UKC

Post 1 - Intro and caveats

2022.  Calendar year 4 with the virus circulating in humans.  

There's a bunch of caveats on the data to keep in mind this week, for various reasons: 

Cases

• It’s been in the news a lot that PCR appointments and lateral flow kits have been hard to come by; no doubt this is suppressing detected cases somewhat.   
• There will undoubtably have been fewer tests taken over the Christmas weekend, resulting in a low in cases and rate constants that doesn’t reflect changes to true infections.
• I'll put a post in comparing pillar 1/2 testing data with the ONS random sampling, in future weeks that may show more divergence.

Deaths

• These had extra reporting lag over the festive period, so there’s a false low and what I think is now a false high in the plot on the splash page on the government dashboard - which I think is deaths by reporting date, not actual date.
• Hopefully the lag is more or less resolved out of the data used for my plots now, but let’s consider the leading edge a bit more provisional than normal until next Saturday...

Hospital occupancy

• Slower discharges over the festive period are likely bolstering occupancy more than we would otherwise have seen, although staff isolation orders impacting care homes may also be extending occupancy beyond the expected discharge point.  So some of the rise in occupancy may be a bit misleading in terms of things going forwards.

Link to previous thread and semi-ccontiniation:

https://www.ukclimbing.com/forums/off_belay/friday_night_covid_plotting_58-...
https://www.ukclimbing.com/forums/off_belay/covid_ctd_since_the_plotting_th...

Post 2 - Four Nations 1

Cases are rising rapidly in all four home nations ow.  The leading edge is messed up by the Christmas break; the very last data marker isn't used for the trendline as it's quite provisional - but it's also very high on all plots - a sign of more growth in the provisional data.  The downturns in the trendline are misleading, from the Christmas sampling low. 

Plot 9x shows cases reaching a peak doubling time of about 10 days in each of the home nations.  I've deleted the right-most part as it's a nonsense consequence of the Christmas low; even the slackening off before that should be considered dubious given that it's in the provisional zone and given all the caveats over cases.

Post 3 - Four Nations 2

Occupancy is now clearly rising in all the nations; the vertical spacing between recent data markers for England and Scotland is large - even accounting for the caveats over discharges, it's going up so rapidly that it's very concerning.  They're getting more spaced out for Wales.  Northern Ireland has only just turned to growth.

Post 4 - England 1

Growth in cases looks to have stalled by the trendline.  I wouldn’t take too much comfort from that; several things going on:

• Festive sampling weirdness
• Difficulty for some accessing testing
• London cases going in to apparently real decay and being a not insignificant contribution to the top level; this is only going to show through for a very limited time period
• Provisional data not used in these plots shows a return to growth.

Hospital admissions have nearly doubled from 1000/day to 2000/day in the last 10 days or so.  The doubling time in plot 9e looks to be backing off - or at least not still getting worse - but that is in the provisional zone.

• Key point - cases are still rising, and so rising daily admissions numbers are locked in for another 10 days or so, at least.

Hospital occupancy

• This is rising - fast - doubling about every 15 days which was the peak we saw as delta became established in the summer.  
• Now, however, it’s doubling from a much higher baseline however and so absolute growth is very rapid.
• As caveated in the first post, some of this is down to issues over discharging patients, themselves in part related to Christmas and in part to staff isolation in the care sector. 
• Recent estimates are that about 1/3rd of patients are “incidental Covid” - meaning they’re in for other reasons, but have Covid.  It’s sort of good news, but whilst Covid still requires enhanced isolation it means they represent more difficulty to healthcare than they would if in hospital without being Covid positive.

Deaths

• These continue to decrease. That trend is starting to become interesting... 
• Looking at Plot 9e - the turn to fast growth has come in the usual sequence to the rate constants for cases (blue), then admissions (orange) and then occupancy (black) but deaths (red) is stolidly remaining in very mild decay - subject to some noise wobble.
• Generally, the characteristic times for occupancy and deaths have crossed between growth and decay very close together in time, and have generally tracked each other pretty closely. 
  • There’s an intuitive explanation to that - the number of people dying (deaths) is proportional to the number of people in hospital (occupancy).  I don’t know how much water it holds, but it’s notable that we’re seeing that phenomenological relationship broken now. 
• This ties in I think with comments from the Health Secretary today and from Chris Hopson [1] that the fraction of patients now going in to intensive care is significantly lower.  
  • We still await a full analysis on how much this is because a large fraction of patients now have significant immunity against severe infection (so immune they were unable to catch delta, now being tested by omicron) and how much is down to reduced innate lethality of omicron.  
• In a rising phase, from this data alone it's impossible to separate an increased lag from admissions to death from a decreased ratio from admissions to death; there's a mathematical degeneracy until the rise in admissions peaks.  Still, it feels pretty promising that the breakthrough infections leading to hospitalisation are indeed less severe as reasonably expected.  The massively important question now is "how much less severe"...

[1] https://twitter.com/ChrisCEOHopson/status/1477263146779910149

Post 5 - England 2

Rolling CFR

Picking up the question of "how much less severe" from the previous post, I have put in a Rolling Case Fatality Rate (CFR) plot.

• This plot estimates the case fatality rate - the fraction of detected cases (not true infections) going on to die.  It does this by dividing deaths on date X (by the 28-day measure) by cases on date X-7, X-14, X-21 and X-28 days and presenting the bounds of those values as a shaded region. 
• Both measures go through a 7-day moving average before this division, and cases go through the “deweekending” process that reassigns some of the Monday and Tuesday reporting peak back over the weekend reporting low.
• This gives a near-instantaneous snapshot in to the CFR by age; the data is broken down so much (age and date) that it's very low numbers; good measurements aren't possible for under 45s.  Lest anyone think I'm complaining, this is a good problem to have with the data.
• My interpretation
  • Omicron infections landed in younger adults first, with London adding significantly to top level cases in that band.
  • The CFR for ages 45-55 is plummeting.  It’s dropped to 30% of its autumn range over the least month. 
  • This is presumably omicron working its way through the system - a lot more cases that are in people who couldn’t catch delta and so who have high immune protection against omicron. 
  • Note that less engagement with testing (e.g. more asymptomatic disease, overwhelmed testing) would raise the CFR, not lower it, so whilst it may be a biassing factor, it's not likely to present false, good news.
  • I don't think this has finished falling - it's going to be another few weeks before most deaths represent omicron infections.  Still, best to assume the worst until the data actually shows it - likewise for older ages; we probably expect the reduction in CFR to be progressively smaller with increasing age...?
  • Caution - the leading edge of this plot is sometimes twitchy and provisional due to reporting lag in deaths; I don't think that there's any reason that would only affect one particular age band, but still - also wise not to commit to believing this till it's moved a week or two away from the leading edge.
• One to watch over the next couple of weeks to see how it shapes up as omicron rises to dominance in the older ages.  In terms of total deaths or inference over hospital demand, a lower 70% CFR would be outweighed by about a week’s exponential growth at current rates, but hopefully the data coming out over the next week or two in older ages will let the modellers significantly tighten their assumptions about this wave.
  • I think the data needed to understand how the next month of spread is going to pan out is just starting to work through; I'd quite like a circuit breaker lockdown from tomorrow for 14 days to take some heat out of the spread whilst the experts crunch all that data.   If we leave it much later, and it turns out this is going to overload healthcare, it will be too late to take a pause and fix that.

Demographics

Plot D1.c shows all of England except for London - the regions where omicron lands later.   Young adult cases aren’t going in to decay in the way they did in London - they also didn’t get slammed to such a high level, with omicron landing later in terms of precautionary measures and press coverage, (and perhaps also without the London underground….?).  Seems like a pretty important datapoint that the London ones went in to decay.  Rapid growth of cases is now established in all older ages.

Regions

Plot 18 - the leading edge of the cases curves is pretty messed up by festive sampling issues.  Still, all regions except London have clear growth.  As comparing the London and not-London D1.c plots shows, the growth is much less biassed towards young adults outside of London, which probably goes a long way to explaining why hospital admissions are doubling much faster outside of London - much closer to the doubling times of cases in those regions.

• Again, those hospital admissions doubling times are short enough to have me reaching for the circuit breaker...  It's out of my reach, however.

Post 6 - London

The plot D1.c shows the young adult spike in cases collapsing pretty rapidly.  The blue indicating decay on the far right of the rate constant plot spreads up through the ages in a diagonal - meaning the turn to decay comes later for older ages.  I don’t think one can tell from the data how much of this is causality by transmission up the ages, and how much is due to successive ages turning the corner later, as a result of more precaution slowing growth, meaning some tipping point is reached later.   The turn to decay is only just starting to land in ages over 70, and perhaps growth is returning to older working aged adults (a very thin sliver of orange on the right).  The data is pretty messed up by Christmas however so it’s quite hard to tell.

What happens next with London is almost the critical information for managing omicron IMO - why did cases max out in younger ages?  is that the end of growth of omicron? Or did a bunch of factors coincide to temporality drive them down?  Understanding this and following the trajectory of these cases through healthcare seems like the key information needed to understand how to manage the next month across the nation. 

• Throwing some brakes on in the form of a two week period of stringent control measure it to give the experts time to get all this data in and analysed seems prudent.  I put Dec 28th on the sweepstakes for a lockdown announcement a couple of weeks ago; leaving it much longer… Well - it might be that everything is going to work out within healthcare limits, but I’d rather we determined that to a much higher degree of confidence first.  

The second plot for London is showing the general and intensive care occupancy between this wave and the alpha wave this time last year.  It’s pretty clear now that ITU is filing up more slowly than in the last wave - a lot more slowly.  Until occupancy peaks, this data alone can’t break the degeneracy  between a lower conversion rate from genera admissions to ITU admissions, and a longer lag in that process.  Knowing as we do that there’s a lot more immunity in the population, a lower conversion rate seems likely; we’ve seen comments on that today from the health minister and from Chris Hopson as linked in an earlier post. This also tallies up with the lack of a rising deaths signal emerging, perhaps.

This is also visible on the Plot 22 for London.  

Having a shift from patients in intensive care to general wards is a great thing for the patients and a good thing for healthcre, although it only buys so much additional headroom in hospitals…. Still, a positive sign that things are moving in the right direction, all-be-it too quickly for comfort.  

Post 7 - International

The phase space plot - showing cases/million on the X-axis and their rate of change in exponential terms as a rate constant on the left Y-axis and doubling time on the right Y-axis.

Basically the more stringent control measures brought in across Europe to limit the spread of delta - that was starting to threaten healthcare due to the larger proportion of older and more vulnerable adults without sufficient immunity - is failing.  Everywhere.

Of the few countries that remain in decay, almost all are turning for growth, despite most of them having more stringent control measures than the UK.

One point to keep in mind - the UK and Denmark appear to be detecting a significantly larger fraction of infections as cases than most of the nations on here.  I had been contextualising this with an estimate of "locked in" death rates using rolling CFR estimates, but omicron has thrown the validity of that out of the window, and it'll be another 2-4 weeks before that can start to get calibrated back in on a per-nation basis.   

• This detection efficacy does not bias measurements of the rate constant or doubling time however, other than as it gets worse with higher cases.

The UKs doubling time for cases is by no means the highest (worst) on this plot, even accounting for Christmas sampling weirdness. 

I find the way this plot is going rather bonkers; hopefully omicron does turn out to be much less intrinsically lethal (i.e. less lethal than delta and alpha for people with no immunity), and hopefully the therapeutics deliver a deus-ex.  I maintain that the data isn't there on omicron having significantly less intrinsic lethality and it's mostly wishful thinking and over-reach so far, but the data also isn't there to preclude this possibility.  Yet.

