UKH

Post 1 - Four Nations (Cases)

Cases are falling everywhere, and halving times very rapid - less than a week in parts of England and for the other three nations, and they're getting shorter everywhere.  

Cases are falling faster than test numbers which suggests it’s a real fall in infection rather than rapid disengagement with testing, as does the change in hospital admissions over the last couple of weeks.

Link to previous thread: https://www.ukhillwalking.com/forums/off_belay/friday_night_covid_plotting_60-...

Post 2 - Four Nations Hospital Admissions

Admissions are clearly falling everywhere except England; this could be down to cases spiking latest in the oldest ages in England outside of London. 

Hospital admissions numbers are now quite misleading however, as lots of people with Covid are going to hospital for other reasons.  As the spike in infection decays, I think the decay of admissions from Covid might fall off a bit more rapidly than admissions with Covid.  Perhaps this will flatten and prolong the peak of hospital admissions (by this measure, not due to Covid) for a week or so.

Post 3 - England

I’m breaking the demograhpic plot down in to London and areas other than London given the different timings.

• Cases in all ages over 10 are now in rapid exponential decay in, everywhere.
• There’s much less of a rebound in school ages than I’d expected a couple of weeks back; at the time Si dH gave their take on why they thought it would be more mild. 
• The young adult ages that saw some off the highest rate growth (most orange) are now seeing some of the strongest decay (darkest purple) - something we’ve seen a few times before; hinting at recently acquired immunity-against-infection?
• The older ages for England excluding London aren’t seeing very strong decay rates - this may be part of the reason why hospital admissions are not going in to rapid decay, unlike cases.
• The next thing to watch for is a potential university associated spike in young adults...  If anywhere can get sub 2-day doubling times, it's omicron + undergraduate halls of residence...

Lots going on in with the regional rate constants in plot 18

• Cases are basically in freewill in all regions, with the halving times still getting shorter.  
• Hospital admissions only ever really hit decay in London with the other regions growth having back off but ended up at a rate constant of ~0, meaning the actual admissions are holding level.  If hospital admissions were given with the same demographic bins (rant rant rant) as cases, it would be much easier to see if this relates too cases in the oldest ages levelling off in much of England.  
  • Keeping in mind that not everyone reported under this measure went to hospital because of Covid, and that the fraction who did seems to be consistently dropping at the moment.  Still - the fact of their having Covid raises the burden on healthcare due to the need for more infection control measures, so it's not a fully discountable effect.
• Deaths are rising in all regions as the wave of infections makes its way through.

The London hospitals plot shows that occupancy of both general and intensive care beds continues to reduce.

Post 4 - International Plot

The plots of cases/million vs rate constants/characteristic times is a real dogs dinner visually - but reassuringly it's gone back from looking like claw marks on growth trajectories to looking kind of swirly as nations start to curve around towards decay in detected cases.  Only the UK has made it so far, Greece is about too.  I don't look for too much meaning on this plot, more to use it and the sanity checking plot to spot interesting things, and to follow up on the OWiD data explorer [1].  Greece seems particularly interesting - no recent rise in the level of control measures, and their cases spiked (harder than the UKs, with their CFR suggesting infections may have spiked even harder) and then going in to a similarly rapid decay. 

Bar charts showing estimates for the recent case fatality rate ("Rolling CFR") are included with linear and log x-axes.  There's still a ~50x difference in the ratio of detected cases to deaths between stand out nations in different parts of Europe.

[1] https://ourworldindata.org/explorers/coronavirus-data-explorer

Post 5 - Final Thoughts

A lot of the data is becoming increasingly uninformative; I'll aim to wind these up with thread 64 unless anything substantial changes by then.

Whilst the virus remains out of control in places beyond our borders, there's a substantially higher risk of a variant arising that brings escape from immunity-against-infection and that's a wildcard to the question I pose below...

A question for people to ponder:  Putting the wildcard to one side, as we move to endemic circulation of the virus in the UK, how many daily infections do we expect as part of that?  How much will these bunch up in to the winter months?  What sort of a baseline level would this present on detected cases?  

I think that we're likely to see detected cases drop well below that baseline due to the synchronised timing of the big push over 3rd doses and of live infection, as that immunity-against-infection wanes summer conditions should help take over moderating infection rates, then what's happened in Europe over the late autumn and onwards gives us a strong hint about when we'll start to see the first truly endemic wave rise.

