reply to didntcomelast:

Yep, I agree a significant number of people have had it, but not enough to make much difference as the how it will spread. Again ball park but if 50,000 people have died after 10% have been infected,  300,000 will die before  we reach anything approaching herd immunity (if that is actually possible)

Report
 didntcomelast 15 Nov 2020
In reply to RobAJones:

Please don’t think I want to get into an argument with you over your figures. I’m 55 with a heart condition and high cholesterol so technically am at a higher risk of death from this damned virus, but you have quoted 50000 deaths for 10% of the population which I’m guessing you’ve estimated to be 5 to 6 million people giving a total death toll of 300000. If you were to use the higher estimate of 12million already infected people that would potentially result in a total death toll of 150000 people which is a huge amount but actually half your prediction. That figure is again based on an estimated number of people infected as in the first few months no one had any idea how many had contracted the virus. If more people had been infected in the first wave than current estimates state then surely the potential is for even fewer total deaths.

what I’m try to say, badly I know, is that is the damage we are causing to the population of this country in terms of lockdown actually worth it? I’m an older person I’ve had a career and earned my pension, the millions of people under the age of 30 haven’t and we seem to be inflicting upon them a terrible legacy of debt and stress. I know it sounds really cruel and heartless but are we trying to save the lives of the old and infirm to the detriment of the young and healthy. My heart goes out to young folk who probably cannot see any light in this long dark tunnel. 

Report
In reply to didntcomelast:

> what I’m try to say, badly I know, is that is the damage we are causing to the population of this country in terms of lockdown actually worth it? I’m an older person I’ve had a career and earned my pension, the millions of people under the age of 30 haven’t and we seem to be inflicting upon them a terrible legacy of debt and stress. I know it sounds really cruel and heartless but are we trying to save the lives of the old and infirm to the detriment of the young and healthy. My heart goes out to young folk who probably cannot see any light in this long dark tunnel. 

Here's how I see the arguments about lockdown, from another thread:

To go right down to the underlying philosophical basis, I think there is a utilitarian consensus on the response to the pandemic: we all (except people like Nigel Farage, but lets leave them out) want to reduce the amount of suffering.

I think it's completely accepted across the political spectrum that deaths is not a good measure of suffering. There isn't really anyone proposing a religious "sanctity of life" type position, that saving lives of people who are nearly dead anyway is more important than ruining the lives of others. I haven't heard that feature seriously in the debate, although we do tend to hear a lot about the number of deaths because it's striking and emotionally salient. I don't believe however that the policy is really being motivated by number of deaths, I think it's motivated by overall damage to society.

The "pro-lockdown" position is that the economic and social pain paid up front in a lock down will save worse economic and social pain in total, because without lockdown policies the healthcare system will crumble, people will have to care for dying family without state support, followed by economic armageddon. So the "pro-lockdown" position gives a net result which is pretty bad, and has to be paid right now, but better than the alternative - which hasn't even been set out in a way that makes it possible to evaluate.

The "anti-lockdown" position is that we can avoid paying the economic and social pain now, and that it'll all be alright somehow. We'll "protect the vulnerable" - but no one knows how. The healthcare system won't be overwhelmed - but no one knows how. Or it will be but it won't matter - but no one knows how. The anti-lockdown position is like a religious belief "somehow, a benevolent force exists that will make the reality we live in, not the reality we live in". 

Both positions value exactly the same things. Both sides can see that schools are pretty much the number 1 priority: the kids need to be there in order to have a decent start to their lives (and even to be fed and sheltered from abuse in the worst cases); and then the parents can work. To be blunt, no one on either side cares about the 85 year olds who'll die from the virus. They're 85 for god's sake. But the "pro-lockdown" position has bothered to consider the practicalities: if someone's 85 and gets the virus, what do we do? They need care in their last days, and they need to be buried. Modern families are not equipped to do this without support from the state. The NHS is the system for this (not the actual burial, but that system has finite capacity too). It is incredibly important that its capacity is not exceeded. The "pro-lockdown" position understands this, whereas "anti-lockdown" just hasn't managed to look it square in the face. What does a family do when two parents are dying of covid and the ambulance does not come? Who deals with it? What happens? What is the impact?

Both positions seek exactly the same outcome: to reduce overall suffering, including grieving those lost to the virus, and job losses and poverty, and social isolation. It all counts, to both sides. The "pro-lockdown" position is prepared to pay upfront to lessen the total impact. The "anti-lockdown" position wants to try to wing it based on faith and dishonesty and a failure to face the facts. One argument is strong, and one is weak.

(There is also the issue of bad faith actors, e.g. Nigel Farage, but that's really a side-show).

Post edited at 14:58
Report
 wintertree 15 Nov 2020
In reply to didntcomelast:

> If more people had been infected in the first wave than current estimates state then surely the potential is for even fewer total deaths.

If you assume measurements of antibodies (“seroprevalence”) correlate somewhat with immunity, then no that potential isn’t there.   Likewise if mild and asymptomatic infection doesn’t confer much immunity - as seems likely, then no, that potential isn’t there.  If there was significant herd immunity from naturally acquired infections, the recent university outbreaks wouldn’t have spread as they did - like wildfire.  Wishful thinking here is barking up the wrong tree for these and a whole bunch of other reasons.

> I know it sounds really cruel and heartless but are we trying to save the lives of the old and infirm to the detriment of the young and healthy

Its not, and never has been, primarily about saving the elderly from covid.  It’s about keeping our humanity by treating those - not just the elderly - who could be saved from it, and about not collapsing healthcare systems in doing so, because an awful lot of younger people will have their lives ruined if healthcare collapses directly or indirectly.

Ask a young person if they’d rather the economy took a hit or their grandparents are left to die at home with the young person trying to spoon feed them some oramorph, as their parents can’t afford to risk the long term health damage they risk from covid.

Really at its core, this all comes down to what kind of society we want to live in.  

Report
 George Ormerod 15 Nov 2020

> what I’m try to say, badly I know, is that is the damage we are causing to the population of this country in terms of lockdown actually worth it?

This has been stated numerous times, but it’s not either health or the economy.  Data from the US States and Sweden vs other Nordic countries show if you have relaxed vs stricter measures you get similar economic impacts, but you kill fewer people with a robust public health response.  Also historically in pandemics areas that had the harshest lock downs had the best economic recovery. 

Report
 wintertree 15 Nov 2020
In reply to George Ormerod:

> Also historically in pandemics areas that had the harshest lock downs had the best economic recovery. 

It’s worth pointing out that this has been known for a long, long time. Healthcare or the economy has been a false dichotomy since day 1 of this crisis.  To give our PM his due, he gave every semblance of understanding this and defending it in his speech announcing this current lockdown-lite.  

Report
 RobAJones 15 Nov 2020
In reply to didntcomelast:

As you say it's guesswork, but I prefer to err on the pessimistic side as we can't be sure. To be honest the latest IFR for the UK is over 1% (700,000 deaths) even that isn't worst case as that assumes the health service isn't overrun.

Are we protecting the old? The currently we seem to be in a position where the health service is just about coping a 400(ish) people are dying every day. Most kids in school are pretty positive and happy they are doing their best to protect their parents and grandparents. 

Financially and age wise  I'm in a similar position to you. I think the likes of us (rather than people under 30) should be paying some of the cost.

Report
 RobAJones 15 Nov 2020
In reply to George Ormerod:

As an example of this. It seems to be generally accepted that had we "locked down" 4 days earlier in March, we could have relaxed restrictions around a month earlier. So fewer deaths and less impact on the economy.

Report
 Si dH 15 Nov 2020
In reply to wintertree:

> You don't happen to have produced a nice list of dates and Tier levels for each UTLA as part of the detailed pouring through regions?  I ask as I'd like to colour a plot...

No, but here are the ones I know.

North East (Northumberland, Newcastle, Tyneside Gateshead and I think Sunderland but from memory not Teeside area, you might know this better than me) was the first region of significant size to have indoor mixing entirely banned on 30/09. (Leicester and Bolton might have had something earlier that was subsequently relaxed again?) Prior to this there were a variety of restrictions in GM which are probably best described as Tier 1.5 or 2 with twists.

The above was replicated in Liverpool region from 03/10 (Liverpool, Knowsley, St Helens, Halton, Wirral, Sefton.)

On 14/10 the Tier restrictions came in. Liverpool region went to Tier 3 and lots of other places went to Tier 2. For the North East this was actually a minor relaxation.

There is a list which tells you where and when entered different tiers 2 and 3 from 14/10 in this Wikipedia article.

https://en.m.wikipedia.org/wiki/COVID-19_tier_regulations_in_England

Post edited at 15:46
Report
 didntcomelast 15 Nov 2020
In reply to wintertree: Alas for me your reply has highlighted my ignorance.  I don’t have a university education I see things from the level of the coal face as it were.  I appreciate that there are a whole host of issues this virus has raised within society and a lot of soul searching will result. I know that as you mention the Virus spread through university students like wildfire, they had no immunity from it but to an extent, with some lockdown restrictions, that rate has fallen dramatically, the behaviour of students though has not changed that much. Shieldfield and Jesmond are little different today as they were this time last year.  It seems a little unfair to phrase your words, for young people to take a hit on the economy versus feeding elderly relatives Oromorph as they die. That hit could be unemployment, homelessness and despair enough to drive a young person to take their own life. 
im sure that in time we will find a solution whether that be a vaccine or so called herd immunity ( I doubt the latter ). My thoughts poorly worded were from observation and an appreciation of younger peoples concerns  

Report
 wintertree 15 Nov 2020
In reply to didntcomelast:

> That hit could be unemployment, homelessness and despair enough to drive a young person to take their own life. 

Which will not be spared but rather worsened by any of the alternatives to strict control measures.  It’s well documented that when people are dying in their tens to hundreds of thousands without access to healthcare that people go in to their own lockdowns with worse economic damage than earlier control measures would have caused.  So all the drivers of youth suicide you give are worse, and there are more - like watching peope you know loose babies without healthcare for delivery and watching grandparents die without help or medical care, and wondering what to do with the body.

> It seems a little unfair

Pandemic or not, what seems unfair to me is how rich the UK is and how many people are hung out to dry - pandemic or not.  The pandemic is shafting more people but we were already failing millions.  I would soul search over why that is, and what we can do about it.  You could argue that politics and selfishness lie at the root cause of hardship amplified by the pandemic.

The economy is built by people, for people.  It’s failing a lot of people and it’s getting worse.  I’ve seen about a hundred covid threads posing the question of saving the elderly now vs the future economy for the young and I’ve seen one or two posing the question about why we don’t change our economy.  

Report
 RobAJones 15 Nov 2020
In reply to Si dH:

> North East (Northumberland, Newcastle, Tyneside Gateshead and I think Sunderland but from memory not Teeside area, you might know this better than me) was the first region of significant size to have indoor mixing entirely banned on 30/09.

I was discussing the graphs produced by wintertree last week looking at the effect of uni. cases with some students. (Tutor time not a lesson) Two of the more interesting thoughts put forward were, that the Newcastle outbreak was started by local infections rather than students bringing it  with them (One of the girls had a brother who has just started at Newcastle Uni) due to their infections increasing later, whereas in Nottingham the increase appears to happen at the same time. The other was that there were greater restrictions in Newcastle  (compared to Nottingham) at the time due to the rate of increase being less.

Report
 Si dH 15 Nov 2020
In reply to RobAJones:

Yes, the main increase in Newcastle happened separately from students, like in Liverpool and Manchester. If you look back through the data on the dashboard (the easiest way to do this is to zoom in on the area of the country you are interested in using the map they provide, then move the slider), you will see that the outbreak there began at the very beginning of September, several weeks before students started moving around. Like Liverpool, they actually showed a drop from early October - the effect of banning indoor mixing I assume. In this period the areas most affected by students were still going up. However the rate in Newcastle turned upwards again a week or two ago.

Nottingham rates were relatively flat until the end of September then spiked suddenly following the student move. Some other places like Brum, Shef and Leeds are a bit less clear, they were already increasing before the student moves but were well behind the worst areas, and after the students moved they started catching up and (quite clearly from the map) spreading infection to the surrounding areas.

Post edited at 16:42
Report
 wintertree 15 Nov 2020
In reply to Toerag:

Nifty plot.  That would suggest another week or so before we see the effects of lockdown in the English case data resolved by specimen date.  The cases data seems particularly shonky at the moment; hospitalisations are rising again - perhaps lagging the return to rising cases after the half term period?  Lots of uncertainty around everything is my sense right now.

> This will be the problem on December 2nd. Everyone will be wanting / expecting lockdown to end, however the live case count is unlikely to have reduced enough to give sufficient breathing space

Yup, as it stands I can’t see them releasing lockdown in about a third of the country unless to an elevated Tier 3.  The decay rate of cases won’t be known for perhaps another 10 days but my suspicious is it’ll be a lot less than last time round.

Furlough isn’t going to make up for the usual burst of income from Christmas meals at pubs etc; really hard times for that sector ahead.  

Report
 didntcomelast 15 Nov 2020
In reply to wintertree:

Again forgive my ignorance but where in recent times has a developed country had a situation where hundreds of thousands has been dying in a short space of time without intervention leading to a self imposed lockdown. 

Report
 wintertree 15 Nov 2020
In reply to Si dH:

> Of course if the results are bad (far higher positivity in PCR tests) then they might not release them.

Not that PCR is that great for pre symptomatic individuals either.  Prevalence seems pretty low in students now the initial mixing events of coming in to halls and houses are well passed. If they can stick to the rather bleak and regimented conditions in halls until their “evacuation” it could be okay.  I’m not sure all universities are going to regiment their departure as the government suggested, and if they hang around after lockdown ends.... ??

Report
 wintertree 15 Nov 2020
In reply to didntcomelast:

People have been locally isolating since the Black Death.  I don’t think people have changed that much since then except that we’re ever more removed from death and it’s consequences, so will take it all the harder.  There’s no shortage of studies over the 1918 flu, and there was copious evidence of people and employers heading towards self imposed aspects of lockdown in the two weeks before the government acted in the UK.  Drug gangs enforcing lockdown in the slums of Brazil in the absence of government action being another example.

You seem to be hinting that somehow we could all go about our cheery way to cafe’s and pubs amidst a wave of mass death and the dispatcher on 999 saying “there are no ambulances”.   It’s not that I dislike that possibility, I just don’t think it’s even a possibility.

Report
 climbercool 15 Nov 2020
In reply to RobAJones:

> As an example of this. It seems to be generally accepted that had we "locked down" 4 days earlier in March, we could have relaxed restrictions around a month earlier. So fewer deaths and less impact on the economy.


everyone everywhere (including me) seems to agree  that we should have locked down earlier in march,  but at the same time everyone  in september seemed to agree that only regions of the country showing higher levels of covid should  be put under strict lockdown restrictions , this seems a contradiction to me, if  locking down early means yo u can come out of lockdown much quicker why was everyone agreeing that we should only lockdown the worst affected areas.

I guess my question is what rate of infection is optimal for lockdown, how long should you wait before locking down, is it always best to lockdown early?   My guess is that if you are going to do a lockdown, than early is always better and locking down a population with minute levels of covid actually makes sense but it is just impossible to get the support of the population when levels are so low.   But this is totally uninformed speculation so i would be interested to hear from someone better informed as to when is the ideal time to do a lock down if your goal is not eradication but just to buy time.

Report
 RobAJones 15 Nov 2020
In reply to climbercool:

> everyone everywhere (including me) seems to agree  that we should have locked down earlier in march,  but at the same time everyone  in september seemed to agree that only regions of the country showing higher levels of covid should  be put under strict lockdown restrictions , this seems a contradiction to me, if  locking down early means yo u can come out of lockdown much quicker why was everyone agreeing that we should only lockdown the worst affected areas.

Not everyone, in August everywhere had low infection rates compared to now. My view was that in September we should have restrictions so R was about 1 anything higher is not sustainable.  With schools opening this was going to be an  estimate, but to me it seemed obvious that infections were going to rise during Sept and Oct. 

> I guess my question is what rate of infection is optimal for lockdown, how long should you wait before locking down, is it always best to lockdown early?   My guess is that if you are going to do a lockdown, than early is always better and locking down a population with minute levels of covid actually makes sense but it is just impossible to get the support of the population when levels are so low.   But this is totally uninformed speculation so i would be interested to hear from someone better informed as to when is the ideal time to do a lock down if your goal is not eradication but just to buy time.

As you say I don't know, but it would be at a rate where people who do, make considered rather than panicked decisions. It would depend on the value of R as well. When R was about 3(ish) four days made a big difference. When it is 1.3 being couple of weeks late is less serious. I get your point about the support of the population, but by now most of us will be aware of people who have died or been seriously affected. Also whether and area is in Tier 1 or Tier 3 seems to have made a difference, I think that has more to do with the restrictions than the infection rate (but I might be wrong)

Report
 Webster 15 Nov 2020
In reply to jkarran

> All the while prevalence starts high and grows fast again as the economy reopens. Factor in the likely spike because of unis/xmas. Something needs to be done to limit growth after the end of 'lockdown' to push the unavoidable problem well into new year.

Why? whats the point of pushing it back as long as possible? your just delaying the inevitable. the shit has already hit the fan and splattered all over the wall with this virus, its just going to keep coming back with a vengeance each time we lift restrictions whatever we do. may as well open up the economy as quickly as possible well we can, at the time of year when is going to mean the most to the most people in terms of moral, and accept that we will be locking down again pretty quickly soon after. kicking the can down the road achieves nothing.

