UKH

Relatively limited interpretation this week as my view is more-or-less unchanged from last week, and as I've spent almost all of today under what feel like Californian skies, and I want to get back out under them sooner rather than later...

Last week I'd noted that the exponential rate constant for cases had dropped across all the English regions and for Scotland.  As we'll see in some of the next posts, Scotland has unambiguously turned to decay.  This is very interesting.

In England, the exponential rate constant of PCR cases is no longer falling and has stagnated at around aa 16 day doubling time.   This feels very much like the "weather effect" to me; with the recent periods of lower rate constants matching periods of higher temperature, and vice-versa, just as was seen earlier in the year before the period where the rate constant rose significantly with dropping of restrictions.   Looks like another glorious week ahead so hopefully that will moderate the impact of dropping many of the remaining restrictions in England in one go...  This makes me slightly less concerned than I otherwise would be.  I'd far rather have seen a progressive dropping of restrictions to give time for re-evaluation of the data over healthcare and outcomes rather than playing Russian roulette, again.  However, we are where we are.

The vaccine plot is very interesting.  How one views the dropping of remaining restrictions in favour of a wave of mass infection will reasonably depend on why vaccination has nearly stalled - lack of demand from people, or lack of supply meaning inability to meet the demand?  Lots of back-and-forth in the last few threads on this, some misconceptions cleared up, no consensus reached.  Why is the plot so interesting?  Because a clear step-like profile is developing in first doses, where each Sunday has a drop from the preceding days - as is common - but then the level for the week ahead remains fixed, rather than rising up, as used to be common.   Something has changed, and in a way that looks very systemic.   Easy to over-interpret but I do wish we had real clarity on this point which I view as critical.  My notional estimate of the delay between doses continues to drop, showing that the rate of second doses is picking up in terms of the number coming due.  This I think is good.

Link to previous thread: https://www.ukhillwalking.com/forums/off_belay/friday_night_covid_plotting_34-...

A look at the demographics of English cases.

Plot D1.c - Demographic Rates (England)

• The rate constant for the age-group 20-25 is close to decay (almost white on the far right of the orange/blue heat map), this looks to me like a collapse from massive spikes at least partly associated with some higher education institutions, and whilst cases are decaying, they’re still above the cases/day/age bin in other age. It’ll be very interesting to see what happens over another wee
  • Do they stop decaying and peg the levels in nearby bins?  SNAFU.
  • Do they keep decaying below the level of their neighbours?  This might suggest the outbreaks have produced enough immunity to reduce the susceptible pool enough to start limiting transmission.  

Plot P1.e - Relative Demographic Distribution of Cases (England)

• This shows the relative distribution of case numbers, with some smoothing.  20-25 spike (gold/white, bottom right) is decaying as the underlying spike in case rates their subsides.  A consequence of this is that a larger proportion of cases are at other ages; as most people are older not younger this means the average age of cases has become older.  

Plot C - Winter vs Summer

• This shows a few sections through Plot P1.e as bar charts (just the outline of the bars drawn for clarity).
• The big demographic shift form winter is clear (blue vs orange/red).
• As with the plots last week, we see the change over the most recent two weeks of data is towards a smaller proportion of cases in 20-25 (red line below orange lines) and raised cases in older people (ages 30-50, red line above orange line).
• So - the demographic shift is reversing slightly; the nuance to this is that it's not due to a failure of vaccines or anything like that, it's just that a giant distorting spike amongst young adults is collapsing back down.
  • What're the consequences of this?  Over the last few months, the ratios of cases:hospitalisations and cases:deaths has been dropping as the vaccines kicked in to ever younger adults and directly and indirectly moderated outcomes.  What the demographic plots suggest to me combined with the vaccine plot is that this has given all it can, and where we are now in these conversion ratios is where we'll stay, or they may even get a little bit worse.  
  • The vaccines never "broke the link" between infection and anything else, they dropped the conversion ratios.  Almost miraculously, but the time for miracles is passed.  Which is why I would expect a sensible, public health driven approach to continue a progressive and measured approach to further unlocking.

The Four nations plots.

Wales looks like decay but I think that may just be reporting lag, if I'd taken the last data point off the trendline would be rising.

Northern Ireland and England are about to exceed their smoothed, de-weekended peak case levels from the last wave.

Scotland is really looking like decay; there's an uptick on the trend line on the far right but that could be an artefact of the noise - that's the provisional zone.   More on Scotland in the next post.

The rate constants for all nations are dropping in the provisional zone; it would be nice if it's a true drop but we'll see what another week brings.  

I've wondered on the previous thread if Scotland was in to genuine decay for some reason, or if this was decay was the consequence of a big outbreak subsiding and wouldn't be sustained.  Outbreaks seem to be dominating more at higher case rates than in previous waves, presumably due to the inhomogeneous nature of vaccination.

The obvious candidate for "why is it dropping" would be reaching herd immunity thresholds for the current level of restrictions.   Feels a little too good to be true right now.

The good news is the turn to decay seems to be translating in to hospital admissions which are looking less steep on the leading edge of plot 7.1s; the rate constants for both admissions and occupancy are slackening off in plot 9s.   I say good news as they're less than 2 doubling times away from exceeding their last wave's peak occupancy which was about 3 weeks but is now looking like 4 weeks or so; if the decay in cases holds, that point recedes further in to the future; if that decay doesn't hold, by the time the authorities know it hasn't held, it'll be time to reign in the spread of infection almost straight away.

I'm taking the peak of the last wave as the crisis level; it wasn't quite so bad in Scotland as England last wave so notionally there is more capacity, but time and again from offline contacts, posts on here and news items, what comes across is the level of exhaustion within the health services, and ramming them to anywhere near the levels seen in the last wave doesn't seem like a good idea.  As well as the hazy question of how far the people in the system can be pushed (hint: if you're a policy maker and you're asking this question, should is a much more important word to people who've done all this for you in horrendous circumstances than can), the rise in self isolation requirements in really starting to hammer service delivery across the board from car markers to public transport to healthcare; it feels like something is going to have to give here very soon, and the pragmatic approach is self isolation requirements for the double vaccinated upon PCR results. 

Yazz's best remembered single applies quite well to the data for England.  The same is probably happening in the other nations (edit: except Scotland!), but with the smaller populations, the statistical noise is more, particularly on deaths, obscuring the trends.

↑ Cases - up
↑ Admissions - up
↑ Occupancy - up
↑ Deaths - up

The rate constant for cases is dropping - moving to a longer doubling time, but it looks to me like the last peak in the rise of cases is now working its way through to all the other measures on plot 9, with admissions, occupancy and death all doubling faster than cases; it looks like the slackening of cases is starting to flow through to admissions, presumably we'll see it flow through to the other measures by next week. 

Cases will soon pass through their previous peak.   Hospital occupancy is about 3 of its doubling times away from the last peak, although I doubt that level can be sustained now (see my comments on the Scotland post) and we'd have to be mad to intend to ram healthcare to anywhere near that level when there's no compelling time pressure (going off the latest modelling posted to last week's thread).  Those three doubling times are a bit over a month; if the wave starts to break that fades in to the future or disappears, if cases rocket following the next round of unlocking it gets closer, and again as with Scotland we're worryingly close to the point that by the time evidence of a rapid rise in infections translates to detected cases, analysed data, presentation to cabinet and policy U-turns, too much hospitalisation is locked in to prevent overload.  

So, it feels to me like we're going to go needlessly close to the wire, again.

On the up side it's by far the least recklessly irresponsible of the three times we'll have risked ramming healthcare - there's actually reasonable odds we won't do it this time.  The looming winter with all its uncertainties does in the eyes of some modellers and public health people present a looming if hard to bound danger - an argument I have a lot of time for and why I've not been aghast at the general idea of an "exit wave" this summer, but when you put a burning oil well out by feeding a lump explosives in to the conflagration to blow it out, you do so exceedingly cautiously and not with utter abandon.  (I learnt that from  MacGyver.)

I neglected to say thanks last week 😳 so I will get my thanks in early this week.

I think the persistently decreasing rate of first jabs looks to be be due to lack of demand. That seems the more plausible than supply issues coincidentally arising about when you'd expect saturation of uptake by 18-29 (ie a bit lower than for age 30-39). 

In Scotland the biggest vaccination centre(s) is/are closing and for first jabs it's more about drop in to mop up those who have missed appointments rather than scheduled appointments.

I think there's half a dozen venues in Glasgow but double that at weekends so it looks like convenience is key or perhaps even the limiting factor.

Until there is a prospect of international travel there isn't much incentive for to encourage the unvaccinated to get jabbed.

Plot A - UTLA cases/100k

• This has had a break for a few week - I think it's really interesting now
• Two of the early outbreak regions are coloured black, the biggest risers beyond those are red, and the rest are orange.  The legend gives the top 14 risers.
• The first areas (black curves) to fall to lowering control measures and the Indian variant had heavily targeted boots-on-the-ground responses including surge testing and more vaccination support including - perhaps critically - in to workplaces where people felt unable to take the time off to travel for a vaccination.  I was at the time heartened by the amount of local knowledge and local public health input in to all of this.
• This intervention seems to have had an enduring effect - both regions went in to mild decay and stayed there which his really interesting; Bolton is now rising a bit.
• As far as I can tell, there's very little targeted intervention in the new areas to fall to the fastest rises now, although I've not been scouring local news so don't take my word for it. 
• I'd be really interested in an antibody survey result from Blackburn with Darwen now...
  • Instinctively it doesn't seem to me like a herd immunity effect moderating cases, if it was, the decay rate should be getting faster and faster by now as it would be well in to the later half of the logistic function.  
  • An alternative interpretation is that there's this "auto-moderation" effect where after a period of rapid growth in local case rates - regardless of wave and so immune fraction - rapid exponential growth is never sustained for long, with presumably local behaviour effects kicking in in response to an obviously worsening local situation.
  • Will auto-moderation work for the "red" areas when it's counteracted by the next round of unlocking?

Plot 18 - Regional Characteristic Times

• The North East in particular looks to be headed in to a decay of cases; I suspect this is not unrelated to an exceptional university outbreak and that it'll revert to tracking other regions.  But who knows, perhaps we lead the way...
• Hospitalisations are rising everywhere, rate constants look to be maxing out with the right sort of lag from the last "cold weather peak" in the cases rate constants.
• A few regions now have enough deaths to return meaningful estimates of the rate constants there.  The North West in particular, where it looks like the deaths were mostly associated with the early stage of rises in cases.  Some discussions on this on factors around this on previous threads.

That does sound like demand limited.  Lots of radio advertising of walk ins here too, which also sounds like demand limited.

If it wasn't so stepped on a 7-day basis I'd have readily agreed without question.  As it is, I really want to see this question given an explicit and detailed answer.  The moral angle is strong and I don't want to jump to a tempting but wrong conclusion.

I've heard quite a few people suggest similar, including pundits on the radio.  In terms of the summer holidays, if someone hasn't had their first dose yet, they're not going to be "travel compliant" for 11 weeks - early October - given the need for two doses and time to bake in the second dose.  Even if that's cut down to the less effective 4-week gap it's still most of the way through the summer before someone getting it now is travel ready.  The one-shot Janssen vaccine would make a big difference here but the news is surprisingly quite on its roll out, it had been mooted for exactly this reason around the time of its approval in May.

Almost forgot the Lissajous plots, too busy watching the sunset and enjoying an IPA; been a long day.

You can see Scotland turning a corner on all measures; if that sustains in to clear decay on all measures, that becomes very interesting.

The recent wave (blue) finally seems to be getting some improvement over the last one (purple) on cases:ITU occupancy in the North West.  Which is important, because they're only ~2 doubling times away from the peak occupancy of the last wave; one report on another thread today suggests it's already impacting service delivery in A&E significantly.  Could be an ugly night tonight on the ground following such an incredibly hot day, assuming "heatstroke and drink yourself in to an ambulance by 5 pm" parties also happen on the other side of the pennies.

I think we're fast approaching a significant milestone that won't be reported; the day will soon come when more people get immunity nasally rather than intramuscularly. Could be this week's puzzle to work out when that might be.

I kind of assumed that point was in the past.  Ever since the Kent variant, the ONS seem more cagey about quantifying the number of new infections per week from their surveys, so I've been leaving out the infections vs cases plots I did way back when; cases seems to work very well as a measure for understanding the current situation.  It wouldn't take much under-detection of infection as cases for those exposed to and infected by the virus daily to exceed those exposed to the vaccine.

But, vaccine exposure is in a controlled location and a controlled dose with controlled genetics.  Exposure to the virus is uncontrolled in location, dose and genetics.  In terms of strength of antigenic response, I don't have a clue how current rates compare.  I don't even know how you would compare them given the different strains embodied in current infections and the vaccine, and given the broader range of possible epitopes exposed by the virus vs the vaccine.  A many-dimensioned comparison, and one for which the evaluation of past events (protection confirmed by immunity from virus vs vaccine) can change in the future based on future events (new variants emerge).  

For all we know, the next variant might have a more "classic" spike but some other mechanism for improved spread, meaning the "first generation" vaccines currently in use work better than a recent live infection, or it might be that the broader response to a natural infection prevents immune escape from an India-derived variant that escapes first generation vaccines a lot more by having further changes to the RBD.

I get the impression there's no level at which this question can be asked where someone doesn't end up licking their finger and sticking it in the wind.

The wind, of course, can change.

