Friday night Covid Plotting #27

New Topic
This topic has been archived, and won't accept reply postings.
 wintertree 22 May 2021

It doesn't feel like a lot has changed from last week really.  The outbreak areas are expanding their influence locally, but there's no clear sign of wide spread failure of lockdown radiating out from them at a great rate - which given the initial very worrying doubling times in those areas, means I still think the local factors including around vaccination and importation events are key to what drove the outbreaks, so I'm still hopeful that they're not going to spread in the way the "Kent" variant did.

Cases in England and Scotland are clearly in growth.  Northern Ireland might be but its in the noise.  Wales continues to see cases decay.  Scotland is going to get its own post this week...

Link to previous thread: https://www.ukhillwalking.com/forums/off_belay/friday_night_covid_plotting_26-...

Post edited at 20:44

2
OP wintertree 22 May 2021
In reply to wintertree:

The England plots are on a log-y axis again this week.

Cases in England look to be stalled, but it’s hard to say; the trend over the last few days is  rising, but we’ve seen that several times before over the least few weeks - I think with surge testing going on in various places perhaps the data is all over the place.  Still, cases definitely aren’t still falling and haven’t been for a few weeks, now hospital admissions look to be bottoming out as well.  I think it's likely hospital admissions won't fall over the next week, given recent behaviour in cases and likely anti-correlation between outbreak cases and vaccination status.  Deaths are down to ~5 / day on average now.  They seem to have had a sudden drop, but it's very low numbers so subject to a lot of noise.

So, there are lots of warning signs in the top level data - the potential is clearly there for cases and hospitalisations to rise, and it seems likely similar behaviour will follow in deaths if this is related to unvaccinated people; that’s certainly the sense the news articles and their “sources” suggest.  On the other hand, once everything has been resolved in the past, I’ve not always been able to square off quotes from some sources as give in news reports and the data (the limitation may be with me).  

The biggest piece of news last week IMO was that there was a significant failure of the test and trace system to release cases to local contact tracing that overlapped with outbreak areas; so hopefully we can and will do better.

The effects of the May 17th unlocking are only just about to be felt in the cases data; another week is needed to see what these do.


1
OP wintertree 22 May 2021
In reply to wintertree:

This is using a log-y axis plots for Scotland except for deaths, where small number statistical noise guffs the plot up.  Scotland has had a slower exponential decay for cases than other regions for some months, and was the first to turn clearly to growth.

We can see that hospital admissions and hospital occupancy are both in to growth.  So far, deaths aren’t.  This is happening at about the typical lag from cases going in to growth a couple of weeks back.

 In England we’ve had news reports that the hospitalisations in an outbreak area driven by the Indian variant were mostly/all people who had not been vaccinated.  Perhaps this is the case in Scotland as well - there’ve been news stories on outbreaks of the same variant south of the Clyde in Glasgow, and perhaps similar factors are at work over vaccination.  If I don't sound very certain it's because I'm not - the situation with the data now is a lot more multi-faceted than it was a few months ago, and I've not got the spare time to pour through all the local reports and statistics to get more certainty on that; we've had some really good discussions from other doing that over the last few threads with regards the North West of England, and I know there's a couple of data-driven Scottish residents out there...

The best place to look to understand this are likely the SAGE minutes; I’ve been short of time for reading them but others on the thread might comment...

The critical question to me is whether the hospitalisations are coming from (a) unvaccinated people, (b) vaccinated people for whom the vaccine is not effective against old strains or (c) vaccinated people for whom the vaccine is not effective against the India strain.  My sense from what I've read in news articles and the various official documents is that (a) is the likely cause; but I can't prove that for Scotland.

I think we can take a good guess at what is keeping members of SAGE up late at the moment; burning questions and an ever more complex puzzle to understand how it all fits together.  They do have access to longitudinal data on hospitalisations though which makes it a lot easier...

If it's (a), this is not too worrying, because there are fewer unvaccinated people every day, and absolute numbers are still very low - by the time they become large enough to start breaking healthcare again, a lot more people will hopefully have been vaccinated.

Post edited at 20:45

1
OP wintertree 22 May 2021
In reply to wintertree:

The UTLA watch plot doesn’t look to have changed much for last week - 3 main outbreak areas, none of which have as much as doubled in number since last week which is something positive... Although perhaps their initial rapid doubling times were partly as a result of test and track playing catch-up to live infections after the systematic issues reported in the press kept contact tracing partially offline for a while.  I’ve done a second version zoomed in cropping the outbreaks off the top.  Quite a few UTLAs have rebounded from exceptionally low case rates (<4/100k/day).  It’s not clear to me how much of this is related to the Indian importation events and pockets of lower immunity, and how much is related to the progressive relaxation of restrictions; the effects of May 17th would only just start being felt in this plot - I expect we’ll see more rise next week.  The turn to growth is by no means uniform.