The situation is South Africa is very interesting; but their routes to immunity, their demographics and their demographically-adjusted scale of excess death are so different to Western Europe that I'd rather take a pause to gather data from our own situation, to apply to ourselves, than engage in potentially wishful thinking inspired by the situation in SA.

This plot uses data sourced from [1] which is a great web interface to their dataset.

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

Thanks! Plenty to think about.

I'm getting the feeling that overall, we just don't know how the next few weeks will pan out and that we still need to know loads more before we can tell. Could be:

• Worst case scenario - occupancy and deaths make previous peaks look like foothills.
• Best case scenario - deaths never really take off again and keep bumbling along in the ~100 range.

And that nobody can yet reliably predict where we are going to be between those.

Is that about right or do we already have "almost certain" tighter bounds?

Happy New Year! Thanks for the updates.

I’m wondering when cases will start to turn a corner. Encouraging that they already have among younger people in London. Hopefully the headline numbers will start to reduce significantly at some point in the next two weeks, although there could be a false drop in the coming days as a result of LFT shortages. It’s also not clear how much of an effect the return of schools and some workplaces will have. Overall, I suspect we won’t see a marked reduction until after mid month. Equally, I suspect we won’t see headline case numbers go much higher - rises in some regions and demographics will probably be masked by reductions elsewhere and LFT shortages. I may be completely wrong…

In reply to Michael Hood: I suspect very death numbers are behind us now (I mean very high relative to the peaks we have seen - you could justifiably say that even 100 a day is still ‘very high’). My fairly uneducated guess is that they will probably rise somewhat but remain below 300 a day on average. Similarly, I suspect ICU capacity won’t be stretched beyond its normal limits as it has been previously, with extra ICU beds being set up in operating units (although normal capacity could get close to getting maxed out). My concern is more around general hospital occupancy, particularly given staff sickness levels and difficulties with discharging recovered patients into care homes.

Re severity for the unvaccinated - I can recall only one early study for Omicron, which suggested c. 10% less severity (can’t remember if that was admissions, ICU or deaths - probably admissions as too early for the other metrics), i.e. materially the same severity. The media headlines are really unhelpful, not least because claiming it’s less severe reinforces anti-vax views. I guess saying ‘it’s less severe for the vaccinated in terms of hospitalisations per 100,000 cases, partly due to booster protection and partly due to there being more mild cases as a result of partial immune escape which we didn’t see with Delta’ is a bit long and complicated for a newspaper front page. 

you mean like this one? https://www.theguardian.com/world/2022/jan/02/new-studies-reinforce-belief-...
good news for mice and hamsters at least, no? 

You've reminded me, I really must refill the ink pad on my 'citation needed' stamp.

This is getting a lot of traction: https://www.gla.ac.uk/media/Media_829360_smxx.pdf

Not an easy read though.

Thanks as usual, some good stuff in there.

In terms of considering possible causes for the fall in London and whether those will be replicated elsewhere, I think the data from Manchester - which followed London up faster than anywhere else - is encouraging.

https://coronavirus.data.gov.uk/details/cases?areaType=ltla&areaName=Ma...

Also this might be worth keeping an eye on. Could answer a few of the unanswered questions in the coming days/weeks. Answers may or may not be those desired, depending on viewpoint.

https://nitter.pussthecat.org/OYCar/status/1477505112738185217

Discussion over the intrinsic severity rumbles on!

The rodent studies are interesting - but mice and hamsters are not people.

The pre-preint LSRH shared is very interesting as quantifies the results on a human lung lining (epithelial cells) which suggests similar things - although if I was being an negative ninny I'd want to see the same sort of assay done on a human bronchial epithelial cell line to see if viral penetration moves in the other direction there - and offers strong support for a change in mechanism.

If this variant really has found a less lethal configuration and manages to supplant delta, that's fantastic.  I've stolidly refused to accept any thinking or planning predicated on this happening as counting on miracles is never wise, but I won't turn it down if it happen.

If it happens.  The proof of the pudding is in the eating - and that really is medical data on what happens in actual people.   If it's both intrinsically less lethal and we're seeing a lot more mild infections due to virus infection people so immune they couldn't even catch delta - and so who retain strong T-cell protection against serious illness from omicron, then. we'd expect hospital admissions and ITU admissions in particular to really crash down....

In reply to Longsufferingropeholder:

Ouch.  That looks like for Australia, Omicron is almost as bad news as Delta in terms of severity amongst the hospitalised.  Which is hard to square off with the comments I've written above on the hints on intrinsic lethality.

That doesn't look promising for some of Western Europe either, although as has been discussed before, different nations have different classifications for what is reported as intensive care, and they have different approaches to pre-emptively putting patients in that space.

A lot is going to become much clearer over the coming weeks.

Each day the intensive care occupancy in London doesn't rise to match hospital occupancy feels a bit like we're moving further from a really bad deaths scenario.

But - that increasing hospital occupancy; if hospitals can't deliver the care those people are going in for; their prognosis will worsen.

All worryingly open to interpretation right now.

In reply to Si dH:

Interesting; thanks.  I've put in a D1.c plot for Manchester.  It's very noisy due to lower numbers, especially in older ages until very recently.  Still, the same expanding wedge shape of blue is appearing from young adults and spreading out up and down the demograhic on the leading edge showing things moving to decay in progressively older ages.

It's really worth flip-booking between the images for Manchester and London - can't be done on the mobile phone interface but arrow keys do it on a desktop. London cases peaked hardest in ages 25-30, and that's where the blue wedge of decay starts.  Manchester cases Peake hardest in ages 20-25, and that's where their blue wedge of decay starts.

The different timings and demographics suggest this is mostly an intrinsic effect of the system - i.e. threshold levels being reached...   You can see that blue wedge making its way up all the ages in both plots over the next week to decay at all ages... 

youtube.com/watch?v=XBZUz4C6kqk&

Of course, schools and workplaces returning add a big chunk to the rate constant and to R, but it does suggest promising things...

On Android, you can just swipe left or right to tab between imagines in a post, once you've selected one of them.

I can swipe on iOS but the image physically slides over to the next.  What I mean is instantly swapping from one image to the other (and back) - this makes the time and age shift of the vertex of the blue wedge really clear.

(when I want to view images like this on iOS from elsewhere I save them to my photo album and can switch instantly between them with swiping the photo roll).

...plus there's the potential for worse outcomes due to covid co-morbidity - we could actually see the situation where covid is a major co-morbidity affecting death rate for other things - which is a good poke in the eye for all those people claiming that covid is only a problem for people with co-morbidities.

Definitely!

Our cases are moving up the demographic pyramid - early December it was mostly schoolkids and some of their parents, xmas eve it was mostly 20-25yr olds, now its into the 30+ on NYE.

No thresholds reached for us, almost certainly due to a lack of previously infected people compared to mainland UK. 7 day rate has stopped rising, but that'll be due to a lack of reporting and change in testing philosophy I suspect - governments chopping and changing things is so annoying. 14 day rate is now up to 3855.

Indeed, see some of the speculative doomsday comments and think a little local knowledge would help rather than looking at simple numbers. You can look at ICU  occupancy for Germany and think the system is nearing collapse BUT you need to know how many are Covid patients and how many of the other patients could be moved out, they are just there because it gives better overall outcomes and the beds are there, in other countries they would never see a hospital bed let alone an ICU.

The operation I had in late summer to remove a blood clot in my leg ( keyhole surgury under local anaesthetic) in the UK according to the NHS is three hours recovery room then discharged home to bed rest,  if bleeding starts then call an ambulance. For us it's 24hrs in ICU for monitoring then 5 days before discharge. That's why Germany has like six times more ICU beds.

I just looked to see how near my local hospital is to collapse, 22 ICU beds, eight with ventilators and another 8 reserve ICU beds. 6 Covid patients, 3 in isolation and 3 in ICU aged 50-77, none ventilated as of the 29th Dec.

A few quick updates.

Plot 18 - Cases

•  The cases rate constant for London is remaining in decay after the Christmas sampling weirdness - decay is unlikely to be caused by running in to the limits of testing capacity, unless capacity is not just reached but exhausted and so lowered.
• The cases rate constants for all the other regions are positive meaning growth in cases continues, but the rate constants are in decay meaning tat the exponential  but not necessarily actual growth is slackening off.  It looks like all but one region is heading towards decay.
  • This  fits with what Si dH and I were discussing about the "Wedge of decay" spreading out through the demographics in London and then Manchester.
  • If feels like there's an intrinsic amount of spread needed before cases go in to decay; and the regions starting later than London are catching up - with a broader and lower peak (slacker rate constants, but for longer).

Plot 18 - Hospital Admissions

• The turn to decay of London cases hasn't translated to a turn to decay in hospital admissions yet - as we expect given that the top level decay in London is driven by younger cases (as per plot D1.c) and is masking a rise in older cases.  Hopefully that blue wedge of decay reaches older ages in the next few days of cases data...  
• Rapid doubling times in all the non-London regions; 5 days in the North East and Yorkshire show admissions are taking off exponentially, everywhere.  Doubling times haven't maxed out yet - assuming they do as a result of the doubling times for cases maxing out and heading for decay.  Absolute admissions will continue to grow so long as the rate constant is positive (i.e. a doubling time) and the non-London D1.c suggests that falling cases in younger adults will mask growth in older ages at least for a while, so looking at London, we probably expect at least another 10 days of sustained high rate constants meaning at least 10 days of exponential-like rising admissions in the non-London region.  That's all from infections already locked in.

Plot 18 - Deaths

• This is all really noisy data from the low numbers (I'm not complaining, this is better than the alternative).   I've expanded the filter window to 11 days for this data.  As the Christmas reporting backlog closes out, deaths in London do look to have been rising consistently whilst omicron era hospital admissions remain in rise.  It looks like this too from the raw data for London from the dashboard.  Still, they're doubling much more slowly than hospital admissions which suggests a changing relationship between admissions and deaths (or a masking effect from the non-zero baseline).  As with the demographic CFR plot and the intensive care plots, tentative confirmation that the additional omicron hospitalisations significantly more mild.
• The north west is also more tilted towards growth than decay, although it's almost in the noise.  It fits perhaps with behaviour on cases and hospitalisations however.

Plot 22 and London occupancy plot

• The latest occupancy figures show much smaller increases over the last two days (right-most red dots are closer together on plot 22) despite rising admissions. 
  • Perhaps this is in part down to day-of-week effects in the admission?
  • Perhaps this is in part discharges picking up in speed?
    • I should really do a calculation for patients-leaving-hospital backed out of the other data series...
• The time-series plot shows that the intensive care occupancy is continuing to rise much more slowly than general occupancy compared to last time.
• Australia is omicron rise to prominence on a similar timescale to us, and is having cases double much faster than the UK, but from a much lower baseline of delta occupancy. They're seeing ITU occupancy rise much faster than us
  • As per LSRH's link up thread 
  • As I noted and jimtitt expanded on, ITU figures are not well suited to international comparisons.
  • Still, interesting to see where this goes...

On the saturated testing capacity thing... worth pointing out it would be both a good and bad news story if true. There seems to be a lot of noise about it from those who love to scream and shout about failures. But if the number of infections truly is anything but a small multiple of the number of cases, and if the latest growth rate figures and ONS numbers were right, London would start running out of people to infect by next week or soon after. So claiming we're missing bajillions of cases is, for some value of bajillions, the same as claiming that it's almost over. I'm not convinced.

London will start to fill up again this week after the Xmas exodus ….. let’s see… 

https://12ft.io/proxy?q=https://www.spectator.co.uk/article/sage-data-sugge...