Another excellent article from Prof (+Dr) John Wright in Bradford 

https://www.bbc.co.uk/news/stories-57866661

Two quotes from a colleague of his:

About half of the patients on the ward today haven't been vaccinated - I have stopped asking them why, as they are clearly embarrassed

"Some have had two vaccines and so have had milder disease - they are alive on Cpap (non-invasive ventilation with oxygen) when without the vaccine they would probably be dead," my respiratory colleague Dr Abid Aziz told me after an exhausting six-hour ward round.

I can see why Westminster govt might deny a shortage but not why devolved governments would go along with that. Hence I take the statements that there is no shortage at face value. However that means those steps make no sense to me.

Very confusing.

Very good point.

68,000 daily 1st jabs Vs 54,000 daily cases (100k realistically?) so we're already there if you use 1st jabs rather than 2nd jabs (180,000 daily).

Not sure if the comparison should be 1st or 2nd jabs.

I just look at the cases plot going off into the stratosphere again and think WTF are the government doing relaxing things until they know exactly what's going on - virtually everything else seems secondary to that imbecility - I despair 

I still think there are limitations in distribution and getting the jab to people's arms. More could be done. Obviously there is a point at which you say "we've done enough" - unless you think the Govt should send people to knock on every front door from a vaccine bus. Personally I think this is now limiting the rates that could be achieved even with fewer remaining people who want jabbing, so they should try a bit harder.

Certainly round here, there is now lots of walk-in capacity if you know where/when to go and have transport/time, but it seems there might be far fewer bookable appointments. The problem is that using a walk-in requires both following local NHS or council social media or checking up on their website every few days, then the ability to take the time to drive there when it comes up. I'm sure this works well for some people, especially groups who live in city centres near the walk-in centres, but equally sure it won't for others.

(Anecdotally - I managed to get my second jab yesterday, but only because I follow the council's social media accounts and drove 30 minutes across the city to get it. I could seemingly have got one at a local walk-in place in a few more days... but that place no longer takes any bookings, it has been closed for at least a month (a couple of months ago it was very busy) and is about to reopen as walk-in place on Wednesdays and Fridays only...)

Edit to add a bit of tentative data evidence to support this - the shapes of the vaccine data curves for Scotland and England are quite different. They both follow the same broad trends but the weekly cycle that we see in England simply doesn't exist in Scotland. Ergo, that weekly cycle must be shaped by the means of distribution and availability of appointments or walk-in centres, ergo, that is still to some extent a rate limiting part of the process.

If you look at data from ZOE, split by unvaccinated and vaccinated, it suggests that the wave in the unvaccinated population has now peaked in the UK. Meanwhile cases in the vaccinated group are continuing to rise.
The two curves have been cancelling each other out for a while, producing a slow down - or even a decrease in some regions.

But ofc the bottom will soon be reached among the unvaccinated whilst the ceiling among the vaccinated is uncertain. Everything else being equal we could therefore see a further pick up as we go into August (as per most SPI-M-O scenarios and others)

Boris and Rishi will not be self-isolating after being pinged - Barnard Castle moment for the app? 

Thanks as always.

So in England 3 more doublings would result in a further lockdown due to unsustainable pressure on the NHS. 

3 more doublings in cases or hospitalisations  would be 400k cases a day, 32k in hospital so that makes sense. 

Probably unanswerable but given we usually use cases as the ‘canary in the coal mine’ with hospitalisations and deaths a few weeks later following on are we going to lose this early warning signal as testing services cannot meet demand/people don’t bother testing to ensure they go on holiday/ Boris doesn’t isolate so why should I etc. ? If this happens, and I’m sure we can’t do 400k tests in a day then cases become unreliable as a marker of exponential growth, they may have already…?

Secondly re Hospitalisations, which are almost always reliable statistics.

Any way we can establish how many people vulnerable to hospitalisation are still out there? 
With cases high and rising I guess everyone will either get 2 jabs or Covid ( or like our health secretary, both)….

Presumably the government has calculated (wishful thinking) that there are not enough people left who are vulnerable to hospitalisation to fulfill the 3 remaining doubles that would take us to 32k  hospitalisations (and a further lockdown)….?

That calculation feels too complex for an easy guesstimate so I suspect it’s a government guess.

Do you think that guess is Russian Roulette with a revolver or an automatic Wintertree..?

The policy is to reopen as soon as possible without overwhelming the NHS. That means, pretty much by design, we come as close as possible to overwhelming the NHS. That means, pretty much by design, that this is going to be a brown alert again.

Can you point us to the data? That would be interesting. I've seen the press release but when I looked at their most recent data reports to government (I'm a contributor, so the full version) I couldn't see any digestible summary of cases data as a function of vaccine status.

It would be great to see cases data reported as a function of vaccine status on the dashboard too. They have the info to do it...

They removed it from the daily report just yesterday I think, they say it's because the sample size of unvaccinated is too low for them now and the CI are too big.

Couldn't find an archive section on their website but found a screenshot from said report on the 6th on Tim Spector's twitter account from the 6th:

https://pbs.twimg.com/media/E58sq5FXoA8DTyC?format=jpg&name=medium

and another one of the 14th:

https://pbs.twimg.com/media/E6Qmo5HXoAMFwoR?format=jpg&name=small

Note : I just noticed that some idiots on Twitter like Toby Young have been using this chart to argue the vaccine doesn't work. Obviously it's total bullshit.

That has to be the quickest u-turn yet .. 

Gove took part in the same 'pilot scheme' when he was a contact of a positive test on his trip to the Champions League final with his son.

One rule for them...

https://www.bbc.co.uk/news/uk-57879730

Knocking on doors with a vaccine bus parked on the street is more effective use of staff time if they deliver more jabs per shift and it might be key for those without transport or child care.

The costs are modest compared to furlough etc etc etc etc etc (several hundred billion) or people getting ill so it's the sort of thing we need to try to see what works.

If someone wrote a comedy series that was a combination of Yes Minister and The Office then I think it would be not that far off what we currently have in government.

Thanks for that reminder that there is way more to demand drops than vaccine hesitancy (logistics, communication, convenience, local supply etc). I still don't trust the data and regard having reliable data on unvaccinated numbers in the two youngest adult age groups as vital information. It's absolutely clear, as illustrated on More or Less, that local vaccination data can be heavily distorted by a high level proportion of student medical registrations (Cambridge was the example they looked at). Vaccination rates might be higher in those groups than NIMS data indicate.

Your pointer to Scotland vs England vaccination data is very interesting. Something different is happening.

I agree with wintertree on behavioural response kicking in, as local numbers go very high. What has put the breaks on my optimism of a month back, and what worries me increasingly, is most people are not stupid about risk, and given current really very high antibody levels in the population, from ONS surveys, case deceleration should have kicked in a week or more ago in England. This is IF herd immunity was happening in a 'vaccinated plus already infected' protective sense. I think delta must be significantly evading previous infection immunity. We know some of the double jabbed can get infected..... it would be political karma if it turns out Javid has spread the virus to ministers who are now spreading it further.

My final point is on the news that this is 'Boris's biggest gamble yet'. It's not. Tens of thousands died on two previous dice throws, when he delayed  implementation of scientific recommended restrictions last September and December.

In Germany vaccination now available from age 12.

What on earth did intelligent people expect from this obvious disaster of a PM? Why didn't progressives in all parties, given their majority of a few percent, work as hard as they could together to stop him gaining a majority? People, often being unfairly accused of being stupid racists, knew enough to work together for their brexit aims; called names by the likes of Liberals in my next door constituency who let in a hopeless tory stooge and lost a really good moderate Labour MP. Plus all those genuine one nation tories who saw their party turn to a dictatorial 'support Boris and brexit or go', losing many of it's best MPs in the process.

I really hope one positive outcome of this horrible pandemic is that this blight of populism is totally exposed for what it is. That needs well educated progressives to admit they made mistakes. We need to remember most people are broadly good, even if some did vote for brexit. I'm not very optimistic.

Meanwhile: 

‘Covid cases are up – 6,000 miles from Paris.’ France baffled by UK quarantine change.

French officials seemed baffled by the move, suspecting UK ministers may have based it on rising cases on the French island of Reunion – nearly 6,000 miles from Paris.

The variant accounts for about one in 10 new infections in France, but the data includes its Indian Ocean territories of Réunion and Mayotte, where the variant is almost dominant.

My theory is that someone told them that with 1% of people infected with Covid any politician who went to a bunch of 'freedom day' celebrations with large amounts of maskless people and did all the photo-ops they were no doubt looking forward to would almost certainly catch Covid.

Then they had an 'Oh f*ck, I didn't think of that moment', and decided the best way to get out of it without looking like total dicks to their followers was to be self isolating. 

It's what you'd expect.  The Scottish peak was most likely caused by specific events i.e. end of school year and the football.   Those events are over, the weather is good, people are on holiday and outside more and vaccination plus control measures is sufficient to get R < 1 in the absence of special factors.

However, tourist season is starting, in Edinburgh a few days ago the trains and bars looked busy and there were too many tourists not wearing masks.   I wouldn't bet on it lasting.

Is that still againist the recommendation of their vaccination advice group (STIKO)?

https://www.zusammengegencorona.de/en/corona-schutzimpfung-ab-12-jahren-wel...

Boris did famously almost die of it though, and didn't Rishi have it too? So probably rules out the self-protection reasoning. It probably is just the universal aghast wtf expression from absolutely everyone on all sides when they said they wouldn't be isolating that made them change their (inexplicable in the first place) minds.

He did indeed nearly die and then after that he took so little notice of the mortal risk that his decisions led to an entirely predicted tens of thousands of extra unnecessary deaths, twice. Your reasoning is clearly inapplicable.

Please stop using Boris (or Johnson) and "reasoning" in the same sentence, there is no credible evidence that shows any significant correlation.

I think there is a sort of flawed reasoning... he reasons rhetorically and selfishly what might be good for him. Sometimes it works and people just raise their eyebrows but too often logic/science/political reality hits him in the face and he U-Turns.

His old tory boss warned us of this. He is an entirely selfish egotist who has no qualms about telling lies. He was always completely unfit for high office.

https://www.theguardian.com/commentisfree/2019/jun/24/boris-johnson-prime-m...

That is an extraordinarily prescient and insightful article. Every Tory MP should read it and weep for the careers.

At the start of the year the Tories were continually slagging the Scottish Government off for falling behind on vaccination and the Scottish Government briefly posted the vaccine delivery schedule they were working to on one of their websites as part of their response.  They were forced to take it down again.

Presumably the vaccination delivery information is either classified as an official secret or labelled as commercially confidential and provided under a non-disclosure in a way which prevents the devolved governments from making it public.

It's hard to fathom on the current information but the US are also increasing restrictions on France. It's discussed here with a chart showing delta is fast heading to dominance.

https://www.connexionfrance.com/French-news/UK-reviewing-red-list-for-Franc...

The UK government position is currently a known concern for not entirely clear reasons.

Exactly.  I think the answer lies somewhere in the middle and tallies with what others have said about big centres closing and the emphasis moving to walk-ins; something of a mismatch between sufficient supply and notionally good demand (notional - as long as it's easy to fulfil?)

In reply to Si dH:

Interesting; thanks.  Something of an impedance mismatch between the vaccination centres and the remaining unvaccinated - presumably many of whom have had the most difficulties accessing vaccination?  A key point of the early response to the Indian variant in the north west was taking vaccination in to work places; perhaps that's one of the way the system could try and access the remaining hold outs.  The leading news stories on people in hospital regretting their choice to decline vaccination might help as well...

It's a real shame there's not more clear and consistent presentation of the longitudinal data with vaccine status.  Given the repeated pieces out of hospitals on the mix of Covid patients it would seem to be a compelling piece of evidence for people wavering over the vaccine.

In terms of the effect Rom describes, I think that's a downstream consequence of the collapse of the giant spike adults aged 20-25 combined with the then-demographics of (un)vaccinated status; a massive spike happened in an age with the lowest vaccine coverage, now that outbreak is collapsing it temporarily masked some of the rises elsewhere in the top level data (never in a breakdown as with plot D1.c).  It's possible that the collapse of those university-style outbreaks were and are masking a more gradual rise within that age bracket as well, which obviously can't be resolved from the demographic data.  

Eventually most infected people will be vaccinated; we knew this was coming from the vaccine efficacies and the estimates at R0 for the India variant.  It's either the start of the endemic phase of the virus that puts the crises truly behind us and leaves us with one more respiratory virus killing people at a similar level to the others, or it's the step that dooms us all, depending on ones outlook.  I'm minded towards the endemic phase, but we'll only really know as we get there.  

In reply to rup:

There's a lot of modelling; the most recent stuff was linked to on last week's thread.  But confidence in the models is low, and the models show a very large range of reasonable outcomes, far too many of which have unsustainable levels of hospitalisation IMO for an unlock on July 21st, and less worrying curves for a later release of restrictions, with the (hopefully realisable) "exit wave" completed well before winter in both cases.  

If you follow the more recent models , we're at worse odds than Russian Roulette.  I don't have great confidence in models right now and prefer to look at the data, and the data is quite clear that we can't tolerate much more rise in the exponential rates for cases without a problem.

In reply to Kirsten:

Like I said last week.... "The wheels on the bus go round and round, round and round, round and round."

Amazing they couldn't see the obvious and immediate backlash coming.  It does name me worry about their capabilities at having to plan more complex and unpredictable stuff, like - hypothetically speaking - managing a successful exit wave during a global pandemic...