Thanks again.

Lots of what looks like good news. 

Any published data on changes to admitted with/for?

Endemic cases - 67M divided by 6 months, so 400k per day. Catch COVID like the common cold - once or twice a year. Less than half detected.

6 months seems to be reasonable estimate for immunity fade now but who knows for years 3-10 of a 2 year old virus.

London data graph shows estimates of the sort of baselines we can expect in days 0-10.

Look what happened last time you tried to do that! 

One of those pesky experts suggested 200-300 deaths per day during winter would = normal covid, in line with a bad flu season with one or the other, rather than both - you can't die twice - getting those that were susceptible at the time. Apols, can't refind story.  At a guess we stop counting cases at some point in the near future.

Looking at the extraordinary spread of the latest wave in a country with low vax rates and no restrictions, I would guess Omicron tends to run out of new people to infect quite quickly.

I've been pondering on how the increased transmissibility of Omicron (but therefore also increased rate of gain in population immunity) will affect the net long term level of cases we should see in the endemic phase. I think it will increase it but I'm not sure how much. I want to put it into a differential equation

Today's journal club paper?

https://www.sciencedirect.com/science/article/pii/S1673852721003738?via%3Di...

Well, that was just talk of winding down the frequency of the posts, not reaching a dead stop. Tempting fate!

The numbers you and Elsewhere suggest seems reasonable; the point is it's going to be big numbers of infections - as it must be for many "cold" causing viruses; especially if seasonality bunches it up.

;In some ways the very rapid spread of waves could be a boon to protecting the vulnerable; the faster it happens, the shorter the period of time that the vulnerable need to isolate for.  The responsiveness of a surveillance system to warn people it's coming is going to need to be pretty tight.

I'm hopeful the "normal Covid" deaths numbers decrease year-on-year; the later in life one is first immunologically exposed, the poorer the enduring immune response I think, and that "average age of exposure" across the population is likely to decrease by about 1 year/year I think.  

I think I'd get lost in the details of that one before long but the top level is very interesting.  If true, it rather drives home the extreme difficulty of an elimination strategy at this point, and the benefits of a robust immune defence against as many different parts of the virus as possible.  Who knows what other variants are in other animals, diverging further and further genetically, waiting to return.   I guess I'm taking the "John Spartan Burger" off the menu at Cassa Wintertree. 

The very rapid fall in cases has been bugging me for a while; I've mentioned a couple of times I thought it looked artifactual, but that testing data and hospitalisation data wasn't setting off alarm bells - allbeit requiring the context over incidental admissions to hospital to keep that calm.

Here's a theory from looking art the rate constant plots.  The last few weeks of data is by far the weirdest ever in terms of how much it's all jumping about.  Still, here goes - with reference to the plot below...

• Omicron growth peaked around mid-December 2021 in England.  
• If things had carried on as usual, the rate constant would have started falling towards decay, hitting that around new year's day, and would have bottomed out at around 12 days or k=(-0.06±0.02) per day.  This is shown as the the green line.
• Socialising around Christmas Day raised R for a while, leading to a rebound in the rate constant as transmission went up - this is the red encircled area.🔴
• As that elevated social connectivity ends, R drops and the cases curve starts to look like one of those almost vertical water slides where you don't really touch the tubes.  It's an unnaturally steep decay though because it's comparing cases now with cases from a brief period of exceptionally high social connectivity.  This also manifests as a region of very negative rate constants (short halving times), encircled in blue. 🔵

So, perhaps we'll see the halving time stabilise around 12 days until wanning effects start kicking in.  That's negative enough that "weather wobbles" should leave it in consistent decay.

Not many chances left to take a punt on what the data's going to do, I hope.

Some sort of Christmas and New Year effect certainly seems plausible.

One thing I'd wondered about is whether there might be a bit of a self-imposed post-New-Year circuit breaker effect, as people see the high case numbers (and look at the number of their friends and relatives who've had covid) and make up for taking risks over Christmas by shutting down a bit in January. This on top of the normal tendency to try to shift the Christmas flab and pay off the Christmas overdraft - I know a lot of publicans will say that January is typically their quietest period anyway.