Need to check their maths thoroughly but on first inspection it looks legit.

Latest variant report has big analyses of relative Omicron/Delta hospitalisation risk and of vaccine efficacy against Omicron.

https://www.gov.uk/government/publications/investigation-of-sars-cov-2-vari...

After adjusting for various factors incl vaccine status:

- risk of hospital admission or emergency dep presentation with Omicron is 50% that with Delta

- risk of hospital admission with Omicron is 30% of that with Delta

(I don't understand the reason for the significant difference between the above figures, which are from the same study)

Total vaccine induced reduction in risk of hospitalisation with Omicron (this includes protection against symptomatic infection) is estimated to be:

- 52% with one dose

- 52% with two doses >25 weeks ago (25 weeks being an arbitrary data cut-off date, not a cliff edge effect)

- 72% with two doses 2-24 weeks ago

- 88% with three doses > 2 weeks ago

Interesting to me as a layman - this seems to indicate how waning of a certain amount of the protection happens over a few months presumably due to live effective antibodies all waning to almost nothing, but a significant amount of protection still remaining for the long term even after a single dose (being no less protective than two doses after 25+ weeks) - presumably t cells.

Edit to add, I do think their one dose risk estimate estimate looks slightly suspect, it seems to use data that says 1 dose is better than 2 (old) doses against symptomatic infection. It's a bit lower (35% vs 51%) against hospitalisation amongst those who are already symptomatic.

From TFA's method:

Drilling into this data is harder given how few SAGE reports come with the data attached (unlike the OBR where every graph has a downloadable dataset). But these can be approximated using scanning software to estimate the numerical data from the graph images. It’s a relatively new tool, and useful in data journalism where sources are reluctant to release the data used for the graphs in public debate.

Speak for themselves!  I've been thresholding images of graphs for 20 years to pull data out and I’m sure people have been doing it with pen and ruler for a century...  The problem with the spectator is how it positively oozes with pre-conceptions - here "reluctant to release the data used" - nonsense; these people are hella-busy and don't have infinite time to prepare, qualify, release and cite every single table used.  Sure, it would be nice - but with the sheer quantity of stuff submitted to SAGE I don't think it's anything to do with suppressing public debate as they hint.  The data is literally there in the plots, it's just been through a numerical transformation from being encoded as numerals to being encoded as the heights of bars.  The data is released, it's just not quite as accessible or precise as if it was released as a table of numerals.

The Spectator is normally pretty tight in their analysis as they need something credible in their articles, and the analysis is the one thing that can be definitively checked, unlike their oozing interpretations and wider cases...

The "hospital departures" rate can be trivially calculated from the admissions and occupancy data.  

• I had a go at deconvolving a kernel between admissions and departures for each wave, but it's not yielding any sane solutions; not surprising using top level data with all sorts of constantly shifting demographics going on and the kernel having a demographic dimension.  
• Sustained rising and falling phases can have the mean time read off the curves; we don't really have a suitable point for that in Omicron yet as there's lots of festive weirdness in the data.

Edit; of course the extra capacity freed up by reduced length of stay is good for about one doubling in occupancy; which is about two weeks of continued growth at this point… Might be enough *if* the return of workplaces, schools and so and colder weather doesn’t lead to a lot more growth of cases….

For sure; then schools, then universities.  Also, the warm spell is about to break (at least up north, have't looked at southern forecasts). 

Lots of pressures adding to the rate constant and to R; it's not that far in to decay in London...

Not sure how the return of universities is going to be managed without giant outbreaks leading to students being confined to quarters unfit for constant confinement.  Still, needs must as the nomination agreements drive, as they say.

There seems to be a common misunderstanding in the public discussion, where several issues are mixed up, namely immune memory, antibody levels and the better ability of T cell immunity to resist immune evasion by pathogen variants.

Antibodies are SUPPOSED to decline after infection or immunization, else you arteries would clog up with them eventually....

What (hopefully) does not decline as quickly following initial exposure plus repeated challenges (either one could be infection or vaccination) is immune memory. However, this is true for both B and T cell responses, and is irrelevant for the issue of better resistance against evasion by the  cell response.

Antibodies need to fit the 3D structure of their target protein surface. The epitopes recognized thus depend on the precise fold of the protein, can be made up of amino acids that are not adjacent in the linear protein sequence, etc.. All of this gives a pathogen lots of options for changes that maintain functionality of their protein while evading immune recognition.

For T-cell immunity, this is much less of an issue. T-cells recognize largely random snippets of proteins presented to them in the context of special molecules (the so called MHC complexes). these snippets can either come from proteins that have been gobbled up by so called professional antigen presenting cells like macrophages ("Hey everyone, look what I found!") or from the continuous sampling of protein production by normal cells ("Hey everyone, look what i just made!").

In either case, because the precise 3D fold is irrelevant, and the snippets are much smaller than a typical antibody target structure, the likelihood to have at least some T cell targtes. that have not been hit by an immune evading mutation is much higher.

thank you for that

So if you have enough Tk cells you won't get sick? 

Thank you!

Do you think it’s limited to the public discussion?  Care to share any comments on this article? 

https://www.bbc.co.uk/news/world-middle-east-59853772

Please correct me here…. As I understand it, whilst B cells and memory B cells persist on successively longer timescales than antibodies, a key mechanism of action against the virus (antibodies particularly released in to serum) and one of their activation routes (antibodies in their membrane hooked up as receptors) mean their efficacy fades faster than T-cells with viral mutation breaking the antibodies faster than T-cell epitopes?  This being more of an issue for spike-only vaccine induced antibodies as the spike is both a key epitope for a neutralising response and a hotbed of mutation (ongoing host adaption?).  So, until B cell response is broadened beyond the spike it’s likely going to fade (in terms of efficacy) faster than the T cell response?

Edit: although the T-cell activation pathway then primed the “from scratch” antibody training response faster once primed T-cells exist?  

That’s the idea - although it’s “less sick” rather than “not sick” - a distinction that gets more important as we age.  Because at the single human level, ageing sucks.

Exactly, as long as you are talking about vaccines being rendered less effective by viral escape mutations. Do not mix this up with declining levels of circulating antibodies or T cells and also but more slowly declining levels of B and T memory cells in an individual patient!

Antibodies work by multiple mechanisms, e.g. clogging up the surface of a pathogen or marking pathogens for attacks by the complement system or nonspecific immune cells. However, the most effective response is by so called neutralizing antibodies that block the physiological effect of their target protein, e.g. by binding to the ACE2 binding site of the spike protein.

Often such binding sites are made from a couple dozen amino acids that are typically on different loops of the protein to generate the required 3D structure, and the more complex something is the easier it is to break!

In contrast, T cell epitopes will be almost randomly distributed along the target, and as long as they are recognized at all by some T cells (which will in part depend on the MHC repertoire of an individual) they will be roughly equally effective. As they are shorter and linear, escaping all useable epitopes while retaining function is statistically unlikely (a virus may have to change several % of all AAs of their spike protein).

Evolutionarily, yes. Practically, I would not worry too much. The spike protein will offer very few sites against which neutralizing ABs can be generated but more than enough potential T cell epitopes (likewise, non-neutralizing AB epitopes). Even if you include other proteins, the neutralizing ABs will almost all be directed against the spike, as blocking of receptor binding, be it by deformation of the spike, masking of the binding pocket or whatever, is the best target for virus neutralization as it interferes with an essential and early step of infection.

I really do not get what is so hot about the inactivated vaccines* that some anti vaxxers claim to wait for. Yes you may also get B cell responses against bits on the inside of the virus, which is probably not that effective, but you lose the MHC-I restricted T cell response (the "Look what I have made" response) you get with vector or mRNA vaccines.

CB

*Except potentially easier transport and storage, which will be crucial in developing countries but is irrelevant at home.

Yup.

I've been getting side-tracked in to the longevity angle of the immune memory cells; it's pretty spectacular compared to many parts of a human.

That's the key bit of the picture I've been under-appreciating I think, thanks.  Those as you say are going to be more stable away from the spike.

A lot more than a couple of dozen if allosteric effects come in to play, where changes on a distant part of the protein alter large-scale vibrational modes changing binding affinities at other locations, such as the RBD...  I suppose that opens up the whacky possibility of an antibody that binds to a low-mutation rate area away from the RBD but closes it allosterically. I haven't done any reading on what the protein modellers have to say on this for Covid.  Hello, new rabbit hole...

With this bit of the spike also being where a lot of mutations are going on, presumably driven by the rapid host adaption phase of the virus....  Which is why I'm so confused by by the reporting on Israel today...  Presumably the membrane fusion part also offers some blocking potential?

I don't want to wait for one, but I did want to get one  as well if it was Valneva - because there are other structural proteins on the virus that are exposed for B cell responses, and by some magic (as with Novovax although that's not a whole virus), there do seem to be T-cell responses usually elicited by de-novo synthesis.  This wasn't historically the case with inactivated virus vaccines?  But something knocking about munches the proteins up in to shorter chains, and the separate adjuvant in Valneva (and Novovax) triggers robust CD4+ and CD8+ responses - at least according to press releases from Dynavax and Valneva; I don't think peer reviewed data is out yet.  That seems like a pretty important advance to me.  It's also where I realise I want to go and spend the next decade doing immunology...  

Dave Garnett pointed out that there are other T-cell epitopes that are only exposed by de novo synthesis such as the viral protease and viral polymerase, and that these can't be done by inactive vaccines, but could be engineered in to a separate mRNA vaccine.  Seems to me like there's a lot of strength in diversity of the T-cell response, especially as we can't rule out some recombinant horror down the line. 

Cheery reporting today from various outlets that London hospital admissions have fallen for two days in a row.  I'd note they've fallen for one day in the North West as well, which has been standing out after London in recent weeks.

It's obviously good news, but I'm not yet convinced it's going to be a sustainable fall.  

1. The latest demographic rate constants plot for London cases is below; week-on-week decay has turned to growth for over 55s (orange sliver on the far right, some of the apparent decay before it will be the Christmas low as well) and decay never landed for ages over 70.  It's hard to understand the impact this will have on hospital admissions without seeing comparable demographic admissions data.
2. I haven't seen a good case made either way if the collapsing cases in young adults is because there's been sufficient infection, or because the stars temporarily aligned over Christmas to reduce R.  The alignment is breaking apart now, so I guess we'll find out soon enough....

The version of plot D1.c for everywhere but London has a dark orange cloud getting bolder in the upper right corner of the rate constants - meaning that the doubling time of cases is getting worse, so cases are growing, lots in these ages.  It hasn't yet turned to decay in longer adults although we saw that in Manchester (plots up-thread) for example.

The plot on London hospital occupancy shows the growth slowing down, and ITU occupancy continuing to rise very, very slowly compared to the last wave.

Plot 18 shows a trend towards decay for the rate constants for cases in all regions - although this will in part be collapsing young cases masking rising older ones - and the rate constants for hospitalisations are starting to max out and head for decay.  Again - there could be masking effects.

In terms of older vs younger cases, the equation is quite different this time around with the demographics of 3rd dose uptake and total vaccine refusal. 

The latest rolling case fatality rate (CFR) plot is interesting - the sharp decay is now apparently coming to ages 55-64 as well as 45-54.  

• Again, a lot of lagged reporting data couple be behind some of this, but it's firming up for ages 45-54 I'd say.
• The CFR for ages 45-54 has now fallen by a factor of 10x from the summer - down from a mid-band value of ~0.3% to ~0.03%.  A factor of about 3x happened before Omicron - presumably down to increasing population level immunity.  We've now had about another factor of ~3.3x which will in part be increasing immunity (including the 3rd for younger adults) and partly omicron infecting people so immune they couldn't even catch Delta.
• If - a big if - we see comparable changes to the older ages as omicron flushes through the pipeline (coming to older ages later) we're definitely living in one of the better timelines.