In reply to tom_in_edinburgh:

Yup; I've been leaning strongly towards outbreaks there, but my interest detector is starting to go "bzzzt zip" over the data.  If it keeps falling, then it's very interesting.  Probably it will return to rise, but you never know....

In reply to Dr.S at work:

A thread or two ago, jimtitt said this was done against the advice of STIKO and that it caused quite the fallout.  Interestingly... In England, parliament are legally bound to follow the advice of JCVI but Scotland are not, they just rubber stamp JCVI decisions by default but can do otherwise if they wish.  The drip-drabs of information suggest JCVI is not going to propose widespread immunisation of those under 16; there's time to re-evaluate that before the return of school terms as they look back in August over whatever-the-hell is going to happen over the next 4-6 weeks.

There was something on the news today about older kids and vulnerable kids might be about to be offered the vaccine. A sensible step if true.

Yes, and that provision seems to suggest (to me) that we're not going to see a recommendation for widespread vaccination of children any time soon.

Interesting one this.... It could be a "told you so" moment for Mainwood's deductions. If he's been right all along, there won't be enough vaccines left of the first, 40m Pfizer order to offer to all teenagers until the second, 60m order starts coming in in August, without putting second doses at risk. BUT there is enough to offer to some of them now. So if we see something like 17yr olds and vulnerable..... It just might be that.

Is there any info you have spotted on the ratio of reported cases to actual infections dropping during this peak? There should be a bigger proportion of asymptomatic infections in the young (many untested) so that will mean we are closer to record infection levels than the current case ratio indicates: the 7 day average on cases being already at 72% of the old record.

Thinking more on the ZOE data... The "ZOE selection effect" has the potential to be highly biassing here; I'd wager a lot of young people not-yet-vaccinated (at the time of the outbreak) were/are in ZOE, but far fewer the people actively choosing not to get vaccinated are...  Quite aside from the very large CIs on their unvaccinated data, I just don't trust it to be representative.

In reply to Offwidth:

I'm afraid not.  I think it's basically impossible to back it out of the data with the change in strain, the vaccination process and the demographic shift.  the ONS seem jittery about consistently estimating the rate of new infections from their data, and as anything I could do would be less well controlled it didn't seem worthwhile. 

Real facts about deliveries aren't published but criticism is public.

 https://www.scotsman.com/news/politics/covid-scotland-health-secretary-warn...

Depends what you call real

https://www.gov.scot/publications/coronavirus-covid-19-daily-data-for-scotl... 

"COVID-19 Vaccine supply data

As of Monday 12 July:

total number of doses allocated: 7,720,270

total number of doses delivered: 6,953,510"

Remember that "squeeze" never happened and the weekly allocation kept increasing steadily throughout.

https://public.tableau.com/app/profile/phs.covid.19/viz/COVID-19DailyDashbo...

Good graphical information showing 68.9% first dosed in 18 to 29 age group with a 2019 population estimate as the denominator.

Wales at 73.4% on the same age band

https://www.bbc.co.uk/news/uk-wales-55855220

There's no real info on how much of the unused vaccine is mRNA and how much is AZ for the stated reason of commercial sensitivity.

A sudden jump in the rate constant for English PCR cases today, doubling time back down to ~9 days measured by looking at the week-on-week fractional change in the measure.

This goes slap-bang against my weather theory as the period it's from is still in the good weather.

My best guess right now is that this is a football spike like the one Scotland saw, this time relating to the Euros 2021 final.   If so, the gender statistics on cases - something I've not dug in to before - should speak pretty clearly on the issue in a few days, if the media comments I've seen on the Scottish outbreak are accurate.

A sudden doubling of the transmission rate isn't what anyone wants to see on the eve of unlocking.  

I think we are seeing the effects of localised outbreaks (I mean big outbreaks at town or city level, not really local) and of previously lesser hit parts of the country catching up with their infection rates and starting to influence the country-level data. The increase in the data in the provisional window (which is big) is far from uniform across the country. Some places that have been high for a while are still holding flatish while other places are rising very fast. I don't think it can be ascribed to any national-level root cause.

This isn't how evolution works. Variants are being produced all the time but the vast majority do not thrive because their modified genome doesn't confer any competitive advantage. However if the host population becomes heavily vaccinated, you can be sure that the variant that becomes dominant will be more resistant to these vaccines than the previous ones.

How do you keep track!? Each account needs a new email!

Interestingly, my daughter and her friends have all been getting Moderna rather than Pfizer.  

The UK doesn't seem to be swimming in Pfizer vaccine in the same way as the US and Germany.

It also makes me wonder about JCVI advice and whether they are taking into account Pfizer vaccine supply schedule, and the timetable of the UK Government on reopening as well as medical factors when they hesitate to recommend vaccinating under 18s.

It's funny, isn't it.

Two new posters are suddenly here on this thread, one with the hallmarks of the repeatedly banned Rom-like poster and the other is starting to show a hallmark of a repeatedly banned covid denier.  I'm sure one of Rom's three current accounts (or more) will be along soon enough to claim it's not them behind the other account.

It's hard to keep up frankly.

I was gong to reply to them, but I shan't.  I'm not sure what point they're making, they've quoted one part of my message to tell me I'm wrong when that part was clearly offered as a counterpart to my previous paragraphs that are aligned to what they're saying.

Suffice to say, variants can arise in more places than just this country, and the vaccination status is very different elsewhere; that could conceivably lead to a new global variant that arrises in places with little vaccine pressure and that is more aligned to the "classic" spike protein, and so when it comes here, our vaccines would give us better protection than recent infection granted immunity.  I clearly gave this as a counterpoint of what could perhaps happen compared to the "vaccine > escape mutant" view, to make the point, as I said "I get the impression there's no level at which this question can be asked where someone doesn't end up licking their finger and sticking it in the wind."

The two most disruptive variants to date have emerged without vaccination really in play in the locations where variants were generated, and at least one of them has more vaccine evasion.  Which isn't surprising when you stop and think about it all.

I'd feel a lot happier if there was a statement on this instead of endless anecdotes and analysis into limited (not sufficiently granular) data.  

I don't mean this as any slight on the anecdotes or limited analysis - these are all we've got and I appreciate all the contributions.  Unusually they don't lead to a single clear determination.

As I understand the role of JCVI - and I may be wrong - it would be strange if they didn't consider all the variables including these.

Been looked at recently by someone you might recognise....

https://mobile.twitter.com/VictimOfMaths/status/1416101293349916676

It would take a lot of coincidence and synchronicity to bring about that level of change from many independent factors without a common trigger all happening to coalesce at once.  

In the past when there's been a sudden excursion like this in the rate constant it's been an artefact of a bank holiday depressing testing on a given date, and there's a low excursion in the rate constant followed by a high a week later; no signs of that here.

I've make a plot of the ratio of male cases in England to all cases using the maleCases and femaleCases API keys; I think this is PCR+LFD data.  There's no filtering or soothing (day-of-week effects are presumably minimal between genders...).  The most recent datapoint is the big positive excursion in the rate constant for cases, and also has a stand-out ratio of male cases.  It was noted that this was the case for the outbreak around the Scotland/England match.  

To try and merge your view and mine, there's probably some of each going on, and the football effect is probably spread all over the country, and might be more visible in the UTLAs with previously lower numbers of cases.

We'll see how the week goes.  We should open a pool on when the next policy u-turn will be.

The 7 day rolling case average has moved from 72% of the previous UK record to 76% in a day! Yesterday it was looking like it was going to take about a couple of weeks to get to record levels  (with fingers crossed for a deceleration). I hope it's noise but that hope is looking increasingly fragile. If a significant proportion with antibodies are spreading the virus it indicates we may be a good bit further off herd immunity than the government think and if so this is going to be bad.... if we grow faster from our high base it's probably already too late to reverse before major hospital impacts become the norm.

The increase is not a one day blip like you get from a bank holiday or something. It continues in the provisional data period.

Looking at the data again, I agree I was wrong - I had obviously just chosen a bad sample of places to look at the recent data for. There are still some places that are fairly flat but enough have shot up in the last couple of days to say it's fairly national (in England.)

If one football match is enough to cause this then we are truly f*cked now that pubs no longer have to distance people. My guess would be that it's also caused by some sectors of society relaxing in anticipation of the rule change.

Edit - and theorising a bit more, purely in terms of the day-to-day change, it might be that with rising cases the weekend effect grows bigger. Have you just added a day to your data feed that picks up a lot of lagged weekend data?

Put it this way.. I’ve ventured to the pub (outdoors) a few times since may. I won’t be doing it again for the foreseeable..

Special edition of More or Less at 11.30 this am. 'Why are we not at herd immunity yet?' is on the agenda.

Got to catch up with Indie SAGE a bit late this week. It's too early for the problem data of the last few days.

youtube.com/watch?v=x-aNWjXOx5M&

Now fronted by Jim Al-Khalili!

Some interesting data showing testing is creaking in some areas. This includes some very interesting historical plots of things like 'medium time for test return' over the pandemic that I'd not seen before. Testing was in a real mess during previous peaks.

They showed the 18-29 first vaccination rate was about 65% (NHS data so NIMS denominator unlike Scotland). With 62% for 18-25.

Vaccination of children on the horizon?

https://www.bbc.co.uk/news/health-57885845

Weekly step changes in vaccination rates - an element of existing periodic triangular shape superimposed on gradual decline with equal & opposite gradient to the rising part of the triangle.

An implausible coincidence of gradients!?

And now I'll start on interpreting tea leaves left over from breakfast...

Yes; that was my point - a one day blip in case numbers from depressed testing (e.g. a bank holiday) when put through the analysis for this plot, produces matching low and high rate constant measurements 7 days apart, and there's no matching low to this high point.  And, as you say, the provisional window...

I just checked; the last day used in the plot was Friday the 15th of July, with Friday's being one of the less sensitive days to day-of-week effects, so it's probably not that although it's a good theory; we do I think expect testing to degrade cases keep rising much more, and as we've seen (cough cough .XLS) there can be non-linear tipping points in the reporting system...

I've extended the gender ratio plot to include the 16th and 17th - provisional data and shown in red, but I imagine that doesn't bias this measurement anything like as much as it does the rate constant one.  The big gender spike lasts for a coupe of days, so we might theorise we're going to see a similar rate constant measurement appear with today's data...  Probably the closest there will ever be to an impulse response measurement of the testing system.

It could well be that people were shifting behaviour with the messaging going on and the big date approaching; and I see from the news that some nightclubs were opening at 12:01 this morning.

 I can imagine that the nightclubs attract a younger demographic than the football, so the hospitalisations from the football could be a bigger fraction of cases.  But with everything going on so close together in time, the data is going to be really conflated for some time.  

In terms of bolting stable door metaphors, it's a really big stable and it's going to be a few days before all the horses have run out...

This morning on the train from Reading to Padders I saw only one person (a passenger) without a mask. Otherwise staff and passengers all wearing them, on the train and in the stations.

A good thing ... although perhaps not much use in the context of nightclubs opening etc...

Pubs are nothing compared to nightclubs in terms of super-spreader potential.  Speaking from experience as we've unlocked completely twice here, the clubs will be completely rammed every weekend and even during the week.

This will be the only thing that helps, and no doubt what the government is banking on! Nice weather keeping people outside, no school, and thus more parents at home and not in the office.

^^^This. The government is now running blind because the incentives for the population not to get tested / report infections is much higher than the incentive to do so. People want to go on the holidays they've booked, they want to hold the parties they've been putting off, they want to be earning money.  Now they don't think they're going to die because they've had some level of previous infection or vaccination they're simply going to hide their symptoms or avoid the realities of testing.  No doubt there's still some believing the whole 'false positives' myth as well.

The big pubs in Wigan town centre are in essence night clubs. They are massive. Last night at midnight they formally celebrated New Year. Reports from people there - it was party time on steroids. A quick google search will show how crazy town centres where. Infections will go through the roof. I don’t have a view really on how ‘bad’ this is - I’ve kinda thrown the towel in, what will be will be. 

Indeed, but guess what, if you come from the Reunion where the majority of the French Beta variants are, you're exempt of the requirement and can enter without quarantine.

seriously? 

The big spike in the rate constant measurements is not sustained, but the trend is still towards growth in the rate constant - so a shorter doubling time for cases - and the trend is now aligned to the rising temperatures, rather than the usual anti-phase response.

The gender ratio for cases has a second day strongly biassed towards males, and is dropping back towards an even split in the provisional window.  Having never looked at it before, I don't know how provisional the leading edge is.

Football or not, a sudden speed up in growth and a departure from the apparent weather-related trends accelerates the timeline for all the measures by a day or so, further eating in to the marginal slack under a reasonable worst case scenario between the unlocking today in England and the U-turn such a scenario would mandate.  

In reply to Rawn1962, Sequal, oureed & Swerstene:

Don't you think it's time you contacted the site owners to discuss your original ban and follow that path to re-instate your original account rather than apparently getting through one account per post?  It's dishonest, manipulative (when posting from more than one at the same time, as you seem to be back to doing) and unnecessary.

I've had nothing to do with your accounts getting banned on this thread BTW.  

It's a good and interesting point over France and Reunion.  I'm all for as much travel restriction as possible/tolerable right now and have made no bones over my views on the dangers of the amber listing.  I'd far rather see mandatory quarantine on all amber list countries, but it's preposterous to have it on France and not Spain and Reunion if the statistics on the relevant variant are as you claim (I've not looked).  

You writing for the Guardian now?

https://www.theguardian.com/politics/live/2021/jul/19/uk-covid-live-news-en...