Number of tests looks lumpy put test positivity has a clearer trend. Decline in cases tracks decline in test positivity. I can't think how that would plausibly* be an artefact.

*I'm assuming the QC has improved since they stopped using an older version of excel limited to 64k records and nobody spotted one lab returned few or no positive results.

https://www.theguardian.com/world/2021/oct/23/covid-testing-failures-at-uk-...

That said, elevated cases over Christmas shouldn't lead to an artificially deflated rate constant afterwards, should they? Assuming that they're actual cases and not some weird reporting artefact (I haven't looked at the testing numbers but I guess they could be if people were being extra-cautious about testing before doing stuff?) then you'd expect to see the natural decay rate but happening from a higher baseline, right?

An excellent post that set the grey matter to work.

I think the answer is "sort of".  If you look at the annotated rate constants plot, the two peak highs associated with the football last summer were followed 7 days later by peak lows.  But, why?  

I think it's in the statistical mechanics, and how infections from a single punctate event segues in to the baseline.

Consider a punctate event (football match, Christmas day).  

• The first generation of infections are very tight in time - all being during that brief event, their detected cases land pretty tightly causing a noticeable spike.  
• The second generation of infections are more distributed over several days (at least) due to lower frequency interactions (the punctate event is over) during the infectious period.  The detected cases are smeared out over even more time.

When measuring the decay of cases between two points in time, a different amount of the first and second or third generation cases are going to be measured on either side of the comparison due to the different widths of their time distributions.  

The key is the decay isn't happening from a "baseline" but from a burst of cases that haven't sloughed out in to a baseline rate yet.  

The week-on-week method rate constants plots I do use no averaging at all but apply to raw data 7-days apart; this delivers a remarkably good plot by absorbing the day-of-week effects, but without any local averaging I think it's going to see some downwards rebound around a week after the cases from a punctate event land.

I hope that's explained clearly!  Having a slow day today - a side effect of sat here knee deep in code with wall-to-wall blue skies outside and not a cloud in sight.  

This should be a real scandal that just seems to have gone away.

PHE should have been running positivity tests on all their data sources using reasonable bounds and should have spotted immediately that this lab was off-the-charts exceptional in terms of what it was returning.

I keep thinking just how profitable it would have been to get the contract, not bother kitting out a lab and just return negative results.  I can't imagine anyone actually thinking they'd get away with with, but it's no less plausible than any other option.

Or, to put it another day, a punctate event rings the system like a bell.

Consider a model world where R=0, and 100% of infections are detected as cases.  (This model works for any fixed infection:case detection ratio however.).  

Figure 1 - individual generations of infection following a punctate event

• 50 infections occurring on day 10 as "Generation 0"
• The next generation of infections is more distributed in time, but occurs distinctly separately to generation 0.
• Over time each generation is more sloughed out

Figure 2 - the baseline number of infections per day in this model world is 20.

• We now add all generations from the punctate event to this.  We see some ringing until the structure of the punctate event is sufficiently blurred out that a new baseline is established.
• You can see how a measure of change over time is going to return all sorts of different numbers based on the periods chosen and the width (in time) of any averaging used.

Figure 3 - the rate constant as measured from the fractional change in cases over 7 days (my week-on-week method).

• It starts with a giant spike which represents the punctate event, and then rings up and down as the ringing structure of the downstream generation works through.  In particular, there's one large, "false" negative value.  It's not really false, but the negative rate constant tells us not that R is negative, but that complex structure is working through the data and that the real fall in cases relates to that complex structure

To translate this to the real world, we'd need to blur everything out because infections from one day are detected over several days, and we'd need to superimpose the general behaviour of the rate constant (falling in this period as immunity thresholds kick in) on to this rate constant plot.

Given the sort of times involved, I probably would expect the real world cases might ring down a bit. as well as rate constants.

But I also agree with your comments that people tend to take it easy in January.  We're also having the mother of all breaks in the weather this year.  

Survey on the impact of pandemic on doctors:

https://www.theguardian.com/society/2022/jan/17/nhs-doctors-tired-impairmen...

Does it though? If I look at my scout troop who are mostly unvaccinated, out of 24 kids fewer than 5 have knowingly had it, and they've all been on bi-weekly or more frequent LFTs.  Do 4/5ths of the unvaccinated population have some sort of immunity?