Perhaps some people in London had a ‘cold’ but didn’t want to have their plans ruined, traveled to the rest of the country, got sicker and got a test, or got admitted..? Of course not, silly me.

My suggestion of building a wall along the London orbital was not well received….

Pretty reasonable hypothesis if you ask me.

“Anecdotally” I know plenty of people in their late 20s/ early 30s who left London feeling a bit under the weather, with parents (50s/60s) who didn’t really care either way, who subsequently tested before seeing more vulnerable elderly relatives. Some tested negative, some positive…

Anyway my self isolation is nearly over.

Fingers crossed that you are right.

Just seen this site. Possibly of interest.  Or possibly I am the last to find it!

https://data.london.gov.uk/dataset/coronavirus--covid-19--cases

For any physics types, there are indeed people doing molecular dynamics simulations to look for non-obvious binding sites to target, including some harmonic stuff.  

https://royalsocietypublishing.org/doi/full/10.1098/rsif.2020.0591

https://chemrxiv.org/engage/chemrxiv/article-details/60c753860f50db53e8397b...

Would love to know if the people doing small molecule screens for drug discovery against the protease etc take any inspiration from this sort of work…

Not helped by the fact you can't get them:

https://www.theguardian.com/world/2022/jan/03/just-refresh-webpage-uk-minis...

That includes NHS staff, who have to spend their off duty time scouring pharmacies for LFT supplies...

I tried finding some FFP2 masks for a long train journey this week. None available in the shops. It doesn't even sound like Boots, Superdrug, etc have ever had any.

I must confess to having used my 4 wheeled bioconfinement bubble instead of a train this year.

Sadly, I don't own or have a licence for a four wheeled bioconfinement bubble...

Having played with protein folding and molecular mechanics (ie modelling stable structures of models while ignoring QM) in a previous life, I’d take those results with an ocean of salt. While the general ideas are likely sound and the techniques and calculations hugely impressive, the level of precision needed to identify the binding energies of a small molecule to even an order of magnitude is likely beyond this type of approach. And that’s just MM (equilibrium steady states). MD (dynamics) is a whole other level. To identify allosterics as relevant  is fantastic, and novel, but it’s a far cry from using these methods to screen small molecules computationally (and will remain so) imho. 
 

Try b and q for masks. 

I have a ScrewFix closer, and i think they carry them. If they have any in stock...

Next day delivery, Amazon

Yep, I got mine from a hardware store.

Not sure what qualifies as an admission and in particular whether that needs to involve an overnight stay but presumably if a case turns up to A&E it’s first assessed and where possible dealt with before being either sent home or admitted. For example, a minor injury dealt with on the spot presumably wouldn’t count as an admission (I maybe wrong). So there must be a certain number of cases which are sorted out by giving oxygen for a little while or some other medical intervention which doesn’t require a stay in hospital.

Has Omicron still not hit hospital occupancy?

"Only one Scots patient with Omicron has needed intensive care since variant emerged.
*Official figures indicate that fewer than 100 people with the new variant have needed hospital treatment."

https://www.dailyrecord.co.uk/news/scottish-news/only-one-scots-omicron-pat...

There are about 1000 hospitalised COVID patients in Scotland so occupancy must be at least 90% Delta rather than Omicron and that Delta occupancy is pretty steady.

*I didn't find the actual official figures referred to on various newspapers.

What's the situation in London?

Anecdote alert: a friend who works as a nurse in A&E confirms this. Increasingly, people who “don’t believe in” vaccines get Covid, call an ambulance, get taken to A&E, but aren’t found to be *that* ill. Some then demand an ambulance home, which gives the staff a bit of a chuckle.

Success. And also picked up an LED batten I've been meaning to get for some time...

That plot you like got an update:

https://nitter.42l.fr/jburnmurdoch/status/1478340581323591687#m

Still needs more countries.

Also feel like this will land well here:
https://nitter.42l.fr/PaulMainwood/status/1478397388607954952#m

It's a reassuring plot.  As we move a question for me is what that plot would look like with immunity against other parts of the virus than the spike - ready for "son of omicron" in (9±3) months time...  Changing rapidly for the better I think...

I'm increasingly impressed by JBMs detailed understanding and communication.  He should definitely become a member of the government's advisory team in some capacity.  With a couple of heavies to force people to listen to him...

The Guardian seem to be changing tone recently - a bit of rear end covering after six months of Private Frazer, lest this crisis sort of evaporate in the next month?  Or am I just getting too cynical and annoyed at almost all the meed?  A story this week publicising someone hawking their book about how the first lockdown was a mistake (IMO they're wrong and hawking the usual, flawed logic, but good way to turn a quick buck now...) and even I think how perhaps the models for the current period might mostly be far too negative...  Funny they've not been shared here...

In reply to captain paranoia:

I'm not too bothered by low test availability as things spike - that was reasonably expected to happen.  What is a much poorer show IMO is the failure to secure sufficient supplies in the right places to keep healthcare staff supplied at the rates needed to prevent unnecessary isolation during the peak.  P^3 - Piss Poor Planning.

One silver lining to cases measures being suppressed by poor testing availability is that the infection > hospital and infection > death ratios are lowering even faster than the data suggests.

In reply to thread:

interesting news stories today - Prof Pollard (Oxford vaccines group and JCVI chair but recused from Covid) taking a different approach to Israel when it comes to a 4th dose, suggesting they should likely be reserved just for the vulnerable. https://www.bbc.co.uk/news/uk-59865108

Plot 22 updates for London and England below - the red dots are bunching up much more for London in recent days showing the increase in occupancy has really slowed down.  The much shallower angle than the previous waves (red vs blue) shows how much more slowly intensive care is filling up. 

Plot 18 update

• No cases plots as the Christmas low is messing up the rate constant measurements at the leading edge, but it looks like top level cases might be rising in London again.
  • London hospital admissions are now in week-on-week decay.  Hopefully this will take some tension out of the system even if cases do return to rising with the return of workplaces, schools and the cold spell etc.
  • Deaths look to be decreasing in all regions; the leading edge can be quite provisional though.  Still, if delta really was finally wrapping up...  Side note - the lack of ITU admissions doesn't necessarily flow in to a lack of deaths once most hospitalisations are in people over 75 or so, as they're less likely to benefit from intensive care.

You know my thoughts on the guardian. Their loyal followers will always jump in and defend them no matter how bad their 'journalism' gets. But that's not for this thread; I already died on that hill.

On testing, found this today:
https://russss.github.io/covidtracker/testing.html  

Thanks for your thoughts

The vibrational allostery literature all seems to be quite theoretical at the moment, with little tie up to experimental tests.  Not much going on with THz spectroscopy yet?  If you'd asked me 5 years ago I'd have said I expected to see more stuff published at that intersection by now.  Shows what a crap futurist I am...

Not quite the same area, but certainly interesting to see Isomorphic Labs become more than an entry in companies house this November just gone...   

IBM's near-term roadmap for qubits is pretty spectacular even before you extrapolate that forwards; and some big names are funding work on translating aspects of MD simulations in to eigenvalue problems amenable to quantum solvers.

Looking at where it's all probably going to be by the end of my careers starts to make me feel like some sort of dinosaur-in-training.  Perhaps time to start planning for some short term ill gotten gains to start that restaurant I've been daydreaming about for the last decade...  (Food Critics will be shot on sight, or at least sent packing with a £10 McDonalds voucher.  Or Wimpy, if I'm feeling magnanimous).

It seems the switch from having tests supplied through NHS employer to NHS staff having to get their own, from wherever they could (or couldn't) happened some time in July.

You’d think they’d have done some basic planning for ultra high cases given  their long standing enthusiasm for the idea…

Speaking of which, the US says “hold my beer” - https://arstechnica.com/science/2022/01/over-1-million-covid-cases-recorded...

Germany looks increasingly like it’s still teetering towards a failure of lockdown with cases definitly not falling in the last ten days and perhaps now rising, and hospital admissions on their way to levelling off from the earlier decay.  They were about the only nation holding out on omicron.

if I'm not allowed to mention a certain 18th century statistician, you’re not allowed to mention qubits. Ok?

China has just hit the red button in a second city. That's only going one way.
Also look at Austria. That surely wouldn't have happened with delta.

It’s not that I have a problem with said statistician, so much as with the houses of cards some build by applying methods to less than appropriate inputs…

Just imagine - Quantum GIGO!

China - buy gold and put it under your bed.

Tempted to start a “what to do with your money when you know a global recession is imminent thread”.

I’m guessing if the Winter Olympics aren’t cancelled that’s the beginning of the end for their zero Covid strategy.

I don't suppose there are any major east-asian bog roll or dried pasta suppliers listed on the LSE? That'd be my plan.

Use with proPRIORty?

anyway, more concerning is the probability of Q-SIGO, surely.

Just think of the H factors that will be built out of the quest for the best quantum perceptron….  Just imagine, all the difficulties of an RNN but quantum! 

I love it when you talk neural.

Today's stats for here - 7 day rate 3890/100k, 14 day 4604/100k (UK 1917 / 3015 respectively, we're still top of the world :-/ ). 7 in hospital from ~1400 cases (admissions 'due to'  / 'with' not distinguished).

Chris Witty said at PMs 5pm press briefing about a third are admitted primarily for something else with COVID and two thirds primarily admitted for COVID.

See youtube.com/watch?v=yl78rcVww4E& at about 20min to 21min.

Here's another stat. 14k nurses left the NHS between April and September 2021

https://www.nursingtimes.net/news/workforce/worrying-signs-register-growth-...

Re the book - that was Woolhouse I think and he made some good points (as summarised in the article) but what he proposed was completely unrealistic for spring 2020.

Germany aren’t in full lockdown though, I think? I’m sure jim titt will correct me. 

China’s bigger financial threat is probably their overleveraged real estate development sector - see the issues at Evergrande. The overall financial and economic impact of Covid has been relatively brief (most economies are now back to 2019 levels or near enough), to the point that investors now seem to largely disregard Covid. Clearly some sectors have been much more heavily impacted and at the individual level many people have lost jobs or income, but on the whole the recovery has been much quicker and stronger than originally expected. Equity and other asset prices have done particularly well, to the extent I’m wondering if there’s another bubble building (of course asset prices don’t have a linear relationship with the economy).

Not sure pasta is big in East Asia. Noodles or rice perhaps. 

About 30-50% lockdown I guess, schools open, eating out etc. Might change this week though, need to get over the holiday statistics hiccup before a decision can be made. One has to remember Germany no longer works on incidence, lockdown policy has to be based on hospital load.

Sure; the control measures are all much lighter than early on, partly a better understanding of what control measures work and partly because they’re all leaning heavily on immunity.  Germany is an interesting example because - by OWiD’s stringency index - they’re one of the strictest in Europe, and this has held omicron’s rise to provenance back about the longest - by sequencing - but it’s starting to look like it’s not quite enough and R for omicron remains > 1, with breakthrough happening now.

This is where the question of intrinsic lethality becomes so important; it’s going to hit a lot more people with no prior immunity over there than here.   Without clear data it’s got to be hard for policy makers to estimate the maximum safe level of omicron cases to stay within hospitalisation criteria.