I mean......
I was thinking more of.... like.... total doses given / 2 or something for the vaccine number, and then cases based on some guess based on ONS numbers. But they're asking the same question I guess.

Any insights from it?  I couldn't make it.

As I've said before the data from Scotland is getting interesting; two more days of data since the start of this thread - two more days of interest.

There's nearly 3 weeks of decay in the data now since their peak in cases, and no policy changes behind that.  There's only so much the football match could give - a spike and decay of less than a week is my guess.  This feels like way too big of a change in rate constants to fit with the effect of the end school term, considering previous term ends.

The gradient of cases/day even seems to be getting more negative which would be a characteristic of the immunity thresholds being reached - accelerating the decelerating decay.

So...... Has Scotland reached the second half of the logistic function, rapidly heading for the herd immunity thresholds for their current restrictions?  Each day of data seems to add weight to that view.

But....  They've moved to their "level 0" today.  It seems like a muted version of the English changes - limited size household meetings, more masks, nightclubs and "adult entertainment" (Billy Connolly stand up shows I presume?) remain closed.

If they really have reached their thresholds, about 3.5 weeks of infections and vaccinations have gone ahead since then, which could well be enough to keep R<1 with the level 0 they're moving too.

If so, England really can't be too far behind.  Although the difference on nightclubs is going to raise the immunity threshold more for us...

It would be reckless to suggest England is definitely going to follow this soon, and it would be even more reckless to drop a significant number of control measures - as I keep banging on endlessly - when there's not much difference between (a) the time it would take to overload healthcare (by previous high water marks, with good reasons to think they're lower this time) and (b) the time it will take for cabinet to solid data in front of them that it's all gone more than a little caca and they need to jump straight to panic level control measures. 

But still.... My hope is getting a bit reckless.  Probably because the alternatives are Not Good.

The uncertainty in the situation actually has a palpable feel of pressure to it this week.  

Edit: Added updated England plots for comparison.  Also, really good to see the drop in cases in Scotland now translating in to a tipping over in occupancy.  As I've said before, if their cases didn't break soon, occupancy was on course to exceed the previous wave's peak.  

Scottish schol holidays are a couple of weeks ahead of English ones,different timetable, not sure if that is relevant.

The stick just came out on vaccines...as I understood it, Boris just announced that from the end of September, vaccination will become a mandated condition of entry for nightclubs and large events.

Yes, got that.  But, as I said, the change in rate constant is way more than is typically seen with the end of terms.

Unless I've missed it, demographic data isn't routinely published for Scottish cases so it's hard to get much more insight in to it.  Likewise they don't seem to give the PCR vs LFD split.

Edit: Re your edit:

Wish I knew!  Would be nice if it contributed a significant chunk of the drop, as then we'd have a saving grace in England to what might be coming.

Wouldn't it make sense to say vaccination required now as all adults have been offered a jab and double vaccination required when all adults have been offered second jab?

One jab not as good as two but better than none.

I suppose that would mean not prevaricating and is therefore wrong.

It would be difficult to present that as anything other than a U-turn.

To be fair, the effect on vaccination rates won't be much different because if you want to have two jabs by the end of September, you need to get on with booking your first jab asap.

Covid tabby has Scottish new cases, hospitalisations and deaths by age and links to sources but I don't know which specific source supplies particular data.

https://www.travellingtabby.com/scotland-coronavirus-tracker/

Fingers and toes crossed that you are right about Scottish case/hospitalisations/deaths indicators turning towards decline.

Total Covid deaths per capita in Scotland are about two thirds of those in England. Since the Scottish population is known neither for its youth nor its healthiness, it seems a reasonable assumption that there have been no more than about two thirds the total number of Covid cases per capita in Scotland - at least up till the most recent wave. Given that the vaccination roll out is pretty similar between the two countries, if Scotland is at herd immunity so soon, why isn't England there already? Can the difference in behaviour in the two countries (before England's freedom day) really account for such a big difference in the herd immunity threshold? Seems rather unlikely to me, but not impossible I suppose.

Not actually disagreeing with you, but I would observe that we've never before had a summer term ending when Covid cases were abundant and rising fast, still less one with the holidays having sustained high temperatures and dry weather. Regarding the Scottish demographic data, cases appear to be falling fastest in the 20-24 year olds, and not particularly fast in school ages. Graphic illustration down the bottom of this page. https://www.travellingtabby.com/scotland-coronavirus-tracker/. Not sure where the data comes from though.

In reply to Rawn1962:

Hi Rom,

I see you ignored my principle point:

Don't you think it's time you contacted the site owners to discuss your original ban and follow that path to re-instate your original account rather than apparently getting through one account per post?  It's dishonest, manipulative (when posting from more than one at the same time, as you seem to be back to doing) and unnecessary.

Nobody has said that they are  an "acceptable long term solution".  I have never said that they are, and so I am not sure why you are replying as if I had.

I should know better than to get dragged in to a giant circular loop of messages with you that end up with you pivoting 180deg and claiming it was your view all along.  Here we go again…

We remain in a very vulnerable place to new variants until a very high level of population immunity is reached and daily case rates plummet.

Why?

• Many people lack any immunity, so a more lethal variant is Very Bad News for them
• Without reading high levels of population immunity a more transmissive variant is Bad News for everyone  
• This is the big one - and the one that ties my views on limiting inbound travel to the current situation and not "for ever" or an "Acceptable long-term solution". The decision to go with a very high number of local cases to reach that point (agree or disagree it doesn't matter as it's happening) further raises our vulnerability to new variants during the surge, because testing is overloaded and sequencing capacity (critical to identifying and responding to variants) is utterly overwhelmed.  If a variant gets in now in sufficient numbers to be a problem, we won't know until it's too late to surge respond to it.

If we're going to do this, we should shut amber list inbound travel for the duration, as we can't risk a variant getting loose before we're done.  

Agree or disagree with government policy, until we either complete or abandon the current project, variant importation and local generation remains a risk.  Local generation is area under the curve stuff and so largely unavoidable with the stalling of vaccination.  Importation is within our control, I advocate that we control it for now, for the reasons given.

Does now really seem like a good time to be advocating for more international travel?  It doesn’t to me.  I rather hope various countries are going to slam the door on us as our cases peak, with the lower vaccine uptake in many of our neighbours, people from the UK are going to represent heightened danger for the next month or so.

Thanks for the comments and the TT link (likewise thanks Elsewhere). 

Your argument is solid as is one Si dH I think has made before over immunity surveys.

It all feels like an unsquarable circle right now.  Hope might hide in the fraction of young infected who have - or have not - gone for testing when symptomatic, and the fraction of asymptomatic but immunogenic cases.  

But why would that be different than for England.

It is very easy to shoot down the herd immunity theory, and that’s why I think it would be reckless to plan on it, but hope springs eternal.

To turn your question (which I quote) on its head; if the differences between the two counties are so small - and infections ran ahead even in England - why are we seeing such dramatically different behaviour?

Edit:

I’ve suddenly got strong pangs to be up in the torridons swimming in the sea morning and night every day and enjoying endless fair weather summits and pub meals.  

I can see how the weather is all the more exceptional up there.

If we all just put 'hi Ron' and otherwise ignore him it would be better for all of us. 

More or Less was just summarising what we know from here. The only new bit for me was Mainwood saying England look to heading to a demand limit of 70% vaccination in the youngest age band. Let's see what the nightclub announcement from Boris of a double jab requirement does to that.

Mainwood said we have real supply issues for accelerating second jabs and doing older adolescents looks impossible until the next delivery expected around the start of Sept. He said we had used half of the remaining vaccine so that makes the missing 30% unjabbed presumably supply limited as well.

In reply to Rawn1962:

You are some obvious! Just stick with one account. 

Social behaviour and antibody immunity both contribute to R.

Apparent herd immunity in Scotland due to restrictions rather than vaccine/infection induced immunity reducing R below 1?

I don't really believe that though.

Another thing not to understand.

The tempting conclusion is that it's hidden in the demographics. The spike in cases in the younger ages being so strongly dominant, its decay now masks all else. Like we've seen in places that start with B. Is it that? Could easily believe that it's got through enough scottish 20-24 year-olds to be starting to look sigmoidy.
I've been staring at travellingtabby and the UK dashboard and your plots and victimofmaths and others and trying to convince myself there's a difference in attack rate in that age group between the two nations, but it's nowhere near obvious. You could believe whatever you want really.

Their second jabs would come due after the next order arrives wouldn't they? So they can go ahead and get jabbed.

Yup.  I’ve made the cases I can, others on here have poked respectable holes in them.  But their arguments would also poke holes I think in a case for the opposite.

I should replace all my interpretation with “What’s going on?  Haven’t a clue”.  Pretty accurate right now.  Unfortunate that the uncertainties dominate at such a critical time, but there we are.  Having such clear, evidence based policy that’s communicated so well in advance along with pre determined thresholds for adding or removing restrictions really helps me feel secure in the knowledge that we’re in control despite the uncertainties…

Could it be that the end of term effect is being amplified compared to previous ones, due to the vaccination demographics? We should probably anticipate a larger effect

In reply to Rawn1962:

Hi Rom,

I see you ignored my principle point again:

Don't you think it's time you contacted the site owners to discuss your original ban and follow that path to re-instate your original account rather than apparently getting through one account per post?  It's dishonest, manipulative (when posting from more than one at the same time, as you seem to be back to doing) and unnecessary. 
 

I shall try and find a simpler way of explaining why I think now is a really bad time to let it in and find out.  

> It is, in the end, a political decision to take that risk and not take others. I just have a different opinion of what is important

Sure, but we are where we are in case rates, how we got here is irrelevant as we can’t reverse that decision and with where we are right now, I don’t want other variants coming in.

As for your opinion, it seems more rooted to criticism of the tories (your choice of words on many occasion, not mine) than to the public health aspect.  I’m actually disappointed at this.

It could be; but the demographics in England have been heavily loaded towards young adults not old children, and the TT link looks similar for Scotland, so I’d still not expect this big a shift.

If it is primarily down to the end of school terms - I’m skeptical also given the lack of a clear effect from half terms - that would be most welcome in England as a counterbalance to today’s policy changes.

To turn your question (which I quote) on its head; if the differences between the two counties are so small - and infections ran ahead even in England - why are we seeing such dramatically different behaviour?

I suppose I could say that Delta kicked off faster in Scotland than England due to lower infection-acquired immunity*, plus our great enthusiasm about Scotland finally being back at the finals of a major football tournament, and that the subsequent decline in cases was due to a combination of our traditionally rapid exit from the major football tournament, school and uni terms ending, hot summer appearing after a freezing spring, and people modifying their behaviour once the country (Scotland) noticed it had the highest case rates in Europe. Current conditions in Scotland might mean we're at a temporary herd immunity but I'm not letting myself believe that it is sustainable, "proper" herd immunity. Hoping that England follows Scotland's completely baffling trajectory soon though in any case.

* not convinced though - local case data trajectories through the pandemic don't really support that, according to my rigorous analysis by gut feeling of my highly fallible memory. e.g. Dundee had high cases in both second and third waves.

I’ve suddenly got strong pangs to be up in the torridons swimming in the sea morning and night every day and enjoying endless fair weather summits and pub meals

There are certainly worse ways of living.... and the border's still open. On a point of order though, referring to The Torridons should probably have the same stigma attached to it as referring to The Peaks.

In reply to Rawn1962:

Hi Rom,

Still not prepared to address the elephant in the room?

Not in the greater interests then?  When my view aligns with my own direct and immediate interests, I try and put it to extra scrutiny or to keep quiet.

Seeing as you’ve put this on the table twice now I’ll bite.  What’s preventing you from seeing them now?  So you have to quarantine for a short period; are you really unable to arrange that?  

> I was OK to accept such harsh restrictions

Assuming you are “RomTheBear”, my how your tune has changed since March 2020.  A near total about face. Seems to me that what it really comes down to is what suits you and what lets you bring emotional arguments against the tories (your choice of words over many posts, not mine).

Re: discussions about cumulative cases and (in past threads) that and antibody data...there is a way to square that with Scotland having higher immunity than England, which would be if prior immunity from disease (alpha or prior) was relatively little defence against symptomatic disease from Delta. Then the effective immunity would just be from vaccination + Delta, in which case Scotland is probably slightly ahead.

I strongly suspect as well that on average Scotland needs less immunity to keep cases down because it's a less densely populated country, more like the South West corner of England than the North West. The natural mean r there is bound to be lower. This doesn't apply in Glasgow though.

(Demographics wise in Scotland, is it just young adults coming down, or everyone?)

Sounded to me like the spare half of the current batch is needed for planned 2nd jabs before then. He was clear new planned jabs would be supply limited. He was speaking at the back end of the 30 minutes of the show.

Scotland has used 2.72M Pfizer doses.

Scotland's share of the 40M original Pfizer order is 3.28M (8.2% of UK order, based on population). 

That suggests a minimum 83% (2.72/3.28) of Pfizer original 40M has been delivered. 

I have no idea how much of the remaining Pfizer 17% has been delivered.

I've not seen if that first 40M order has been fully delivered or if it is still "on schedule" according to a schedule that has not been published. The best I can see is PM spokesman said in 2019 that "The majority of 40m that we have ordered are expected in the first half of next year" which based on Scotland (see 83% figure) has been met.

I conclude current/future Pfizer supply (eg second Pfizer order for 60M) should be "good" and "on schedule" but I do not know what the schedule is. 