In the UK mainland, a group of 24 kids in which only 5 had ever had covid would be fairly exceptional, I think. Already back in September before school started up SAGE were estimating that over 50% had already been infected (far fewer recorded cases) and it's rocketed since then - with Delta in that age group there was a very large caseload through autumn even before Omicron came along. I think from memory of your posts that Guernsey's historical prevalence is lower though?

Another theory I have as perhaps being partly behind the super quick turn down of Omicron rates is the effect of self isolation rules. Historically in each wave the time period over which significant rise has taken place up to very high levels has been the order of weeks and the turndown, you would think, has been driven mostly by behavioural trends, restrictions and rising immunity - things that last a relatively long time. In that scenario the number of people isolating simultaneously is never too large. With Omicron, the rise was so fast that we have put huge numbers of people in to self isolation very quickly at the same time. This may be the reason that the fall is more sudden than we might otherwise expect (or at least one reason.) If it is, then the fall will either stop before rate sheet very low, or/and there will be a rebound.

Of course self isolation rules have relaxed in various ways over time, and I could well be wrong.

Covid moves in ever mysterious ways.

The top level rate constant for England has almost stopped decaying by the all tests measure (newCasesBySpecimenDate) but is showing rapid decay still by PCR only tests (newCasesPCROnlyBySpecimenDate).

I'm more minded to think the "all tests" data better represents trends in infection, given recent changes to rules over PCRs.  

I'd said I thought the decay might slacken off to a halving time of ~14 days; it's blown through that and is at 21 days or so (top level), and doesn't look like it's found its slowest decay yet.  Is this part of the bell ringing, or are we headed for pretty lackadaisical decay in cases?

The demographic data isn't the most illuminating; this stops a couple of days before the slowing of exponential decay.   Particularly beyond London, cases are rising in under-10s, and the baseline is looking pretty stubborn in over 90s.  The faster other cases decay, the sooner these areas will come to dominate the top level measures, perhaps that's what's happening.  Time will tell.

What is interesting is that this suggests people are continuing to engage really well with LFTs despite the very optimistic tone in the media.  I say that having just got a bit red "T" line on mine today, so I'll be spending the next week or so staring at a reduced number of walls.  I'd hoped to hold out against getting infected until the next set of lowering of self isolation rules so I could at least go for a walk whilst voluntarily isolating...  

The regional rate constants plot has lots going on...

• We see the slackening of decay in cases in all the English regions.  
• Hospital admissions look to be turning to decay after a while at a plateau (rapid growth - high rate constant - then no growth - zero rate constant).  The plateau here without one in cases is a bit odd; perhaps explained by the demographic shift towards older cases as the wave went on; perhaps conflated by the sheer messed-up-ness of the cases data over Christmas.
• Deaths look to be heading in to decay at the right sort of lags from hospital admissions.

So, it looks like the attention shifts back to cases once more...

Thanks as always. Got my t-line as well this week. Mild symptoms so far, hope yours are too. Fingers crossed there is no line on days 5 and 6 and we can be released!

Thanks; I'll raise a glass of hot squash for mild symptoms in Club T-Line.

In reply to thread:

Scotland is seeing a similar slackening of decay to England, although not as extreme (yet?).

Which is what makes it notable that Scotland are moving to dropping a significant swathe of control measures - https://www.bbc.co.uk/news/uk-scotland-60042339

Pretty impressive high test result numbers from France yesterday.

and record highs pretty much everywhere, although the absolute numbers aren't quite so mind boggling as that 

Yet news on health service pressures (the key pandemic control measure), is still seemingly worse in the UK.

The ONS survey today is the latest good news on UK infections.

https://twitter.com/ByDonkeys/status/1483782129968422913?s=20

Are you sure?? Where are you looking?

Presumably not here:
https://thehill.com/policy/healthcare/588427-france-says-some-covid-19-posi...

or here:
https://ourworldindata.org/explorers/coronavirus-data-explorer?zoomToSelect...

or here:
https://ourworldindata.org/explorers/coronavirus-data-explorer?zoomToSelect...

Yes I'm aware of all those. I never said the healthcare pressure in France is low. I'm just pointing out as bad as things are there the situation in the NHS is currently worse, and has been so for a long time now, for all sorts of reasons (mainly government errors) on top of covid.