On that front, Johnson has a new pet phrase “ride out” - hopefully he’s up for a good kicking today over and insistence on doing this as fast as possible; lots of reasons to slow it down a bit right now…

Roy Lilley has been incredibly angry in his first posts of the new year. He sees the latest set of critical incidents as leaving an appalling legacy, given other constraints on the NHS. He firmly blames the government for poor planning and being in thrall to backbenchers. Link below (which includes a classic rant):

"If the NHS had a leader, if the NHS had a chief nurse, if the NHS had a medical director, if the NHS had a board of directors, if the NHS was properly represented, if patients had a voice, if staff could tell what is really happening, if there was an effective opposition in Parliament… they would all face-down this hopeless rag-bag, of a rotten government."

https://myemail.constantcontact.com/A-price-worth-paying.html?soid=11026658...

Various health experts have opined that a very worrying factor is away from London the NHS has much less flexiblility to deal wth critical incidents.

Hold on tight, its going to be a rough ride.

Vaccinations going the way I expected now the low hanging fruit is nearly all picked... talk now is of shifting staff back into the main stressed NHS areas locally. Not finishing the job is of course a risk for all of us, as those most likely to end up in hospital through lack of a booster are those stuck at home (and the huge numbers in deprived aras without any jab, partly because of no government coordinated efforts beyond slogans and insults). 

If, per the widely reported Warwick research, we've basically locked in the current peak - restrictions needed to be applied ten days ago to have a big impact, is what's been said - then to some extent the peak is going to be little impacted by jabs happening now/soon as well - particularly first jabs which take longer to have an effect than boosters. It seems logical that at some point the best outcome will come from redeploying staff into caring for those admitted to hospital rather than supporting further jabs.

To some extent the risk to all of us has already crystallised - the future societal risk from those jabs forgone is largely being addressed/crystallised by infection anyway now.

I think there's more time than the Warwick paper suggests - it's not tracking reality well, and we're still seeing rate constants hold in the older, more vulnerable ages.  Model or no model, if rate constants are holding at high levels, transmission is high and control measures can slow that; there's about a 6 day lag between infection and rate constant measurement, and to address that I'd want to see a model fit to the most recent data; without such a model saying things have peaked in the last 6 days, hope springs eternal and I wouldn't want stale modelling used as an excuse for inaction...

Jabs for the less vulnerable was always a numbers game against transmission - and a very short term numbers game as it relies on a very high antibody level to compensate for the poor binding affinity / match with the new variant.  That high antibody concentration is going to come down pretty quickly...  The lower 3rd dose uptake is in younger adults, cases have peaked there and so - more or less as per your comment - I think more doses there will do very little, unless cases rise rapidly when things return over coming days and weeks.  For these younger adults, the doses were much more about transmission than protection from severe health infection which is already robust in younger adults.

I'm minded to think carefully about what Prof Pollard has to say over jabs and to think carefully about the safest way to enter truly endemic circulation, which requires antibody protection to be allowed to fade in the less vulnerable, to allow their immune systems to refresh their deeper protection against divergent variants.

Forgot to say - totally agree, especially with the demographics of the 3rd dose.

That was the pre Omicron plan, wasn't it - jabs for the over 40s, natural infection for the rest? And I'd only expect that age criteria to rise over time as repeated exposure hardens immune systems.

I reckon the far bigger crises will be in social care. All those care homes and all those personal carers going door to door doing care work at home are going to be in a rough position over the next few weeks.As the sector is fragmented ( unlike Trusts etc) its going to be difficult to get a true picture.

I'd agree they will be hit equally hard but the country has got used to the care system being in a terrible mess; so in these times of parsimonious response and planning the government are much more worried about the NHS, as it could cause them serious damage very quickly. We simply can't cope for any time as a society without functioning hospitals but we have all accepted a seriously failing care system for decades.

That was my understanding of what was happening and the implication of some cabinet Comms quite a while back (although I thought over 50s). I don't think it was ever explicitly stated though.

I broadly agree but if that pans out it's going to be interesting politics how the government will shift emphasis.. The mixed messaging from Boris is hardly encouraging those still due to be boosted to hurry up; and he has no coordinated plan to deal with the home jabs let alone the unvaccinated.  I'd add that Boris and his MPs have rather pinned everything to vaccination being the only way practical way out of this.  Also, as the Prof from Warwick said, on the beeb this am, things may be on the bad side of the modelling and there may be no choice.

On the plus side:  we have ventilation units, masks and testing in schools (at last!...so hopefully a little less educational disruption for the nation's children); there is still no sign of a significant increase in deaths due to omicron; trusting public LFT positives to free up PCR capacity for where it's most needed is sensible.

Worth a read of the Chris Hopson Twitter thread that LSRH linked upthread, it touches on both critical incidents and the current impact on care homes of Omicron outbreaks (seemingly not severe.)

Offwidth, I'm not sure if it's how you present it to us, but Roy Lilley just comes across as an old man who is angry at the world. Those sorts of people generally have poor judgement. I'm not inclined to go and read his stuff at all.

(Edit)

Non sequitur...I'm more and more starting to think that if deaths and ICU numbers still haven't moved much in another 1-2 weeks, we should be considering a radical change of tack.

Yes, I've not seen it spelt out, but that was what I thought the plan was.  

Is that because of the way he writes every other line in bold?

Its the staffing issue due to illness/isolation that I reckon will casue far more issues in the care sector.

What's that, 5x the usual turnover rate?  (And so presumably the usual training rate).

Ouch.

Offwidth, start a healthcare thread, posts like CPs will get a wider audience than those who plod through the data threads over the initial posts...

To what, out of interest?

I'm assuming you start from a "they seem surprisingly flat" position and are therefore saying a lack of movement from that over the next two weeks would prove more definitively for you a far milder outcome - but what would you do in response? Further reduction in self isolation requirements, to reduce staffing pressures?

You're free to ignore Roy but I'd say its a major mistake. A lesson I learnt long ago is you don't have to agree with everything a well informed person says to gain useful information. He is the best informed commentator on NHS management in the UK and at a time when the govenment are actively suppressing NHS bad news he is an especially vital information source.

Start one yourself, if you think its important, I'll contribute. My posting history contains many OPs on the subject where I see it as most appropriate. I've explained repeatedly why I post here (I think the government induced interrelated NHS and Care System structural problems are the main factor now in the English pandemic control, not the covid hospitalisations).

They have, and that's worked really well at getting over the hump in terms of vaccination numbers. But I guess looking slightly further ahead - by say March or April I see little merit in further expending tons of energy on widespread full-coverage vaccination campaigns any more (absent further variants that demand it) - the gaps will have been mostly filled by infection by then, so it will move from a societal problem to an individual problem for the non-vulnerable - and there are plenty of societal problems left to fix, like the care system or the state of the NHS, and plenty of thinking to do about what the right way to live with covid actually is for the vulnerable and the non vulnerable and what changes we need to make in response.

Huge focus on all-adult vaccination was never going to last forever - just like "stay at home" had to morph, this phase will also pass.

One example:

Time is rapidly running out to avert a period of miserable isolation for undergraduates when term returns.  Some universities managed 3-day doubling times pre-vaccination.

Isolation in dense appartement blocks with no "private" outdoor space and with "catering" provided by the universities (packets of crisps and bananas in some examples last year...) for 7 days is going to be as miserable as ever.  I think the students have been kicked about enough over the last 18 months that they could at least be allowed out to go for walks and to meet someone outside the facility to take delivery of a take-away etc.

(Edit: I don't think the data will be in to rationally drop isolation by the start of term, which is why I'd have planned for distance learning until 1st Feb at least, followed hopefully by a return with less strict isolation requirements).

Yes, these are all factors, but - as you've pointed out to others before - these are notionally data driven threads on the primary data around infection and consequences, not the other issues you are raising.  I'm not saying don't discuss or raise them here, but that I think they deserve their own thread with a coherent narrative to tie it all together and to make the points - and they'll sit better there as the points are as inextricably linked to politics as they are to the pandemic data.

I really do think you should start a thread.

I've monopolised too much discussion as it is, and I don't have the insight or awareness in to many of the issues that you are better informed on.

Remove the requirement to self isolate after a positive test and make it guidance only would be my guess at the next step of the data is as good as it looks.

We have nowhere near enough data to make this call yet, but that would be the obvious next step.

Start to change people’s mindset and significantly increase population immunity by Omicron infection.

Not just Covid but many other pathogen that realise said vulnerability ; we've got rock solid evidence now that we can control the risks from influenza far more easily than we can from Covid; and I believe we've never had as effective vaccines against influenza.  

An idea I'm warming up to is giving the NHS App access to centrally calculated, individual risk score for various respiratory nasties, based on individual medical records, and tied in to a higher powered, many-pathogen surveillance (shunting some of the current PCR capacity from P1/P2 to an expanded ONS random sampling survey) to give risk tailored to the present time and geographic region of the app user.  This could also tie in with "Air hygiene ratings" on restaurants, pubs and workplaces etc.

We've generally - in terms of proactive control measures, not follow up medical care - let flu sicken and kill old people even whilst flogging symptom suppressing medication with the advertising message that you can go on the bus to your office whilst infectious.  Covid and other diseases are two totally different strands, and they need knitting together cohesively over the next 24 months; wouldn't it be great to see lessons learned over Covid being used to empower vulnerable people to live lower risk (and therefore better quality / long) lives against all such risks?

Another one is facilities level air handling to avoid weapons grade viral loads.  I look back at the kind of cold/flu > chest infection that's not uncommon in lecturing staff at the end of pre-covid winter terms and think it's just madness to let that continue.

I don't have a fully formed idea of what I think should happen but yes, the obvious need is to reduce staffing pressures in critical infrastructure, particularly the NHS and social care. If the risk of death or being ill enough to need ICU is very low for everyone involved, then at what point do we start allowing covid-positive (but asymptomatic) staff to work in areas that are unable to provide their usual range of services? I would like to see something more targeted than VS suggests above, although I think his is probably a better guess at what would actually happen.

I'm also thinking about how we would have looked at covid differently in March 2020 if the first few weeks of the pandemic had gone like these last few weeks have, with little rise in deaths or ICU and just a big wave of relatively mild hospitalisations. I suspect the lockdown would have stopped at the first checkpoint and people would have gone back to normal. It was the daily deaths figures that really got the public onboard with the need to control covid in the first place.

But like I say, not fully formed and a lot could happen in the next 1-2 weeks.

On a website apparently last re-structured in about 1998?

Yes, I would have thought you could adopt some of the wording about exercise from the lockdown legislation into the self isolation legislation - there are already exemptions there for some things like shopping for food (2.3.b here: https://www.legislation.gov.uk/uksi/2020/1045/made - although presumably there's a more recent version that allows for the earlier release criteria); some sort of lockdown-style "solo exercise" wording is presumably there on the shelf already.

In reply to VSisjustascramble and SidH:

Given delays in testing turnaround and testing saturation, isn't self isolation becoming (unknowingly, perhaps) more voluntary anyway, in practice? The link above, plus articles I'd found previously but can't refind now, had led me to understand that the legal requirement to self isolate starts when you get a positive test result, not when you get the symptoms or order the test.

Not arguing, for clarity, with the idea that changing the public guidance would result in a step change in people's approach, since I suspect most people are still isolating from point of symptoms as per the guidance - same as I suspect a lot of people will still self isolate if pinged by the app.

Yes, although I haven't checked recently, the previous legal requirement for self isolation began when you were instructed by a member of test and trace - obviously following a positive result or being traced as a contact that must isolate.

Do people really use the app and respond to it? I don't know anyone who would isolate after being pinged by the app. They might moderate their behaviour slightly but more likely they take a test then carry on?

I guess much of the force came out of it when vaccinated close contacts no longer had to isolate (August, was it), but prior to that I would have assumed so - that was the whole point of the pingdemic, right?