Scotland has used 179k Moderna doses suggesting UK has received a minimum of 2.2M doses of the original orders (7M+10M?). I conclude Moderna deliveries are dribbling in as that seems the most plausible.

Scotland has used 4.06M AZ which suggests a minimum of 49.5M AZ doses delivered to UK of the original 100M order. 

In terms of being "on schedule" according to the unpublished schedules, Pfizer looks the best and Moderna the worst.

That's wonderful, gives over 10 weeks to get some decent superspreader events going before doing anything to moderate it. Talk about closing the stable door after the horse has bolted.

I think you’ve got a near exhaustive list of the “Scottish Factors” there.

I think that’s a sane approach to it.  Your “level 0 day” is going to give you a lot more wiggle time if it ends up with R>1.

I’ll raise a pint to that.  Not in the pub though!  Like minimike our occasional forays to the beer garden are going on hold for now.

We’re going a county north in to Northumberland for our holidays this year.  Hopefully next year…. Living on the west coast is one of my dream ideas, but then again so was working from home and home schooling in order to spend more quality time with them…. until I actually tried it for an extended if unexpected period!

Ah, like with The Lakes?  I’ll show myself out.

In reply to Rawn1962:

Hi Rom,

So you’re not going to address your abuse of multiple accounts and you’re not going to explain why a mild quarantine requirement prevents you from travelling under the circumstances you outline?

Yes they can be heard, and I’ve heard what you’ve said and I’ve heard your diversions to avoid justifying your position.

The best I can see is PM spokesman said in 2020 (not 2019) about AZ that "The majority of 40m that we have ordered are expected in the first half of next year" which based on Scotland (see 83% figure) has been met.

Alpha conferring little protection against Delta was not an idea that hadn't occurred to me at all - thanks. Not very convinced by the population density argument though. I suspect the bulk of the Scottish population is housed in similar density to the bulk of the English population. The low average population density is caused by large, near empty areas in the north and west.

Cases are falling in all age bins in Scotland, more or less simultaneously, perhaps slightly delayed in oldest age groups. Not much decline in those above working age though. Fall slower than rise in all cases

For me the north-west coast is the best place I know in spring, but it is somewhat harder to love in winter.

I’ve only ever been in august.  A mix of perfect days and just dismal rains but there’s always something to be found.  Gutted that the Melvaig Inn has gone.  Swimming in Gairloch as various different big sea mammals come by is just fantastic.  

Just listened to that more or less. Apart from here, I've been getting most of my analysis from James Ward and Mainwood. So yes, no surprises there. But I presumed it to mean we could jab more teenagers if uptake remained low in young adults. That's the impression I've got from what I can piece together from the insane and unfathomable way of reading that Twitter have designed.

So sorry - my previous reply to your post contains an inadvertent double negative in the first sentence that could well read as unpleasant sarcasm, which was absolutely not intended! I'd edit it only that option seems unavailable at present for some reason.

Haha, no worries, I understood what you meant anyway from the post as a whole

Mainwood explicitly says there is a supply problem for further stages in vaccination for older adolescents in More or Less and for shortening the  2nd jabs period below eight weeks (from 24 mins on the link). You can't blame Twitter for that. Do you think he is wrong?

https://www.bbc.co.uk/sounds/play/m000y49w

In other news more leaks from Cummings... Boris just didn't care about people in their eighties dying in the autumn 2019 wave. Just so stupid and callous it's almost unbelievable even for him. The point of restrictions was protection of hospitals (something Boris said he didn't buy!!??)...not caring about dead relatives will bite Boris in the end.

https://www.bbc.co.uk/news/uk-politics-57854811E

No, I've been quoting Mainwood on these threads for ages. You wouldn't believe any of it, but now he's on more or less it's suddenly gospel?

If you read what he writes, he maintains there is "a very specific supply constraint". Namely the gap between Pfizer orders that means you can't run out of the 40m before ~sept. That's different from "there's a supply problem". I took it to mean if uptake in 18+ stayed low there's scope to extend to teenagers.

Show me where I disagreed with Mainwood. I always thought he has been one of the heros of the pandemic. He is in that position because public information on supply is so scandalously poor. I have disagreed with your slant on what he actually said on a handful of occasions.

The supply problem is hardly very specific, it's the main current problem. All 2nd jab acceleration and any extension to adolescents needs non AZ supply and there simply isn't enough extra until September.

In reply to Rawn1962:

Easy said....he was at the centre of government and no-one has shown he was never wrong about anything he has said about Boris, which if he were a clown would be very easy to do. I cant stand Cummings but most of the public damage of what happened after Barnard Castle was down to Boris not sacking him.

When Mainwood writes on twitter, he appears to be doing the best he can with poor data? There are still potential errors. When he talks on more or less, he can only be doing the same.

Is it worth me listening to for half an hour? (That's not rhetorical.)

The question of capacity for vaccinating teenagers is moot because the JCVI said yesterday that "At this time, JCVI is of the view that the health benefits of universal vaccination in children and young people below the age of 18 years do not outweigh the potential risks." The decision is not made on supply or logistic constraints. More details here:

https://www.gov.uk/government/publications/covid-19-vaccination-of-children...

Indeed. Maybe s/poor/limited, but yes. Thing is though, he shows his working, and lists sources, and they're all quite official, even though they're not very verbose. If you follow it all through it's hard to reach any different conclusions.

Yes, it actually probably is. James Ward makes an appearance too. Not sure how I feel about the BBC pulling in "armchair experts" (their words) for these shows, but they've picked some good ones. Although still not our resident one, sadly.

JCVI will surely be factoring in the supply constraints and the 'making best use' arguments, won't they? I thought they were allowed all the info to make the best decision, and to word their statements to maintain public confidence and make it look like their hand wasn't forced by any outside constraint.

In reply to Offwidth:

Edinburgh, Glasgow and Dundee in particular has a lot of housing with common stairs, often poorly ventilated. Population densities are often confusing in Scotland where there is some degree of polarity between densely populated urban areas and the huge acreage of near empty MAMBA. 

According to the statement I linked, no. The argument against is made plainly on the basis of health considerations, with some additional consideration of the benefits to education (but, this seems to be given relatively little weight, partly because of the planned change to guidance in schools from September.) The only operational point made was that vaccinating children would put additional pressure or challenge on other routine vaccination programmes in young people. The point was explicitly made that JCVI few the benefits of those programmes to be greater than a covid 19 vaccination programme in young people.

In reply to Rawn1962:

A lot? The advice remains under review but not for that reason. Read the link.

It's worth adding an important piece of context: an important consideration seemed to be the high vaccination uptake in the UK in adults, reducing the indirect benefits of vaccinating kids on the rest of the population (except for household contacts of the immunosuppressed.) This provides a basis for the advice being different from that in some other countries.

This might be a first, but I'd agree with our #1 recidivist on this. I don't see anything in the text that commits the JCVI to recommend a strategy that can't be implemented for operational reasons. I think we'll go down the ages when we can see the next Pfizer order on the way.

Edit: There'll also be a bigger pile of data from other countries by then, so there's more than one reason, but I think we will be vaccinating teenagers by the winter.

Fair enough. I appreciate it's very variable. I was over simplifying a lot without good knowledge. In my mind, my thinking was:

- I would equate Edinburgh to somewhere like Bristol. Sure there there are lots of student and a few small deprived areas, but not on the scale of Glasgow or a number of large English cities.

- Glasgow is a bit of an outlier, being more like some areas of the north west of England with high population, a mix of highly affluent bits but a lot of deprivation.

- Other cities in Scotland I know little about but are basically all small (Dundee only has 150k people according to Wikipedia?)

Hence my simplification of Scotland being equivalent to the south west of England plus Glasgow. (It's also similarly reasonably well isolated from the rest of the country.)

Ah, the long awaited update is out!

A very thorough and clearly written set of reasoning on how they reached their position.

With the emergence of low probability serious side effects in young adults (which some how get none of the arguments AZ did) and the small numbers of children in the clinical trials, I am happy with their position.  Trials are ongoing, and offering the roll out to vulnerable children where there is much more immediate benefit will scale up the data as well.

Another data point to confuse the issue of vaccine supply/demand/mismatch and the first dose rate.  It is good that this decision could be made on direct grounds only.

I suspect the situation will have moved on significantly by the time they have more trials and side effect data to revisit this decision.

The thrust of the conclusions is that they think the risks currently outweigh the benefits, hence my quote upthread. They would need to change those conclusions in order to reach a different recommendation. This would require:

- more safety data giving a lower risk picture,

- more benefit, perhaps from a new covid variant that affects kids worse or emerging worse data on long covid (although they seem fairly sure about that.) You could also get more benefit from vaccinating kids if adult immunity waned...but you'd address that by giving booster shots to adults.

Operational capability to increase doses doesn't cut it. I'd be surprised like you if they hadn't thought about that, but the written basis for the decision is clear.

I'm hearing what you're saying. I'm just not convinced that it's not a factor in the decision. Obviously they wouldn't write it up in the public domain, and quite rightly. The reasons they've given are sufficient. No need to give any more when it would just lead to damaging and unhelpful messaging.
I don't think it changes anything here, but I do think it's a strong factor in their decision making that obviously they wouldn't draw attention to.

Given a few members of the JCVI have now done interviews in the past 24 hours and have been asked the question “would you rather your child is vaccinated or gets Covid” and answered “gets Covid”, it’s a relatively safe bet that logistics/ supply isn’t driving their view at the moment.

They have also pointed out in the same interviews that the Uk is not an outlier in this approach and that there are other countries adopting the same reasoning.

Interesting. I haven't seen the interviews. I did think that was the implicit message in some of what was said in the statement.

I'm really pleased to see their view on the risks in kids of long covid. That's quite reassuring, since data on the topic is very hard to come by without an agenda behind it.

I think similar.  LAs in Greater Manchester have seen cases numbers fall rise fall rise constantly in this last wave.  Bolton now has the lowest case rate.  But our numbers are still high - we just keep reinfecting each other because we are so densely populated and with more than our fair share of COvid factors (poverty etc etc).  I’ve never known as many people with Covid, all have been double jabbed.  
Screen shot from Manchester Evening News 

Some updates to the most interesting bits illustrated with quotes from contemporary culture.

Scottish cases - continuing to drop.  

• To quote the first episode of South Park, "do your impersonation of David Caruso's career".   Let's hope that unlike Caruso's career, there's no rebound.

English PCR rate constants and gender ratios (including LFDs).  

• Looks pretty compellingly like a single-event spike with both the gender signal and rate constants collapsing back down.  
• This is gong to bugger up the the weather correlation data.  Soon enough this will probably land in the media as with the Scottish one.  
• "It's coming home, it's coming home, Covid's coming home".

Should be right back on track next week though; if anyone's in a stuffy indoor space this week, it's not through choice.

Here's one for you - looks like that student-led(?) outbreak that happened in Durham has done something for local immunity levels.

(Those blue MSOAs are the ones that were hit worst and have dropped hugely week on week.)

Edit - if I stare hard enough at the graph in the second image, I can almost convince myself that the north east is about to turn over. Looks like cases this last weekend will remain lower than the weekend before.

It would be nice if I'm right because cases there have got extremely high.

I want to believe....  

... but I'm left pondering the issue of the population denominator for these case rates; the 2019 MYE population for the city centre MSOA is about 15,000.  The university itself has about 15,000 undergraduates, most of whose presence waxes and wanes with the terms. A lot of those are located in the city centre MSOA and the one to the west.  If the population estimates include the students, then the actual "on the ground" case rate could just about be the same as elsewhere, but the rate normalised including the absent students would be halved.

Also, for the effect to genuinely persist after the undergraduates departed, there would have to have been significant bleed through of cases from the students to the local community.  Whilst that's certainly the view angrily expressed on the ground in many of the surrounding areas,  the senior management have been at pains to point out in the local media that there is no evidence of this having happened.

I'd not thought about the issue of undergraduate heavy MSOAs before and the population estimates.  Certainly be nice this is true; after the successive outbreaks the students there must be closer to any immunity thresholds than most other places...

Speaking of population estimates, this was on the dashboard today: (Take note, Offwidth and LSRH)

On Thursday 22 July, vaccination uptake for the UK, nations and Scottish local authorities will be updated to use the mid-2020 population estimates.

This I presume will have the biggest changes on the very old and the age range of migrant workers.

Edit in response to your edit: 

If Scotland really has, we have to be close.  But we won't have the benefit of three more weeks of vaccination and infection generated immunity having our back when restrictions were dropped.  

Gone from about 3,000 a couple of weeks ago to 1,600 today.

I agree both Scotland and England have seen a football related spike.  But most likely where Scotland has now fallen way back the Tories will have stopped that happening the same way in England by sticking to their 19 July date.    Reopening during the spike will amplify it and make any fall off slower. 

 Even if they were dead set on reopening they should have waited until the spike passed before doing it.

Scotland - second doses exceed first doses for Pfizer and even for Moderna that entered the vaccination system last.

It looks like first jab demand from ages 18 to 29 is about saturated/completed unless supply is tight and 2nd jabs are being prioritised, however...

"Pfizer supply remains consistent, and we have every confidence that the manufacturer will continue to deliver, as it has done, according to the delivery schedules."

https://hansard.parliament.uk/Commons/2021-07-19/debates/143115b3-a71b-46e5...

As ever, delivery schedule unknown.