The demographic data for England looks a lot like an extension of last week - the rate constants for cases still turn to growth for > 85s in England over the last couple of weeks, but this is not reflected in the rate constants for the worst outbreak areas.  It’s likely that a small change in the low case numbers is driving this somewhere else in England but I haven’t found it yet; my low-effort attempts at searching the data didn’t find the source.  But somewhere there’s been one or more care home outbreaks I think.  Given the demographics, this could perhaps be driving the hospitalisation signal more than the growth in cases in the outbreak areas - which still looks to have a younger age distribution than previous waves - perhaps reflecting vaccine uptake?

he difficulty from where I sit is understanding the relationship between cases, the new variant and vaccination status; SAGE and PHE are putting out a lot of minutes on this and are giving these questions very detailed attention.  I think we can infer far less from the data I present now than at the end of 2020 as the situation is just so much more fragmented.

What is reassuring is that none of the outbreak areas have continued their highest exponential rates - we’ll look at this in another post…

Plot 18 shows the strongest return to growth for cases in the North West - driven by the outbreaks centred around importation of the Indian variant.  The highly provisional far right of the cases heat map looks like it's going orange for growth - we'll see what another week brings.  Both London and the Midlands (NHS regions, not English regions) have hospitalisation signals returning to growth as well as the North West.  Looking at the raw data on the dashboard, these are likely both real but represent very small numbers - perhaps an additional 5 people / day.  Perhaps one of these corresponds to the rising cases for >85s in England.

So, quite aside from the outbreaks it looks a lot like we're on the cusp towards more widespread growth in cases; with the effects of May 17th yet to really land in the data, I think we can see where this is going in terms of case numbers. 

For the very old, the vaccine does't necessarily prevent hospitalisation but improves outcomes.  So, even though we have rising hospitalisations, the context perhaps around where cases have grown in the demographics and with regards gaps in vaccination means I'm still keeping a somewhat positive outlook - despite all these growth signals, I hope that it's not a sign we're heading back in to seriously bad territory.   But really I'm grasping at straws from increasingly disjointed data and we'll have to see what SAGE make with their privileged access to it

Post edited at 21:00

1
OP wintertree 22 May 2021
In reply to wintertree:

These are new plots using the localised case measurements from the government dashboard's API.  These include both LFD and PCR cases.  These lots are done by looking at the ratio  (cases on day x+7)/(cases on day x) to measure the exponential rate constant, as a way of minimising day-of-week effects without other processing.  Raw data is shown and a polynomial filtered trendline.  

At this level of upper or lower tier local authorities and with cases so low, the data is very noisy.

To my mind, the data for Bolton is encouraging.  The very fast doubling times have not been sustained - perhaps partly through surge testing catching up, but also because control measures still work, and when there is local grapevine and news coverage as well as increased local intervention, the rate constant rapidly drops.  This is probably not a coincidence.

So, hopefully the other regions will soon see some moderation of growth rates etc.


1
OP wintertree 22 May 2021
In reply to wintertree:

Grab bag plots.

Vaccines plot - green are first doses that are followed up - if everyone gets doses in the same order.  I've made another plot with second doses hanging down; it's cleared on this one that the leading edge of second doses is about to have to cover the big spike in first doses from mid-March.  Perhaps this is related to the significant stock of doses it has been suggest are being retained.  Then, in another 4 weeks or so, there will be very few second doses due so first doses in younger people can go ahead at quite the pace.

Variants plot - this hasn't changed from the update last Thursday.  All the usual riders apply - this appears to be by reporting date not sample date, surge testing and prioritisation is involved etc.  It was disappointing to see the latest data point for the Indian variant land on the 5.1 days doubling time line, but I'm not panicking yet as cases in Bolton appear to have dropped to a lower rate exponential, and there's suggestions from various places that there's quite a delay (more than a week) on sequencing data, so hopefully this is partly so high from the surge testing over a week ago.  So, I really hope to see next Thursday's data point fall to a longer doubling time; if it doesn't I'm going to have to reevaluate a lot of my optimism. 

The week-on-week rate constant measurements for English PCR cases shows a clear return to growth; much of this I think looking at the numbers is driven by the outbreak areas; assuming they back off growth or go in to decay, I think especially with the effects of the May 17th unlocking we will see more growth; so all eyes turn to the hospitalisation signal a couple of weeks down the line from that, and to what SAGE have to say on the longitudinal data on vaccination, infection and hospitalisation in the outbreak regions.

I don't think we can afford to have test and trace stuttering as well as having such an open approach to travel; at some point our luck - such as it is - will give out.


1
In reply to wintertree:

There's been a lot of special factors in Scotland the last few months.  Kids went back to school and the older kids have been doing an absolute ton of tests.  Although there are no official exams the schools have been told they will need evidence for predicted grades so there's all the usual exam pressure.   