Report
 wintertree 15 Nov 2020
In reply to Webster:

> Why? whats the point of pushing it back as long as possible

We either pogo in and out of lockdown with cases low or cases high.

If we keep cases low, people can go to cafes and people can go to the GPs and go to hospitals for arthritis surgery and for cancer surgery.  Schools mostly run as normal.

If we keep cases high, cafes shut and routine and cancer surgery is cancelled as hospitals get overloaded.  Schools and parents are highly disrupted by class after class isolating after one kid tests positive.  Thousands to tens of thousands more people die and tens to hundreds of thousand more people have bad illness and miss work and potentially income and many will go on to have longer term health damage.

> kicking the can down the road achieves nothing.

It allows more of the economy to remain open and allows non-covid healthcare to remain open.  It saves medical staff from burnout and minimises public health damage and death.  This delivers us in to the vaccine enabled future with far less economic and health damage.

Post edited at 18:13
Report
In reply to Robert Durran:

> the irresponsible behaviour at Christmas (which I agree is going to happen whatever)?

That.

Report
 Webster 15 Nov 2020
In reply to wintertree:

cases are high, and are going to stay high between now and christmas whatever we do. they are never going to come down to the low single digit thousands unless we have a full on winter long lockdown. and im talking no school or uni, no work, nobody leaves the house for more than an hour walk a day as per the french lockdowns. We know that is never going to happen, therfore we know we are going to need another lockdown sooner or later in the new year. may as well enjoy xmas while we can and accept it

Report
 George Ormerod 15 Nov 2020
In reply to didntcomelast:

> Again forgive my ignorance but where in recent times has a developed country had a situation where hundreds of thousands has been dying in a short space of time without intervention leading to a self imposed lockdown. 

In a way the public's response to COVID-19 in March.  The country was self locking down 1-2 weeks before the Government bowed to the inevitable and instituted a lock down.  The company I work for took action in the UK in that timeframe before the 23rd, with everyone working from home.  It doesn't take too much to imagine what would have happened if the Government hadn't had an official lock down:  Lots more dead people and similar economic damage.

Report
 didntcomelast 15 Nov 2020
In reply to wintertree:

I struggle with the example you have given being the flu pandemic of 1918. That implies that in the 100 years since then nothing has changed, at the time of that pandemic the world was on the very first steps to recovery from the most costly war in history in terms of loss of life, serious injury and financial loss. Many countries were almost bankrupt. We had no National Health Service and far less knowledge of virus infections as we know today. Medical treatment in terms of medication and procedures have developed massively, the world was simply not as capable of dealing with the pandemic as we are now. 
You also mention that I seem to believe we should all be cheerful and carry on as normal, well, we cannot carry on quite the same but it’s not a reason not to have a cheerful outlook on life. If as was mentioned by another earlier we estimate that between 5 and 12 million people have been infected already by COVID and some 50,000 have died. That means millions have not died, they have survived and are living their lives to the best they can. We should be optimistic in the face of adversity. We will get through this but dwelling on the negative aspect is not the way forward. 

Report
 wintertree 15 Nov 2020
In reply to didntcomelast:

I really don’t understand your point at all.  I gave more than one example but you refer to only one.  I gave two examples from this very pandemic of people going in to isolation in the absence of government orders.

The fundamentals of infection control in a pandemic haven’t changed in 400 years.  Life is more valuable now than in 1918, our supply chains are more fragile (more global, more “just in time”).  Some of our productive people are far more reliant on healthcare to remain productive (because they can live with illness that once would have killed them).   Pandemics hurt the economy worse in some ways than they did.  

> We should be optimistic in the face of adversity. We will get through this but dwelling on the negative aspect is not the way forward. 

I think you misunderstood me quit severely.  If the disease got out of all control people would not be “optimistic in the face of adversity”.  They’d be caring for dying relatives, mourning friends and family lost in healthcare free home births, worrying about that lump they can’t get checked out, and going in to self imposed lockdown until normal service was resumed.   This idea that we can all just say “oh f**k it” and got on with life pretending there isn’t a pandemic on is just risible.

You suggest we are capable of dealing with this pandemic now - despite a century of medical advances we were reduced to a panic driven lockdown in March, and all the control measures and new treatments didn’t prevent that happening again.  In private houses people are the same as they were a century ago, but with closed windows and central heating instead of open fires drawing fresh air in to the house.  Up until the point a person hits hospital little has changed from a century ago.  Once they hit hospital they stay there longer than in the past as we can keep them alive trying to heal them. 

Post edited at 18:57
Report
 didntcomelast 15 Nov 2020
In reply to wintertree:

True you gave two examples. The Brazilian one is a somewhat different scenario to the U.K.  they have a leader who is completely dismissive of the pandemic, that also have something of a more lawless society to our own. The favelas you mention are ruled by drug gangs due to an absence of regular law and order. I suspect some of the motives behind the drug gangs lock down are not citizen welfare as we know it. 
I am bowing out of this conversation now. I feel you and I will never see any common ground over this. I am not dismissing some form of control measures to reduce the infection rate but I firmly believe we are doing damage which we haven’t realised yet. I would be happy to be wrong though. 

Report
 wintertree 15 Nov 2020
In reply to didntcomelast:

I and another poster also gave examples of the UK in this pandemic.

> but I firmly believe we are doing damage which we haven’t realised yet.

I disagree.  Those making decisions know exactly how damaging they are.  I know how damaging some are to my household and my business.  I shared a link a few days ago on how damaging they are to child mental health.

But people make this decisions, and I abide by them, *despite* the damage because the evidence - old and new - and logical reasoning say this is the least worst choice.

Report
 NathanP 15 Nov 2020
In reply to kaiser:

> A lot will depend on Carrie now she's in charge.

> My guess is that she'll be much keener on lockdown than Dom was.

I would have thought that too but - not read through the rest of the thread yet so this could be redundant - I heard yesterday that the reason it all blew up over Lee Cain and Dom Cummings was that they had been pushing hard for an earlier lockdown and were suspected of leaking the possibility to the press to force Boris' hand. 

Report
 RobAJones 15 Nov 2020
In reply to wintertree:

 > Schools mostly run as normal.

Just heard, two of my old ones are shut next week, due to staff self isolating, but at least they're the exceptions at the moment.

Report
In reply to jkarran:

That’s a very good point about the behavioural impact of vaccines being rolled out and what happens if say 20m people get vaccinated but the other 50m don’t. I’ve been thinking about what this means for my ‘return to the office’ for example (have been WFH since March and no desire to go back any time soon even if BoJo encourages it). We have a fairly young workforce, average age around 30 at a guess, very few people over 45 and pretty much no one over 55. This means no one will get the vaccine any time soon, apart from those who are medically vulnerable. Hypothetically, a return to the office in a non-SD way would need to rely on herd immunity through infection and not too many people dying or getting long Covid etc. In this hypothetical scenario, most people would be ok but a few will get hit badly and die or get long Covid. Hence I don’t think a return without SD will happen any time soon, although I expect people will be encouraged to return in an SD way (which means once a week due to massively limited capacity) from about May / June.

I doubt vaccinating the vulnerable and health workers would in itself reduce the prevalence rate in the rest of the population if restrictions are relaxed to what they were back in August. However as you say there would be a lot of pressure to relax them and the government may well do that. Not sure the resulting spring wave would be as deadly as what we’re seeing now, as in theory the vulnerable should be protected, but it would still involve thousands of deaths based on 50m not being vaccinated and some of those who are vaccinated succumbing as well. A lot depends on how soon we can get to the required level of herd immunity through vaccination. The cynic in me expects that the government’s inability to organise an efficient mass vaccination programme will be the next scandal...

Report
In reply to Bobling:

A sensible idea but will it happen? 

Report
 didntcomelast 15 Nov 2020
In reply to wintertree:

I couldn’t resist replying when I said I wouldn’t. Yes you and another stated that the U.K. self imposed lock down in March. As I recall though the government had already started to impose conditions on businesses prior to the official lockdown so technically it wasnt solely driven by the public. 
I too abide by the rules imposed on us contrary to what I’ve written may imply. I have to the best of my ability followed every instruction and rule. I can and do so easily, I’ve worked throughout this crisis, I’ve even worked more than normal and on many occasions turned down work, I’m unsociable enough not to be concerned about not being able to socialise and I live with the only family I have other than one child living in Canada.  Life has not be too difficult relatively speaking. I imagine it’s been far harder for others and as I said in the first instance young people have suffered far more.  Let’s hope the actions we have taken so far are not seen in the same light by young people as the actions taken a few years ago by many of the older generation which will deprive young people of so much more. 

Post edited at 20:49
Report
 wintertree 15 Nov 2020
In reply to didntcomelast:

There was mass individual and corporate action far in excess of what the government was imposing in the two weeks before lockdown.  

>  Let’s hope the actions we have taken so far are not seen in the same light by young people as the actions taken a few years ago by many of the older generation which will deprive young people of so much more. 

I think our actions will be looked at negatively - because there are examples of other countries acting more swiftly and decisively and strictly with risk control measures, and they’re the ones will less severe economic damage, long term health impacts and death and less total time under onerous restrictions.  There are contemporary examples and there are historic examples.  There’s guidance from SAGE and independent SAGE to that effect.  It’s an abrogation of responsibility by leadership and a betrayal of those who suffer the most from protracted lockdown measures - which will be longer as a result of waiting until too late.  Again.  

Report
In reply to kaiser:

I thought Dom was actually pro lockdown, more so than Sunak. 

Report
In reply to didntcomelast:

At 50k a day it would still take 3 years to get everyone... and there’s no clarity on level of immunity after mild or asymptomatic infection. 

Report
 wintertree 15 Nov 2020
In reply to climbercool:

A big difference between now and March is that we now have both symptomatic and random testing for the virus giving us a good handle on how many infections there are, where they are and what's going on in general.  Back in March we had no real idea how many cases were out there or how to treat those that ended up in hospital, and hospitals were filling up alarmingly quickly.  As well has having a much better understanding of where the virus is this time around, the time it takes for cases or hospitalisation to double before lockdown was several times longer than it was back in March.  All of which

> I guess my question is what rate of infection is optimal for lockdown, how long should you wait before locking down, is it always best to lockdown early?

It depends on what your goal is.  If the goal is elimination, you need to lockdown as soon as there is community transmission, or preferably as soon as it's a risk and before it's proved (which takes precious time).  For the UK, elimination is off the cards.  So why are we locking down?  Because other risk control measures have failed to be effective enough to protect healthcare.  I think effective measures would have been within our grasp with clear, consistent, pro-active messaging and planning from government.  Perhaps I'm wrong.  

Regardless, we are where we are.  

So, when to use lockdown?   The doubling time for cases indicated we lost control around 18th August with ~1,500 detected cases a day.  There's no point locking down before this as things were reasonably under control.  Call it 3,000 infections per day at that point or 60,000 in the population for a prevalence of about 0.1%.    Hospitalisations and deaths from that level of infection were well within the ability of the systems to cope.  So, that's the point I would go for lockdown - pushes the cases back into the zone where things are under control (doubling times are very large or cases aren't growing).  

At the time it seemed like individual super-spreader events were being a lot of what tipped things over the edge; so instead of going for T2/T3 locally, a much harder - and potentially even more geographically confined - lockdown could be used as soon as prevalence passes 0.1% locally.  Whack-a-mole isn't really the right analogy, it's more a case of spotting bush fires early on and putting them out before they spread and merge.  If this new mass-testing works out well then such local lockdowns could be wound up much faster given sufficient public cooperation to help identify symptom free individuals.

Report
In reply to didntcomelast:

I get the sentiment but there are two issues. First, it is not clear if those who have mild or no symptoms can get it again. Second, it’s not just about deaths, it’s also about long Covid and lung damage. This is what concerns me. I’m 39 and healthy as far as I know, so not likely to die from it but the risk of long Covid is not inconsiderable. 

Report
 wintertree 15 Nov 2020
In reply to Misha and NathanP:

There were some suggestions the first time round as well that Cummings had been pushing for a lockdown earlier on.  I guess he thought the empty roads would help with his drive north with bad eyesight.

It is quite terrifying that DC may actually have been giving the PM the best advice; I get the impression the PM is getting his ear bent off by some of the remaining senior tories who are a bit more partial to the libertarian nonsense; but his speech announcing the recent lockdown I though made it very clear where he falls on this, and I actually found it quite comforting to hear him clearly recognise and refute the fallacies in that view - not exactly directly but in a very clear message about why we had to act.  It's maddening - he's clearly got the potential to do a good job at this.

Report
 jkarran 15 Nov 2020
In reply to Si dH:

> Edit to add, on the point of what happens after vulnerable people are vaccinated - Van Tam even said at the last briefing that he would hope at that point restrictions could be relaxed. I really don't think Govt will see any justification for significant restrictions to prevent the rest of us getting it if the death rate is low.

With only 20M people vaccinated the death rate (and more importantly the hospitalisation rate won't stay low for long if we relax restrictions significantly after vaccination. Assuming (hopefully) the reported 90% efficacy is across the board, not 95% effective in the young, 50% in the very old then at least 2M of the most vulnerable will still be vulnerable. It'll spread much less freely in semi-closed institutional communities but we're talking about this as a tool to open up society including care home visits etc. Also there will be ~10M aged 50-60, most still working, fairly vulnerable and exposed but not vaccinated. The CFR for that group seems to be in the 0.5-1% range, split the difference and assume maybe 1/3 of them get covid in the post vaccine rush for economic normality, that's another 25k dead. If the CRF for the vaccinated group is 5% (probably lowball), vaccine efficacy 90% (probably lowball) and maybe 1/5 (?) are eventually exposed in the rush for normality that's another 20k of them dead too. That doesn't look like the first wave of vaccine (given to the elderly and priority key workers) delivers any kind of real break from restrictions in Q1 or Q2 of 2021. We need more doses, or maybe a different strategy (informed by the vaccine performance and the as yet unstated objective, minimising deaths or maximising economic performance, normally a false dichotomy but maybe not here) because as is unless we're very careful there's every chance we get a big delivery of a good vaccine and a lot of deaths and no economic recovery.

jk

Report
 jkarran 15 Nov 2020
In reply to Webster:

> Why? whats the point of pushing it back as long as possible?

From the government's perspective, it's to give the highstreet the month it dines out on and to avoid the 'Grinch!' headlines.

> your just delaying the inevitable. the shit has already hit the fan and splattered all over the wall with this virus, its just going to keep coming back with a vengeance each time we lift restrictions whatever we do.

Probably, yes.

> may as well open up the economy as quickly as possible well we can, at the time of year when is going to mean the most to the most people in terms of moral, and accept that we will be locking down again pretty quickly soon after. kicking the can down the road achieves nothing.

Ok if you're willing to kill a lot of people for the sake of Christmas and to risk quickly overwhelming your healthcare system and thereby cancelling Christmas.

I do sort of agree with you here but there's a very fine line to be walked. Ending the lockdown with a return to something like tier 1 nationwide (give or take hotspots) just because 4 weeks is a round number isn't sensible. A January (maybe Feb) lockdown looks pretty inevitable but if we're going to pay the price for Christmas we absolutely have to make sure that lockdown isn't needed until after it!

jk

Report
 Si dH 15 Nov 2020
In reply to jkarran:

You might be right but really you need to be arguing with the politicians. I'm not saying we should definitely open up completely once the vulnerable are vaccinated - I'm on the fence - but I think that is what will happen. They aren't going to keep places shut to protect 50yos, or at least I strongly doubt it. It depends strongly of course on how quickly they could vaccinate them. 

Post edited at 22:05
Report
In reply to jkarran:

> From the government's perspective, it's to give the highstreet the month it dines out on and to avoid the 'Grinch!' headlines.

If only we could just cancel Christmas. It would probably only be one year and we've cancelled a whole load of other stuff like the Olympics and so on. It wouldn't be the end of the world.

Report
In reply to Si dH:

I suspect it will be as you say. It will then be down to individuals to decide what’s right for them. That’s probably ok in the context of relatively few deaths due to the vulnerable having been vaccinated (big caveat: assuming the vaccines are effective for the elderly, which we don’t know yet). Some people will happily go to pubs etc. Others, like myself, will be far more cautious, until they themselves get vaccinated. We will still have SD and face masks and large events will still be banned until well into next year, I should think.

Edit - it also depends on NHS capacity. I suspect it will take a while to reduce the number of Covid patients to a manageable amount after the winter. 

Post edited at 22:51
Report
 wintertree 15 Nov 2020
In reply to Si dH & jkarran :

> They aren't going to keep places shut to protect 50yos, or at least I strongly doubt it. It depends strongly of course on how quickly they could vaccinate them. 

Some estimates here, probably only good to within a factor of 2.

I think policy will depend on how quickly it would tear through the unvaccinated under 65s  if the mandate to follow all controls was dropped.   There’s potential for perhaps 100,000 deaths and 450,000 hospitalisations in the age range 50-65 if the virus ran riot in that age bracket.  To be able to process that many people through hospital needs the infection to be spread out over at least 8 months, which would require perhaps Tier 1.5 level measures for those 8 months.  I’m not sure how well it would go down with peope in that age bracket to find out their deaths are being strung out to be processed all tidy and orderly style so everyone else can have T1 instead of T2 for the short period until more vaccines and more doses are available.

100,000 dead in that age range removes about a half million person years of positive economic contribution which flips some of the calculations vs older groups.

As you say it depends on how fast the vaccine can be rolled out; I don’t think it would be at all sensible economically or socially to open up to pre-pandemic levels until vaccines are available to everyone over 50 or even 45.