I didn't disagree with Mainwood. He didn't say demand was saturating at around 50% for the 18-25+ age group nationally then on July 6th  (the foolish statement in that double thread that really annoyed me, and that you used Mainwood to try and back that foolishness up). Local NIMS data simply isn't reliable as a denominator in the 18-25 bin as so many students are counted twice and I've not seen anything yet that totally convinces me that national numbers remove all the double counts. There were real local supply/logistics issues being reported by some here on July 6th and more in the media at the time but some had fingers in their ears and were not listening.

Mainwood is saying clearly now on More or Less that he expects it to saturate around 70% nationally. Wales and Scotland are a good bit over 70% already and still first jabbing so I can accept his view that it will be around 70% but I think on the high side of that range. We were already at 62% first jabs for 18- 25 last Friday (on the NIMS denominator).

I have no idea why you keep playing this tiresome semantic game.  Anyone who claimed then that it was saturating at 50% should be apologising to the thread (as I did when I misunderstood a data bin label)

In reply to Si dh

Yes it's worth a listen but it was a bit disappointing for me compared to their best work on covid. Mainwood was worth it alone from about 23 mins in. The big scandal is that we need Mainwood, it should all be public.

You’ll like this one too. Stick it all on to the airlines. 

https://www.theguardian.com/world/2021/jul/20/border-officials-told-not-to-...

no mention of France… 

Noted. That'll be interesting.

The new number is almost identical to the old number.

https://www.nrscotland.gov.uk/statistics-and-data/statistics/statistics-by-...

I’m so happy I’m going to have a little party.

They’ve shown such responsibility, driving the importation events of the first wave and this one…

Oh, good point. Unless... Aren't the Scottish la ones NIMS now? Too many denominators to keep track of.

Called in on the co-op on the way home from some proper 'scorchio' climbing. The shop must have been at their defined limit as we were in a short queue, one out, one in.  Nearly half the shoppers were not wearing a mask. I really hope this isn't representative. Clear signage asking 'please wear a mask' and a big security man on the door (saying nothing to the unmasked).

Nope, the national 18-25 first dose vaccinated percentages linked above are with a denominator from 2019 population data (as per the link, and which for convenience I copied here). I agree its a bugger to track, which is part of why I praised the very clear Scottish data.

So..... lemme get this straight. The local ones are staying as they are (NIMS) and the nations are changing from 2019 ONS to 2020 ONS? That's a lot less exciting.

To be fair, when we travelled a few months ago they were very thorough and if you didn’t have the paperwork you didn’t get on the flight. 

Now I have to figure out how to avoid driving through France for the grand total of 15 minutes and winning me a 10-day quarantine. 

My wife went to a local cafe and then shop yesterday, and said none of the staff were wearing masks at either. The cafe is one we've been to lots of times and their covid measures were pretty good before. Today for the first time since last spring, when I picked my son up from preschool his teacher wasn't wearing a mask.

Admittedly the hot weather makes it especially hard to wear them for long periods.

For Scotland yes, it would be lovely if the UK government used 2020 population estimates for Englands national figures so we could know NIMS isn't causing a data issue on national numbers and we could then directly compare the seperate nations.

Today's news is full of farcical contradictions in policy; one minister says you don't have to isolate if pinged; No10 says 'but you really should'; NHS website says you should isolate if pinged, or if contacted by T&T...

Two obvious reasons for Government to switch from legal requirement to 'personal responsibility':

1) blame shifting

2) not having to pay support for isolation

Nightclub owners, having been thrilled by 'freedom day' are now furious at the announcement of vaccination passports to be introduced in September. Rather than now, when we have cases growing to record levels.

Oh, and then there's more dirt from Cummings, saying Johnson was taking a "sod the elderly; we don't need to lock down" stance last autumn...

Government seems to be in chaos at the moment... [edit: even more so than the last 18 months...]

I’d add a third - keeping the wheels on the NHS over the next few weeks against the effect of rising pings.

Its hard to see even their more devoted supporters putting up with much more of this chaotic confusion.  (Or am I getting carried away with wishful thinking again?).

Unfortunately I think it's wishful thinking. Too many government supporters only being two out of the three monkeys ☹️

I'm afraid I'm going to agree and it's not just down to intelligence. Went for a walk last week with a friend and a couple of his house mates from university. They both thought that the best was being done under the difficult circumstances, mainly concentrating on the vaccine roll out. I think it's easy to underestimate the percentage of the population who rely on the likes of the telegraph and mail for their "news" 

I'm starting to wonder if there's a mind control slug at work when people tell me this during in-person discussion.  I start surreptitiously finding excused to look at the back of their neck and behind the ears.

(The one case I solidly agree with this on is the vaccination program).

The friend, whom I have climbed with for over 30 years, engineering degree, was trying to explain to me, me a couple of months ago why it wasn't obvious that the Kent variant was definitely more transmissible? At least the delta variant has changed his mind. 

I think virtually all posters on here will agree, but we seem to be squandering the benefits it should have provided. I suppose at least the recent  disagreements on here (by posters with a history) , show the best course of action isn't as obvious as it was in September and December 

I think we've all been reminded that "obvious" is subjective.

No; that's a sensible reason to keep legal enforcement. My two are reasons Johnson is using...

I don't think we're squandering the main benefit - immunity isn't being lost because of our actions - but we're not capitalising on it half as well as we could.  Or are we?  The near future is not very predicable at the moment IMO, and perhaps we'll luck out as maybe Scotland is, so perhaps it's better to say we're not stacking the cards in our favour as much as we could...

Indeed, I think there's the most spread in view points between the regular posters on the subject that I recall.  It's often easier to spot the worst course of action than the best, mind...

A fair point, I probably should have said obvious to me. The consequences of decisions made then were, to me (and many others) at least, predictable. The effect of the current one's, to me, are less so. 

Basically the government have taken a huge gamble that enough people have been vaccinated, and that the vaccines are effective enough against the latest virus strains, so that letting it rip amongst the remaining population (and those vaccinated but still susceptible) is going to lead to sufficient herd immunity, and isn't instead going to lead to NHS overloaded and yet another massive peak of extra deaths.

Trouble is they're gambling with every one of us, and from all the discussion (*) on this thread it's fairly obvious that nobody can definitvely say whether the gamble will work or whether we'll end up with the third avoidable Covid disaster within 2 years.

* - I know several people on these threads are diligently looking at all the data that's available (thank you), but it still seems to come down to "nobody knows".

That’s basically the first half of my take.

The second half of my take is that there is no time pressure to have dropped all remaining restrictions yesterday, and that the “football spike” shows we’ve got a ways to go yet to get to the point immunity starts quenching cases under current circumstances let alone with mass gatherings in full swing, and that we could have had some measures dropped yesterday and saved the rest for 2-3 weeks time to give slack in the system to delay their release if the gamble turns out not to have been taken catastrophically soon.

A massive peak of extra deaths seems pretty far fetched from where we are now, but healthcare overload is perilously close.  It looks like cases turned to decay in Scotland in time to avoid it.  In England?  Well, we’ll know soon enough.

I newspaper headlines tomorrow reporting SAGE informed story saying  " three weeks to avoid covid restrictions"

https://www.tomorrowspapers.co.uk/

I see the Mail, Express, Times and Telegraph are leading on the migrants and Patel paying off the French...

What has it come to when the most germaine headline is in the Star...?

the issue is that the former could easily be the result of the latter.

Very true. 

Yes, an odd choice of where to use the word “crisis” in some headlines when you look at how one is a clear and immediate threat to the nation that’s got the potential to quake universal healthcare and the other is a few people crossing the sea in boats.  Look over there, a squirrel!

A foreign squirrel...

430 = a few?

Would this logic apply if I say we only have a handful of covid deaths don't worry about it?

I think comparing a fixed point early on in an exponential process to a non exponential process wore thin over a year ago, didn't it?  You can't possibly be making such a deceptive comparison out of ignorance can you?

youtube.com/watch?v=bghbxemp4kQ&

https://en.wikipedia.org/wiki/Exponential_growth

So you're standing by the statement that 430 does indeed equal a few?

I stand by my comment that one holds the potential for exponential growth and one doesn't.

One is causing on the order of a million people to self isolated right now and is taking healthcare to the limits, again, the other is not.

One has killed around 150,000 people in the last year, one has not.

One is disrupting supply chains across the country leading to food shortages (again) and all sorts of headaches for businesses already hard hit by the pandemic, one is not.

If you asked me which is worth of the word "crisis", it seems staggeringly obvious.

I think it's pretty sick actually that you've tried to numerically equate one migrant entering the country (a life moving around) to one Covid death (a life being prematurely extinguished).

I've been giving you the benefit of the doubt for a long time, but frankly I don't think you are here in good faith.  

In terms of arrivals in the UK, it is a very small number, so a few seems a reasonable description.  Do you disagree?

I will. 430 is a few. There's no numerical or mathematical definition. A few of the stars are in our galaxy.

In 2015, 250,000 people per day were using Waterloo station, I don't think you'd notice another 403.

https://www.standard.co.uk/news/transport/revealed-london-s-busiest-train-s...

I'm saying quite clearly if I misrepresented 430 as a few that would be unfair, so equally if you do the same that is also unfair. Scale is not a factor, 430 never equals a few. 

If you want to fly into a childish rant because you can't admit that was disingenuous then try and make out I was saying immigration is directly equal to covid that is your prerogative. I certainly can't assume your acting in good faith with a response like that.

You need to do some research in why the two numbers are not equivalent.

https://www.youtube.com/watch?v=bghbxemp4kQ&

I think we need a new term for a certain type of poster- "A Rom".  Whether they are in fact Rom-classic, or a follower of genre doesn't really matter.  There are quite a few around.

Do I really? 

I might adopt your view actually, all the tragedies in the world now seem so small and insignificant. 

Oops shouldn't have said that, now you'll claim I think immigration is a tragedy!

How about if we look at the number 403 as a percentage of the UK population?

It starts to look a lot less than a few, I think 'miniscule' would probably be more appropriate.

"Quite a few people are under 20".  Clearly TomD89 regards that as meaning less than 430

A few people died on 9/11. Does that sound right?

Honestly I'm coming around now, silly that I've been using 'a few' wrong this whole time. I'll follow your convention going forward.

I suppose you can't see the difference between 'quite a few' or 'a few hundred' and 'a few'?

You must be right and I will adapt accordingly. 

Apology accepted

Phew

Somewhat back on topic, can I have a bit of a grump about how the media mostly seem to be acting like the "pingdemic" is some unexpected twist that the pandemic has suddenly thrown at us, and not a completely predictable consequence of a policy that says "we don't care about case numbers so long as the NHS isn't overwhelmed" which they could have been challenging the government on as soon as it became apparent that that was the policy?

I think the word few might have been used to emphasise the difference in scale between the two issues. You could spend all day arguing about how you interpret a few. I'm sure we can all present cases where few can mean something different. I'm also quite sure the point trying to be made was that covid is more of a crisis than 'some' people crossing the channel.

If you think the migration issue is on a par with covid, then argue that case, rather than how many a few means.

The media still haven't got their head around cases tomorrow / cases today = cases today / cases yesterday. Every time it goes predictably upward it's a "new peak".

A worthy rant.  I second it.  You'd almost think perhaps easing down isolation requirements/exposure thresholds on the double vaccinated so as not to disrupt all businesses, education and healthcare would have been a more appropriate next step than re-opening nightclubs.

My issues was and remains the inappropriate drawing of an analogy between numbers from a non-exponential process and an exponential process.  

This is a far more pivotal disagreement that the game of "semantic silly buggers" you've tried to play over the meaning of the word "few".  Good way to side-track the debate from my core problem with your take on it mind.

Having thought on it, there is a more basic problem with your analogy as applied even to a  steady state.  430 Covid deaths/day in England would now equate to around 50,000 people (or more) in hospital with Covid.  This would be a disaster for universal healthcare for all people in the UK.  

I repeat my earlier comment, I don't think you can be this uneducated about the data or the nature of an exponential process, not after your year of  engaging on UKC with these topics.  That doesn't leave many positive options open to interpret your motivation now, does it?

I stand by the fact that about 8500 total people in 2021 = about 40/day = a few = way less than our euro counterparts = Patel is an awful person. 

Another update on the plots illustrated with some popular music.

The tension mounts over the uncertainty in what's happening with cases. [1]

There's tangible decay in both the English and Welsh data.  It's in what I consider the provisional zone of these plots - the raw data points are unlikely to change much with future days of data, but the trendline can still end up going up where it currently goes down, if future rises are big enough.

England:

• It's not a "day of week" effect for England as shown by the 7-day measurement for the doubling time being the closest we've seen to stagnation since mid may (stagnation being exponential rate constant = 0, doubling time = ∞ ) - this measurement should be pretty invariant to ratio-based day of week effects.
• I think it's likely just a return to more "normal" daily case levels after the likely football spike.
  • II've been arguing with myself about if this is a sensible interpretation, as more cases just raises the base from which the next step of exponential growth happens, so my interpretation doesn't fit with that, but...
  • I think it might happen from such a sharp impulse event when there're lags involved; daily numbers might drop again for a few days as there was a period where people were infected but not yet infections;  then the numbers rise again as the next round of infections downstream of those hits; by that point things probably desynchronise enough to become part of the rising baseline signal.  This may be poorly explained due to heat addlement.
• Still, this is just how the enduring downturn in the Scottish data began... [2]
• But, if we just hit that point, we don't have the benefit Scotland did of the three weeks between the turning point in cases and the dropping of more restrictions, and we dropped a lot more in one go. [3, 4]

Wales:

• Same comment as England on day-of-week effects
• Do people in Wales watch England in the finals? The world of football is a mystery to me.