As yet another example, the latest news today is in the care system staff vacancies have doubled in 9 months (in the context of brexit, bans of the unvaccinated, increasing wages) and no government actions to counter these factors is making any real difference yet. Then we wonder why we have bed blocking.

What's that coming over the hill, is it a rebound?  Is it a rebound?

Visible in both raw data on the dashboard and 7day rate constants is a turn growing cases at the top level, driven by a return to growth in about half the English regions (so far).

The "ringing a bell" process where the festive spike settles in to the baseline couldn't explain this - it causes the decay rate to ring down from too-low to not-low-enough but always maintaining decay.  So, something more is happening...

Looking at demographic cases for the South East - which as always stop 2 days before the regional rate constants - shows cases in continued growth in ages 0-9, turning to grow in 10-14 and heading towards growth in 85+ and ages 35-45 (parental ages of the school growth, and so driven by this?).  Given the rapid decay elsewhere in the demographics it won't be long before these ages are enough to drive top level decay; if it wasn't for ages 85+ that wouldn't be much fo a concern given where we are now; hopefully as general societal prevalence continues its plummet there will be a downstream turn to decay in those looked after by younger adults.  

Be interesting to see how this develops - especially given the imminent dropping of masks in secondary schools.

Hospital admissions continue to fall.

Certainly looks that way. The provisional data for Monday looks pretty bad - reversing most of the decline we had seen since the Monday before.

Of course it's a much higher proportion of LFTs and presumably that's mostly to do with schools.

Well I've decided to dutifully register my positive LFT result, and I learnt Test and Trace info can be completed online without the tedious phone call I got when we had PCRs before Christmas for my son's bout of Covid, so that's good. Still tedious and clearly systems are not well integrated as you key in a lot of info they should already know multiple times but much better than the phone call!

And then they phoned anyway……. 

🤣

An interesting read here - around three quarters of reported adverse effects from vaccination are also present in the placebo control arms of trials. Things like headaches etc.

 https://arstechnica.com/science/2022/01/up-to-76-of-covid-vaccine-side-effe...

I wonder how that “nocebo effect” scales when one particular vaccine has been receiving a lot of bad press….

 

A friend of mine was amused that his online contact tracing thing ended with a standard "customer satisfaction" type question that the people implementing it presumably tacked on without thinking about it: "would you recommend NHS Test and Trace to a friend?" His response was along the lines of "I bloody hope not!"

Last weeks IndieSAGE slides are out... including more stuff than usual on hospital pressures (bad except for ICU data, which is pretty much back to pre covid) and on schools pointing out  absences (mainly covid) are a record for kids and staff (and this at the start of term.... before we drop mask mandates). Fingers crossed Boris got lucky again with restriction changes but it looks high risk....and on that subject, still no change on NHS staff vaccination deadline (only 2 weeks away now and in London that's still around 10% of front-line staff).

https://www.independentsage.org/wp-content/uploads/2022/01/WeeklySlides_14J...

Risking 10% of the healthcare workforce we are so desperate to hold onto in order to make a marginal-at-best dent in transmission, in a highly vaccinated and prior immunity population. Stunning. Healthcare staff are more likely to have had it, recovered, have immunity and have access to the best PPE, testing, ability to determine antibodies etc.

I have to maintain it's just a threatening ruse by government to increase uptake and that by April 1st it'll be dropped, otherwise it's just too stupid a move to make sense of. Anyone supporting this forfeits their faux-concern for healthcare overwhelm, NHS worker wellbeing and dismantling/privatization of the NHS. I suppose some will still argue these frontline staff who have first hand experience with this for 2 years, probably had it themselves multiple times, still healthy and working hard to save lives are ignorant and selfish people who need to lose their jobs for the good of the whole. Awful.

Yes if you blanketly enforce mandatory absences for any detected cases, irrespective of severity, in a extremely not-at-risk group (children), you're going to see record absences. 

Anyone would think there's a global pandemic or on something.