The other big worry I’d have about cancelling all restrictions are the viral load effects.

Report
In reply to Robert Durran:

> If only we could just cancel Christmas. It would probably only be one year and we've cancelled a whole load of other stuff like the Olympics and so on. It wouldn't be the end of the world.

If, and it's a big if, they could be confident about vaccination reaching the right stage by then, they could perhaps suggest families celebrating Christmas at Easter this year.  It's a bit broken in a Christian sense but most people celebrating Christmas aren't really doing it with a Christian viewpoint.

However, pushing it out to there *then it not being possible* would be a big vote-loser.

Report
In reply to wintertree:

I suspect having been in ICU himself he’s got a different perspective compared to the libertarian loons. Some people don’t seem to take Covid seriously until they get it themselves or someone close to them gets it. 

Report
 wintertree 15 Nov 2020
In reply to Toerag:

> I looked at my stats and the live case count last night and rate of increase has increased again

I've updated my doubling times plot with today's data release.  As ever, the right hand side of all the curves on this is very twitchy and can change with future days data.  The estimates of doubling time stop 5 days in the past, as that's the last point for which the data is relatively final and not subject to much reporting lag.

The doubling times for all three measures are decreasing, meaning that the exponential rate is increasing for all.  So it looks like the plateaus on hospitalisations and deaths were temporary; not surprising as they were likely derived from the levelling of of cases around half term, an effect that didn’t last.

I've added a vertical line for when lockdown might start to show an effect on these using your 14-day guide.   It looks like all measures could increase by quite a bit before lockdown starts to bite.   In terms of hospitalisations in particular, that's not good news.

Post edited at 23:50

Report
 Rigid Raider 16 Nov 2020
In reply to wintertree:

Er... this vaccine....  BBC reported this morning that the vaccine has to be stored at minus 80c meaning it's unlikely you will be able to nip to your local surgery for a shot. You will have to travel to a regional cryogenic facility, twice at the right interval, ensuring you get the same vaccine as last time if others are available by then. All this will need to be organised next summer before winter 2021.

Report
 jimtitt 16 Nov 2020
In reply to Rigid Raider:

-80C isn't cryogenic, it's just coldish. The vaccine has to be stored overnight in a normal fridge for use the next day anyway.

Report
 Wainers44 16 Nov 2020
In reply to Rigid Raider:

> Er... this vaccine....  BBC reported this morning that the vaccine has to be stored at minus 80c meaning it's unlikely you will be able to nip to your local surgery for a shot. You will have to travel to a regional cryogenic facility, twice at the right interval, ensuring you get the same vaccine as last time if others are available by then. All this will need to be organised next summer before winter 2021.

Dry ice. Beers in one half, vaccine in the other. Nothing too special from what is being reported. 

Sheer numbers is the issue and making sure that's what is left of the non covid health service doesn't grind to a complete standstill has to be the focus.

Maybe use the army...with a drive through concept taking over sections of motorway to stack the queues and have the space for post vaccine monitoring afterwards?

Report
 Billhook 16 Nov 2020
In reply to Rigid Raider:

> Er... this vaccine....  BBC reported this morning that the vaccine has to be stored at minus 80c meaning it's unlikely you will be able to nip to your local surgery for a shot. You will have to travel to a regional cryogenic facility, twice at the right interval, ensuring you get the same vaccine as last time if others are available by then. All this will need to be organised next summer before winter 2021.

Its not difficult.  It already happens for farmers.

My daughter is one of the many people who go around collecting sperm from prize bullocks and also goes around inseminating cows.  An ordinary car will do fine, the sperm is stored in what are ,in effect vacuum flasks packed with dry ice.    

The logo on the side of her van might need changing but I'm sure she wouldn't mind a break from dealing with cattle.

Report
 TomD89 16 Nov 2020
In reply to Misha:

Why you think gyms of all places are the least likely to open? I don't agree with that assessment personally, nor do I think it's likely to see an extension beyond 2nd December (unless the gov really want to throw away their last scraps of credibility).

Report
 Si dH 16 Nov 2020
In reply to wintertree:

> > They aren't going to keep places shut to protect 50yos, or at least I strongly doubt it. It depends strongly of course on how quickly they could vaccinate them. 

> Some estimates here, probably only good to within a factor of 2.

> I think policy will depend on how quickly it would tear through the unvaccinated under 65s  if the mandate to follow all controls was dropped.   There’s potential for perhaps 100,000 deaths and 450,000 hospitalisations in the age range 50-65 if the virus ran riot in that age bracket.  To be able to process that many people through hospital needs the infection to be spread out over at least 8 months, which would require perhaps Tier 1.5 level measures for those 8 months.  I’m not sure how well it would go down with peope in that age bracket to find out their deaths are being strung out to be processed all tidy and orderly style so everyone else can have T1 instead of T2 for the short period until more vaccines and more doses are available.

> 100,000 dead in that age range removes about a half million person years of positive economic contribution which flips some of the calculations vs older groups.

> As you say it depends on how fast the vaccine can be rolled out; I don’t think it would be at all sensible economically or socially to open up to pre-pandemic levels until vaccines are available to everyone over 50 or even 45.

> The other big worry I’d have about cancelling all restrictions are the viral load effects.

I think my reply to JK yesterday was probably on an incorrect basis as the published draft vaccine priority list goes down to age 50. So I would assume from that that people aged in their 50s will get vaccinated as quickly as possible after older groups, but those under 50 may not. So I think it's unlikely someone in their late 40s or younger will get a vaccine before restrictions are removed completely; people in their 50s might.

I can't dispute your figures but I think they would fail to really gain traction with the public. I think what drives people's response is how vulnerable they feel to either themselves or someone they love getting the disease and dying from it or being seriously incapacitated. The number of people who feel that way (even of those aged 49.5) will drop massively once older groups are vaccinated. And I do think that once you are talking about substantially lower death rates and very low chance of actually overwhelming a hospital, the impact of restrictions on wider society and mental health does at least have to be very carefully balanced against the reduction in Covid deaths - I don't think that balance will be obvious even before the economy is accounted for. All this assumes the available vaccine(s) to be highly effective of course.

Post edited at 08:02
Report
 climb the peak 16 Nov 2020
In reply to kaiser:

Having read through and agreed with the more scientific and detailed posts, I think this simple explanation could be the the most likely!

Report
 wintertree 16 Nov 2020
In reply to Si dH:

I pretty much agree with all that.  Once we're vaccinated down to ~50 years of age, then it's unlikely we'll see restrictions continued, assuming the intent is for covid to become an endemic seasonal illness (vs going for near-elimination for example).

My worry is that there's been so much talk about the over 75s being the most vulnerable that it can easily skew how people approach this; this is true at an individual level in terms of death etc but in terms of vulnerability of the health service, the age bracket 50-75 is just as important where the lower hospitalisation and fatality rates are balanced by the much higher number of people.  I think there's going to be both popular and political pressure to move too fast, too soon vs what has been the guiding principle for restrictions - protection of healthcare. 

Report
 wintertree 16 Nov 2020
In reply to Neil Williams:

> It's a bit broken in a Christian sense

Well, it was never really their festival anyway...

Report
 jkarran 16 Nov 2020
In reply to Si dH:

> You might be right but really you need to be arguing with the politicians. I'm not saying we should definitely open up completely once the vulnerable are vaccinated - I'm on the fence - but I think that is what will happen. They aren't going to keep places shut to protect 50yos, or at least I strongly doubt it. It depends strongly of course on how quickly they could vaccinate them. 

I agree with the latter part of that, we will likely yet again relax to far too fast for too long but if the fag packet numbers in my last post are remotely reasonable that third post vaccine wave, if ignored, exceeds wave one, possibly one and two combined if it's stubbornly ignored. Thing is we already know deaths in those numbers are unacceptable to the public, probably even more so once the 'they were nearly dead anyway' argument is harder to make (for the 50-60 group).

Whether with the vaccine we likely have coming we could open up our economy quite soon, quite completely and quite safely, I'm not sure. It seems to me that what natural immunity exists does so mostly in the young who didn't take the same precautions and in urban workers who couldn't. I wonder if their immunity plus the vaccine for 20M non-seropositive workers (including all those working directly with the vulnerable) could be enough if targeted cleverly to get R under 1 with very limited additional restrictions. Probably not but it's a thought. Anyway, it wouldn't be popular and it'd look inhumane but if it could work it might be sensible given what could happen anyway if vaccinating the old causes us to lose control of the virus in the rest of the population anyway.

We don't have enough doses pre-bought for the <60s assuming it will be distributed to key frontline workers and the oldest first. We may be able to buy more or we may have to wait for the next wave of vaccines, either way it'll be months after the first. That's a long time to live with restrictions keeping R at or about 1 while hardly anyone is dying.

jk

Post edited at 10:57
Report
 jkarran 16 Nov 2020
In reply to wintertree:

> The doubling times for all three measures are decreasing, meaning that the exponential rate is increasing for all.  So it looks like the plateaus on hospitalisations and deaths were temporary; not surprising as they were likely derived from the levelling of of cases around half term, an effect that didn’t last.

Is the 'random' testing being trailed in Liverpool and the new rapid test data included in that do you know? I expect a significant change to how we look for covid will change how much we find and against a moving backdrop that all gets quite unhelpful.

jk

Report
 wintertree 16 Nov 2020
In reply to jkarran:

> I wonder if their immunity plus the vaccine for 20M non-seropositive workers (including all those working directly with the vulnerable) could be enough if targeted cleverly to get R under 1 with very limited additional restrictions. Probably not but it's a thought. Anyway, it wouldn't be popular and it'd look inhumane but if it could work it might be sensible given what could happen anyway if vaccinating the old causes us to lose control of the virus in the rest of the population anyway.

Yes; the true meaning of "herd immunity" has been rather lost over the last 9 months.  It means people without immunity are protected by those with, regardless of how that immunity is aquired (infection, vaccine, etc.).  Especially if the vaccine doesn't work well on those in their final decades, they may be best protected - statistically speaking if not individually - by using the vaccine on others and using that and natural immunity to set R<1 to keep prevalence low.  

There are strong arguments either way I think economically and socially with gnarly morals throughout.  Which way the pendulum swings is going to depend very much on the details of the Pfizer Phase 3 trials.  

As you say, it would look "Inhuman".  Twice now our government have failed to clearly communicate how earlier action is better in the long run, and have abrogated responsibility to the media reporting rising case numbers, hospital overloads and rising deaths.  Should the approach you are considering be the best one to move forwards I don't have much faith in the government to explain this hard choice to people.     

Report
 wintertree 16 Nov 2020
In reply to jkarran:

> Is the 'random' testing being trailed in Liverpool and the new rapid test data included in that do you know? I expect a significant change to how we look for covid will change how much we find and against a moving backdrop that all gets quite unhelpful.

A very good point.  I just looked at the text on the dashboard has been updated at some point to read "Number of people with at least one positive COVID-19 test result (either lab-reported or lateral flow device)".  I don't know when that change came in, but it makes the cases date even harder to interpret, especially when it's rising.  That being said, changes in testing tend to be progressive and not create sudden jumps in the data, and the doubling time in the cases data has always been quite disconnected from that in the other measures.

The new rapid testing of symptom free individuals shouldn't affect data on hospitalisations or deaths much I think.

These tests may not be the magic bullet the government were looking for to send undergraduates home...

https://www.theguardian.com/world/2020/nov/13/covid-test-for-mass-uk-screening-could-miss-up-to-half-of-cases-say-scientists

Report
 jkarran 16 Nov 2020
In reply to wintertree:

> As you say, it would look "Inhuman".  Twice now our government have failed to clearly communicate how earlier action is better in the long run, and have abrogated responsibility to the media reporting rising case numbers, hospital overloads and rising deaths.  Should the approach you are considering be the best one to move forwards I don't have much faith in the government to explain this hard choice to people.     

Listening to Any Answers the other day it's certainly clear this isn't well understood. It'd have to be explained exceptionally clearly by someone capable, credible and widely trusted. Weird suggestion: Attenborough, maybe too BBC.

jk

Report
 neilh 16 Nov 2020
In reply to Misha:

I am expecting it to be extended for another 2 weeks.That way Boris will be Ok on Christmas.

Report
In reply to kaiser:

> My guess is that she'll be much keener on lockdown than Dom was.

I thought he was the one pushing for it.  Wasn't it him behind the leak designed to force Johnson's hand, which was what in turn what led to the big bust-up?

Report
 wintertree 16 Nov 2020
In reply to wintertree:

> I suspect lockdown is going to have to be held, or released into T3+ for most locations, until shortly before Christmas.   

News story this evening: https://www.bbc.co.uk/news/uk-54965279

Discussion of the differing regional effects of the tiers and the likely need to release from lockdown into something stricter than T3.

Report
 RobAJones 16 Nov 2020
In reply to wintertree:

If this is true (from the linked article)

"And clearly the impact has been undermined by the spike in cases last week when the daily number jumped by 10,000 to over 33,000 on Thursday.

That rise has been linked to a last bout of socialising before the lockdown came in."

Opening hospitality in any area looks very risky. 

Report
 wintertree 16 Nov 2020
In reply to RobAJones:

>  this is true (from the linked article)

It's true but is seriously crap "analysis" from their Health Correspondent.

Below is a plot showing the number of positive cases per day, based on the day the specimens were taken.   I've added a filtered trend line, and I show the difference between the smoothed trend line and the data points on the middle plot.  I've coloured weekend specimen taking days in blue and Monday's in red.  

What's going on here is that far fewer specimens are taken on a weekend, and then there's a surge in specimen taking on a Monday.  

  • Given today's announcement about the Royal Mail preparing to empty post boxes every day, I'm wondering if what's actually happening here is specimens are being assigned the date they are removed from the postbox and not the date they were taken....  Groan.
  • This specimen taking lag means that there's an apparent "spike" of 10,000 cases every Monday.  Scary stuff until you look at a trend line and the patterns around it.
  • There's also a reporting lag as the time it takes to feed the spreadsheet files from the testing labs into the PHE database seems to vary dramatically, and sometimes the clearing of a reporting lag coincides with the Monday catch up spike in by-specimen-date data to make a big scary looking jump in the cases by-reporting-date.  

I don't know if the BBC's health correspondent made up the unreferenced link between the  spike and a last bout of socialising, or if he's speaking to people whose day job is to understand the data but who aren't doing a very good job.

The only thing one can infer from the spike is that the testing system is that testing is down on weekends and up on Monday's.  The idea that a bout of socialising would translate into a 1-day wide spike some time down the line is risible - even if there were no weekend/weekday effects but testing ran smoothly without any 7-day cadence I would be highly skeptical of such a suggestion because the natural variation in people will "blur out" a 1-day spike of transmission events into a several day window of detected cases down the line.   

Aside - the bottom plot is those residuals normalised to the "expected" statistical variance if there were no systematic effects, only random ones.  These plots should be randomly spaced in a normal distribution with a standard deviation of 1.  The values are a lot larger and they're growing which implies more systematic effects interfering with the "natural" flow of people into specimens.  I'm taking some hope from the signs that this distribution has started tightening in the last week.


Report
In reply to TomD89:

Gyms were among the last places to reopen back in the summer. I would argue that pubs and restaurants are higher risk but gyms are still relatively high risk compared to shops for example. So I'm not entirely optimistic.

Report
In reply to Si dH:

I think you're broadly right in that once the vulnerable groups are vaccinated (over 60s/65s and those who are medically vulnerable), there will be a general (not total) reopening. It will then be down to individuals to act in line with what they're happy doing. This assumes that the 'tail' of Covid patients in hospitals will have reduced to a manageable level by then. Realistically, I don't see this general reopening happening before May and probably not before July - and even then I think SD and face masks will be with us for the rest of the year or longer. Then if it turns out that significant numbers of younger people are ending up in hospital (low % of a larger number), additional restrictions would need to be brought in again - T1/2 sort of stuff.  

Report
In reply to wintertree:

> I pretty much agree with all that.  Once we're vaccinated down to ~50 years of age, then it's unlikely we'll see restrictions continued, assuming the intent is for covid to become an endemic seasonal illness (vs going for near-elimination for example).

I would have thought that with a variety of effective vaccines available (I'd be amazed if it's actually 90-95% but it does sound promising), near elimination would be a feasible goal to aim for over the next 2-3 years. I do have a concern that even younger people can have a fairly serious bout of illness, which in some cases turns into Long Covid. It's not something I'd like to experience.

Report
 RobAJones 16 Nov 2020
In reply to wintertree:

Thanks.

So that rise of 10000 was only 1000 or so more than expected due to it being Monday?

There is a small increase in cases around that time that might have been caused by  socialising prior to lockdown?

Report
 wintertree 16 Nov 2020
In reply to RobAJones:

I think the meaning of 1,000 cases here or there is lost in the noise at this point; you can’t draw any conclusions from it.  At least some of the asymptomatic lateral flow testing is apparently feeding in to these numbers as well which makes it even harder to interpret small changes.  

Post edited at 22:08
Report
In reply to RobAJones:

The easiest thing to look at is the 'cases by specimen date' graph (especially the rolling 7 day average) on the official website. https://coronavirus.data.gov.uk/details/cases (ignore the last 4 days of data as that's incomplete).

This shows that there was gradual rise over the last 4 weeks. It will be interesting to see what happens to the graph for this week as this week's tests should mostly be for post lockdown infections. We won't have reasonably complete data for this week till next Tue / Wed of the following week though.

Of course this doesn't capture all the cases. The estimates for that based on various data gathering methods come out on Thursday / Friday each week (for the preceeding week).