There was also the Holyrood election and then we had these absolute knobs in Glasgow having their post match victory riot.   Not surprising the incidence in Glasgow went up.  Probably need to wait and see a bit to distinguish trends from one-off factors.   The school effect is going to cool off now exam season is finishing.

2
 Misha 22 May 2021
In reply to wintertree:

Thanks. As you say, lots of factors at play. Today’s dashboard map includes cases up to Monday and there’s a lot of yellow in the more rural parts of England (particularly further south), as well as in a few urban areas. This just goes to show that there’s a masking effect going on, with low prevalence areas offsetting growth areas. As your UTLAs plot shows, base line growth is back across a large number of areas but significant growth is confined to only a few areas. As you say, we should have a better idea of what’s going on in a week’s time, except that the numbers will be impacted by the impact of the 17th coming through… I suspect B.1.617.2 is well seeded now but it’s taking time to show though. The government has suggested that there should be an update on how much more transmissible it is in the coming week.

 Misha 22 May 2021
In reply to wintertree:

SAGE modelling of R for B.1.617.2

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/...

They think around 1.6-1.7 currently but as they point out this is based on current spread so could change as the situation evolves, so I’m not sure how much weight we should attach to this at present. However I thought the graph of imported vs local cases on page 2 is interesting - not that many imported cases…

 Si dH 23 May 2021
In reply to wintertree:

Thanks again.

Ref SAGE, the most recent variant technical briefing has now been published:

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/...

A few interesting bits of data. Probably most interesting bits are the vaccine effectiveness data at the end (supports today's news reports - reduced (~33%) effectiveness after one dose and probably similar effectiveness after two doses when compared with b.1.1.7), plus some interesting comparisons of growth rates between S gene positive and negative cases in different regions. S negative cases are still declining in all regions except, marginally, East Midlands, and I strongly suspect the growth indicated there is all down to the big school outbreak in Erewash which has since been successfully squashed.

There is also some new information on the new VUI - it's another E484K job.

Post edited at 07:41
 Offwidth 23 May 2021
In reply to Si dH:

It was nice to be reading positive news on variants, then I found this.

https://www.theguardian.com/world/2021/may/22/no-10-tried-to-block-data-on-...

Of course Cummings is now insisting herd immunity was the initial government plan and their response in September led to many needless deaths. Hancock insists they are on track for June 21st but I'm yet to see a single SAGE expert interviewed who agrees.

https://www.theguardian.com/politics/2021/may/22/dominic-cummings-claims-mi...

1
OP wintertree 23 May 2021
In reply to Si dH:

Thanks for the link; to add - that’s the change in effectiveness against catching it and becoming symptomatic.  A reduction in the single dose protection of that from 50% to 33% is going to reduce the chunk vaccination has taken out of exponential growth.  Given the strong signs of vaccination working to drive decay in the demographic data, and given how the progressive relaxation of control measures strongly leans on vaccination to keep R<1, this is not great news.   It means we need to vaccinate faster to maintain the status quo.

Hopefully a single dose will have less reduction by this variant in terms of hospitalisation and serious illness/death but that’ll take some weeks yet to determine I expect.  I think this is a reasonable expectation.

1
OP wintertree 23 May 2021
In reply to Offwidth:

> Of course Cummings is now insisting herd immunity was the initial government plan

That was strongly my opinion very early on, and that of more than a few other scientists.  Opinion formed from the governments (in)actions and statements and speculation, not strongly evidenced stuff.

But, to date, absolutely no documentation - eg slides, recordings, briefing notes - has been leaked to evidence this as far as I know.  The early SAGE documents once released showed IMO a strong over confidence in the then under-constrained modelling back in those early days, with government response at the time closely tracking the advice derived in part from the models.

So, if there was a shadow conspiracy to all this, I hope DC can produce significantly more evidence than his word (...) or an edited blog post.  Either way, I imagine he’s watching his back more than normal right now...

1
 MG 23 May 2021
In reply to wintertree:

I think its difficult to know whether Johnson or Cummings is less trustworthy. 

 Offwidth 23 May 2021
In reply to wintertree:

There was clear top level evidence from Sweden that they were initially working with the UK and NL on a herd immunity approach (and felt let down when those governments changed their mind). Our government insist a herd immunity plan was never there but in my view they are almost certainly dishonest in this (it would be a bizzare lie from the swedish position). The public urgency of Ferguson back then would seem odd if herd immunity or something close to it wasn't being considered. I think the main question is who and when, not if?

There are enough independant leaks to pin the blame on Boris for the September and December delays that together needlessly costs tens of thousands of lives and associated suffering and additional economic damage.

I see Cummings as just the latest criminal witness turned supergrass if he helps expose serious government failure with hard evidence. There is no need to like him.