I'm a bit narked that our school are grading attendance in a way that directly implies absences are bad - we have a "satisfactory" for 2.5 days missed.  Should we be sending unwell children in to school then and not bothering to follow the guidance on symptoms, isolation and testing?  I'm going to politely enquire if this is a school led approach or if it is imposed from the top down; either way seems pretty daft that anyone whose child actually had detected Covid and a 10-day isolation period is going to get a bad attendance score.  For doing the right thing and for following the guidance/rules/law...

Edit - I'm mostly narked on behalf of others.  I'm not one to worry about what "the system" thinks or what's written on a piece of paper somewhere, but I know others who get sleepless nights from that sort of thing, and it's just self-defeating to create perverse incentives counter to the public health side.

As their slides make clear, about half the admissions are now "with covid" not "because of Covid".  Admissions are falling in all regions and cases in more vulnerable adults are generally falling.   We'll probably see some rebound from the return of schools but then when that fades, it's hopefully a background issue through the spring and summer at least.  Yet I don't think the pressures on hospitals are going to return to pre-Covid levels; rather Covid has accelerated a lot of change that was already in motion.  I'm sure the current cabinet would be all too happy for Covid to take the lions share of the blame and not the top-down trajectory over the last decade.

Yet another new brand of LFTs today… 

I agree but the 'with covid' shows either they are struggling with infection control or, more likely, a lot recent admissions with infections but for something else, where covid may be a confounding factor in the hospitalisation. As one consultant pointed out recently: 'I've not seen any condition in hospital that covid doesn't worsen". Infection control eats resource in any case.

Waiting times: into A&E, in A&E, getting a bed once admitted,  and waiting to leave if care packages are an issue, are all way too long. BTW (if you can excuse the sarcasm) Javid has been very clear the NHS is doing wonderfully well under the tories so you must be wrong on that trajectory!!??

On the general prospects of reaching an endemic state I'm more nervous than you as the government have snatched further ongoing problems from close to the jaws of victory several times now. My optimism is fatigued. More cautious messaging and mask mandates were a cheap no- brainer with infection levels so high and cases declines slowing....but the back benchers needed red meat to help save Boris.

It's not a ruse, it's serious idiocy and the deadline is in two weeks, not in April, as you can't get two jabs by April unless first jabbed by the start of Feb. All sorts of very pro-vaccination organisations in and around the NHS are urging a pause.

Our new ones turned up today and are nose only, not tonsils.  You'd think they'd measure the diameter of younger children's nostrils and offer a kit with a correspondingly narrower bog brush.

I was idly wondering what the year or so the LFT regime has done to the nations gag reflex.  I bet nobody has gathered data on that!  (Mind may be going slightly off the rails from some combination of infection and isolation)

Masks worked great over December when cases were spiking didn’t they?

Scotland with stricter rules definitely had lower cases than England right?

Give it a rest - it’s winter and hospitals have got silly Covid protocols to follow - that’s why they’re under pressure.

It’s time to drop all measures (including self isolation and non-ONS testing) and get on with things.

As I predicted at the start of the Omicron wave - “the pandemic will end in March due to high population immunity from infection”. (Ill dig up the post later).

I don’t think I’ve been wrong yet?

As will any respiratory infection on someone ill enough to need to go to hospital here.  I imagine Covid is going to continue to be worse for older people for some time as they're going to struggle to build the immune base that younger people will age with.  On that front - I've not seen in the news any reports yet on outcomes of the MAB or anti viral trials in the UK. 

It does; as we move to endemic status that's going to need some reevaluation - perhaps with more embedded control measures and less segregation, and particularly with relevance to influenza as well for winter 22/23...

What little I've seen of hospitals in the last 18 months isn't promising from a ventilation / air handling perspective.  Drive by and see open windows in winter where there should be full air purification systems ensuring comfortable, sterile air supplied over every bed in every ward.  A major infrastructure program, no doubt.  But from people I've spoken with, not much sign of even portable HEPA units.

Yes.  

However, I'm not convinced controlling the spread of Covid infection and recharging the pandemic potential is a suitable fix to this, more a way of deferring the problem - with compounding interest.

We don't have enough headroom for what - in terms of loading - is not exceptional compared to some flu seasons.  This is a far more fundamental problem than controlling Covid cases; if we follow your thoughts on this thread through to a bad flu season, we'll be using significant society wide NPIs to enable the NHS to run cheaper and smaller than otherwise.  I'd rather pay more tax...