In many ways, the more important (and certainly more reliable) data is hospital admissions. I believe average time from infection to hospital is around two weeks and there's a bit of a delay on reporting the numbers (particularly over the weekend, eg today we have numbers for last Friday in E and Thursday in W, S & NI). So again we should look out for a reduction or at least a levelling off in the numbers being reported early next week in respect of admissions at the end of his week.

Interesting to note that the numbers for E show a significant drop from 1,666 to 1,433 from Thur to Fri last week. However I wouldn't pay too much attention to one day's data as it could be due to non-reporting from a particular NHS trust / area (it's happened before), or it could be a natural day to day fluctuation.

It's really frustrating that the Welsh admissions numbers are almost meaningless as they count suspected cases. As a result, their numbers are relatively high and fluctuate a fair bit as a % of the average. It would have been great to see a reduction following their fire breaker but you just can't see it in the gibberish which is reported there. Good to see though that the Scottish numbers have reduced from their recent peak in late October - evidence of the restrictions there working, at least to some extent? 

Post edited at 22:26
Report
 kamala 16 Nov 2020
In reply to Jon Stewart:

Jon, may I copy your text to that person I mentioned on another thread? It says exactly what I was trying to convey to him but hopefully far more effectively.

My words were so ineffective that he replied: "I can see that you think that it is justified to destroy people's livelihoods and businesses and make them homeless. If significant numbers of long covid patients have kidney failure do you think it acceptable to forcefully remove kidneys from other people in order to save them?" Been a while since I've been so angry - both that he could read that in my message and that I couldn't explain more clearly to him...

Report
In reply to kamala:

> Jon, may I copy your text to that person I mentioned on another thread? It says exactly what I was trying to convey to him but hopefully far more effectively.

Even better, if you can be arsed, change it a bit and pretend that someone they like said it - much more convincing that way.

> My words were so ineffective that he replied: "I can see that you think that it is justified to destroy people's livelihoods and businesses and make them homeless. If significant numbers of long covid patients have kidney failure do you think it acceptable to forcefully remove kidneys from other people in order to save them?" Been a while since I've been so angry - both that he could read that in my message and that I couldn't explain more clearly to him...

I don't even know what that response is supposed to mean, but if I can help shut the f*cker up in any way, I'll be delighted.

What do these people think the motivation behind the lockdown policies is? The situation, and the policies are so obviously terrible for everyone, that unless you're getting paid directly for every additional hour spent on netflix or whatever, then there's no chance you're gaining from it. There is no plausible "lockdown corruption" conspiracy.

Basically, the "anti-lockdown" perspective comes from people who have a vastly overestimated their understanding of the whole interconnected health/economic/social pickle we're in. They think that the scientists and policymakers and government simply haven't noticed the downside of virus control restrictions, or have given them much too low a weighting in their consideration. But the "anti-lockdown" crew are much cleverer: they have seen the true impact of lockdown, and can see that it is much worse than the harm that would occur without the restrictions. They can't show us how they know, but they know, because they're the clever ones. Much cleverer than all the people in the world who've spent their entire lives getting to the point where they're responsible for advising governments on how to respond to a pandemic.

If you think it's worth engaging, use whatever material you can get your hands on. But sometimes, it's a better use of time to do something else instead. Listen to some good music, or a podcast, watch a good film, and don't get wound up by the way other people fail to understand the world because of normal, human flaws (and a dash of infuriating stupidity).

Post edited at 23:15
Report
 wintertree 16 Nov 2020
In reply to kamala:

Is this your offline contact with a scientific background?

From their world-view, this pandemic is all a big misunderstanding by hundreds of scientists who are falsely ascribing sniffles, hospitalisations and deaths to Covid based on their misunderstanding of a bog-standard diagnostic test.  In calling for control measures we all accept are economically damaging to counter a threat they believe is non-existent, from their view you are advocating for clearly wrong decisions.  Arguing against that is arguing against a symptom, not the root cause of the divergence between you both. 

They've had to suspend all critical thinking to get this far.  If you can't flip their view over those suspensions of thinking, you'll get nowhere with the downstream arguments.  If you can get them to re-evaluate their position on what's going on, the rest may fall in to place naturally..

  • Why are there so many hospitalisations with covid symptoms when there's almost no influenza or other viable cause detected?  What is putting so many people into ITU?  There's plenty of reputable news coverage over this. 
  • There are about 250,000 samples being taken through Pillar 1 & 2 a day rigthtn ow, and 25,000 detected cases.  If this is all false positives, it would need a 10% false positive rate.   But on August 1st there were 130,000 samples being taken and 700 cases implying a 0.5% false positive rate if there were no actual cases.  So, either cases have gone up massively or the false positive rate has inexplicably risen by 20x in the last few months.

Challenge them on an evidence basis - How do they explain the evidence over hospitalisations in the absence of influenza (PHE surveillance reports)?  How do they explain the 20x rise in sample positivity and the absence of whistleblowers screaming about a 10% false positive rate (which, frankly, would be totally ludicrous and is way far removed from the already unevidenced, unreviewed claims from Henhegan).

If you can't get some common ground on the evidence their view is built on, you're pissing in to the wind arguing over what control measures should happen.  

Post edited at 23:09
Report
 kamala 16 Nov 2020
In reply to Jon Stewart:

Thanks Jon, very kind! And thanks for more useful material. He's not a close friend but it would be a shame to cut ties since his heart's in the right place even if his brain isn't, at the moment. 

The original question was whether it was okay to cause harm to prevent harm, and I was trying to explain that interconnectedness of society, and that letting granny die wouldn't necessarily save all those livelihoods. As you say, he seems to assume that all will be well if we just carry on as normal (minus a few grandmothers). He's even bought into the "cases not increasing, it's just positive tests increasing" line. I think he's been hearing real life tales of economic and social distress from clients but not been touched by the virus himself.

Anyway, I shall let myself cool down before putting something together with your text. And if that doesn't get through to him, I'll disengage. I've been in strict quarantine the last two weeks so should be bursting with energy for a fight but somehow it doesn't seem to work that way...

Report
 wintertree 16 Nov 2020
In reply to Jon Stewart:

> Much cleverer than all the people in the world who've spent their entire lives getting to the point where they're responsible for advising governments on how to respond to a pandemic.

Some of them are certainly better at getting some politicians to listen to them than to their scientific advisers.  It worries me that more than a few Tory MPs in in the sway of these people - look to the USA to see where it ends up - with a government advisor telling the people to "rise up" against their state government's covid control measures.  

https://arstechnica.com/tech-policy/2020/11/trump-advisor-tells-michigan-to-rise-up-aginst-covid-restrictions/

Report
 kamala 16 Nov 2020
In reply to wintertree:

Hi wintertree, yes it is, well remembered. I think he's an optometrist or something along those lines and I guess he's been hearing 92-year-old clients telling him they'd die happy if they could only see their grandchildren again, plus experiences of real economic hardship. So he's been posting updates from ZOE, interpreted to mean cases aren't rising, and following the numbers about false positives etc. from there.

As you say, it's the evidence on which the rest of the edifice is built that must agree, otherwise all the critical thinking in the world won't come to the same conclusion as if one was working from a different starting point. You've given me some useful places to start and Jon has helped to express the more "moral" side of the argument. 

Once I start thinking about all the possible arguments the whole thing just seems overwhelming. If it wasn't someone I'd like to keep in touch with, I'd never have engaged at all...

Report
In reply to kamala:

> The original question was whether it was okay to cause harm to prevent harm, and I was trying to explain that interconnectedness of society, and that letting granny die wouldn't necessarily save all those livelihoods.

We all know it's OK to cause harm to prevent harm - everything we do is based on that principle. You go to the dentist to have a tooth out: it's suffering that will save greater suffering long term. You vaccinate your child who won't like having an injection, but it's better than exposing them - and then wider society - to a devastating disease. There is universal agreement that it is OK to cause some harm to prevent greater harm, that's how the world is run.

Sure there are extreme cases where we're not sure if that balance is acceptable, but framing the covid policy case as one of these is misunderstanding the problem. The pandemic forces us as a society to look at at a very unfamiliar scenario: the brink. The policymakers are concerned that we do not come face to face with the brink, so they instigate restrictions; the "anti-lockdown" proponents have simply failed to appreciate the reality of the brink. There's no meaningful moral or political argument to be had about the brink. Nobody, no matter how young or healthy or economically virile, would like it - only those intentionally going out to profit from disaster would benefit (this is where Nigel Farage comes in).

> As you say, he seems to assume that all will be well if we just carry on as normal (minus a few grandmothers). He's even bought into the "cases not increasing, it's just positive tests increasing" line. I think he's been hearing real life tales of economic and social distress from clients but not been touched by the virus himself.

You can try Wintertree's tactic of using evidence to debunk falsehoods, if you think it'll work. 

> Anyway, I shall let myself cool down before putting something together with your text. And if that doesn't get through to him, I'll disengage. I've been in strict quarantine the last two weeks so should be bursting with energy for a fight but somehow it doesn't seem to work that way...

I've been off work with covid, then "holiday" time off, plus lockdown, and the less I go out and less exercise I do the less energy I have for anything. There were some upsides to the spring lockdown in the good weather, but this really is absolutely shite. There are non two ways around that.

Post edited at 23:40
Report
 kamala 16 Nov 2020
In reply to wintertree:

Re instilling belief in people: I just saw a tragic you-tube clip discussing a tweet from an American nurse (Jodi Doering, if anyone wants to look it up).

She said the patients who stuck in her mind were the ones who "still don't believe the virus is real. The ones who scream at you for a magic medicine and that Joe Biden is going to ruin America. All while gasping for breath on 100% Vapotherm. They tell you there must be another reason they are sick. They call you names and ask why you have to wear all that 'stuff' because they don't have COVID because it's not real. Yes. This really happens. And I can't stop thinking about it. These people really think this isn't going to happen to them. And then they stop yelling at you when they get intubated..."

I for one have assumed that people who get ill would finally acknowledge reality. This is unbelievable, and heartbreaking and highly disturbing.

Report
In reply to wintertree:

> Some of them are certainly better at getting some politicians to listen to them than to their scientific advisers.  It worries me that more than a few Tory MPs in in the sway of these people

But many Tory MPs (and MPs generally) are absolutely thick as mince, and utterly dreadful people. They can't understand - or don't have any interest in - science, or philosophy, or anything really. They're just people who have grown up to believe first and foremost that they should have a high status in society and that being an MP is befitting to that, and they've pursued their path because it's been the line of least resistance in the wealth of opportunity open to them. Some of these people have truly rotten souls - they know the harm they're doing and they don't give a f*ck. Others aren't that bright.

What use is it presenting science to these kinds of people? I'm glad that's not my job. Obviously some journalist or think-tank chief with aligned motivations and who presents a "front" that reflects them in style and culture is going to be more successful at gaining these politicians' trust. These people don't give a f*ck about the facts.

"There are some bad people on the right".

Report
 kamala 17 Nov 2020
In reply to Jon Stewart:

That's some of the angle I was going to try - the balancing of harms. But that concept of the brink is a good one, I shall borrow that.

Hope you've recovered well from the covid and that your energy levels get back to normal soon. 

I was hospitalised in March. I tested negative but I suspect that's because I had what I thought was my annual asthma cough for a few weeks beforehand, so any active virus would have been long gone. Luckily I see so few people in normal life that I'm certain I didn't pass it on. It just seems too coincidental that I'd be laid low by unrelated pulmonary embolism exactly when a clotting-related disease is sweeping the country. But covid or not, I'm still very breathless and a bit brain-fogged and it's no fun. (The psychedelic hallucinations when coughing were interesting, but not worth the rest of it!) 

Report
In reply to kamala:

> I was hospitalised in March. I tested negative but I suspect that's because I had what I thought was my annual asthma cough for a few weeks beforehand, so any active virus would have been long gone. Luckily I see so few people in normal life that I'm certain I didn't pass it on. It just seems too coincidental that I'd be laid low by unrelated pulmonary embolism exactly when a clotting-related disease is sweeping the country. But covid or not, I'm still very breathless and a bit brain-fogged and it's no fun. (The psychedelic hallucinations when coughing were interesting, but not worth the rest of it!) 

Sounds rough!

Report
 kamala 17 Nov 2020
In reply to Jon Stewart:

Realistically, I'm probably lucky to be alive, thanks to some good friends, and very switched on doctors. Some comedic moments though - like when the ambulance reached the hospital on Thursday evening. The ambulance guys were keen to drop me off and get off shift but it took half an hour to get me in. And why? Because the entrance was blocked by half the police cars in North Wales and a shedload of people milling around, "clapping for the NHS"! You should have heard their comments...

Report
In reply to kamala:

> Because the entrance was blocked by half the police cars in North Wales and a shedload of people milling around, "clapping for the NHS"! You should have heard their comments...

That is classic.

Report
 CurlyStevo 17 Nov 2020
In reply to jkarran:

cfr is pretty much a worthless stat as it’s based on known cases. What you want to know is IFR, which is always an estimate. The current IFR across all age groups in England is thought to be about 0.3%, in the under 55s it’s significantly lower than that probably by an order of magnitude. At age 55 you have roughly the all age groups IFR, it is effectively the break even point. As treatments improve the IFR will fall further, it was much higher at the start of the pandemic, particularly in England.

Post edited at 04:21
Report
 Si dH 17 Nov 2020
In reply to wintertree:

> > I suspect lockdown is going to have to be held, or released into T3+ for most locations, until shortly before Christmas.   

> News story this evening: https://www.bbc.co.uk/news/uk-54965279

> Discussion of the differing regional effects of the tiers and the likely need to release from lockdown into something stricter than T3.

This could be correct, but I think you have mis understood what she said (I watched it), probably due to selective quoting in the article. When she said 'Tier 3 plus' she was referring to what we already have had, not something new. Presumably the 'plus' is just the closures that were beyond pubs. The gist of what she said was that Tier 3 had worked to reduce infection, Tier 2 had flattened it in some areas but not others, and they had seen little effect from Tier 1, so there may be a need to strengthen the tiers. The clear implication was not that she would advise strengthening Tier 3, but rather that she would advise strengthening Tier 1 and probably Tier. 2 because you need to be at a minimum of equivalent to Tier 2 in order to flatten the rate.

Report
 Si dH 17 Nov 2020
In reply to Si dH:

Just to add,  may be noteworthy that the question she was answering had actually been asked directly to Hancock about likely policy, but he immediately shifted it to her and asked her to explain the likely scientific advice, without saying anything himself to respond. So whether what she said is remotely supported by govt, who knows. Her answer was also prefaced with a bunch of caveats about seeing the infection rate on 2nd December. The most interesting thing I suppose is that she said she would expect to see the effect of lockdown in the data within the next week. It was all consistent with what people on here have been able to work out from looking at data ourselves.

Post edited at 07:44
Report
 GrahamD 17 Nov 2020
In reply to Misha:

> I would have thought that with a variety of effective vaccines available (I'd be amazed if it's actually 90-95% but it does sound promising)

The real question is what is the effectiveness of a vaccine for those in the vulnerable categories.   I always thought vaccines were supposed to work best for those with stronger immune systems and less well for those with suppressed immune systems (aka the vulnerable).  I'd be interested to hear what someone working in immunology has to say.

Report
 wintertree 17 Nov 2020
In reply to CurlyStevo:

> The current IFR across all age groups in England is thought to be about 0.3%

I'd say that's the lower reasonable bound for IFR right now, with an upper of about 0.6%.

Below is my plot of CFR vs the detection>death lag for UK level data; as we are apparently in a near-plateau phase for (detected) infections, these are all converging - to about 1.6%.   The various random sampling surveys differ by a factor of about 2x in the number of actual infections, between about 5x and 2.5x the number of detected cases.  So that gives bounds of 0.32% and 0.64% for the IFR.  Either or which is still too high for uncontrolled growth in cases, but likely significantly better than it was in March/April.

>  As treatments improve the IFR will fall further, it was much higher at the start of the pandemic, particularly in England.

Indeed - although if you view the key problem to be controlled for as hospital overload rather than death, somewhat counterintuitively improved treatment can reduce deaths but worsen hospital overload, because dead people don't need a hospital bed.  I think one of the reasons it was so high at the start was the forced discharging of patients into care of when the patients weren't provably covid free (either not tested, or testing with significant false negative rates especially when someone is just post-infection) when discharged 

Post edited at 10:44

Report
 jkarran 17 Nov 2020
In reply to wintertree:

> I think the meaning of 1,000 cases here or there is lost in the noise at this point; you can’t draw any conclusions from it.  At least some of the asymptomatic lateral flow testing is apparently feeding in to these numbers as well which makes it even harder to interpret small changes.  

I see yesterday they've started reporting where cases are based on the address given with the sample, not one scraped out of an out of date NHS database. Cue a big bounce in university towns' covid prevalence. At one point (post student migration) apparently 12% of cases were tied to the wrong address. You couldn't make this up, we're nearly a year in!

jk

Report
 Si dH 17 Nov 2020
In reply to jkarran:

> I see yesterday they've started reporting where cases are based on the address given with the sample, not one scraped out of an out of date NHS database. Cue a big bounce in university towns' covid prevalence. At one point (post student migration) apparently 12% of cases were tied to the wrong address. You couldn't make this up, we're nearly a year in!

> jk

That's really annoying. Any more information about the impact or whether they are going to fix any of the data provided to date? 

Report
 RobAJones 17 Nov 2020
In reply to wintertree:

I had read in a few place that whilst an IFR 0.3-0.6 was a global estimate, it was likely to be a lot higher for our older/less healthy population?