Post edited at 09:54
1
OP wintertree 23 May 2021
In reply to Offwidth:

There is a difference though between considering that “option” and it being policy.  Still, if DC just produces testimony not backed by hard evidence, perhaps next year’s public enquiry will lay it all out for us.

[...]

Post edited at 13:36
1
 Offwidth 23 May 2021
In reply to wintertree:

Sure, but if so why not just say they were considering that option but didn't go with it? It wasn't a total catastrophe in sweden in the first wave, where they still had control measures (and a massive benefit of significant public trust in their government). Everything seems to point to herd immunity being policy and there being a cover-up (and if so, some senior government scientists must have been complicit).

3
 Si dH 23 May 2021
In reply to wintertree:

After our discussions last week about how outbreaks spread, I decided to take daily screenshots of the MSOA map around the outbreaks in the North West (and some other areas) for a week and see what it looked like. (I'm not aware of any other way to do that, there used to be a date slider on the dashboard map but it only did weeks and now it seems to have gone completely.)

These first images are taken from the dashboard every 2 days since 15th May showing weekly average case rates up to 5 days previously. They are marked by the date at the end of that 5-day period. You can see that in this period the Sefton (Formby) outbreak subsided. Spread from Bolton/Blackburn seemingly reached Preston to some extent. Although the spread out from Bolton has happened in all directions, it has happened much more to the north and west than to the south and east (ie, not much spreading in to Manchester.)


 Si dH 23 May 2021
In reply to Si dH:

I also looked at comparisons of high case rates in Bolton with possible causal factors at MSOA/LSOA level, in the form of maps.

The first image below is just the same as the last one above, zoomed in but from yesterday. I used that one because a few more areas turned purple overnight making it harder to visually pick out the differences in today's map.

The second image is a screenshot of that MSOA-level vaccine take-up plot that was in a BBC news article about a week ago, ie prior to the vaccine bus effect.

The third image is a screenshot of a map of Bolton showing the locations of the most (red) and least (blue) MSOAs. From https://www.boltonjsna.org.uk/

The fourth image is a screenshot of a map of Bolton showing the proportion of residents from BAME backgrounds. There is no scale but from the website I can explain that the darkest areas are around 90% BAME and the lightest under 10%.

The original epicenter of the outbreak was I think in Rumworth, which has a fairly low vaccine take-up rate, high deprivation and high proportion of BAME population (mostly Asian.) However, my fairly simple conclusion from these images is that the way that the outbreak has subsequently spread is not clearly correlated with any of these factors. I think that's perhaps quite important to understand when considering how outbreaks should be contained.

Post edited at 17:34

 Si dH 23 May 2021
In reply to Si dH:

I also took daily screenshots of some other areas. This one is from Leicester down to Luton. You can see the spread of the Bedford outbreak (the region west of Bedford, part of Northamptonshire, incidentally also has one of the highest rates of Indian variant sequencing in the country.) However at MSOA level there is a lot of noise. The outbreak has now spread as far south as Luton (sorry, first few images don't show this bit of the map) but at a macro level it doesn't seem to spreading a lot. Interestingly, nothing spreading into Milton Keynes. I included Leicester just in case anything happened there, but it didn't.

Post edited at 17:24

 Si dH 23 May 2021
In reply to Si dH:

And this one is from Nottingham up to Leeds. You can see the outbreak in Erewash in the south subsiding (it was mostly just lots of cases at a school I think), and a new outbreak possibly appearing in Kirklees LA (around Dewsbury.)

I also took some screenshots of the London area but nothing happened. Rates are higher in the west around Heathrow.


 MG 23 May 2021
In reply to Offwidth:

Given the general chaos at the time, I can well imagine it was discussed/considered/evaluated etc without being policy, so quite possibly no one is lying here for a change. 

In reply to Si dH:

This a really nice set of investigations that deserve a big thanks. But...

> The original epicenter of the outbreak was I think in Rumworth, which has a fairly low vaccine take-up rate, high deprivation and high proportion of BAME population (mostly Asian.) However, my fairly simple conclusion from these images is that the way that the outbreak has subsequently spread is not clearly correlated with any of these factors. 

With a bit of local knowledge I'd cast a bit of doubt on this. Not much, but a bit. I think the boundary lines are hiding some of the truth. The lines on the map don't match the lines on the ground. And it's hard to convey just how stark they are in Bolton. The gradients are like nowhere else I've seen, and not all where that map has lines.

In reply to Longsufferingropeholder:

That said, the people with all the data will know this, so if there is any more information to be gleaned they should be able to make good choices.