And the whole irregular flutters of media concern over mass deletion of the app suggests that at least some people think there's a downside to being pinged - if you're not trying to avoid a perceived adverse consequence of being pinged, why delete it?

https://www.bbc.co.uk/news/uk-59878823

Some freeing up of PCR capacity announced - no need for confirmatory PCR for a asymptomatic positive lateral flow whilst cases are high.

Good to see some moves to free up PCR capacity, even if it does mean that data is loosing it's consistency.  It'll be oh so very interesting to see the next ONS update when it comes out.  Think we'll hit double digits?  

Scotland have also reduced self isolation from 10 days to 7 going off the BBC Live page (aside -those pages are a real PITA for linking to specific content, a shame each item doesn't have a permalink.)

Whatever grumbles people have over the politicisation of the Covid response, I'm thankful it's not an election year in England.

https://www.bbc.co.uk/news/world-europe-59873833

French President Emmanuel Macron has been accused of using divisive, vulgar language after he used a slang term to say he wanted to make life difficult for unvaccinated people.

"I really want to piss them off, and we'll carry on doing this - to the end," he told Le Parisien newspaper.

Instantly reminded me of an early South Park episode...

youtube.com/watch?v=fynWOio9jBo&

Pfizer and Moderna were talking about Omicron specific mRNA vaccines in about 3 months.  So even if the protection from mk1 Moderna and Pfizer doesn't last it only needs to be a bridge to mk2 becoming available.

If the UK government say no more jags and the US / Germany are going a different way my feeling right now is I'll get out my credit card and buy whatever Pfizer/Moderna/Valneva are selling.  I'm happy to spend a hundred bucks a year for a few years playing whack-a-mole with jags vs mutations.

If you look at his CV he last worked in the NHS in 1998, it is one hell of a long time ago. I know that you clearly hold him in high regard.

I regularly get targeted by people of similar ilk in the manufacturing sector especially in exporting. When push comes to shove, they are just out of date and in the end not well informed, the number of times I have to pull up these so called commentators on what is really happening is unreal.

His useful life on commentating probably ended about 5 years after he left. By then everything has moved on.To him its now a consultancy /speaker/journalism fee generator to nicely top up his pension.

That assumes that vaccinators are all clinical staff; they aren't. It also assumes that vaccinators, if clinically trained, will return to hospitals; many won't, having come our of retirement to support vaccination.

I hold him in high regard as he gets information about NHS management issues out in public faster than the serious press. If he was indeed as moribund as some here suggest how on earth do they think he does that and gets all that wide range of senior people reading his blog?  By magic??  His style is typical of him being an old bloke. I disagree with him on quite a few things (most notably: the benefits of a national pay spline; and how to best to work with staff side reps).

He spends all day reading lots of other sources  you do not know  about?

just because he has a blog does not mean lots of people in the right place read it. They probably have better things to do.  

The only thing I had assumed was that the "where possible" clarification was too obvious to need mentioning!

Clearly that was an error on my part. I shall try not to assume so much in future...

Hmmm..  man on internet forum calls out someone who ran Trusts, advised government, led Inquiries and devoted his retirement to encouraging and communicating the spread of good management practice in the NHS. Man on internet who thinks 300,000 sign ups to the smallish world of  NHS management blogs in a rather old fashioned style is trivial?

As I said  last worked in the NHS nearly 25 years ago--- thats a generation in my book.

I appreciate you hold him in high regard.

But what has he done on grit?

(Probably more than me, I hate the stuff - all about finesse and technique and not brute force and ignorance...)

Some pretty clear language from the select committee today recognising the severity of the situation and the risk of a positive feedback loop on staff departures.

https://www.bbc.co.uk/news/health-59888785

But in its report published on Thursday, the Commons health and social care select committee said the health service was hugely understaffed and was facing an "unquantifiable challenge" in tackling the backlog. 

While it welcomed the government funding, the committee said NHS England was yet to publish "how it plans to meet its workforce requirements", despite this being promised by Health Secretary Sajid Javid by the end of November. It said it wanted the government to draw up a recovery plan for services by April.

The report said many NHS staff were feeling under pressure and could quit unless they saw the "light at the end of the tunnel" of more staff coming.

Jeremy Hunt - the former health secretary who now chairs the committee - said the NHS was short of 93,000 workers and there was "no sign of any plan to address this".

He described the staffing crisis as "entirely predictable", adding: "The current wave of Omicron is exacerbating the problem, but we already had a serious staffing crisis, with a burnt-out workforce.

"Far from tackling the backlog, the NHS will be able to deliver little more than day-to-day firefighting unless the government wakes up to the scale of the staffing crisis facing the NHS."

Well, there's only one question really now - "Is the threat from Covid over?"

I'll be disappointed if in two weeks we're not looking at the data and saying "Well, that was that." and watching tension bleed out of every part of the pipeline.  

The threat to high risk individuals is going to remain, and it will never really be "over" any more than other circulating diseases, but hopefully the virus' ability to overwhelm healthcare will be retired at least to the thread level we've long lived with from influenza.  (Although if the worsening issues across healthcare aren't addressed, a bad flu or covid season could be a disaster...)  

So, that's my optimists hat on!  The cautious option is to wait and see.  Two more weeks of data, please...

It'll be much longer before the cases and hospitalisations wind down to more truly endemic levels, and then there needs to be a discussion on what kind of additional death rate this represents over the baseline pre-covid all cause mortality, and what level of control measures that is or isn't going to require, and how the other control measures are wound down in a precautious way.

I'm sure some will disagree and as ever I could be utterly wrong, but I hope not as I've had enough and I think everyone else has too...

I've put in rate constant plots measured by 7-day change in cases, split for three demographic bands (roughly: school, working age, retired).

London:

• Rate constants peaked for all ages between Dec 14th and Dec 18th.  They've more or less been decreasing since and school and working age had moved over to decay.
  • There is a "false" low on Dec 25th clearly caused by fewer people doing tests on Christmas day.  I say "false" as cases are cases, but it's deceptive in terms of underlying infections.
• All bands then have a resurgence in rate constants, peaking on Dec 29th.  
  • This took working ages over in to growth for a day, but then they went back to decay.
  • This Dec 29th peak is also present beyond London; I take it to be infections downstream of Christmas Day - the timing is tight, but this is by specimen date, and fits with the ongoing speculation over Omicron's shorter generation time...
  • The most recent data point is for Jan 1st; this is 7-days after the Christmas sampling low, which means the rate constants are falsely over-estimated (in terms of relevance to infections) because the comparison is with falsely low case values.
  • I reckon that without this bias, all the datapoint for Jan 1st would be in decay.
• So, to me, it looks like trade constants have been heading consistently for decay since the mid-December peak, upset by festive sampling weirdness and a pretty mild (in exponential terms...) flurry of infections downstream of Christmas.

So - a pretty robust tendency to decay in London, indicating an inability to hold R>1 most of the time.  By the time many workplaces and schools return there'll have been almost another 3 weeks of infection granted immunity realised - at high actual rates.

• So. it seems reasonably likely London is not going to see a return to rising cases over the next few weeks.  If it doesn't, hopefully it won't be for long...

England excluding London  - this is the same plot, but for all of England excluding London.

• The growth rate in the retirement age band was plateauing, give or take the Christmas sampling low.  It then has the Dec 29th peak (downstream infection of Christmas Day?), and that peak in rate constant is reducing - except for the "false" high on Jan 1st, another sampling artefact.  The rate constant is decreasing much more slowly than for London - a more spread out peak of infections and more moderate doubling and halving times?
  • The absolute numbers of cases are rising pretty rapidly as the rate constant remains positive - I put in an actuals plot for these age bands.
  • Best case there's 3-5 more days of data to come before the rate constant turns negative and the actuals go in to decay...
• The growth rate in working ages peaked around December 15th and the rate constant has been slowly heading for decay ever since give or take theDec 29th peak and sampling artefacts. 
  • Best case this might be in decay in a few days
• Growth came much later to school ages and is now turning to decay.  More thoughts on this below...

Hospitals Plots

• Admissions in London have really levelled off - maybe even nudging in to decay.
• ITU is still not filling up either in London or England (no space for that lot)
  • It's a bit weird how admissions have shot up and ITU occupancy hasn't really changed; from a noddy maths model based on conversion ratios and lags that makes no sense; I get the impression it might be in part a long tail of Delta cases?

General Thoughts:

• London is probably more rebound-proof than the rest of England for now...  
• School ages cases are probably going to explode with the return of schools; I suspect over the break they've been following parental age cases through household infection.  When they explode, that's likely going to drive a wave of infections in parental ages.
  • Why explode?  They've not seen so many cases - especially beyond London, and most children are not getting the 3rd dose - which grants what I suspect will be pretty short lived protection against infection with (and so transmission of) omicron.
• University return => giant outbreaks, miserable self isolation in cramped quarters with poor institutional planning over meal supply?  Perhaps enough of them got it over the break to moderate things.

Pandemic over. You heard it here first folks.

Cheers wintertree your plots have been illuminating in these dark times.

May I politely suggest you don’t fire up the laptop this Saturday and take a well earned break.

It's not over till the data says it's over... 

Lots of risk of a rebound with the return of schools and workplaces, and the immunity-against-infection boost from the 3rd dose is gone to fade pretty quickly I think (but not protect from serious illness), so worth watching it for a few more weeks.  Aim is to put the laptop down after Saturday Jan 22nd when we've hopefully seen how all that pans out, and seen that it does so controllably...

Until this year's Christmas variant?

Me too. Or it could burst them all when everyone who's put off seeking treatment decides it's time to get that bulging, throbbing fourth testicle looked at...

Happy to oblige. On the 'London leads the way' train of thought, have a look at this:
https://nitter.42l.fr/PaulMainwood/status/1479082096694018050#m
and note the way the three regions plotted followed each other so faithfully in the earlier waves.
Anything could still happen in not-london.
But ITU numbers go a long way to reassure. I'm keen to agree that this (Omicron at least. In the UK at least.) will be carved onto the bottom of the list of viruses that cause the common cold before much longer. All the crystal balls say peak has to come and go in the next 10 days.

Or if that's even a choice.
For those who like to keep score:
https://nitter.poast.org/ArielKarlinsky/status/1422478206389661703#m
(In case the legend is tl;dr, I recommend looking at the numbers in black)

This is worth highlighting for a slightly different reason. We've all seen headlines about numbers of children going into hospitals being relatively higher, but might not have immediately reconciled that with unvaccinated people making up more of the hospitalisations than ever, and children are unvaccinated people. Read it as: There are fewer adults going to hospital relative to children, because they're vaccinated.

Aside: I was looking back over the Warwick and LSHTM models earlier today. I read them differently this time. You can tilt your head slightly and see the message that restrictions would have done close to nothing for omicron in the absence of time-travel. When you interpret the "we don't need any more restrictions" as "we'd be very close to just as f**'d anyway so there isn't much point", it makes it easier to assent to the decision to stay open. It's not for me to judge which phraseology was used behind the wood paneling of no.10.

Well, most of us are going to have broad and enduring immune protection against serious illness by then, so here’s hoping it’s more like a bad flu season at worst…  

How great to have differing opinions on if we can really almost say “endemic phase achieved” instead of differing opinions over how bad it’s going to get.

Something I’ve not plotted but that is looking decidedly hairy is hospital occupancy in some of the regions.  I’m certainly going to take extra care not to plant it in a tree on my way to the tandoori takeaway tonight….

If the data had been developing even marginally worse I don’t think there would have been time to curtail spread in time to avoid serious repercussions; not a very sane way of doing things…

" - the timing is tight, but this is by specimen date, and fits with the ongoing speculation over Omicron's shorter generation time..."