Imperial college

"In high income countries, the estimated overall infection fatality ratio (IFR) is 1.15% (95% prediction interval 0.78-1.79)."

Also fag packet stuff, but you graphs of CFR +22 is about 2% for the last 6 weeks (with about  half the infected people being identified??)

Post edited at 11:15
Report
 wintertree 17 Nov 2020
In reply to Si dH:

Thanks for the clarification from the broadcast.  I only managed to tun in when they got to the lateral flow testing part.  A false positive rate of "less than" 5-in-1000, 100,000 non-symptomatic people tested and 700 positives.  Anyone want to do the numbers on that?...

> The most interesting thing I suppose is that she said she would expect to see the effect of lockdown in the data within the next week. It was all consistent with what people on here have been able to work out from looking at data ourselves.

Yup.  It's a nervous wait; looking at the measured doubling times, whatever effect caused cases, then hospitalisations and then deaths to slacken off their growth starting by late October appears to be wearing off.  It's tempting to ascribe this to half-terms...  (Although the timing in my plot seems to early, the measuring of a doubling time is delocalised by ±8 days or so in a trade-off between immediacy and a curve that's legible not dominated by noise).

It looks to me like T1/2/3 took the doubling times from ~10 days to ~30 days, and then half term relaxed them even further, and the exponential growth rate is now rising in all the measures again.   It's worth keeping in mind that we got the slower growth rates against higher absolute numbers so the rises in absolute terms has always continued, meaning all absolute numbers are higher than ever for this "wave".  This is important as the main "red line" driving policy is hospital admissions and occupancy levels.  With the doubling times apparently decreasing on all measures and some time to go before the effects of lockdown translate in to hospitalisations, the pucker factor is high. 


Report
 wintertree 17 Nov 2020
In reply to RobAJones:

> I had read in a few place that whilst an IFR 0.3-0.6 was a global estimate, it was likely to be a lot higher for our older/less healthy population?

I find it kind of meaningless to dive in to too granular a detail - it depends on demographics, health, healthcare, ability and willingness to isolate, care homes vs care at home.  The demographics shift continually as infection progresses through any country so actually narrowing down a meaningful measurable quantity is nigh on impossible.

But it's a bad time to be old and obese.

> Also fag packet stuff, but you graphs of CFR +22 is about 2% for the last 6 weeks (with about  half the infected people being identified??)

Yes; I think +22 days from detection to death is too long but it's in the plot to give a range for interpretation.   I think it's diving down to meet the others because as we remain in a "plate phase" for the measures, the lag becomes irrelevant and all the values agree; right now the +22 day value is dividing deaths from the plateau phase by a lower cases level from the earlier rising phase.  My interpretation of this plot is that it's telling me +22 days is too long a lag.  It's complicated though by the unknown time dependance of the friction of infections detected as cases...   If the plateau holds for another 5 days I think it'll end up on about 1.6%, the same as the others.  

There's not much point gong to a longer lag as there's the 28-day cut-off from case detection in the reporting for deaths.  What is maddening is that the data must exist for every individual on time from detection to death and this could be published in averaged form.  I don't know what fraction of deaths are detected initially by pillar 1 on hospital admission and what are caught earlier by pillar 2.  All of this tells me it's probably pointless to spent too much time thinking about it.

Report
 Si dH 17 Nov 2020
In reply to wintertree:

> Thanks for the clarification from the broadcast.  I only managed to tun in when they got to the lateral flow testing part.  A false positive rate of "less than" 5-in-1000, 100,000 non-symptomatic people tested and 700 positives.  Anyone want to do the numbers on that?...

Apparently everyone who tests positive with a lateral flow device also then gets a pcr test to confirm it. This seems reasonable to me and to be honest I think quashes some of the concerns raised about the programme. However I don't think any data on these specific pcr tests has been published, which is annoying because it's obviously important to understanding accuracy of lateral flow testing for future. 

> Yup.  It's a nervous wait; looking at the measured doubling times, whatever effect caused cases, then hospitalisations and then deaths to slacken off their growth starting by late October appears to be wearing off.  It's tempting to ascribe this to half-terms...  (Although the timing in my plot seems to early, the measuring of a doubling time is delocalised by ±8 days or so in a trade-off between immediacy and a curve that's legible not dominated by noise).

> It looks to me like T1/2/3 took the doubling times from ~10 days to ~30 days, and then half term relaxed them even further, and the exponential growth rate is now rising in all the measures again.   It's worth keeping in mind that we got the slower growth rates against higher absolute numbers so the rises in absolute terms has always continued, meaning all absolute numbers are higher than ever for this "wave".  This is important as the main "red line" driving policy is hospital admissions and occupancy levels.  With the doubling times apparently decreasing on all measures and some time to go before the effects of lockdown translate in to hospitalisations, the pucker factor is high. 

I think I agree that on a national basis its reasonable to infer that the half term caused a slowing of the average rate which then sped up again. However by looking at local or regional data its also clear that the half term effect is small in comparison to the tier 3 effect. 

Report
 wintertree 17 Nov 2020
In reply to Si dH:

> Apparently everyone who tests positive with a lateral flow device also then gets a pcr test to confirm it. This seems reasonable to me and to be honest I think quashes some of the concerns raised about the programme.

Thanks; I hadn't seen that level of detail; it makes sense and seems an appropriate way of tightening the LFT results.

> However I don't think any data on these specific pcr tests has been published, which is annoying because it's obviously important to understanding accuracy of lateral flow testing for future. 

Yes, this would be very good to know.

Although the key item I think to understand is the false negative rate - the more medical people involved in the evaluations are being careful to point out that a negative result means "not likely to be currently infections" and qualifying that it does not rule out someone in the early stages of incubating the virus, who will go on to be infectious.  This suggest to me that: 

  1. The LFTs followed by RT-PCR conformation are a powerful tool for pulling asymptomatic infectious people out of circulation through broad trawls of the population - a powerful way to help turn around hot spots.  
  2. They may not be a magic bullet for clearing undergraduates to return to the family homes in December, and that the results given to the students need to be clearly and precisely contextualised about what they do - and don't - mean.  Ideally students would quarantine in the family home for another 5 days followed by another lateral flow test to determine if they have gone on to become infectious from a transmission even shortly before they left halls.

> I think I agree that on a national basis its reasonable to infer that the half term caused a slowing of the average rate which then sped up again. However by looking at local or regional data its also clear that the half term effect is small in comparison to the tier 3 effect. 

I agree, but in terms of the red line over hospitalisations, the decay of cases in former Tier 3 regions seems slow, so they remain near the local red line for hospitals - especially given the various lags in the system.  Meanwhile the areas that didn't have the benefit of Tier 3 are generally still heading towards that red line as they haven't yet got to the point the effects of lockdown translate into reductions instead of growth (in data which is always ~5 days old.)   So it's could get very close to the wire before it starts improving.  Hospital capacity can be shared on a national basis.

Post edited at 11:47
Report
 RobAJones 17 Nov 2020
In reply to wintertree:

> All of this tells me it's probably pointless to spent too much time thinking about it.

My wife would certainly agree, at least all the figures are of the same order of magnitude. In the future, a better understanding the IFR of say 50-60's might affect the vaccination strategy?

Report
 wintertree 17 Nov 2020
In reply to RobAJones:

>  In the future, a better understanding the IFR of say 50-60's might affect the vaccination strategy?

I think the powers that be must have that data from the NHS, and a combination of that data and the data on the efficacy of the vaccine on people aged 75 and over are going to be some of the most important factors in the vaccination strategy.

Report
 jkarran 17 Nov 2020
In reply to CurlyStevo:

> cfr is pretty much a worthless stat as it’s based on known cases. What you want to know is IFR, which is always an estimate.

Yes, IFR, sorry, I was sloppy. My fag packet estimates of lethality were based loosely on this (I sort of split the difference between the M and F groups with and without comorbidities since we're pretty unwell as a society by late middle age) https://gh.bmj.com/content/bmjgh/5/9/e003094.full.pdf

jk

Report
 jkarran 17 Nov 2020
In reply to Si dH:

> Any more information about the impact or whether they are going to fix any of the data provided to date? 

https://www.gov.uk/government/publications/covid-19-geographical-allocation-of-positive-cases/geographical-allocation-of-positive-covid-19-cases

Dr Yvonne Doyle, Medical Director at Public Health England, said:

We have updated the way we record the location of people who test positive for coronavirus to prioritise addresses given at point of testing, rather than details registered on the NHS database.

This better reflects the distribution of positive cases in recent weeks and months, particularly among younger people of university age who may not have yet registered with a GP at their term-time address.

I think from the reference to recent weeks and months they've backdated the change for the publicly accessible data but honestly, not sure! It's not finest piece of science communication.

My wife was reading a press digest of this to me at seven this morning while the baby screeched and bashed my half asleep face with a rattly toy. I thought it was the situation making things unclear, apparently not, just a vague statement.

jk

Report
 Postmanpat 17 Nov 2020
In reply to Si dH:

> Apparently everyone who tests positive with a lateral flow device also then gets a pcr test to confirm it. 

>  

Are you sure about this? The figures for the Liverpool programme (6/11-15/11) apparently show positivity of the LF test at 061% and the PCR test  at 2.98%. If they were only retesting LFT positives that wouldn't make any sense.

  In any case, on the face of it it is an enormous discrepancy. The PCR result is roughly in line with the ONS random testing (PCR)for the North West. The lateral flow result is roughly in line with the zoe numbers.

  Either one or both of the tests is very misleading or there are completely different sample groups. Or something else? Any clues?

Post edited at 14:26
Report
 jkarran 17 Nov 2020
In reply to Postmanpat: 

>   In any case, on the face of it it is an enormous discrepancy. The PCR result is roughly in line with the ONS random testing (PCR)for the North West. The lateral flow result is roughly in line with the zoe numbers.

>   Either one or both of the tests is very misleading or there are completely different sample groups. Or something else? Any clues?

Well we know they're different sample groups. You're supposed to be symptomatic to self select for a Pillar II PCR test. Anyone can and is encouraged to take an LFT. Pillar I is a messy mix of diagnostic and screening tests.

jk

Report
 Postmanpat 17 Nov 2020
In reply to jkarran:

> Well we know they're different sample groups. You're supposed to be symptomatic to self select for a Pillar II PCR test. Anyone can and is encouraged to take an LFT. Pillar I is a messy mix of diagnostic and screening tests.

> jk

It's not really clear (to me). It doesn't seem to be that those who test positive with an LFT then have a PCR test.

   The Liverpool City site (entitled "Mass testing programme")divides numbers into LFTs and PCR and says that the sources are CIPHA and Pillar 2 in the community but doesn't clarify which is which or whether there is overlap in the samples.

The BMJ says "Participation in this pilot is voluntary. There is no call or recall. All participants receive two tests, the standard PCR test and the rapid turnaround (within 1 hour) lateral flow Innova test. Those with a positive result in either test are asked to self-isolate and are registered with the national track and trace programme, which initiates contact tracing. Key workers, health and social care staff, school staff, and children aged 11 and over are being targeted, but anyone can get tested, preferably at least twice within two weeks."

  That doesn't appear to tie in with the fact that different numbers have been tested for each type of test test.

  Have they just co-opted the existing Pillar 2 testing programme (PCR) and called it part of the new mass testing programme without really explaining that?

Is there an explanation somewhere?

  Either way, you're still left with a big disparity between the ONS random survey positivity (2-2.5%) and the LFT numbers (I think?).

Post edited at 16:36
Report
 Si dH 17 Nov 2020
In reply to Postmanpat:

> It's not really clear (to me). It doesn't seem to be that those who test positive with an LFT then have a PCR test.

https://liverpool.gov.uk/communities-and-safety/emergency-planning/coronavirus/how-to-get-tested/mass-testing-faqs/

Under "what happens if I test positive", it's very clear that LFT positives get a confirmatory PCR (but obviously have to isolate in the meantime.) I saw it originally on the council's Twitter thread, which seems to be their main way of releasing data.

>    The Liverpool City site (entitled "Mass testing programme")divides numbers into LFTs and PCR and says that the sources are CIPHA and Pillar 2 in the community but doesn't clarify which is which or whether there is overlap in the samples.

> The BMJ says "Participation in this pilot is voluntary. There is no call or recall. All participants receive two tests, the standard PCR test and the rapid turnaround (within 1 hour) lateral flow Innova test. Those with a positive result in either test are asked to self-isolate and are registered with the national track and trace programme, which initiates contact tracing. Key workers, health and social care staff, school staff, and children aged 11 and over are being targeted, but anyone can get tested, preferably at least twice within two weeks."

>   That doesn't appear to tie in with the fact that different numbers have been tested for each type of test test.

The BMJ description you quote above is misleading.

Getting a second test is voluntary and you are recommended to have it a week after the first, so I don't expect many people will have had it yet. Would I bother after having had a first? Quite possibly not. They are focusing quite a bit of social media effort on persuading people to do this though.

>   Have they just co-opted the existing Pillar 2 testing programme (PCR) and called it part of the new mass testing programme without really explaining that?

That's an interesting question. The latest data presented in an unhelpful way is here:

https://mobile.twitter.com/lpoolcouncil/status/1328729964808712192/photo/1

The total number of PCR is very approximately the same number of PCR tests that you would expect to have been undertaken for symptomatic people if you look at the number of positive cases since the programme started and assume a positivity of 3%.  Which means the number of asymptomatic PCR tests are either similar to symptomatic ones, or else there are very few and they are doing what you say.

I know there is a mixture of LFT and PCR for asymptomatics; as a minimum there are up to 40000 home tests apparently being delivered as part of the mass testing programme that each contain three PCRs for a household, but I don't know how many of these results are back, possibly few. My impression was that a few days ago asymptomatic people were also being assigned to mobile PCR test centres (as opposed to the fixed ones, which there are plenty of in the Liverpool area for symptomatic people.) However it now looks like they are no longer taking bookings for the mass testing and instead are just telling people to turn up at one of the fixed sites, so this must have changed.

> Is there an explanation somewhere?

>   Either way, you're still left with a big disparity between the ONS random survey positivity (2-2.5%) and the LFT numbers (I think?).

I think your number must be symptomatic people, or else the LFT test has a very high false negative rate - hopefully that's not the case.

Edit to add: it's notable that the LFT positivity rate is significantly higher for non residents. Presumably that shows the increased risk of getting Covid for people who have to travel to work (since you have to either work or live in Liverpool to be eligible.)

Another edit to add: there is more information here and at the linked PHE pdf:

https://liverpool.gov.uk/communities-and-safety/emergency-planning/coronavirus/cases-control-and-testing/cases-by-area/

It implies strongly to me that you are correct that they have just co opted all of pillar 2, which makes it impossible to compare outcomes from PCR (majority symptomatic) and LFT (all asymptomatic). Hopefully, maybe, someone somewhere has this data and can use it to inform policy.

Post edited at 17:41
Report
In reply to RobAJones:

> I had read in a few place that whilst an IFR 0.3-0.6 was a global estimate, it was likely to be a lot higher for our older/less healthy population?

I can't remember my source, but I found this on IFR a couple of weeks ago:-

Below 55 negligible

0.4% at age 55

1.4% at age 65

4.6% at age 75

15% at age 85

Report
In reply to wintertree:

> Nifty plot.  That would suggest another week or so before we see the effects of lockdown in the English case data resolved by specimen date.

Stolen from one of the Thomas Pueyo articles .

Can you start a 'Wintertree's covid stats thread' to put all your graphs in so I don't have to keep trawling various threads to find them?

Report
 Postmanpat 17 Nov 2020
In reply to Si dH:

> It implies strongly to me that you are correct that they have just co opted all of pillar 2, which makes it impossible to compare outcomes from PCR (majority symptomatic) and LFT (all asymptomatic). Hopefully, maybe, someone somewhere has this data and can use it to inform policy.

>

  Thanks. It's no wonder there is so much confusion around when they can't even present data clearly.

Report
In reply to GrahamD:

Agree, that is a key question. Not all non-vulnerable people will agree to be immunised (I think if uptake it's high enough, whatever that % needs to be, it should be made quasi compulsory unless you have a medical excuse, eg £10,000 fine and can't renew passport / driving license, that kind of thing). Besides, 90% effective at preventing getting sick with Covid is not the same at 90% effective at preventing transmission. If you have say 10% of the population where the vaccine won't be particularly effective, 80% uptake and 10% who won't / can't have it, and it's only 50% effective at preventing transmission, that's 50% still able to transmit. Which is too much for herd immunity for the 10% for who it's not particularly effective. I'm not an epidemiologist so this is a vast oversimplification and the numbers could be quite different but this is broadly what the issue could be. We'll just have to wait and see...

Still, a vaccine which is reasonably effective for most people who choose to take it is better than no vaccine, so it's a start. We might have SD and face masks plus contact tracing for years yet but at least life could gradually return to something approaching normal.

Report
 CurlyStevo 18 Nov 2020
Report
In reply to Misha:

So pretty much you want mandatory vaccines or people cant work etc? I thought if lets say you had it, you wouldnt get it and be safe , but then if i didnt only i could get ill and not pass it to you?  Im happy to take it as long as i got a full ingredient list and side effects etc and the option to sue if any serious side effects happen (which you cant do due to the new laws on vaccines)

Report
 wintertree 18 Nov 2020
In reply to CurlyStevo:

That’s great.  Your post said the IFR is thought to be about 0.3%.  I said I thought 0.3% was a lower bound and an upper bound might be about 0.6%; I got both bounds by working through the “official data” - case numbers, death numbers and random sampling surveys.  A difference of 2x is important and should be qualified.