OP wintertree 23 May 2021
In reply to Si dH:

Lots of great maps, thanks.  The way the failure to hold R<1 is spreading on those from the initial outbreaks has very little resemblance to how “Kent” spread it seems (although I’ve not visualised that at MSOA level).  Very different set of factors as work.  The more rapid growth between Bolton and Blackburn fits with a ”superposition of influences” model although I dare say there’s other factors at work, but it’s an area I know very little about.  Although there’s a car for sale in Bolton that’s beckoning me for a test drive...

1
 Tonker 23 May 2021
In reply to Offwidth:

> Of course Cummings is now insisting herd immunity was the initial government plan and their response in September led to many needless deaths. Hancock insists they are on track for June 21st but I'm yet to see a single SAGE expert interviewed who agrees.

He's going off on one on his twitter account with more tweets added today.

Whether or not the actual words of Herd Immunity were in any plan this is an inevitable outcome of the governments original plan to manage the epidemic in one wave before the winter of 2020/21.

How could it not be? If you want it over in one wave then you must build the level of immunity required to prevent a second wave and if vaccines are not available then the immunity must come via infection.

Watching Marr let Jenny Harries and Priti Patel off the hook today was disappointing. Any half decent interviewer could have nailed them by pointing out what I have written above.

What I cannot get my head around is how did they think the NHS could ever hope to cope with such a policy when in March 2020 we had circa 4500 ICU beds. Even basic bag of a fag packet maths told you this would be breached well before the peak of any wave where you wanted the epidemic done in one wave.

That said our influenza pandemic plan we've had since the Blair government firs created one up was the same. Our NHS could not have coped with that sort of pandemic either. Alan Johnson admitted in 2009 that a serious flu pandemic would cause ICU capacity to be breached x10.

1
In reply to wintertree:

It's an area I know a bit about. (I say a bit.... A lot but not super current). It's an unusual place. I didn't realise that until I left. Anyway, if you asked me to stick pins in a map where I think is most connected to rumworth, I'd put them in Blackburn, Bury, Cheetham Hill, possibly Oldham. This might go some way to explaining my nodding along in last week's thread.

Post edited at 19:46
OP wintertree 23 May 2021
In reply to Tonker:

That’s a lot of tweets.  

I thought JH’s statement as quoted on the BBC was very carefully worded.

Well, it’s going to be an interesting week.  Let’s hope everyone involved makes it through without any unfortunate accidents.
 

1
 MG 23 May 2021
In reply to wintertree:

> Let’s hope everyone involved makes it through without any unfortunate accidents.

!?

OP wintertree 23 May 2021
In reply to thread:

A very cheerful plot - yesterday saw the second highest number of total vaccine doses given. 


1
OP wintertree 23 May 2021
In reply to MG:

> !?

It times and places less enlightened than modern Britain, this sort of political bunfight with threats of damning testimony against senior cabinet members sometimes seems to end abruptly and early with some misfortune or other.

1
 MG 23 May 2021
In reply to wintertree:

> It times and places less enlightened than modern Britain, this sort of political bunfight with threats of damning testimony against senior cabinet members sometimes seems to end abruptly and early with some misfortune or other.

Well yes. I can't say I'd be too sad if any of the participants had an accident,  but don't think it likely..

 Dr.S at work 23 May 2021
In reply to wintertree:

like a Mig-29 escorting your flight to  'firendly' city because of a bomb onboard?

OP wintertree 23 May 2021
In reply to Dr.S at work:

> like a Mig-29 escorting your flight to  'firendly' city because of a bomb onboard?

Yes, pure coincidence I posted an hour before that story broke.  I’m still struggling to process the enormity of that news story and what it means.  Unreal.  

In reply to Dr.S at work:

> like a Mig-29 escorting your flight to  'firendly' city because of a bomb onboard?

Or accidentally brutally cutting off your own head whilst shaving...

 SFM 24 May 2021
In reply to wintertree:

Something that has been bothering me about this latest outbreak is the possibility of vaccine overconfidence. We know it takes  roughly 28 days for the full compliment of initial immunity to kick in. How many newly vaccinated folks are taking increased risks before this time and is that reflected anywhere(Perhaps not counting “as vaccinated” until after 4weeks)?  
I read somewhere that 1 dose gives you 30-50% protection against the Indian strain but if you think you are bullet proof then is this in part fuelling these localised spreads?

 Offwidth 24 May 2021
In reply to SFM:

That percentage is against catching the virus. The effectiveness against serious illness is much higher. The main risk isn't to the irresponsible vaccinated individual.

 Offwidth 24 May 2021
In reply to wintertree:

Yes Jennie did indeed word things very carefully, as every SAGE member speaking in a private capacity has said June 21st looks too risky at present. She works for the government and not only does she represent them she has a shiny new job for demonstrating efficient cover. She even deflected on the science... her answer that some regions don't have majority Indian variant currently wasn't a scientific answer to the question "is the Indian variant now dominant", it was a political answer.