I thought the understanding was that England is closer to the destination than any comparable nation.  There's little doubt we're nearly there, no risk of that not happening.

The risk - as always - remains avoidable damage along the way.  

LOL. Paper surgical masks all round still. No change since March 2020. 

I wasn't sure if they were required to prove they have begun the process? Or simply if they weren't fully vaccinated by the time the regulation comes into force in April then they're out? I've not come across a situation that involves policing the prerequisites of a regulation before that regulation has come into force? Do you know this is the case for certain? eg. someone gets the boot if they can't demonstrate jab one by start of Feb?

Anyone who isn't vaccinated by now is not someone who simply hasn't got around to it, they clearly object for one reason or another. Pausing will only buy time to try and pre-emptively recruit +10% to make up for those you will inevitably lose once un-paused. Anyone in that 10% unvaccinated category will have seen the large font writing on the wall and have begun jumping ship anyway. Most potential new recruits will be from the younger age brackets, who have the least vaccine uptake. 

I was surprised when making a visit to a facility not long ago to be asked to bin my brand new and well fit FFP3 and replace it with a surgical mask, and to then entirely see people in paper masks.  At least I didn't see a single nostril unlike the wider world where they poke out on a regular basis.

There was a good experimental paper that Bruxist posted a while back showing the reduction in transmission from well fit FFP2s/FFP3s.  More or less a dead stop against transmission.  Then I went down a rabbit hole of reading about how these apparently magic masks work, and ended up finding out how candy floss is made. 

No comment! ;D

https://www.channel4.com/news/factcheck/factcheck-tens-of-thousands-of-nhs-...

Most permanent staff won't get sacked immediately they will get moved away from the front line (or there would be an almighty show-down with the powerful unions). Most 'bank" staff will lose their jobs and quite a few on other contracts. We need all these staff for the ridiculous current level of front line pressures.

I agree with wintertree we are close to the 'exit' (subject to no nasty new variants) but plenty of experts see potential bumps on the way in the next month or so, given the very fast relaxation of restrictions that only the tory party seem to want. There are easily enough vulnerable to omicron out their for another rise in hospitalisations. We yet again rely on the public paying attention to covid risk. I share the view of a lot of leading figures fighting this pandemic that the UK population have been very good at making judgements (irrespective of the minority of covidiots), especially given the often terrible government political messaging ... makes me pretty proud.

Both the Sunday and the Monday are showing week-on-week growth now in some regions...  Tuesday is looking pretty big given how provisional it is...

What days are the LFT reporting spikes on?  I know it's not a different day-of-week pattern to the PCR data but haven't properly looked.

As well as plot 18 and a local demographics plot for the south east, I've done a plot of cases in the south east split in to two groups based on which have a rising or falling week-on-week behaviour from the most recent day of data.   School ages and adults aged 35 to 39 which looks to me like the tip of a wedge in ages 3o to 50 or so.  Growth in school children that eventually forces visible growth above wider decay in parental ages; so the question then becomes how long before the growth in children reverse, and do absolute cases become large enough in parental ages to force visible growth elsewhere?

Plotting total cases in the South East for the rising and falling ages is interesting, top level cases turn to rising not when the curves cross but when the gradient on the growth (red) curve becomes steeper than the gradient on the decay (blue) curve. It's going to take a pretty big change in the next two days of demographic data to deliver what we're seeing in the less lagged top-level numbers, which is why I wondered about the day-of-week pattern for LFT cases.  

The Max Planck paper? Here, if anyone wants it: https://www.pnas.org/content/118/49/e2110117118

Some updated plots with the ONS data and dashboard case on dual y-axes, and a comparison of rate constants measured from them.

Remember ONS infections decay more slowly than the symptomatic component of dashboard cases as the former are "live" virus and the later are "new virus", and the virus lives for longer than it's new.

Looks to my eye like testing has held up pretty well through the peak, so hopefully no surprises in store from gross under-detection of infections around the peak - all the more reason to have confidence the decay in hospital admissions should continue.

I can report I've had my busiest week, discharging CV+ and stepdowns back home. We're typically a week or two behind the curve as these are people leaving hospital.

Children have parents, you know...

My colleague is isolating because his children brought it home. His wife now has it, fortunately, not too bad at the moment.