From the link you gave:

This article presents data from two models estimating daily infections in England, deriving recent IFRs estimates of 0.30% using the MRC unit’s data and 0.49% using ONS data.

So I stand by my comment that the 0.3% is a reasonable lower bound: both your link and my suggestion for an upper bound - derived from “official” data - come out quite close.  Theirs is 1.63x larger; mine 2.0x larger.  As there’s a big factor between the bounds I like to give both so I’m not baking in either optimism or pessimism in what I use.

There is nothing “official” about the CEBM whom you link.  They are not government, they are not NHS.  They appear to have no elevated access to information above your or I.  Their outputs are not peer reviewed.  Their output is not endorsed by anyone official.  Henhegan, as boss of the unit, has been expressing anti-lockdown views since the start and has produced repeated “analysis” in support of that view with no basis in evidence.  I am not disputing the non-official data from them you linked but as a matter of course I work through anything they say as I consider the units output tainted.  The post linked below and the 5 up-thread ones go through testing one of the worst CEBM outputs - their claim that the rising cases in August were explained by false positives.  How wrong they were, and it directly fulled the rising conspiracy nonsense.  Have they put out a statement clarifying that right or wrong it’s no longer relevant to the situation as actual infection is clearly driving numbers now?  No. There was no evidence used in the false positives PCR claim, just presentation of a graph and a bit of a gish gallop.  The CEBM has produced a lot of informative collation of information from around the world on covid but I no longer trust a single one without working it through independently and I sure as hell don’t consider them official.

https://www.ukclimbing.com/forums/off_belay/death_scenarios_could_be_four_times_too_high-727345?v=1#x9328353

Post edited at 07:25
Report
 wintertree 18 Nov 2020
In reply to wintertree:

Here is an early, evidence free gish gallop from the head of the CEBM that traded off his and his groups standing to make its case, and presented no scientific analysis of any data (aka evidence) and no analysis of historical pandemics (aka evidence) in support of the position he put forwards.

https://www.cebm.net/2020/03/covid-19-the-tipping-point/" target="_blank" rel="noreferrer noopener nofollow" id="guid-5fb4d39f8d8b1" class="counterlink">https://web.archive.org/web/20200406021115/https://www.cebm.net/2020/03/covid-19-the-tipping-point/

From the very start the boss of the unit has been:

  1. Calling it totally wrong with regards the pandemic - which is within his field of expertise 
  2. Making no use of any techniques - recognised scientific methods or otherwise - to justify the opinions he puts across on the situation
  3. Putting forwards economic arguments - in the same outputs - which is not within his field of expertise.
  4. Doing all this whilst doing so under the reputation also umbrella of an Oxford University unit callings itself the Centre for Evidence Based Medicine.

You can draw your own conclusions from this.  I would consider myself professionally compromised if I was doing all of those items from above together, and if I was still doing so after 6 months I’d be doing some soul searching.

You’ll note I had to use the WayBack machine for that article - someone has been deleting some of them it seems from their website...

Post edited at 08:01
Report
 wintertree 18 Nov 2020
In reply to wintertree:

My WayBack link got garbled.

I’m not so sure the blog was deleted - whilst the URL was redirecting me to the home page an hour ago, now it’s there. Some sort of glitchy website?

https://www.cebm.net/covid-19/covid-19-the-tipping-point/

Edit: deleted the WayBack machine link from this post as UKC appears to garble them.

Post edited at 08:55
Report
 deepsoup 18 Nov 2020
In reply to Misha:

I almost feel bad posting a Daily Mash link to such an erudite thread.  But it does seem like they might have called it. ;-)

https://www.thedailymash.co.uk/news/health/next-lockdown-to-immediately-follow-this-lockdown-but-to-be-a-separate-different-lockdown-20201117202634

Report
 CurlyStevo 18 Nov 2020
In reply to wintertree:

"> That’s great.  Your post said the IFR is thought to be about 0.3%.  I said I thought 0.3% was a lower bound and an upper bound might be about 0.6%; I got both bounds by working through the “official data” - case numbers, death numbers and random sampling surveys.  A difference of 2x is important and should be qualified."

So the gist of your post is your 'figures' are as valid as the Centre for Evidence-Based Medicine methodology. 

Actually their lower bound on the MRC credible interval was 0.22%.

No offence but your methodology of:

"The various random sampling surveys differ by a factor of about 2x in the number of actual infections, between about 5x and 2.5x the number of detected cases.  So that gives bounds of 0.32% and 0.64% for the IFR.  Either or which is still too high for uncontrolled growth in cases, but likely significantly better than it was in March/April."

Doesn't seem anywhere close to as good as the statistical method they used in the report / paper.

Post edited at 11:52
Report
 wintertree 18 Nov 2020
In reply to CurlyStevo:

In your first post you said "the current IFR across all age groups in England is thought to be about 0.3%" without any reference to a source.

I replied with a simple numerical approach to say that I agreed with at 0.3% in so much as it might be a lower bound, but I thought the upper bound might be about 2x higher.  I made this contribution because it worried be to see you presenting 0.3% in isolation as different ways of linking cases and infections produces significantly different values and 0.3% is towards the lower side of what the median values for the different surveys suggest.

You replied with a reference that you called "official" which is in no way official but which also gives a higher upper bound based on the median value of a different random sampling survey that's not so far from what I suggested.

> Actually their lower bound on the MRC credible interval was 0.22%.

Yes, that is the lower bound using the credible interval data for the MRC dataset.  I was talking about the median value and the bounds on that achieved by comparing different sources of data (MRC, ONS etc), so there is no disagreement here.  Taking the lower credible interview from the data source that gives the lowest median IFR would be compounding the bias, wouldn't it? 

> So the gist of your post is, your stats are as valid as the Centre for Evidence-Based Medicine. 

The gist of my first post was

  1. I trust nothing out of CEBM without working through it myself for reasons already given
  2. Before you gave the CEBM as a source for your 0.3% figure I considered that a lower bound and I gave my estimate of a higher bound along with my method for doing so.  You're free to poo-poo my method all you want but instead you choose to effectively dismiss my comment by calling to an "official" source.  The irony being that, said non-official source did very much what I did - look at deaths and relate them to infections via several different random sampling surveys and modelling based analyses (the MRC nowcast), and the number you gave was the median from the most optimistic input set, and they also gave a second number from the most pessimistic input set that wasn't so far from my second number.   So you dismissal of my comment with a call to an "official" source actually reinforces my point.

> Doesn't seem anywhere close to as good as the statistical method they used in the report / paper.

I think my approach is broadly similar as in your source actually - but let us be clear, they are not "official" and their "report / paper" is a non-peer reviewed blog post.  We both use the same inputs, and different but not incompatible methods.   I've estimated a couple of numbers by reading values off different plots and dividing them to link cases to estimated infections, they've sourced the estimate infection numbers and done the calculation with a bit more precision.  The other difference In methodology is that they have assumed a statistically distribution of times from infection to death in their measures, and I have measured CFR for a range of times instead, and drawn an observation from the current plateau phase in measures which indicates that in this period the statistical distribution "drops out" of the measurement.

So, I stand by my initial observation that the value you gave was on the sunny side, I stand by my estimate for a pessimistic value, I note that your reference you then gave is not offical and uses the same data I do, and I refute claims that my method isn't "anywhere close to as good as the statistical method they used".  

In terms of my final arrogant refutation:

  • You speak of "statistical methods" but there is no use of statistics in either my method or theirs, other than a distribution of times from infection to death in theirs.  What I am doing is a pretty noddy level analysis and I don't actually think theirs is much better.  They're performing no statistical tests on the data to propagate the noise caused by the natural variations in deaths through to their measurements - something I do in some of my other analyses.  
  • I think my approach of testing different lags in the CFR has more validity in some ways than their use of a single convolutional kernel for the infection > death impulse response function.
    • I am visualising and testing the sensitivity of my analysis to a "free parameter" to understand it's importance and how it could bias the results.
    • They're pulling something out of a paper and apparently blindly using it without doing any tests or visualisations to see how sensitive their results are to this "free parameter" or how it is biassing their results.
    • In a rising exponential phase, this lag or convolutional kernel is an absolutely critical parameter.  It's the parameter an analysis is most sensitive to. 
    • So in this instance at least I argue that their method is not as good as the one I am using.  Especially as I'm drawing my observations in a plateau phase where this sensitivity drops out, and their August blog you link was not.   
  • I have eyeballed a couple of numbers from plots; this will be accurate to about with 15% I think.  Yes, this is less robust than taking the numerical data and using it, but the difference it makes is small compared to the credible intervals on those datasets and the time-variation in the measures as the demographics shift with the growth in cases, and the variations from a bunch of other things like which week's data I choose to use from the survays.  I was up front about this.  

I've been contemplating getting a numerical dataset together from the MRC and ONS surveys to let me do a tighter job of this make an IFR version of my variable-lag CFR plots, but for me it's just my personal way of engaging with the news to understand what's going on; I'd expect significantly better from an academic PI with some PDRAs under them and funding than a few Excel plots.  

Further, in terms of their use of a convolutional kernel to link infections and deaths - I have done analysis here that shows this to be a fundamentally flawed approach.  I have done this by attempting to find the convolutional kernel between hospitalisations and deaths and there is no such fixed kernel, it is changing all the time - the same demographic shifts.  The idea that a fixed kernel translates from infections to deaths is provably wrong and bad science - yet they use it.  The only ways to resolve this are to look at data in a "plateau phase" as I have where the kernel drops out, or to look at longitudinal data for each individual death based based on the date their case was detected and to link deaths by date of detection (as a poor proxy for date of infection) to the random sampling data and MRC nowcast.  Neither the CEBM nor I have access to that data and so we're both reduced to high level navel gazing.  I take a professional pride in testing and communicating the assumptions I make in my navel gazing.

Edit: I realised that I've banged on a lot here, and I'm trying to make my case not to have a go at you.  I didn't explain the full detail of what I did in my first reply; it's dripped out here and there over various UKC threads, but methodologically I stand by it and welcome all constructive criticism to tighten it.  The next step for me is to integrate the numerical sources for as many of the random and nowcast datasets as I can to tighten the link from my CFR estimates up to IFR estimate.

Post edited at 12:34
Report
 RobAJones 18 Nov 2020
In reply to CurlyStevo:

I would worry about someone citing an, article that contains

A Bayesian inverse problem approach applied to UK data on COVID-19 deaths and the disease duration distribution suggests that infections were in decline before full UK lockdown (24 March 2020)

to justify their findings no matter how good their methods appear to be.

Report
 wintertree 18 Nov 2020
In reply to RobAJones:

Thanks for prodding me to dig in to their source.  As I explained at length above I disagree with using a single time-distribution relation between infection and death as it is not compatible with the data in the UK so I didn't dig in to their reference.  

The CEBM blog post cites [Wood 2020] for the parameters of their lognormal distribution, but that paper itself just lifts the value they used from another paper [Linton et al]

Linton et all is a paper from mid-February that fits a distribution to 34 patients largely I think drawn from Wuhan Province in January 2020.

Clinical care has changed dramatically between January 2020 and the CEBM blog post in August.  The data they use is sourced from a demographic incompatible with the UK.    It's drawn from a very small number of deaths before almost anyone knew almost anything about treating the disease.

The concern is that the wrong choice of distribution massively biasses the resultant IFR during a rising exponential phase of cases and deaths.  

Garbage in, Garbage out.

I note this is not to deride Linton et al in any way - that was a very important dissemination of information very early on that shaped understanding of the severity this disease posted and I have no qualms with their methodology given the scant access to information at the time.  

[Linton et al] - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7074197/

Post edited at 12:48
Report
 CurlyStevo 18 Nov 2020
In reply to wintertree:

"You replied with a reference that you called "official"

come on! "more official" != "official""

As far as I can tell some of your figures are back of fag packet, specifically how you go from CFR to IFR.

Also 0.3-0.49% is "about 0.3%", in reference to the post I was replying to above, where numbers being quoted were several times higher.

Post edited at 12:56
Report
 wintertree 18 Nov 2020
In reply to CurlyStevo:

> As far as I can tell some of your figures are back of fag packet, specifically how you go from CFR to IFR.

You could have said that the first time around rather than calling to a "more official" source which is in no way official, and which I consider deeply methodologically flawed.

I sat down, looked at case plots and looked at the plots for the various random sampling surveys, noted the multipliers between them and applied those.  Different multipliers for different weeks, so I rounded them off a bit.   It's estimation, and I was clear on that up front.  The different ways of estimating this make little difference to the range, measure the multipliers for one week and the numbers go up a bit, measure it for another and they go down a bit.

As a method this worries me far - far - less - than CEBMs use of a fixed statistical distribution between infection and death.   That can massively bias their findings when applied - as they did - during a rising exponential phase.   Small differences to that kernel can make massive differences to the inferred IFR.  Following RobAJones' comment if you dig in to their data source for this absolutely critical parameter in their analysis it turns out to be wholly inappropriate and unjustified.  Where-as I explore sensitivity to that kind of parameter to understand how it may bias things.

My method was certainly good enough to flag to me that your post was giving only a lower bound and not an upper one, and your source gives something similar to me for the upper bound.

Edit to your edit:

> Also 0.3-0.6% is "about 0.3%" in reference to the numbers being quoted that I was replying to that were several times higher.

I take your point but to me they're not "about" the same at all.  They're both small numbers, but as they act by being multiplied onto something, what matters is the ratio between them, which is large.   To ground it in the real world, the worse case consequences of that difference are about 110,000 extra deaths.    

I stand by my original comment that the value you gave - without source - seemed to me like a low side value.

I know it seems like a small point, sorry!  I think it's quite a big difference.

I'm happy for you to point out that I'm not doing the most robust approach to sitting from CFR to IFR - and I've openly acknowledged that.  I'm pretty miffed at having it suggested that my approach is "not anywhere near as good" (to paraphrase) as theirs.  

Post edited at 13:10
Report
 CurlyStevo 18 Nov 2020
In reply to wintertree:

well in which case your upperbound of 0.6% based on an estimate of IFR to CFR is a big difference to 0.49% presented in the article. You can't have it both ways

I was mostly replying to correct RobAJones who quoted "50000-60000 have died, most estimates put the infection fatality rate for UK between 0.5 and 1.2." which seems very high to me

Post edited at 13:14
Report
 CurlyStevo 18 Nov 2020
In reply to wintertree:

By the way the who say the world wide median COVID-19 infection fatality rate is 0.23% (corrected)   (inferred from seroprevalence data) https://www.who.int/bulletin/online_first/BLT.20.265892.pdf

Post edited at 13:17
Report
 CurlyStevo 18 Nov 2020
In reply to wintertree:

I don't know if you watch that dr John Campbell on you tube, whilst I acknowledge he isn't a credible scientist as such he often quotes around 0.3% as the IFR for C 19. (now a days)

Post edited at 13:20
Report
 wintertree 18 Nov 2020
In reply to CurlyStevo:

> well in which case your upperbound of 0.6% based on an estimate of IFR to CFR is a big difference to 0.49% presented in the article. You can't have it both ways

In terms of saying “Hey your reference free 0.3% seems on the low side of the possible and here’s why I think so” I don’t see a problem.  Both their upper value (which you didn’t quote) and my upper estimate are much larger (as a factor) than the number you gave, and both upper bounds are twice as close to each other (as a factor) than the number you gave.

So I stand by my post suggesting I thought the single value you gave was a lower bound and making a suggestion of an upper one.

I suspect that CEBMs use of a time-to-death distribution from early on in the pandemic as applied in a rising exponential phase is causing them to lowball estimate all their IFRs.  They have something of form for having assumptions in their analysis that end up producing optimistic numbers.

Today’s evening data release should get us to the point the approximate plateau phase has gone on long enough that this lag parameter drops out of the analysis over all reasonable ranges of lag from infection to death, so I’ll be revisiting my numbers soon with a full set of workings.

Post edited at 13:29
Report
 CurlyStevo 18 Nov 2020
In reply to wintertree:

see my post above from the WHO

Post edited at 13:22
Report
 CurlyStevo 18 Nov 2020
In reply to wintertree:

https://www.youtube.com/watch?v=ofGMSH5tjYk& "IFR 0.3% roughly the same as the UK" at 20:30

Post edited at 13:28
Report
 wintertree 18 Nov 2020
In reply to CurlyStevo:

> By the way the who say the world wide median COVID-19 infection fatality rate is 0.23% (corrected)   (inferred from seroprevalence data) https://www.who.int/bulletin/online_first/BLT.20.265892.pdf

Submitted May 13th.  That was pretty early days.  Fag packet maths for seroprevalence in England (where I assume we are both discussing) is rather higher at around 1% for the first “wave”.  I wouldn’t expect it to be that high for what’s happening now.

> I don't know if you watch that dr John Campbell on you tube, whilst I acknowledge he isn't a credible scientist as such he often quotes around 0.3% as the IFR for C 19.

I can’t cope with YouTube videos.  I’m not saying that 0.3% isn’t credible, just that I think something around twice that is as credible.

To come back to my original comment - there is a lot of difference between the different surveys trying to measure “true” infection level.  There’s a lot of uncertainty out there right now.  The source of that uncertainty - and by extension which of the random population surveys is most accurate - is the real question of interest; understand that and the range of possible IFRs tightens. 

Post edited at 13:31
Report
 CurlyStevo 18 Nov 2020
In reply to wintertree:

Dr John is a pretty respected guy if he's happy four weeks ago to say "IFR 0.3% roughly the same as the UK" I don't really know why you are making such a big deal about it. Dr John always refs all the sources in his videos.