I spoke to you a bit behind the scenes - partner and friends in bar Jan 1 evening,  all positive and symptomatic around the same time on the evening of Jan 4 , so fits.  If anyone can suggest when I can breath easy again in terms of (not) testing positive myself, please do ... 

Well, yes, but also maybe no. You never know..... While it's unlikely, we've not *really* seen a proper immune escape variation yet. And.... did someone say H5nCoV2??

Sweepstake on lifting of plan B:
I'll have Jan 31st.

Yep, that Mainwood plot I linked doesn't have a very appealing future.

I'm not that sure there really ever was. It's been the most serious contender for spread-despite-lockdown.

Drive safe.

Do you think we know why? I'm left wondering whether it's driven by something different about Omicron, so might be expected to apply universally, or whether it's a "wall of immunity" feature that means places like us (and SA) are likely to fare better than our continental cousins with their generally greater proportion of immunonaive... I guess maybe cases will be more universal but the consequences of them less so, but I'm guessing wildly here.

That’s a very good question.  

Several papers pointing towards a different cell entry mechanism potentially moving the centre of infection up the respiratory tract but I’ve not seen clinical confirmation from actual people (yet?).  Or it could be the wall of immunity - one that was going untested until omicron came along.

The answer to the question is more important for some other places than for us and for SA it seems…. 

My guess is we’ll know a lot more in a month or so…

Yeah, you’re probably right, time will tell.
 

I guess it feels like maybe the flattening of ICU cases is at the better end of what reduced severity might have expected, but I feel like we spent ages saying “well, it’s got to burn through everyone young in London soon, and then once it gets into older ages the volume of case numbers could easily outweigh that…” - and then before you know it younger ages are in decay (did it burn through, or was immunity just holding up and limiting re infection), and then suddenly old ages tipped straight over and followed them quick as you like, like someone let the air out.

I can’t help worry it’s all a false lull, because it feels like we’ve caught too lucky a break (the nhs is in a mess still, obviously, before Offwidth jumps in to remind us!) in a way that we haven’t done at any other point in the pandemic…..

There's another theory doing the rounds that the propensity to infect people delta wouldn't is giving access to the low hanging fruit of household transmission, which drives both the case count up and the apparent generation time down. Can I bet 50p on 'all of the above'?

I think we might have some entanglement of thought processes going on here, very much agree to all that.  

The lack of a really big post-Christmas spike in the rate constants (c.f. that Euros match) is reassuring to me that this is more likely to be a case of "sufficient immunity" rather than many stars aligning to reduce R for a bit.

In reply to Longsufferingropeholder:

I can definitely agree that Something Has Changed.  

In reply to both:

France is perhaps the best comparator to look towards on OWiD?  Similar timings of omicron rising to prominence, similar lack of apparent interest in limiting the rise of cases, likely to have much larger gaps in immunity.

Short term, yes. Also Austria, Netherlands. Medium term Australia, HK, China. More as contrast than comparator.

I think Jan 26th is the next formal review, so I'll take that. I can't see Boris taking another back bench beating if he thinks it's unnecessary.

WT: I think your interpretation of the 1st Jan data is a bit off as we know there is a big false low on that date as well as on Christmas day. The same happened last year and made all the data through to about 5th Jan biassed either up or down (by about the 6th it had all worked through.) I don't think you can take anything from the 01/01 data or from the week-on-week comparisons throughout this period without a big slice of judgement. However... I still agree with your broad conclusion, especially when the slowing cases are viewed together with the low ICU and deaths data.

(Edited a bit)

Also tend to scepticism on this. School kids had a huge round of Delta infection more recently than any other groups and I've seen at least one study reporting that the antibody response produced by kids in response to Delta was strong even from very mild and quickly removed infections. I think it's highly likely a very significant majority of school age kids have already had the virus - SAGE were estimating over 50% I seem to remember even back in September. So, although most are unvaccinated, I suspect they still have a large number of pre existing antibodies from infection - much more so than any adult groups. Taking this together with the apparent reduced severity of Omicron and the moderate nature of the illness in kids before, I think the vast majority of infections in kids over the next few weeks will be either asymptomatic or very mild colds. As such, my bet is that primary age reported cases will go up a bit but stay relatively low and secondary age reported cases will increase more but mostly driven by the level of asymptomatic LFTs being done. We'll see. I'm certainly expecting a much less plague ridden half term than the last one in my house. Fingers crossed.

You could be right, I'm leaning towards a bigger low on Christmas Day than New Year's Day, but the testing numbers are only just appearing in the provisional zone, so I am jumping ahead of the data.

The slice of judgement employed (or rather the 2 thin slivers):

1. Looking back at raw cases data, the Christmas low was significantly larger (fractionally speaking) than the New Year's low last winter.
2. The synchronised jump up of all rate constants looks very much like an artefact to me; we'll see if they all jump right down again with tomorrow's update...

> [...] the preliminary data for 3rd and 4th Jan is certainly not skyrocketing in the way it was a fortnight ago.

Yes, the provisional zone is quite comforting.  

I was mulling over the latest ONS infection survey update and its implications - around 7% with Covid for the last week of data; perhaps order of 12% by now?  (This being the live infection count).  Pretty sure with a number that low (in absolute terms) even including what came in the weeks before, the decay of cases is leaning heavily on the very high antibody levels resulting from recent 3rd doses; I think there could be a very long tail to the decline of omicron cases as those antibody levels fade over the next few months.

The thing is, the vaccine and delta induced antibodies that happen to have neutralising ability are a poor fit to the omicron spike it seems, so successful prevention of infection needs mega-high antibody levels.   The degradation pathways for antibodies are going to be mostly stochastic, so the decay of antibody levels is going to be - to a crude approximation - exponential with time.  So, how long will the super-high levels needed to prevent omicron infection last?  Back to my long-tail comments above...  Although I think once primed some T-cell stuff can activate antibody production PDQ on exposure to the virus, and kids immunity is something else (I've started to think I should grow a clone me and joint the tech bros cult of transfusing young blood for health and longevity...)

In terms of illness, absolutely less plague ridden, totally agree.  As you say, that's going to cross over in to how much is (or isn't) symptomatically detected, and the impact of asymptomatic LFTs.  

I should have said "explosion in infections" not in cases.  Will be interesting to see what policy changes go on there at the next review date...

Thanks for the reply. I hope the curry was good!

My key point is the last month has been mostly about the NHS being left in a mess  by govenment inaction and mistakes and that only part of its pressures are due to covid. With the NHS in some imagined better state in 2021, after what any genuinely responsible government should have done tactically  (let alone strategically, in terms of emergency improvements in staffing and funding for the NHS and Care and specific areas like local public health) after the first year of the pandemic ( in the background of brexit), restrictions might not have been needed for omicron. Jeremy Hunt was speculating on his committee report on the risk of an NHS semi-permanently locked into massive waiting lists. Great news if you are a libertarian who wants to see the UK importing more of the expensive ( in funding and public health terms) but highly profitable US approach to healthcare, as better off people will pay to avoid queues and will buy insurance policies to even out risk.

Such a good curry;.

The same family have been cooking my curries for 20 years now, but sadly their restaurant business is a casualty of the pandemic; they've converted to take-away only which looks a lot more profitable but that's one of the post-pandemic endpoints I was really looking forwards to gone...

Nothing like a good curry however to make me feel my age and relative sloth - time was I'd demolish that much food and be just about sated.  Now half of it's in the fridge and I'm not ready for it yet...

Its already happening, and that option on going private has been in place for years and years via BUPA ( UK not American) anyway and is just being amplified.

Of course the power/amazing skills etc  that consultants have as one of the drivers to the private UK gravy train is often pushed to one side to avoid upsetting the applecart.

Anyway its time you started an NHS thread as wintertree suggested.

Jermey Hunt is far more interesting to listen to these days.

Your key point belongs in the politics forum. In a thread entitled "preaching to choir"

It would be nice if the choir sang instead of distracting themselves..... more people might stop and listen. The NHS is in the worst condition it has ever been in in its history and despite being regarded as a national Institution, most here (and even more so in the population in general) may as well be hypnotised. The populists are winning hands down because way too many intelligent people are distracted, when they should be campaigning.  Things look even worse in the US where (building on the majority of white college educated men who voted for him) Trump's lies still haven't been buried. Populism is also a threat across much of the EU. I fear for the future. I never expected the majority of the educated in a modern democracy to behave like nero.

As far as I know, nobody here has tried to distance themselves from issues over healthcare.  But it’s also not the primary thing people have been coming to these threads to discuss.   I think this will include quite a few posters who either work in healthcare and/or who have close family who do so.

Yes, some covid issues are inextricably mixed with healthcare.  Some parts of covid are not, and some parts of healthcare or not related to the covid data (eg how it came to be as it was by February 2019). 

https://www.ukclimbing.com/forums/post.php?

Covid data>NHS>USA voting pattens>Trump>Populist threat to EU.

I think a little focus might be needed.

Lunch time procrastination, I tried it out.  It needed a black background for the colour mapping to work with any reasonable colour map.  It just colours by the average stringency index over the same period that's plotted (3 weeks) - perhaps it should be for a 3-week period a bit earlier on in the timeline.  It looks like a dogs dinner - in part because everything is so bunched up on a "main sequence" right now.  Squinting through the noise, some quite takes on I:

• The big surprise for me was Iceland; they're really going for it with (presumably) omicron having had very little spread of delta before.  Likewise Australia; if anything they're the two places I'd expect to be a bit more cautious until they know what they're letting themselves in for.
• Vietnam, Austria and Germany are in a rising phase for the rate constant (exponential situation getting worse)  despite their much tighter control measures
• Greece - something is wrong with their stringency data in either the download or my pipeline..

The other big surprise was Germany's position on the rolling CFR analysis plot; assuming their survival rate is comparable to other Western European neighbours it just looks like they're loosing interest in testing - which is reflected in their positivity data vs ours, especially pre-omicron.  

The PCR testing has dropped drastically since end Nov (KW 49/50) as most of the activities which previously required a PCR test are now closed or restricted to only vaccinated/cured.

More positively the ICU occupancy has halved through December, plenty of hospitals are recording no Covid ICU patients.

Well known newspaper reporting today that incidental admissions are now up to 40% of those in hospital. Seems to be further increasing - I assume as Delta wears out completely and omicron incidence continues to go up.

And yet France is predicting peak over the next few weeks. we think Germany can be special at times, but there is something not right when there is havoc all around. 

Timing.  I suspect it's all in the timing - the strong measures in Germany have brought delta infections crashing down, and follow on measures including ITU occupancy are now crashing down too.  Those measures have also delayed omicron, but it looks like it's starting to break through.  

Which is great news for Germany because everyone else is being a flock of canaries in the coal mine for their policy towards omicron...  This isn't the kind of race anyone wins by being first...

In reply to Si dH:

Interesting to think about incidental admissions - I think it can bias interpretation in different directions depending on what prevalence is doing.  It suggests the hospital > ITU conversion ratio isn't as rosy as it looks, but thankfully there's a much bigger margin developing there...  

Indeed it looks like luck in the timing, Germany was crashing because of the Autumn Delta wave and clamped down in October then increasingly in November before Omicron was really on the table, they could also look how it was hitting the UK so could anticipate to a certain extent. The death rate in December was still high as a fallout from the Delta wave but is falling off as that peak is passed. It's also fortunate that our vaccine mix is better placed to reduce the effects of Omicron as the boosters and childrens vaccinations start to reach effective numbers.

Todays lockdown measures should help a bit as they are a bit more stringent but the quarantine shortened for many to help combat staff shortages.