Now the government are behaving as if omicron has already gone with their latest waiting list plans. Good to see Matthew Taylor (NHS Confederation) and Chris Hopson (NHS Providers) two of the most sensible public NHS voices in the pandemic, speaking common sense in the face of such daft targets.

https://www.theguardian.com/society/2022/jan/20/ministers-and-nhs-bosses-cl...

Historically the LFT lows tend to have been on Fridays and Saturdays rather than the weekend, with Sunday sometimes being the highest day of the week. Sunday is usually noticeably higher than Saturday, in contrast to PCR numbers. I have generally assumed this was down to school testing regimes with kids doing a test the evening before school, and maybe also work testing regimes to some extent. I'm not sure how much this stands up now though. It feels like use of LFTs has evolved and the same pattern may no longer apply...

Another thought, now that false positive LFTs are not generally being removed by PCR confirmation, how much do we need to account for that? 

Fag packet calc for the 17th Jan in England - there were about 59000 LFT positives of which 9000 were PCR confirmed.

On the same day there were ~1.2m LFTs reported (numbers not yet finalized but probably won't change much.) False positive rate of 0.3%? (This was the first number we ever got from a study, I've seen others since but don't know if they are better.) That would mean we should expect around 3-4000 false positives.

So to answer my own question it's worth considering but not high enough to change the picture really.

I've been going over the changes announced this week in my head and I can't find an angle from which it makes sense to drop masks in schools or to encourage people back to their offices before dropping isolation requirements.   This just creates an isolation lottery without any intent to stop the virus from spreading - perhaps it's meant to put some breaks on the process and stop it going too fast, perhaps it's about mitigating peak viral loads, perhaps it's just not joined up policy.  Answers on a postcard.

"Spread the virus don't spread the virus".

In reply to stealth_mode_rob:

Interesting, interesting.  A good kind of busy compared to the folks bringing them in...

Probably want to keep the infections high so we can 'burn our way to herd immunity'. I mean, that's been the policy since July, hasn't it?

Not really; there is no enduring herd immunity against this virus.  There’s just the endemic state, which we’ve been moving towards since the very beginning.  The current vaccines don’t deliver us to that state, but they take a lot of the consequences out of the next step that does…

Policy since June looks to me like letting the virus spread at a rate moderated by control measures, in order to keep raising population wide immunity towards endemic levels including those not vaccinated at all.  Although not AFAIK ever stated, a key part of this would seem to be allowing younger people to develop immunity against the full compliment of live viral material.

But what doesn’t make sense to me is dropping all measures against spread *except* isolation orders; this seems like a recipe for maximising disruption.  I’m not sure what the right phasing of relaxations is but this doesn’t feel like it.  

You're assuming government policy is based in science. I'm not convinced it is.

A lot of science goes in to the policy making machine…

Yes... But so does a lot of swivel-eyed lunacy... Quite which comes out is questionable...

It's politics. They've probably done focus groups and polls that show getting rid of isolation would be unpopular with the dicks who voted for them whereas the same dicks hate masks.  Maybe nobody wants to work beside somebody that's got it or a lot of people like being able to claim they need to isolate every now and then (like Johnson himself).

What the Tories wanted was for everyone to catch Covid in one massive spike and after a month with lots of death for it to be over so they could boast about England being first to exit the pandemic.

What actually happened is the isolation rules thwarted their plan by putting an automatic firebreak in the system.  They ended up with so many people isolating they may as well have had the restrictions. 

It would be really interesting to see the graph of number of people isolating and compare that across countries with different levels of restrictions.

Yup, today's update to the demographics has a lot of growth landing on the Sunday.

I've put in a plot of raw case numbers for the south east, binned according to if their age bucket is in week-on-week growth or decay on the most recent day of data.   The legend gives the age bins assigned to each curve.  Looking at this and the demographic plots it looks very much to me like we're seeing household infections follow from school age infections.

So, nothing we've not seen before - and the expectation that the growth in adults will mostly be confined to around 30 to 50 year olds, and will turn to decay shortly after the capacity for growth in kids is exhausted.   Perhaps the coupling is weaker than in some past waves; a lot more recent immunity-against-infection about. 

The latest vaccine surveillance report is out BTW - it's got a lot more plots in it including some on seroprevalence broken down by origin (vaccine vs infection).