In reply to MG

I don't believe you can be working with two other nations and be confused. They almost certainly went for herd immunity and were forced to change their mind. The truth will come out and this government will happily say information demonstrating that clearly is misrepresenting their heroic efforts. Just like Pritti Patel did when asked about India vs Pakistan numbers and red list decisions on Marr. That she can get away with that on Marr shows how pathetic holding the government to account has become. It was also convenient the position raised by Labour about closing airports until the home office can prevent mixed traffic light queues ran out of time.

Post edited at 08:30
1
OP wintertree 24 May 2021
In reply to SFM:

It’s a good point that the vaccination could have - temporary - regressive effects.  I don’t know how consistent the messaging is around “this is not a superpower jab and takes time to work” is; I didn’t get any such message with mine.  I got it from knowing a little about immunology, from data in SAGE minutes and from discussion from others who know more immunology than me.  I don’t know what I’d assume otherwise.

Not the first time there’s been potential for perverse incentives in the presence of unclear messaging (eg what an LFD clear means).

In reply to Offwidth:

> That percentage is against catching the virus. The effectiveness against serious illness is much higher. The main risk isn't to the irresponsible vaccinated individual.

Yes, the behavioural side of this isn’t so simple; although how those two protection levels scale up from zero after vaccination I don’t know; the one plot I recall seeing suggested hospitalisation tail off over about 21 days from first jab, so it’s still wise to be cautious for some weeks, even if only taking the selfish perspective.

1
 Offwidth 24 May 2021
In reply to wintertree:

There isn't enough data to give reliable detail yet on actual hospitalisations with the main Indian variant? From the Kent variant very few people double jabbed were hospitalised and those were from particularly high risk groups. The latest research is said to gives high confidence the Indian variant will follow a similar pattern at a slightly higher risk level.

OP wintertree 24 May 2021
In reply to Offwidth:

Sorry yes I should have been clearer; the data I recalled on the hospitalisations reducing after first dose was pre the Indian variant.   I was following on from your comment that the wider risk is not with the vaccinated individual by noting that they are still at risk of hospitalisation for some weeks from jab 1; this was with existing variants.  As it seems there’s somewhat reduced efficacy against catching the virus from the Indian variant after one dose (table 9, technical report 12 linked up thread by Si dH) it seems reasonable to assume that risk of hospitalisation if exposed to the virus soon after a jab is going to go up for this variant.

1
 Tonker 24 May 2021
In reply to wintertree:

More tweets from Cummings today

Claiming the government did not realise that 250k deaths from a herd immunity policy would lead to the collapse of the NHS and from this economic implosion. It was only when this actually sunk in that they changed plan.

Staggering lack of joined up thinking if true.

I hate the bloke but if he brings down Johnson I'll support him!

1
OP wintertree 24 May 2021
In reply to thread:

Good news/bad new update.

Good news - Sunday just gone's total number of doses given is the 3rd highest for a Sunday, and the highest since the March bumper period.  This is typically the low water mark in the weekly cycle, so a bumper week ahead?  Looks like we're gearing up to be ready for the bumper crop of second doses coming due relating back to March/April.

Bad news - more growth in the week-on-week take on doubling times for the England PCR only data.  Five consecutive days of growth and 13 days of growth in the last 21 days.  I've added a very filtered polynomial trendline to the plot; the far right of it remains provisional.  It's going up.  Hopefully the vaccines mean the corresponding curve for hospitalisations doesn't go up as much in a few weeks time, or that if they do the supply of unvaccinated people has shrunk enough that the growth can't be sustained for long against a background of decreasing control measures.

Post edited at 16:55

1
In reply to wintertree:

> Good news - Sunday just gone's total number of doses given is the 3rd highest for a Sunday, and the highest since the March bumper period.  This is typically the low water mark in the weekly cycle, so a bumper week ahead?  Looks like we're gearing up to be ready for the bumper crop of second doses coming due relating back to March/April.

Yes. This week should be/needs to be a big one. Following week big too. Looking at the scant evidence that exists and some chicken bones, and concluding that Mainwood's analysis does it all and does it really well, it's easy to see that it will be. Not sure which week will be the peak of the whole rollout but most likely one of the next 2. 

Edit to add: now is a good time to check nerdle

Post edited at 18:07
 minimike 24 May 2021
In reply to wintertree:

Ignoring all the caveats that should be and are applied to that polynomial, the inflection occurs in first week of March.. that’s EARLY! Before lockdown easing.. before Indian variant (supposedly). Hmm

OP wintertree 24 May 2021
In reply to minimike:

> Ignoring all the caveats that should be and are applied to that polynomial, the inflection occurs in first week of March.. that’s EARLY! Before lockdown easing.. before Indian variant (supposedly). Hmm

The inflection but not the zero crossing.  There is inertia and momentum to the rate constant though, phenomenologically speaking. It won’t suddenly cross the axis without first heading for it, changing course if necessary.