Certainly you're looking better placed for a low risk exit than we are.

To beat an increasingly broken drum...

This is efficacy against transmission; with Omicron and any currently approved vaccine that efficacy relies on very high antibody levels, which will fade over time.  If omicron isn't allowed to spread significantly within that wanning time, then the scoop of the problem recharges.  I think the effective wanning of immunity-against-infeection from all approved vaccines will be faster for omicron than for delta because of this higher antibody level (needed to compensate for the worse antibody binding affinity).

So this different mix doesn't reduce the effects of omicron, it pushes them a few months down the line. 

All the vaccines offer very similar health protection from serious illness.  This also wanes over time, but much more slowly.

At some point the cork has to come out of the bottle, and the effects that matter with Omicron are far more about the people with no vaccine in them, than about people with multiple doses of any approved vaccines.  The scale of that problem is what the FT plot Longsufferingropeholder posted is showing - for a couple of places, anyhow.  .

So, the question is, will Germany allow omicron to spread this winter and fill in the significant total blanks in their immunity, or are they going to keep it bottled up until spring/summer, when the spread will be easier to bound, and when the vulnerable can be better protected with winter seasonality removed?

Big Questions!  Not a bad situation to be in compared to the places that are failing to keep a lid on omicron and that have significant gaps in immunity where it matters.

Shortened quarantine is all the rage recently!  Does this apply to health care workers as well?

The million dollar question! As it seems the objective is to push as far down the line as possible (there are some here that disagree with this but they are most certainly wrong). Economically Germany can cope, Omicron specific vaccine will hopefully be available soon, effective medicines are becoming available, the pressure on the non-vaccinated continues to increase (and the waverers becoming converts), we have a government which doesn't have to worry about elections for a few years and a health minister who understands all this stuff (and isn't worried about his political career).

Distracting ourselves with covid plotting? Yeah, you're right, that's bang out of order on this thread.

We know about the state of the NHS.
We will not fix it by ranting about it on the internet to people who already agree.
All that does is fill up this thread with the same rant you've already ranted a hundred times.

Pretty much every conceivable approach is being taken to this pandemic across the world; Germany is certainly taking up a very different position to the UK in that.

Of two things I've no doubt:

1. Learned scholars will argue about interpretation of the end score sheets for a couple of decades
2. Some governments are going to learn better from the results than others...

JBM touched on this. https://nitter.1d4.us/jburnmurdoch/status/1478348889514258433#m

The stringency tails didn't look as dramatic as I was hoping. I guess it just doesn't change colour as dramatically as I thought it would. Thanks for giving it a go.

Edit to add this: https://nitter.1d4.us/VictimOfMaths/status/1479396101803638786#m

I was speaking more generally than UKC and especially more than on this particular series of threads. Everyone supposedly knows about the major NHS problems but where are the major campaigns for something to change, given how vital it is to the UK?. Plenty of regulars on the covid threads post useful stuff on the NHS here and elsewhere; Wintertree in particular posts exceptionally well on the subject.

In respect of this thread: data issues outside standard covid stats are very much part of the current covid response, especially NHS staff absences, NHS PPE and testing supply and NHS discharge failures due to over stretched care homes.

Like Roy I think the NHS subject is absolutely rantworthy as it's maddening so little is being done: it's like 'the emperor with no clothes' where people get ill and often die. The political situation impacting covid response is worse still in the US, thanks to Trump; and still a key issue in parts of the EU.. Political failures too often seem likely to dominate the west's exit from covid. I'd add antivax shit is still being spread on social media despite promises of better action on this from the companies and governments.

First data update from Indie SAGE for three weeks is very thorough (first 30 minutes of the link).

youtube.com/watch?v=phK_8X4l26Y&

Also lots in the latest ONS insights today:

https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/con...

I find myself rather more sympathetic to Offwidth's pleas for attention to the state of the NHS. Of course it's a situation we all know about in general, and agree about, and furthermore we all agree that it precedes and goes beyond the present pandemic.

But there is extraordinary reason to pay special attention during this wave, as this wave is the one with potential for much more wide-ranging and systemic damage to society, given that it seems Gov has adopted the Let It Rip strategy (except where it somewhat inconsistently & puzzlingly hasn't, e.g. masks & the odd filtration device in 1/3 of schools, etc.) It seems to me that the potential healthcare damage of this wave lies in covid morbidity and non-covid mortality rather than covid mortality, and the contributory factors include non-ICU capacity, staff illness, supply chain breakdowns, ambulance service shortage, social care staff shortage... Non-healthcare damage is also increased by widespread illness among public service workers in transport, food... perhaps even in parliament as there's quite an outbreak there at the moment (though thoughts on whether or not the latter poses a threat to good social functioning probably does belong on another thread).

I'm willing to bet that not one of us who reads and contributes to these threads does so because of a purely disinterested and scholarly interest in the abstract pleasures of mathematical modelling. We're here because we have skin in the game: we want to understand what direction covid is taking, and what practical effects it's likely to have on our own lives. When WT starts singing his Ode To Ventilation for the nth time, I'm sure nobody will complain that we've heard it all before: it's an as-yet unresolved issue that makes a material difference to how the pandemic progresses and, on these threads, we're surely interested in all factors that make a difference, no?

You just posted a link to one of Colin Angus (at Sheffield Uni)'s plots. Worth noting that he's recently started trying to plot NHS capacity against staff illness. In the absence of a dashboard documenting local incapacity to deliver healthcare, his new plots seems to me a useful index of where & when covid is doing most damage to the chances of anything resembling 'normal life':
https://twitter.com/VictimOfMaths/status/1479484966232862726

It's got attention, and it's had a lot of comments on these threads, but if Offwidth wants the attention it clearly deserves, I think starting a thread on the subject will work much better.  More clarity of focus, more readers, more comments.

As you say it's an important issue - a critically important one for the next few weeks - and there's an awful lot of reasons to be concerned.

• Pucker factor is backing off a bit as signs of a post-new year rebound aren't queuing up at the door yet, and as hospital admissions look to be heading for clear decay everywhere.

The point I keep raising over Covid is that if we have to have control measures because the government has failed to prepare healthcare "to live with the virus" - or at this point even for a bad pre-Covid flu season - we all need to be abundantly clear that this is because of a failure to properly fund and manage healthcare as much as because of Covid.  Accepting significant control measures to allow a government to skimp on maintenance of a fundamental pillar of society is very different to accepting signifiant control measures because of a brand new global pandemic we know nothing about.

There's a lot of people who've made incidental comments on these threads who would I think have more to say on a dedicated thread.  Far more people than are bothered by my pet peeve over ventilation.  Although having said that, teachers up here in the northlands are waking up to the idea that opening windows is a bad solution in proper cold weather.

https://planetradio.co.uk/metro/local/news/north-east-teachers-call-for-bet...

As Offwidth said: Like Roy I think the NHS subject is absolutely rantworthy as it's maddening so little is being done.  Give the rant a bit of form, and give it a proper airing as a thread.  I think there's a wide ranging discussion about why this is all happening and particularly if Hanlon's Razor applies or if there's other forces at work towards long-term profit extraction by wanting-to-become vested interests.

I think at that point it's teetering towards the new-normal and we have one more co-morbidity out there; this has been the endpoint we've been aiming for since we found the vaccines worked.  Seems kind of obvious to me that "learning to live with the virus" is going to involve additional work for healthcare - quite a lot for a few years as the bell rings down from a pandemic clang to an endemic hum and then a bit more due to an additional disease in circulation.  That's the real scandal IMO - that we've had the CRG ****wombles wittering on about "learning to live with the virus" whilst objecting to anything that could ease the transition to that and not having any initiative over healthcare.

But I'm not saying anything here I've not said a dozen times before.  

A dedicated thread of its own would get a lot more input and a lot more views.

As far as it goes on these threads, we're stuck in a loop of the same comments being posted with no real progress.

Well put simply, not here.

Chanting "the NHS is banjaxed because people voted for things I wouldn't have" does not add value to a covid-specific thread.

We all know it is. I don't think Sajid is reading this thread though. And I'm not convinced that you getting all speakers' corner in here would make him suddenly see the light and change his ways.

The politics forum is over there ->

Good data, thanks. Surprising amount of regional variation in VoM's graphs.

How we laughed: https://www.theguardian.com/world/2022/jan/07/the-rise-of-lateral-flow-test...

When this government buys something and it is surprising that government is buying it and how much government is paying for it, then it will be one of the Tories' pals selling it.  They are very consistent about that. 

Arch-brexiteer Dyson makes air purifiers.

https://www.tes.com/magazine/news/general/heads-shock-over-expensive-dfe-ai...

It's the thing though. We all know your threads have been a centre of gravity - but you can't divorce the wider picture from the stats. You don't restrict yourself to the stats either.

Who on earth would read them bar nerds if it wasn't for the fertile multiplicity of supportive views coming from all angles?

To tell people to "go drink at another bar"* is counter-productive in my opinion.

I value Offwidth's perspective as I do your own hard work - his is a vital perspective that informs and contextualises yours.

These threads are too valuable to allow a silly pseudo-animosity to derail them.

I would rather have all knowledgeable input in the same conversation. So, chill!

* paraphrase

Confounder warning:
Not sure how long since we last mentioned it but it had faded from my thoughts that the figures reported in the primary diagnosis supplement are acute only. Was just reminded by the twitterers' reminders.
Quite likely the proportion of 'with covid' is higher in the others but difficult to quantify that, or the variation on that.

I have (or so I thought) very carefully not been saying that, largely for the reasons you give.

I’m obviously not making myself clear.

I’m not saying Offwidth would stop posting on the subject here.

I am saying it’s time for a proper thread of its own, and that that’ll get a wider audience I hope.  

Offwidth said they want the choir to sing instead of distancing themselves.  Nobody here is distancing themselves from it IMO, it’s just not the front and centre topic of conversation but one of many pieces of context and there’s only so many times everyone else can agree.  To claim the other contributors are distancing themselves; I know a fair few are in it up to their necks either working in healthcare or supporting partners/family who do.

Offwidth clearly wants this to get more focus; that focus isn’t emerging on these threads.  I’m not saying “drink at another bar” so much as “go to town”.

MG has started an NHS thread so I guess we will see how it goes...

The data on the last half year of my NHS related OP's was hardly encouraging

https://www.ukclimbing.com/forums/politics/an_end_to_another_nhs_management... (no replies)

https://www.ukclimbing.com/forums/politics/poor_quality_housing_impacting_o... (29 replies none on the NHS angle)

https://www.ukclimbing.com/forums/off_belay/dr_philip_hammonds_new_covid_bo... (6 replies)

https://www.ukclimbing.com/forums/off_belay/another_poor_judgement_by_boris... (a proper extensive debate but as a part 2 of someone elses thread).

https://www.ukclimbing.com/forums/politics/more_or_less_have_fun_with_hanco... (1 reply)

https://www.ukclimbing.com/forums/politics/prof_gupta_on_bbc_news_again-734... (a good sized thread but with a few defending the scientific monster)

https://www.ukclimbing.com/forums/politics/test_and_trace_a_licence_to_tax_... (3 replies)

Another segue I see as important:  the horrendous harassment of those providing an important public health message on covid (this time an international example).

https://www.theguardian.com/world/commentisfree/2022/jan/08/as-a-scientist-...

I doubt Javid reads that, either...

Preparing tonight's update, there's either some very good news about case numbers in over 65s, or there's a reporting glitch.  The change in the data looks rather artifactual.

I don't suppose any of the "Twitternauts" have comes across any mention of a reporting glitch or other interruption to case numbers for over 65s?

Nothing where I'm looking