Report
 CurlyStevo 18 Nov 2020
In reply to wintertree:

"Submitted May 13th.  That was pretty early days.  Fag packet maths for seroprevalence in England (where I assume we are both discussing) is rather higher at around 1% for the first “wave”.  I wouldn’t expect it to be that high for what’s happening now."

Revised version received: 13 September 2020 – Accepted: 15 September 2020 – Published online: 14 October 2020)!!!

Report
 wintertree 18 Nov 2020
In reply to CurlyStevo:

> Dr John is a pretty respected guy if he's happy four weeks ago to say "IFR 0.3% roughly the same as the UK" I don't really know why you are making such a big deal about it. Dr John always refs all the sources in his videos.

I’m not making a big deal out of that.  I don’t disagree but I stand by my comment that significantly higher values (around 2x maybe 1.5x) seem just as credible given the significant uncertainty on the true scale of infections.  Determining the “true” IFR for the present time is probably going to take a year to do well.  In the mean time, I like to keep the uncertainties on values under consideration.  The reference you yourself gave was in agreement with my post - which you dismissed based on your reference - that 0.3% is likely a lower bound.

I did make a big deal out of suggesting CEBMs method is way better than mine.  I disagree for reasons I have given extensively.  I’ve no problem with the query you eventually raised with my method, but it took 3 goes round to get to it after two dismissals.  As it happens I think it makes little difference vs my issue with CEBM’s but I don’t expect you to accept that and the onus is on me to tighten up that step.

I also think all non-privileged (ie without longitudinal  data from access to all NHS death records) analyses face the same lag problem as the CEBM analysis and that the near plateau phase we are passing through will resolve that for the first time.  Given all this uncertainty I don’t extend trust to any source giving a single number.  To be fair to CEBM they did give the range although they did not recognise nor test the critical sensitivity to their lag model.

Edit: As you’re picking various sources aligned to your 0.3%; let me pick one from a credible group that is giving a value for high income countries close to twice my estimated upper bound  of 1.15% although that was first wave and might represent closer to 0.8% for where the UK is now - https://www.imperial.ac.uk/news/207273/covid-19-deaths-infection-fatality-ratio-about/

As I said, there’s a lot of uncertainty out there.

Post edited at 14:03
Report
 Blunderbuss 18 Nov 2020
In reply to CurlyStevo:

> By the way the who say the world wide median COVID-19 infection fatality rate is 0.23% (corrected)   (inferred from seroprevalence data) https://www.who.int/bulletin/online_first/BLT.20.265892.pdf

The worldwide IFR is not relevant to the UK which has 12m people aged 65+ and a third of the population obese.....

Report
 Blunderbuss 18 Nov 2020
In reply to anyone:

Belgiums death rate by population is already 0.127% so if 0.3% is a correct IFR this would imply 42% of the country has been infected.....wouldn't this high a prevelance slow down the epidemic there considerably, even if they had no restrictions?

Post edited at 15:32
Report
 CurlyStevo 18 Nov 2020
In reply to Blunderbuss:

You are making the classic mistake which had already been made on this thread by thinking Case Fatality Rate is the same thing as Infection Fatality Rate. 

Report
 Blunderbuss 18 Nov 2020
In reply to CurlyStevo:

Where did I mention CFR?

Report
 CurlyStevo 18 Nov 2020
In reply to wintertree:

Imperial college have done so well with their other predictions / models right Lets face it no one knows the number of actual infections in the first wave and in any case they are trying to estimate the overall IFR not the current IFR which is sort of more interesting.

What do you think is going to happen in Sweden? My guess is they are going to end up in full lock down after a much worse second wave hits home hard, as they will be very reluctant to take the measures France, UK, Belgium, Italy etc has had to take. I think they are already really at over 50 deaths a day if they could actually process them like most other countries seem to be able to. 

I really don't like the way they lag their data by 10 days slowly filling up the values!

Personally I've always remained skeptical of their claims that they are at heard immunity already (since about may 2020)

Post edited at 15:37
Report
 CurlyStevo 18 Nov 2020
In reply to Blunderbuss:

Oh sorry I misread that.

Yeah but then maybe the IFR on the first wave was ~1% when the hospitals got full up, people were put on ventilators, and we didn't have any good treatments. It doesn't mean right now its not more like 0.3% for example.

Report
 Blunderbuss 18 Nov 2020
In reply to CurlyStevo:

> Imperial college have done so well with their other predictions / models right Lets face it no one knows the number of actual infections in the first wave and in any case they are trying to estimate the overall IFR not the current IFR which is sort of more interesting.

Which Imperial models are you referring to?

Report
 Blunderbuss 18 Nov 2020
In reply to CurlyStevo:

> Oh sorry I misread that.

> Yeah but then maybe the IFR on the first wave was ~1% when the hospitals got full up, people were put on ventilators, and we didn't have any good treatments. It doesn't mean right now its not more like 0.3% for example.

What post are you making this point in reference to? 

Report
 CurlyStevo 18 Nov 2020
In reply to Blunderbuss:

"Belgium's death rate by population is already 0.127% so if 0.3% is a correct IFR this would imply 42% of the country has been infected.....wouldn't this high a prevelance slow down the epidemic there considerably, even if they had no restrictions?"

no it wouldn't, as 0.3% is the approx (lower bound) for the UK's current IFR not the 'overall' IFR for Belgium

Report
 RobAJones 18 Nov 2020
In reply to Blunderbuss:

Lombardy 0.194

Manaus 0.28

Report
 Blunderbuss 18 Nov 2020
In reply to CurlyStevo:

> "Belgium's death rate by population is already 0.127% so if 0.3% is a correct IFR this would imply 42% of the country has been infected.....wouldn't this high a prevelance slow down the epidemic there considerably, even if they had no restrictions?"

> no it wouldn't, as 0.3% is the approx (lower bound) for the UK's current IFR not the 'overall' IFR for Belgium

0.3% is the lower bound, yes.....so my point is if 0.3% is correct (and I don't think the demographics of Belgium are that much different from the UK) then >40% of their population will have been infected already.

Report
 CurlyStevo 18 Nov 2020
In reply to Blunderbuss:

~0.3% is the CURRENT IFR lower bound based on the current situation in the UK. Many of the deaths you are asking about in Belgium happened earlier in the pandemic when the IFR was higher. You would need an estimate of the OVERALL IFR for that.

Post edited at 16:19
Report
 Blunderbuss 18 Nov 2020
In reply to CurlyStevo:

Ok. I get you now. 

So if it is currently 0.3% Belgium must have been experiencing around 67k new infections per day 3-4 weeks ago.......and the UK around 140k per day.

I'm calling this as bollox

Report
 jkarran 18 Nov 2020
In reply to CurlyStevo:

> Oh sorry I misread that.

> Yeah but then maybe the IFR on the first wave was ~1% when the hospitals got full up, people were put on ventilators, and we didn't have any good treatments. It doesn't mean right now its not more like 0.3% for example.

I keep seeing this about the ventilators and improvements in treatment but I'm not convinced the picture is as rosy as it might seem.

Hospitalisations and ICU use may be down as a fraction of infections vs the first wave but that is in part because of who had been catching covid (the young) and it could in part be because care is being rationed now the disease is better understood with the very old and frail being given basically palliative care in homes not hospitals so they don't end up filling ICU beds, ventilated and dying anyway when those beds may save younger fitter patients (UK ICU's didn't wholesale choke in wave 1 so this represents a potential improvement in outcomes should things get significantly worse but not a real benefit in wave 2 over wave 1).

The whole thing about keeping people off ventilation, it's not that the ventilators kill patients, it's that they're resource intensive (in and post use) and fairly pointless in a lot of cases as by the time many older sicker people get to needing mechanical ventilation with Covid their chances of recovery are very poor.

I'd be surprised if the IFR really has dropped very much now the Covid is back in every strata of society, yes there is the compound effect of a small number of small improvements in treatment but then advanced treatment is also apparently being rationed. For the rest requiring hospitalisation they are all quite small improvements in prognosis so far, there doesn't seem to be a real game changer treatment yet.

jk

Post edited at 16:37
Report
 wintertree 18 Nov 2020
In reply to CurlyStevo:

I think we've largely been in agreement you know, I just got narked to be slagged off compared too CEBM which hits some residual professional pride.  Nothing personal as I'm just a geezer on UKC and they're from hallowed Oxford so I take on board the perspective of someone reading my rants.  

Anyhow, so narked I did a more methodological approach for my upper bound estimate, taking the median value from the last few ONS pilot infection surveys for new infections per day and doing linear and quadratic fits to the data (taking the mid-point of each week reported as a data point) to get a continuous time-series for an IFR vs lag analysis.  For deaths for this analysis I used reported deaths for England.  The reported deaths data is filtered by a polynomial filter to smooth over the day-to-day random variations.  All these are plotted so you can get a feel for what they do.

I then analyse the IFR for the recent period where deaths have somewhat plateaued and likewise cases.   The dashed lines show the data points used, and the subscript in the corresponding legend is the lag.

My previous estimate of 0.62% looks right on the ball or even a bit small.

It's debatable how it's appropriate to deal with the low temporal resolution of the ONS data; unlike CEBM I have presented how I did that and explored the sensitivity to different reasonable approaches to doing that, they have not.    I have also explored the sensitivity to different lags rather than running with a specific model.  When I get time I'll type in a few more weeks of past ONS data and perform this analysis vs a sliding time window for the date of deaths.   It'll need some consideration about how to interpolate the data, probably with 1st and 2nd order fits to each date using the nearest several ONS data points.

I think an appropriate way to interpolate the sparse resolution of the ONS data might be to re-cast detected cases using the ONS data to inform a time dependant low order model of the efficiency of treating (% of cases caught) but that seems a bit too much like hard work.

Edit:  This is using the "death within 28 days of a covid positive" measure; we know that the (eventual) data by death certificate is higher so that could suggest my estimated measures of IFR are lowballs.

The real question that remains for me is why is there so much disagreement between the MRC Nowcast, the ONS data, REACT and ZOE? 

> What do you think is going to happen in Sweden?

Hard to say I think.  The social factors that helped them out first time should still help them out this time, but it seems pretty clear what people interpreted as a quest for some herd immunity in Tegnell's policy hasn't materialised - unsurprisingly.   If the people went in to this winter with some sense of exceptionalism they're in for a rude awakening and that's going to shift how they respond to government guidance.  So far I've been impressed that Tegnell seems to be recognising the reality around him, admitting mistakes and generally responding to the actual situation.  I think they'll do the pragmatic thing, but one thing I am very bad at is predicting government decisions.

> Personally I've always remained skeptical of their claims that they are at heard immunity already (since about may 2020)

Indeed - I think we've generally had similar takes on this from the early days, sorry I went off on one - it was very much about the CEBM comparison I took offence to.

Post edited at 16:48

Report
 wintertree 18 Nov 2020
In reply to jkarran:

> and it could in part be because care is being rationed now the disease is better understood with the very old and frail being given basically palliative care in homes not hospitals so they don't end up filling ICU beds, ventilated and dying anyway when those beds may save younger fitter patients (UK ICU's didn't wholesale choke in wave one so this represents a potential benefit should things get worse, not  areal benefit in wave 2 over wave 1).

A while ago I dove down into some demographic hospital figures from an NHS dashboard.  It look to me like > 85s that are hospitalise are generally not going in to ICU beds either.  After a point it becomes the worse option I think.

Report
 wintertree 18 Nov 2020
In reply to wintertree:

The same analysis over a longer period, smoothing the ONS data by fitting a 2nd order polynomial at each point using the nearest 3 or 4 ONS data points and then smoothing the output with an SG filter.  This inclusion of more data pulls cases lower towards vs the left side of my previous plot.

As with my CFR plots up-thread, the results are very dependant on the lag used but are starting to converge in our current "plateau" phase.  To my reading this points to an IFR a bit over 0.6% using the ONS dataset.

The precise value of the "lag" is a bit nominal as the ONS data is not very localised in time, being typically given for a 7-day interval.

Post edited at 17:49

Report
 Si dH 18 Nov 2020
In reply to thread:

Looking at case data over the last two days for England, I tentatively think we are starting to see the start of the lockdown effect. Tomorrow's reported data will be key because it'll report many of the cases that were taken as specimens on Monday this week, which are always highest. Thursday has been the day we've seen the highest reporting peaks. If tomorrow's number is substantially underneath 30000 (last Thursday it was 31000 I think), or even better below 25000, then we will be on a fairly clear downward path. How steep and how consistently (where) will take a while longer to work out. Maybe another week.

Post edited at 18:27
Report
 wintertree 18 Nov 2020
In reply to Si dH:

I was just looking at the beginnings of Monday’s peak and wondering “How high will you go”?  The residuals about a trend line for cases have been looking tighter this past week suggesting the weekend/Monday distinction is reducing -if  so that’s something to try and interpret carefully looking at the final weekend dip as the reporting lag runs out.   I think you’re right that we’ll start to see it change.  The question is - “by how much”?  

Deaths look to be decreasing quite convincingly as of today; I don’t think that’ll last and expect them to rise again in 7 days or so before they dip again as lockdown works it’s way through.  This is from looking at the pre-lockdown plateau in cases then hospitalisations.  This plateau being some mix of half terms and T2/T3 biting in some areas whilst others rose ready to take over prominence. Total numbers in hospitals continue to rise so the worry there remains that this is the most likely short term crisis.

Report
 CurlyStevo 18 Nov 2020
In reply to jkarran:

I take your point, but I think the outcome of people admitted to hospital in the first wave of covid in the UK was pretty poor. I think the estimates of the percentage of people that have asymptomatic / mild infection have gone up too, which would also bring down the IFR.

Post edited at 18:44
Report
 wintertree 18 Nov 2020
In reply to Toerag:

> Can you start a 'Wintertree's covid stats thread' to put all your graphs in so I don't have to keep trawling various threads to find them?

I'll have a go; maybe just update it once a week on the same day for consistency given the various week/weekend effects.  I'll start one this Friday, as by then another ONS update will be out and I've finally motivated myself to do an ONS derived IFR estimate.

Report
 wintertree 18 Nov 2020
In reply to wintertree:

There was a mistake in the code for the plots on my 17:31 message.  The IFRs didn't agree with the previous plots because of a registration error between the red and black curves.  I've reviewed and fixed and check it, updated plots below.  I should have spotted this the first time round.

Edit: Now I'm wishing I'd pulled in the ONS data ages ago.  The third plot compares the ONS estimates of infections/day with the data for cases in England coming out of the government dashboard, from Pillar 1 and Pillar 2.  There's nothing in it that's not already been commented on, but it's quote something to actually see.  

Post edited at 21:03

Report
In reply to Juicymite86:

If there is insufficient uptake, I expect it would be made mandatory (except if someone can't have it for medical reasons), and rightly so. It's about protecting yourself but more importantly it's about protecting others. There will be people out there who don't respond well to vaccines and there will be a few who can't have them for medical reasons. These people will be relying on herd immunity through vaccination.

As for ingredients etc, I'm sure you can read up about that if you like but you'd probably need a degree in immunology or similar to understand it. If it passes all the relevant approvals, that's good enough for me. May be there's a tiny chance that someone will die as a result of the vaccine, 1 in 100,000 or whatever, but that's a lot fewer people than would otherwise die from Covid. I'd take the tiny risk of a vaccine over the not inconsiderable risk of Long Covid or death!

Realistically, vaccines are the only way back to a normal life for us but that requires enough people to take them.

Report
In reply to deepsoup:

I think the Daily Mash analysis hits the nail on the head.

I think they'll open up non essential retail though. Not convinced about gyms and walls. 

Report
In reply to wintertree:

Regarding IFR, let's consider the position at a high level based on known numbers.

The are just short of 70m people in the UK.

Depending on which metrics you use, there have already been around 60k deaths.

That is almost 0.1% of the total population. At the rate the second wave is going, unfortunately we'll easily get to 70k / 0.1% by the time it's over (probably long before it's over).

Only a small proportion of the population has been infected so far. Antibody studies done earlier this year suggest it's somewhat under 10%. That could be an underestimate and more people have been infected since those studies. Perhaps it's more like 20% now (personally I doubt that) but it's not going to be anywhere near 50%, for the simple reason that if it were more like 50%, we would know a lot more people who have had it and been symptomatic.

On that basis, 0.3% seems like a reasonable estimate for the bottom of the range and it's likely to be towards the higher end of your range of 0.3% to 0.6%.

Another way of approaching this is to consider that we've going to get to 0.1% soon and that's after two relatively brief waves (well we don't know how long this second wave will be - if it lasts all winter, we'll get to a lot more than 0.1%). If the virus is left unchecked to sweep through the whole population, which it has not done thus far, the death rate would logically be several times higher. Again, that supports the IFR being towards the higher end of the range.

In short, if we look at how many deaths we've already had vs how many people are likely to have been infected, it's clear that the IFR isn't 0.2%.

The other thing worth bearing in mind is that people who get mild infections might not get lasting immunity, so the IFR 'per person' as opposed to 'per case' could be higher than your 0.6%. Food for thought.

Report
In reply to Blunderbuss:

> So if it is currently 0.3% Belgium must have been experiencing around 67k new infections per day 3-4 weeks ago.......and the UK around 140k per day.

Good point, this is another way of looking at it at a high level. We are on around 400 deaths a day and rising. Estimated infection numbers (ONS, React 1) have been around 50k a day for the weeks running up to lockdown. That's the best part of 1%.