Thats around when schools returned, and around when the incredibly strong correlation with the central England temperature melts away.

Make of that what you will.

1
OP wintertree 24 May 2021
In reply to Longsufferingropeholder:

> Edit to add: now is a good time to check nerdle

The motion stabilisation going wrong?  The ship and the early morning fog seem to be anti phase; I started to feel motion sick watching it.

1
In reply to wintertree:

The tower grew

In reply to wintertree:

> Bad news - more growth in the week-on-week take on doubling times for the England PCR only data.

I don't see any attempt to have local rollback of the lockdown release in the areas where numbers are now exceeding 1000+ cases/100k (they're ~1400 and ~1800 as per 19th May).

To me, this just seems to be crazy, dogmatic, head in the sand nonsense.

In reply to wintertree:

> Yes, pure coincidence I posted an hour before that story broke.  I’m still struggling to process the enormity of that news story and what it means.  Unreal.  

It would be interesting to know whether the NATO fighters based in the Baltic states were put on alert.  Apparently at the time of the request from air traffic control the Ryanair plane was closer to Vilnius than Minsk.  

In reply to captain paranoia:

> I don't see any attempt to have local rollback of the lockdown release in the areas where numbers are now exceeding 1000+ cases/100k (they're ~1400 and ~1800 as per 19th May).

> To me, this just seems to be crazy, dogmatic, head in the sand nonsense.

It's crazy, especially if they are getting ready for a large number of second doses.  The research says you need two doses to get good protection against the Indian variant.   It takes 3 weeks for the second doses to be effective. 

Obviously the strategy should be to slow the Indian variant spread down and give the vaccine time to get ahead.  The only downside is loss of political credibility.

In reply to tom_in_edinburgh:

> The only downside is loss of political credibility.

"Data not dates"... 

In reply to captain paranoia:

The local politicians have drawn the battle lines and dug the trenches already. Not sure they realise what they're advocating.

 Tonker 25 May 2021
In reply to Longsufferingropeholder:

43 COVID patients now in Bolton hospital. This is up from 18 on 16/05

On the postive side the 7 day rolling case rate in those aged 60+ seems to have peaked on the 15/05 and is now dropping slightly so we might expect numbers in hospital to peak in the next week?

I suppose this also depends on the cases rates for those in the 40s and 50s though.

 Tonker 25 May 2021
In reply to Longsufferingropeholder:

> The local politicians have drawn the battle lines and dug the trenches already. Not sure they realise what they're advocating.

What do you mean by this?

 Si dH 25 May 2021
In reply to captain paranoia:

They promised data not dates, but they also promised that Tiers were a thing of the past. If they are to keep to their word, unless good evidence becomes apparent that serious illness is prevented by the vaccine, then they need to roll back across the country.

From a purely epidemiological point of view, local restrictions will do nothing more than put a slight delay in to the problem anyway. The variant is already widespread across the country. Last week there were only about 4 known local hotspots where the Indian variant was highly prevalent (Bolton, Blackburn, Bedford, Sefton), since then that number has doubled (see today's news, in which I am also certain the list of areas under new guidance is not comprehensive), in another week the number of areas with local hotspots is likely to have doubled again. This rise in Indian variant outbreaks may be partly driven by travel from the worst hit areas but is also driven by the relaxing of restrictions in large areas of the country where despite low infection rates, data shows that most of those infections are the new variant already (eg the South West where SAGE papers show a higher proportion of infections being thought to be b.1.617.2 than in the North West or London, despite there not yet being any noticeable outbreaks in the area.)

This morning's headlines seem to imply the Government is taking the middle road of putting in place additional local guidance but not announcing it. Presumably this is so that they don't have to take the political heat caused by having to take a step back in the roadmap, but can still point back in future and say they tried to do something. Idiots.

Post edited at 07:55
OP wintertree 25 May 2021
In reply to tom_in_edinburgh:

> It would be interesting to know whether the NATO fighters based in the Baltic states were put on alert.  Apparently at the time of the request from air traffic control the Ryanair plane was closer to Vilnius than Minsk.  

There’s a map on the BBC this morning; it looks like the plane was spitting distance - minutes of flying time - from Lithuanian airspace and twice as close to Vilnius as Minsk.  Waiting until the last moment severely limits the chance for anyone else to get planes in the air.

1
In reply to captain paranoia:

> "Data not dates"... 

Maybe I am missing something but to me a graph that looks like this is saying there is a huge issue with the Indian variant in schools.

https://twitter.com/dgurdasani1/status/1396181139044421634/photo/1

Post edited at 08:00
OP wintertree 25 May 2021
In reply to tom_in_edinburgh:

Maybe.  The age distribution has shifted down noticeably in these outbreaks.  But with vaccination reducing susceptibility and having a strong age correlate, we’d expect cases in older people to be much lower, shifting the distribution younger.  