Northern Ireland:

• Still rising
• They've generally been "behind" the other home nations in this wave.

Edit Another pessimistic interpretation would be that lag in the testing system is going up as a result of both rising cases and the rising number of isolation orders; logistic disruption is spreading everywhere right now, and that this is contributing apparent decay that'll resolve eventually. [5]  I'm not too sold on this, but it is probably wise to treat the leading edge as a bit more provisional than normal.

[1]  youtube.com/watch?v=-aMCzRj3Syg&
[2]  youtube.com/watch?v=XBZUz4C6kqk&
[3]  youtube.com/watch?v=SrlhLaNClgw&
[4]  youtube.com/watch?v=wmin5WkOuPw&
[5]  youtube.com/watch?v=X_-q9xeOgG4&

I won’t mention the fact that any time variant process can be modelled as exponential (with varying rate constant as we have used extensively here) and any such process can also be modelled as non exponential (linear with varying gradient)... so, what’s an exponential process in the context of variable rates of change?

//coat

If we're going to get all terminological about these things I'm a Taylorist when it comes to looking at functional forms.

I'm afraid I'm not actually clear what you're getting at, the heat (and its effects on Jr) has rather addled my brain today.

Not such a useful tool as it has been, but introducing a logistic function gets a bit esoteric for the audience I think.  If I was being an optimistic I would see the difference in the last two waves in the Scottish data between...

• Last wave: Simple exponential decay with a fixed, negative rate constant derived from control measures as applied to a mostly susceptible population
• This wave: The really quiet different and more Gaussian like form you get from differentiating the logistic curve as herd immunity is approached - modified by a bit of ongoing vaccination.

The difference is not something I can test with Chi Squared or a P-value because of the absolutely bonkers and unknown (unknowable?) noise statistics on the cases data  - that's been the bane of my life and why I fall back to verbose interpretations and not simple hypothesis tests.

Fully agree, and I’m more than happy with the analysis approach you’ve used. 
 

my point is more about the semantics.. everyone is talking about the pandemic in terms of an ‘exponential growth process’, which it is in the case (and only in the case) of a fully susceptible population (ie. on day zero, when R=R0). After that instant, the exponential rate constant changes, so it’s not a *simple* exponential growth anymore and simplistic descriptions of it as such lead to wild over estimations of case growth (like continuous 5 day doubling for months) which don’t materialise and fuel the ‘scientists don’t know what they’re talking about’ loons..

Testing lag can be observed through the dashboard graphs like the one I showed yesterday and today's update of the same one (below.) For example in the graph below we might expect a small number of additional cases to be added to the data for Monday 19th, but not much, and very few for any date before that. (That's consistent nationally.)

It's holding fairly well at the moment and not responsible for the behaviour we are seeing. As per yesterday's post, North East definitely dropping in to decay over this last weekend I think, from this graph - it's not so clear yet for any other regions so good to get some corroboration from your data. But how long it might last, who knows...

Yes; sufficiently qualifying what I write without loosing the plot of what I’m trying to say is an endless challenge.  It’d be much easier not to qualify it and let people get carried away.  But I’ve been very careful from the start as - hard as it is to believe - I sometimes get a smidgen of an impression that some people deliberately try and misrepresent data and analysis for various ulterior motives.  I know, hard to believe and I should probably just join the Loan Gunmen of the Apoclypse and set up a conspiracy board.

Still, pretty exciting development to be talking about life beyond the exponential, as hypothetical and optimistic as it currently is.

One definition of exponential is that "the rate of change of the quantity is proportional to the quantity".

That definition does not say that the "constant" of proportionality must be a fixed constant rather than varying proportionality changing with lockdown, social distancing, masks, vaccination and infection induced immunity for example.

Does definition of exponential (of which there are more than one) have to have fixed proportionality?

Yes I don’t disagree. 

If you allow the constant of proportionality of the growth (the rate constant) to change, you can model literally anything (including linear growth or decay) as ‘exponential’ which makes the term rather lose its descriptive power. I would say a process is only ‘growing exponentially’ for as long as the initial rate constant is, um.. constant.

Sorry; I detect a risk of talking at cross purposes here; I didn’t think you did, I was just pontificating.

Perhaps it is time to say uptake of first jabs for adults across the UK is saturated/completed for practical purposes because first jabs are down to 40k per day. That's less than a tenth of what daily jabs were.

Alternatively supply is a fraction of what it was despite government statements saying supply is not the limiting factor.

Yes that is true, but only if you purposely describe something with "the rate of change of the quantity is proportional to the quantity" when there is a better description such as "linear".

Thanks for the take on the provisional data - I agree that it doesn't look like it's struggling, and it seems more likely a systemic issue would act as a false ceiling rather than lowering points (e.g. the XLS debacle). 

Carrying on the musical interpretation theme -  youtube.com/watch?v=wB9YIsKIEbA&

Scottish school holidays will be a factor - pupils doing home tests are part of the overall numbers and this element would reduce with end of term. I haven’t looked at the dates closely but it seems that the numbers started dropping before the schools broke up there. Perhaps it was a case of the football spike receding followed closely by end of term. Also the Scottish numbers are a fair bit closer so more prone to noise, though this is clearly beyond any noise effect. 

As you say, it’s a curious one for the English cases. Again, it’s worth noting that school holidays start this week so some LFDs will have dropped out (hence might be worth looking at PCRs only). Testing delays due to system overload could also be an issue. I’d want to see a few more days’ data for that reason. Still, it’s possible that English cases stabilised around 40-45k a day before the impact of the 21st. Of course 40-45k a day is still a lot but it’s a lot better than 100k or 200k. Unfortunately due to the 21st I doubt  that cases will stay at 40-45k for long.

* I meant to say the Scottish cases are a fair bit smaller in number.

My guess is they are getting what they ordered and the total number of units being delivered is as expected but the mix of vaccine doesn't match requirements because they needed to change policy and not use AZ for young people and it is young people who are left.

Possible but I doubt it. I don't think Hamza Yousuf has mentioned Pfizer issues since mid June. 

The rate constant plot is PCR only data and has an exceptional slacking off in the exponential rate, and the end of Scottish term could only cause a direct fall (stopping of LFDs) for one week or so, not the 3 seen.  It’s hard to believe schools are behind that much transmission to make such an enduring difference, going off demographics.

If (big if) we were getting close to the HIT for the old restrictions we won’t be *that* far from the threshold for new restrictions; maximum uncertainty ahead.

Today's Jersey news - their test& trace system is now pretty much overwhelmed and they're looking at implementing more restrictions to go with the compulsory mask wearing they brought in earlier in the week. They currently have just over 3k active cases in their population of ~107k - 2.8% of the population. UK live case rate is currently 0.7% by my calculations, so the UK has some way to go to crash T&T, assuming it has the same relative capacity as Jersey's.

We had our first 'community seeded' case (no obvious source from a traveller) of our 'exit wave' yesterday, and Alderney had their first at the weekend. It will be interesting to see what happens for us compared to Jersey and UK, we've better vax rates, but a population much less previously infected. Still no on-island restrictions, but contacts are made to s-i until they've had a negative test I think.

The increased modelling complexity should have led to a more cautious unlock when case levels were so high. R has always been a population estimate of lots of sub-spread from the fast super spreaders to the slow 'barely spreaders'.

In my head from watching peaks across the world decline from well below herd immunity levels I expected the balance of infectiousness and immunity to have led to an earlier deceleration than we have seen (hence my earlier optimism). Its good to see some slowing at last but far too early to call the English peak as being in decline.

I'm pretty sure of the following :

The positive test rate is lower than it has been in earlier peaks due to more untested asymptomatic spread.

Testing capacity is getting stretched in a few places in the UK but not quite as much as in previous peaks.. Indie SAGE looked at this last Friday.

There must be more spread in the previously infected than in the double jabbed (and that's still significant) but the major problems come when that spread hits pockets of the older unvaccinated (especially in some ethnic groups in deprived areas).

The unlock spike should likely be a bigger problem than a football spike. Public indoor behaviour has changed noticably for a large minority on July 19th from most anecdote I've heard.

Regarding the unlock spike potential, my wife and I both work in retail, she works for M&S and I work for Asda. From the 19th she said there were roughly the same number of people wearing masks as last week, in My store about 30-40% were not!! 

Curiously I also saw a larger number of ethnic minority customers of all age groups shopping in our store without masks compared to last week. 

My stores customers tend to come from more low income backgrounds, we are near the west of Newcastle whilst my wife works just outside Durham which tends to be more affluent, this may have a bearing on the attitude of customers but I overheard a maskless couple talking to someone masked in store and they said they’d stopped wearing masks because they’d both had Covid and so were immune. It was clear that they had not considered that mask wearing was for the benefit of others. 

It has also been interesting that about a month ago I had a spike of deliveries to Covid+ customers, at one point probably a fifth of deliveries which lasted a couple of weeks, whereas this last set of shifts it’s dropped down again to one or two.

For anyone interested in outputs from the Zoe study (briefly discussed on the last thread as they had temporarily suggested that it looked like infections in unvaccinated people had peaked) they have just updated their methods and it has made quite a big difference to their estimates of prevalence:

https://covid.joinzoe.com/post/covid-estimates-updated-vaccine

Their methods are obviously quite susceptible to changes in the underlying population they are trying to model.

Zoe gives useful information but I've not trusted their predictions for a long while now.  I'm glad they have acknowledged their recent predictions with large error margins didn't stack with current actual data.  Even changing their modelling may not be enough if they have a population sample that is too unrepresentative to be meaningful.

How many people would you need to put into isolation to cause a decay in cases?
I'm not convinced we're there yet, but would be interesting to get a handle on how many people in proper, full isolation you'd need to effect signs of decay. Maybe it is some indication that T&T might be enough to tip the balance when we're close to the tipping point?
Or is the heat getting to me too?

100%!!!! 

It makes no difference to R if the rest of us interact with the same number of people when those in isolation are all replaced by someone else.

That's a bit of an exaggeration but I think isolation of a million or two is less important than everyone reducing R by taking precautions (eg previous lockdowns).

It clearly isn't 100%.

They aren't though.

Again, precautions are being removed. And we're at 7% of our workforce isolating at last count. I don't think that's representative, and I don't know whether it's enough to make a dent, but I could well believe it would be.

If 50% of all infected people entered isolation for ten days (after becoming aware), and if by doing so the total number of other people each of them infected was reduced (on average) by 50%, then r would reduce by 25%.

The theory and the maths are straightforward. Obviously in reality there are lots of variables involved with peoples' lives.

To answer the original question, well any level of self isolation makes a contribution to reducing case numbers, and it will still be doing so just at the point when r=1 and tips over. Like many other things you can say with 100% confidence that doing it will lead to case rates decaying sooner than they would without, even if you isolated only ten people in the whole country. The question is how significant the contribution is. I suspect that at 600k per week just from the app, the effect certainly won't be negligible. (1% of the population - and obviously more of them proportionally would be expected to have covid than everyone else.)

Sorry, not followed the whole thread as I’ve been working.

Bit of anecdotal from the North West:

A&Es are packed. Queues on corridors as bad, or worse, than at any point in the last 18 months. After weeks and weeks attending very few covid cases, I’ve been back to two or three a shift

The system is being overwhelmed and we haven’t the resources to attend the calls coming in. The only positive is I’m not seeing the severity of presentations in the covid patients that have been vaccinated, but they’re still often sick enough to require hospitalisation.

All in all, it’s grim. Thank Christ this block’s over and myself and a few workmates are escaping to the hills this evening.

You may have missed the bit where I said "That's a bit of an exaggeration but I think isolation of a million or two is less important than everyone reducing R by taking precautions (eg previous lockdowns)".

However 100% has some justification as if you merely isolate 99.99% of the population but have 0.01% on a hedonistic plague island (or simply "normal" life with slightly insufficient precautions) you still have growth in cases and R>1 for the 0.01%. 

Alternatively fewer people isolate (eg only those who test positive and their contacts) and everybody takes the obvious precautions because there are symptomless but infectious cases, false negative tests, people who don't test and people who don't isolate if they test positive.

The last daily update on the cases vs times plots until Saturday, as for the next couple of days, weekend data emerging through the reporting pipeline salts the trendline; that can't be accommodated until this Monday's data comes through.

Daily cases continue to decay in Scotland, England and Wales and to a squinting optimist, those in NI are starting to peak.

Another stand-out low day in the rate constant for English PCR cases.

• This could be effects of the stand-out high temperatures starting to work through.
  • But, it's soo tempting to draw a wider conclusion coming of what's happened with the Scottish and Welsh data.  That's the beckoning lure of optimism, but there's plenty of other interpretations we've hashed out in previous posts.  
• This analysis compares cases on day X with those on day X-7, which is why it's still showing growth where the most recent days in Plot 6e have turned to decay, if decay is sustained for a few more days, we'll see that flow through in to the rate constant measurement.
• Conflating factors lie ahead - the downstream infections from the point-like (in time) football event will start to land soon, rapidly followed by any increases from freedom day.
  • (I'm inferring the football event I'm cases, the gender statics and past coverage of post-event analysis of similar data for Scotland)

Have a great time in the hills, you so deserve it. Hope you can get a rest and feel refreshed.

The unlock thing is different in that it will have both a spike from 'celebratory' behaviour around 'freedom' day but also a continuous increase in R from the removal of restrictions.