Report
 Dr.S at work 18 Nov 2020
In reply to jkarran:

> The whole thing about keeping people off ventilation, it's not that the ventilators kill patients, it's that they're resource intensive (in and post use) and fairly pointless in a lot of cases as by the time many older sicker people get to needing mechanical ventilation with Covid their chances of recovery are very poor.

Whilst I agree with most of what you say, ventilators certainly can kill patients, and gung ho early ventilation could increase mortality.

Report
 jkarran 18 Nov 2020
In reply to Dr.S at work:

> Whilst I agree with most of what you say, ventilators certainly can kill patients, and gung ho early ventilation could increase mortality.

I may have phrased it poorly, my point is that it's not in general the ventilators causing poor outcomes, it's that a lot of Covid patients who would die without ventilation will probably also die with it.

I'm not a doctor so I can't speak to how gung ho they are or were about ventilating Covid patients but I presume it's approaching last resort given the high number of patients and low number of ventilator beds coupled with the generally poor outcomes.

jk

Report
 wintertree 18 Nov 2020
In reply to jkarran:

> I'm not a doctor so I can't speak to how gung ho they are or were about ventilating Covid patients but I presume it's approaching last resort

My understanding is that assisted breathing with externally applied positive pressure - “CPAP” - has since been found to work sufficiently for many patients where initially experience prior to covid suggested only ventilation would.  This seems to be because the problem isn’t reduced ability to draw gas in to the lungs but reduced gas exchange across the membranes.  It’s sort of shaping up like this is more a vascular (endothelial?) disease than a respiratory one, and it’s messing with the gas exchange on the capillaries.   CPAP doesn’t create a mechanical infectious pathway into the lungs, doesn’t require sedation and is easier to recover from.  A real break.

Although I’m not the medical kind of doctor so I wait to be corrected!

Report

Papers saying 5 day, 3 household relaxation over Xmas. This will probably change but the more significant piece is that apparently there would need to be 5 days of lockdown for each day of relaxation. Journo speak of course as it’s not clear what lockdown and relaxation mean here. Still, the principle must be right. If anything, I hope this time they will lock down straight after NY (or Xmas, though personally I’m only interested in NY as my family don’t celebrate Xmas), without waiting for the numbers to go crazy.

If the ratio is anything like 5:1, I’m not convinced it’s worth it... However the reality is most people will still have family get togethers, so no point criminalising a large proportion of the population. The point that people would be more likely to break the rules for a second time is also well made. 

Report
In reply to Misha:

Im happy not to get a vaccine , its over a 99%  survival ratio if you get it . Im eligible for the flu jab but never get it , im happy staying like this but if they try to force it, i think many wont react well

Report
 GrahamD 07:29 Thu
In reply to Juicymite86:

> Im happy not to get a vaccine , its over a 99%  survival ratio if you get it . Im eligible for the flu jab but never get it , im happy staying like this but if they try to force it, i think many wont react well

Probably not.  It's a pretty selfish society these days.

Report
In reply to Misha:

> Papers saying 5 day, 3 household relaxation over Xmas. 

> If the ratio is anything like 5:1, I’m not convinced it’s worth it....

As someone who can't really be arsed with Christmas at the best if times, suffering heavy restrictions on getting to the hills for most of January so that people can spread the virus recklessly for a few days would be pretty galling.

Report
 mik82 07:50 Thu
In reply to Juicymite86:

I expect that although it won't be compulsory, going unvaccinated will place restrictions on people anyway. It's quite likely that a certificate of vaccination will be needed for foreign travel for example. 

Report
In reply to mik82:

> I expect that although it won't be compulsory, going unvaccinated will place restrictions on people anyway. It's quite likely that a certificate of vaccination will be needed for foreign travel for example. 

I hope so. It would be good to see the anti-vaxxers squealing.

Report
 wintertree 16:29 Thu
In reply to Si dH:

> Looking at case data over the last two days for England, I tentatively think we are starting to see the start of the lockdown effect. Tomorrow's reported data will be key because it'll report many of the cases that were taken as specimens on Monday this week, which are always highest

This looks pretty compelling after today's data release.

Report
 jkarran 16:33 Thu
In reply to Juicymite86:

> Im happy not to get a vaccine , its over a 99%  survival ratio if you get it .

Not for everyone you might pass it on to though and of course poor uptake makes for a longer harder slog out of the economic quagmire.

jk

Report
 Si dH 17:03 Thu
In reply to wintertree:

Yep, 23000, very positive 

Report
 RobAJones 17:50 Thu
In reply to Si dH:

I'll be interested in the cases amongst school children.

Pupil attendance data published on Tuesday showed the number of pupils self isolating has now roughly doubled from between 3.2 to 3.7 per cent in the week after half term, up to between 5.8 to 6.7 per cent last week – meaning up to 552,000 pupils were not in school.

but I'm not sure this a result of significantly more positive cases??

The average number of pupils isolating per confirmed Covid case increased from 17 pupils on November 5, to 28 on November 12.

Post edited at 17:52
Report
In reply to jkarran:

It just seems abit unjust if it is forced on us , its supposed to be a society where you have say on your own life etc can choose religion,  trad or sport . If something is then made mandatory and forced on you,all them rights you have, have been lost surely?

Report
 Si dH 19:10 Thu
In reply to RobAJones:

There is some data on infections by age group in today's ONS report. None of it looks significant to me and they haven't highlighted any of it particularly, but it's not a breakdown I've particularly followed to be seeing trends.  Under 19s are still the lowest affected age groups by case rate for the pandemic as a whole.

The mass testing in Liverpool has included high schools (except where parents opt out) and the recent data for positive tests in Liverpool doesn't seem to include a particularly high number of young people (see link to pdf at the bottom of the link below), so I think you can be confident there weren't many asymptomatic kids in the schools they've tested at. I have no idea how many schools have been tested yet.

https://liverpool.gov.uk/covidcases

Edit to adds that in the period you are talking about, the general population infection rate has increased a lot, particularly in areas that previously had low rates, so you'd expect to see an impact on school attendance across those areas of the country. In the north west you'd expect it to be slowly getting better.

Figure 8(b) of today's ONS report is particularly arresting as to the effect of different area restrictions in the weeks leading up to lockdown, by the way.

Post edited at 19:17
Report
 RobAJones 20:31 Thu
In reply to Si dH:

Thanks

> There is some data on infections by age group in today's ONS report. None of it looks significant to me and they haven't highlighted any of it particularly, but it's not a breakdown I've particularly followed to be seeing trends.  Under 19s are still the lowest affected age groups by case rate for the pandemic as a whole.

I suppose my pessimistic way of looking at the positive reduction in cases in general, is that if cases in schools age children rise, that might cause a problem in the future.  SAGE has said, there is evidence that, since schools opened older children have

“significantly higher role in introducing infection into households”

> The mass testing in Liverpool has included high schools (except where parents opt out) and the recent data for positive tests in Liverpool doesn't seem to include a particularly high number of young people (see link to pdf at the bottom of the link below), so I think you can be confident there weren't many asymptomatic kids in the schools they've tested at. I have no idea how many schools have been tested yet.

Not sure either,  a national study of  testing in around 100 schools has started, but I don't think any results have been published yet.

Part of my concern (paranoia?) is that there has been a dramatic increase in positive school cases locally. One of my previous schools has reportedly had 13 staff and 30 students test positive (although the school hasn't confirmed this). The school is closed and there is now a mobile testing unit there, so we might find out if the outbreak was more widespread. Hopefully the results from Liverpool would indicate that this is unusual case.

> Edit to adds that in the period you are talking about, the general population infection rate has increased a lot, particularly in areas that previously had low rates, so you'd expect to see an impact on school attendance across those areas of the country. In the north west you'd expect it to be slowly getting better.

> Figure 8(b) of today's ONS report is particularly arresting as to the effect of different area restrictions in the weeks leading up to lockdown, by the way.

Yes, there is no doubt some of the measures have been effective, a couple more weeks of a national (so tiers not confusing the issue) "lock down" and schools being open (no half term) should give a clearer picture.

Report
 jkarran 20:40 Thu
In reply to Juicymite86:

> It just seems abit unjust if it is forced on us

The virus or the cure? 

> its supposed to be a society where you have say on your own life etc can choose religion,  trad or sport . If something is then made mandatory and forced on you,all them rights you have, have been lost surely?

Mandatory vaccination in the UK is about as likely as the second coming of Christ. Grow up, stop whining, take some responsibility.

Jk

Report
 wintertree 20:58 Thu
In reply to Si dH:

> There is some data on infections by age group in today's ONS report

Am I missing something?   The latest release on https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/bulletins/coronaviruscovid19infectionsurveypilot/previousReleases says "Released: 13 November 2020" and is last weeks release?

(Not that I'm sat here mashing "refresh" for the ONS report...)

Whilst I'm here, here's my coloured version of the dashboard cases data with a trend line.  Monday would normally be excluded as its still a bit provisional but I've put it back in from a sense of optimism.  Both weekend days look to be genuinely quite low.  I'm going to have to re-think my doubling time plots now there's a halving time apparently in play for detected cases.  

Latest UTLA plots below as well using the 7-day rolling sum data from the dashboard.  A lot of UTLAs still have growth in cases over that time period although for most of them it is slowing down.  

Post edited at 21:10

Report
 Si dH 21:48 Thu
In reply to wintertree:

Sorry, my bad, I had been looking at the PHE weekly surveillance report, not ONS.

https://www.gov.uk/government/statistics/national-flu-and-covid-19-surveillance-reports

Agree that the 7 day rolling sum isn't showing any significant effect of lockdown yet, most of the falls are still ex tier 3 areas. I think we are seeing it in just the last few days of individual days' data given the trend that looks apparent on a whole-country basis, but until the 7 day rolling sum data starts changing we won't know whether the rate of fall is reasonably consistent around the country, or how fast it is.

I'm also not sure whether we should even expect the places with lower prevalence but higher r to turn over as fast as places who currently have higher prevalence but r close to 1.

Post edited at 22:06
Report
In reply to Juicymite86:

Do you think the risk from a vaccine which has been tested on thousands of people is higher than the risk from Covid (bearing in mind it's not just death, it's lung damage and Long Covid you have to worry about as well)?

If not enough people take it, they'll have to make it mandatory because the world can't go on dealing with this shitty situation forever!

There are many things which are 'forced' on you, like the speed limit or, in our little climbing world, the fact that you shouldn't bolt grit. These things restrict our (hypothetical) freedom but they are there for good reasons.

Post edited at 22:38
Report
In reply to Robert Durran:

They should already be squealing from seeing what a world with a deadly virus and no vaccine looks like. But they're not because [insert ridiculous conspiracy theory].

Report
In reply to wintertree:

It's great that cases are coming down, which is as you would expect - we are now getting results from testing post lockdown infections. We always knew that infections would reduce. There's a question by how much - I suspect we'll still be in the tens of thousands a day by 2 December. And I also suspect that the contact tracing hasn't miraculously improved. Which all goes back to my original question - what comes next? Some kind of Tier 4 for most of us, I suspect. 

Report
 wintertree 23:03 Thu
In reply to Misha:

I’m on the fence about how much of the improvement we see today is down to the effects of T3 really kicking in in and how much is from the lockdown.  I think we’re on a tipping point where it moves from the former to the later.  

> I suspect we'll still be in the tens of thousands a day by 2 Dec

Agreed, and I think gov have more or less committed themselves to ending long lockdown then, through their messaging.

>  And I also suspect that the contact tracing hasn't miraculously improved.

It seems to be gradually improving.  I’ll be back with an updated plot after tomorrow’s ONS report.  It’s a looong way from being effective enough to make sufficient difference though.  Interestingly the data I think hints that the problem isn’t just the absolute number of cases but also a fundamental efficacy issue.

> Which all goes back to my original question - what comes next? Some kind of Tier 4 for most of us, I suspect. 

I think it would be extremely foolish to release anywhere to less than Tier 3 with the unavoidable socialisation (above or below the board) expected over Christmas; healthcare headroom has to be preserved for the potential consequences of that eventuality.  This weeks PHE report shows the flu season as still not having started noticeably.  If that continues to be the case into December the pressure on hospitals is lower so the government can relax more if their sole guiding principle is keeping hospitals from overflowing at a national level; and this is the only principle I can really see behind their actions to date.

Report
 wintertree 23:07 Thu
In reply to Si dH:

Agreed with your comments on the 7-day averages; they lag the current situation.

> I'm also not sure whether we should even expect the places with lower prevalence but higher r to turn over as fast as places who currently have higher prevalence but r close to 1.

Yes; the “dogs dinner” plot I did suggests there’s an almost invariant trajectory locally where cases will rise at a high R before reduction starts.  Lots of possible interpretations to this from local naturally acquired herd immunity to people just not taking it seriously until there’s a local buzz driven by local consequences; I’m leaning towards the later but the dependence of R on prevalence is fascinating and goes right back to the start of this pandemic.

Report
 TomD89 07:25 Fri
In reply to Misha:

There's an ethical difference between enforcing a speed limit and forcefully inserting genetic material into people against their will. I hope you can appreciate that.

If you personally get the vaccine your chances of catching the virus, whether people around you choose to vaccinate or not, is extremely low.

You'll not overcome human psychology here I'm afraid, the more voluntary you make it the higher the chances of people taking it. The more you try and force the situation, the more resistance you'll come up against.

Report
 RobAJones 10:10 Fri
In reply to Misha:

I don't think there will be much scope for relaxing restrictions. Allowing more people (6?) to meet outside and non essential shops to open (I'd like to add gyms/walls, but think that is me being selfish). For me that would be about it. Any more would result in r>1. Now that significant improvements are only a few months away my balancing of economy v health service is for r=1 or just below. Schools need to stay open so unfortunately hospitality need to stay closed (I also think if pubs are open, mixing in households will increase, even if it not allowed. If we can meet in the pub, what's wrong with meeting at home?)

We don't however, seem to have learned during this pandemic, so what will probably happen is a free for all, for a week around Christmas, followed by a lockdown more like the one in March/April possibly including school closures.

Update from 17:50 yesterday. After all the staff and nearly all the students were tested. 17 staff (about 20%) and 78 students (about 10%) tested positive.

Post edited at 10:12
Report
 Monkeydoo 12:51 Fri
In reply to Misha:

Easter ! !

I'm  Still not eligible for furlow money  / won't make it past January never mind unti Easter!! 

Report
 Si dH 13:39 Fri
In reply to RobAJones:

.

> Update from 17:50 yesterday. After all the staff and nearly all the students were tested. 17 staff (about 20%) and 78 students (about 10%) tested positive.

Wow, seems pretty high in a single school. Which LA is that in?

Report
 RobAJones 14:48 Fri
In reply to Si dH:

Cumbria.

I assume they are the majority of the 100 or so cases in Carlisle over the last couple of days

and in another, but very different, Cumbrian school.

It was confirmed on Monday that 53 Sedbergh School pupils had so far tested positive for Covid-19 over the weekend, and that mobile testing units would be testing all pupils and staff on Tuesday and Wednesday in an attempt to catch the spread.

Post edited at 15:12
Report
 CurlyStevo 20:17 Fri
In reply to Misha:

you do mean overall ifr ofc its also pretty normal for ifr to apply to the current situation not overalls stats.

Report
In reply to TomD89:

I don't really see the ethical difference because it's about protecting other people. No vaccine is going to be 100% effective and not everyone can have it for medical reasons (for example people with a suppressed immune system, such as those who are on immunosuppressants following an organ transplant). So there will always be vulnerable people around us and we owe it to those people to get vaccinated for both their protection and our own.

The vaccine, if approved, will be signed off as safe by people who are experts in the field. That's good enough for me (and for most people, I suspect). I can see that at the moment we have about 400 people a day dying from Covid. I very much doubt a vaccine would be approved if evidence showed that 4 people a day might die from it. 4 people a year, may be.

I agree it's better if people take it willingly and so it will be voluntary initially. If uptake is not high enough, it will be made pretty much compulsory one way or another. I hope that won't be necessary but it may well be.

Report
In reply to RobAJones:

I agree, except I suspect the post Xmas lockdown will be similar to what we've got at the moment, for 2-4 weeks, depending on how early it's introduced and how bad the impact of Xmas will be.

Report
In reply to CurlyStevo:

You may be right that the IFR has reduced somewhat due to better hospital treatment. Too early to say.

Report
In reply to Monkeydoo:

Sorry to hear that but unfortunately I think the reality is we won't see a significant relaxation of restrictions until Easter. Everyone will be very happy if things pan out better but realistically I think that's unlikely given a significant roll out of the vaccine is still some months away. 

Report
 veteye 23:22 Fri
In reply to Misha:

" No vaccine is going to be 100% effective and not everyone can have it for medical reasons (for example people with a suppressed immune system, such as those who are on immunosuppressants following an organ transplant)."

This is only a small part of your comment, but I disagree. The vaccine is not based on an attenuated live virus, so there is no danger of a vaccine virus. The vaccine is based on an envelope spike protein, and I feel that those who have compromised immune systems, should be able to have the vaccine. They may not respond as well as those with more typically balanced immune systems, but if they mount some response to the vaccine, it is a start.

Report
In reply to veteye:

I am not an expert on vaccines - my understanding was not everyone can have them but perhaps that is different with the new technologies which are being used now, so perhaps you are right. However it is certainly true that no vaccine is going to be 100% effective, so it's important that we have a high level of herd immunity through vaccination to protect the 5-10% (or more) for who it won't be effective.

Report

Please Register as a New User in order to reply to this topic.