The increase in R from removing restrictions is going to have a multiplicative effect on the secondary infections from the football spike AND the freedom day spike.  Acting against that is the test and trace system making large numbers of people self isolate.  The question is will the Tories allow that to go on when they clearly want people back at work and don't care about infections.  My guess is they'll do something to neuter test and trace and the outcome is going to be a 'double peak' curve with the second peak much higher than the first.   

Seconded!

Yes, if the interpretation around the football is right, then the timing couldn’t have been worse.  Still, at least cases will have been falling, not rising, when it all lands.  I think the uncertainty over what it’s going to look like is big; my rather uneducated guess is that cases will keep falling in Scotland but at a lower rate (I won’t say exponential rate as I think it’s in to a more Gaussian like decay now) and England is - as you suggest - going to see a double peak.   Higher?  I hope no but fear yes.

I still worry the current English drop might be due to something else, like some sudden increase in positive case processing delays. Indie SAGE reported big problems with accessing PCR testing in parts of the UK last Friday. It just looks way too sharp a change. However please don't think about such things while you take your well earned break.

One last popular culture interpretation -  youtube.com/watch?v=0AEj3LA2vSo&

I think those kind of factors are more likely to limit growth in cases but it feels like a stretch for them to explain the kind of recent monotonic decay.

The shape of a HIT based decay is going to be very different to the exponential decay previously seen

• In previous waves, the decay of cases/day rate remains constant in exponential terms and ever smaller in actual terms (wave ended by control measures setting R<1 against a mostly susceptible population)
• Upon reaching the herd immunity threshold for current restrictions, the decay rate in both exponential and actual terms gets faster with time for a while after cases peak. I've got a plot set up to illustrate this with the current English data, but we need another few days of such decay before I'd give it any significance and share it; I'm not sure we'll get that before things perhaps return to rising.  There's going to be a bit more skew from the monotonically rising number of vaccinated people adding asymmetry towards decay.

I think this is coming through strongly now in the hospital admissions data for Scotland - their decay has been going on for long enough to translate in to hospitalisations, where the data is not subject to the kind of concerns you raise about cases.  Look at that curve plummet.  We've heard from several healthcare professionals on various threads (Stichtplate on this one) on various threads this last week about how things are going in England, this couldn't come too soon for Scotland and I hope that last round of unlocking doesn't delay it happening for much longer in England.

This feels like a major milestone to me - that Scotland and now perhaps England can hold cases in decay at pre-freedom day restriction levels in fair weather.  Not seen any coverage of it elsewhere.

Or..... youtube.com/watch?v=t9wmWZbr_wQ&

Edit: or maybe still youtube.com/watch?v=qZtJxhp7mUw&

New yellow card summary is out:

"This safety update report is based on detailed analysis of data up to 14 July 2021. At this date, an estimated 20 million first doses of the Pfizer/BioNTech vaccine and 24.7 million first doses of the COVID-19 Vaccine AstraZeneca had been administered, and around 12.1 million and 22.8 million second doses of the Pfizer/BioNTech vaccine and COVID-19 Vaccine AstraZeneca respectively. An approximate 1.3 million first doses of the COVID-19 Vaccine Moderna have also now been administered. ."

https://www.gov.uk/government/publications/coronavirus-covid-19-vaccine-adv...

So what's the interpretation? It would seem there is insufficient supply to give any more Pfizer first doses (and still have enough to provide the committed second doses) unless the next order is due to start arriving in the next few weeks.

I had basically accepted your and others' argument a couple of weeks back that supply was not at the route of the reduction in 1st jab rate (but thought the delivery system could still be better). Given how tight these numbers are, I suspect there probably were supply constraints involved in limiting the number of appointments given after all. Otherwise, it just seems far too much of a happy coincidence that we can give everyone who needs it their second dose just before we get to when the second order is due, no?

One (likely imo) interpretation is the next order is due to start arriving in the next <8 weeks, as widely expected.

Edit: stories like this add weight https://www.thetimes.co.uk/article/gps-sending-vaccine-doses-back-as-demand...

I don't think it's fair to give all the credit to coincidence. There's a possibility that the timing of the shift from 12 to 8 weeks was carefully thought through (chill out Tom, this is a scientists' decision, not the Tories, so it's ok for it to be a good one). Remember there's moderna too, and while it doesn't look like big numbers, it's critically enough that it would have got us over the net in the situation you suspect. 

Paul Mainwood:

"So, I think timings of the new order must be well confirmed internally, and the system will "bridge" to the new order with decent stockholding.

To be clear: there *would* be supply issues if we took gaps <8 weeks, or saw a sudden demand spike. But things look fine as they stand."

It's demand-limited, pretty sure. I don't think anyone outside the walled garden of actual data can say how close we came to the supply limit, but it doesn't look like we hit it.

Edit again: holy s**t, France. Do something!

Yer, what I meant was that it seemed too much of a coincidence for it *not* to have been thought through and planned. The point being that we have injected fewer people than we'd have liked for several weeks now, but, despite that, the Pfizer numbers suggest that there is only just enough to second-jab  the people who have already had their first jab, before the next batch starts to arrives. It logically follows that the thinking through and planning was making sure that Pfizer 1st jabs were limited over recent weeks, in order that a supply crunch did not occur. Obviously there is some additional Moderna availability but I'd be surprised if we had significant sticks of it lying around that had been offered and not taken up?

(Re: second batch, it may be that Pfizer have simply agreed to continue supply at the existing (fairly steady I think) rate up to a total of 100m rather than 40m, in which case distinguishing between batches is arbitrary.)

Obviously a well planned use of "Just In Time" methodology, or maybe some planning plus some luck.

Which one's your money on?

I'm not convinced first jab appointments have been limited. There's been a bombardment of ads for buses, pop ups, walk ins etc. I know your experience was trickier but there really has been a push to get people to come forward. 

There's stocks of all the vaccines. They aren't publicly acknowledged (why?) but the EU exports and Scottish/Welsh supply numbers show they're somewhere. 

The second order was always slated for August/September. At current rates the first order would dry up (edit: I mean stop arriving, not run out) some weeks before that. It's all unknown to they public for bullshit weasely 'commercial' reasons (my arse) but it's not unreasonable to think there will be a gap. But you're right, there may not be. No way of knowing that.

More selective quoting. Mainwood said on  More or Less there is no current supply slack for 1st vaccinations for older adolescents nor for speeding up 2nd vaccinations in England. He seemed to think the next batch would arrive in very much a JIT manner. If everyone in the younger age groups had turned up for a first jab there would clearly have been supply problems. None of us know what is going on in detail as it's kept secret as you say for bullshit reasons.

Nice.

Full thread here, if I've done Twitter linking right.

https://mobile.twitter.com/PaulMainwood/status/1418314755916320768

Pretty sure I haven't misled.

Mainwood appeared on the high profile radio show last week to make two main points. The first point was 18-25 first jabs were demand limited and would probably level out around 70% (way above the 50% being argued here by some, a few weeks back). The second point was increasing the rate of second jabs or adding first jabs for adolescents would be supply limited.

That's..... Yes. Yes, those things. What are we arguing about?

Just some non-scientific observations. I have children at 2 different schools. Both schools have abandoned the bubble system and have been advising children with +ve LFT to take a PCR. If the PCR is positive then T&T take responsibility for directing isolation. However the drive to take a PCR was to allow the child back into school (both schools did this) and most children took it. Often the rest of the household were taking PCRs too. Now there is no benefit in taking a PCR but significant risk of missing out on holidays. A previous 100% conversion of LFT+ve to PCR test (pre 19th July) is now 0% (in my children's social bubbles). As an aside one school is not passing parents contact information to T&T citing legal concerns over GDPR.

Second observation is at work where every cough and sneeze would be a day off and a PCR test. Again everyone is now paranoid at missing their summer holidays and pretty much no tests are being submitted. 

I'm sure we can look at the data to see if this is likely to be a significant cause of decline (i.e. % tests +ve etc) but I'm not sure its truly due to receding cases. We await the next REACT data.

People avoiding testing, the collapse of the football event spike, exceptional weather - lots of things that could be creating a misleading low in the cases data.  

If the unlock leads to rapidly rising infections, then I think if this is a genuine and short lived drop in new daily infections, it won't really show in either the REACT or ONS data sets; as these show live infections not new ones, which blurs our the shape of the new infections curve by the length of live infections; this would smooth out and fill a brief valley in new infections.  

If this has been a real drop in live infections, I think the most likely place we'll see it reflected is in a slackening off of hospital admissions - sharp features in cases are blurred out in that plot, but even if cases turn to rapidly rising again, if it's a genuine drop in infections it should show through probably not as a decay by a brief plateau.

There have been far fewer sites doing appointments recently than when people in their 30s and 40s were getting first jabs? However I would agree the number of pop-up should ideally have compensated for this. The issue in my eyes with that was that the pop-ups should have been available at a greater range of times and better advertised if we really want to vaccinate the hesitant and not-very-bothered bits of the population.  But - what I said this morning has nothing to do with that, it was entirely based off your figures.

What we now know is that, despite the reduced number of first jabs in recent weeks, we only have just enough pfizer left to give those people their second jabs as they come due before the second batch starts to arrive. Therefore, to my mind we can say with 100% confidence that one of the three following statements is true:

1) we were unknowingly supply limited over the last few weeks because the government knew it needed to reserve pfizers for second doses, or,

2) we were demand limited (edit - or distribution limited), but actually there was almost no more to go around anyway, or,

3) as per 2, but the second batch of Pfizer is due to start arriving earlier than we expect, such that people who have already received their first jab will get their second jab from the next batch (and the government had confidence in this at least a few weeks ago)

One of those things has to be true. My bet is that government have been limiting the availability and marketing of distribution sites in order to keep a lid on the jab rates while being able to state their are still enough jabs in the country as a whole to go around and blame the drop entirely on demand, but I may be wrong. That's also consistent with them closing down the national website telling people where pop-up sites were to be held after only a single week of it operating at the end of June. It doesn't really matter now, but it puts the lie to stuff that was said about us having loads of supply to go around over the last month or two.

Seeing vaccine buses all over the place now, including outside the climbing wall/ sports centre. What I’ve yet to see is anybody using them…. 

We'll, quite. This is my point about distribution and marketing. People can only get vaccinated if they know where and when to go! Given that we will undoubtedly now be looking to pick up people who are a bit hesitant or are just not terribly bothered and not going out of their way, it becomes more challenging and more important to get it right.

(Another case in point - when I got my second jab at a walk-in last week, it was a massive centre with loads of staff and was running well below capacity...but it was the only one running on the city that day (I drove 30 minutes) and to know about it you had to look for the council advertising dates and times on their social media. That approach won't work with people who aren't desperately keen will it?)

Anyway...it's got more interesting to watch case rates again now. Back to work in the meantime for me...

I think if it were 1) we'd have seen people able to book a first jab but unable to find an appointment for their second at 8 weeks, rather it would be at 9 or 10. That would have been the pragmatic way to play that situation, and the vaccine rollout planners have been good at making those calls throughout. So I think it's both 2) and 3). Although we can't be definitive for reasons already covered, I'm less sceptical than you on this occasion (but don't take that to mean I am otherwise).

Yes true, I agree they could have told under 30s to wait 11-12 weeks if necessary which supports your point.

Even 9-10 weeks would almost certainly have bridged it, and you could probably get away with nobody noticing that that was policy if it just looked like there weren't any appointments. Moot though thankfully, or sadly, depending on viewpoint.

Population details for 18- 24:

NIMS 5.239 million

ONS  2020 estimate 4.708 million

That would make current vaccination percentage numbers in the 18-24 age group  based on ONS 2020 population estimates (in line with Wales and Scotland data presentation) a multiplication factor of 1.11 higher.

Yep, it would move things the right way if it is the more correct number, which it likely is for that age group.

This level of uncertainty in the estimate of adult populations just boggles my mind.  

It must be interesting to have the kind of freedoms and problems caused by being in the half a million people who the state can't agree exist or not.

Certainly a long way from the repressive totalitarian state some of pop up posters on here are painting as the natural progression of Covid risk control measures.

What I can't get past is that most of the error with NIMS is put down to double counting, but then when you read about it NIMS is "blah blah something something people with an NHS number yadda yadda", so surely those are unique and could be deduped?

Mainwood's NIMs based estimate of saturation for first jabs is around 70% of that 18-24 (inclusive) age group. In ONS 2020 population estimate terms that would become around 78%, which is above current levels in Scotland and Wales. Current English first jabs must already be around 69% on an ONS population denominator.

Yep, all good news if correct, and no reason to believe any of that won't be, in the fullness of time.

In reply to Si dH, Offwidth et al:

And another thing.... that change in MHRA guidance to extract 6 doses instead of 5 just came up again. If we've been doing that, which we presumably have (because look at the Wales numbers), then the 40m first pfizer order presumably could be good for anywhere up to 48m doses, presumably. At what level that's baked into the maths I'm not sure.

I think a lot of the problem seems to be NHS # for those who are deceased. Though if they're using NHS Digital to verify NHS#, it's probably eliminating a lot of living people, too...

Yep. I've read the wording in full a few times now and I can't get my head around why it can still be so far off. 

800,000 people with Covid last week

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

Hopefully not for half a million in the 18- 24 group!? One might think in a functioning democracy there would be a data pack explaining the differences, since some UK nations do use the ONS number.

Latest ONS infection survey data:

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