A shame an informed analysis has apparently been politically censored... 

1
In reply to wintertree:

> There’s a map on the BBC this morning; it looks like the plane was spitting distance - minutes of flying time - from Lithuanian airspace and twice as close to Vilnius as Minsk.  Waiting until the last moment severely limits the chance for anyone else to get planes in the air.

This is interesting in itself.   They had a Mig 29 escorting the Ryanair jet according to the newspapers.    If the Mig started to move before the jet was told to divert it would suggest a set up.

In reply to wintertree:

> Maybe.  The age distribution has shifted down noticeably in these outbreaks.  But with vaccination reducing susceptibility and having a strong age correlate, we’d expect cases in older people to be much lower, shifting the distribution younger.  

Just the steepness of the rise in the graph is far too fast to take chances assuming this is for a reasonably large population.

> A shame an informed analysis has apparently been politically censored... 

Shameful in my opinion.   We have idiots censoring the data before reasonable people can see it because they are worried that reasonable people might disagree with them.

1
 mondite 25 May 2021
In reply to tom_in_edinburgh:

> If the Mig started to move before the jet was told to divert it would suggest a set up.

Off topic but I dont think there is any question about it. Its reported someone watched them board and then it was just time to create a fake threat either in the hope of actually getting away with it or just to help keep things confused for long enough to get the plane down.

 Si dH 25 May 2021
In reply to wintertree and tominedimburgh:

The biggest single school outbreak in recent times appears to have been the one in Long Eaton (part of Erewash LTLA.) See graphs, it didn't show up in Wintertree's analysis because it's part of the much larger Derbyshire UTLA. Bolton's outbreak began in young adults but now covers all ages. The Erewash outbreak was b.1.1.7.

Post edited at 08:36

 Tonker 25 May 2021
In reply to the thread:

Does anyone know where you can get cases rate by age band for each LA?

When I look at the COVID dashboard it only gives it by 0-59 and 60+.

 Si dH 25 May 2021
In reply to Tonker:

If you go to the dashboard for that area and look at the colour chart I copied above, you can click on any part of it to see the case rate for that age band on that date.

 aksys 25 May 2021
In reply to Si dH:

> If you go to the dashboard for that area and look at the colour chart I copied above, you can click on any part of it to see the case rate for that age band on that date.

Looking at the latest demographic data for Bolton on the 19 May case rates were highest in school age children.

Age/Cases/Rolling sum/Rolling rate

5-9/25/179/886.6

10-14/13/220/1171.1

15-19/21/147/886.2

That’s 546 cases in the week up to that date, roughly 1 in a 100 kids with covid. All down to Lateral Flow testing I guess.

Surprisingly little on this in The Bolton News  https://www.theboltonnews.co.uk/.

On the plus side cases overall seem to have peaked on the 18 May and there were only 4 cases in the +70 age range in the week to the 19 May. Strong evidence that the vaccines work.

I would be interesting to see the age demographic on hospital admissions in the same area but that isn’t available.

OP wintertree 25 May 2021
In reply to mondite:

> Off topic but I dont think there is any question about it. Its reported someone watched them board and then it was just time to create a fake threat either in the hope of actually getting away with it or just to help keep things confused for long enough to get the plane down.

I was expecting someone to start a thread, but nobody has, so I've started one in Off Belay.  It's outlandish.  I'm surprised a transcript of the conversation with ATC has not been released.  I hope the western authorities release the cockpit voice recorder for the event.

Post edited at 10:24
1
OP wintertree 25 May 2021
In reply to wintertree:

Something is wrong with my Saturday 20:52 post - the two plot D1.cs are the same.  Replacements below (from the same date).


 aksys 25 May 2021
In reply to aksys:

Therese Coffey did well on GMB this morning on the government’s latest communications failure.

https://mobile.twitter.com/ukiswitheu/status/1397100407567040512

Staggering incompetence!

OP wintertree 25 May 2021
In reply to tom_in_edinburgh:

> Just the steepness of the rise in the graph is far too fast to take chances assuming this is for a reasonably large population.

After the first few days, the demographic rate constants (see my post immediately previously) look similar across the age ranges, so much of that steepness comes from the scale of the outbreak rather than a school-age specific boost to transmission, perhaps.  It looks to me like if there is a school-associated increase above the baseline fast rate of exponential growth, that effect is secondary to the wider pattern. 

> Shameful in my opinion.   We have idiots censoring the data before reasonable people can see it because they are worried that reasonable people might disagree with them.

Indeed.  I can see why in some circumstances you want to control the release of information to support a coherent plan of messaging and control measures that work together to achieve a result through clear, consistent messaging.  That's not us right now - this morning we had a local public health director (North Tyneside) on the radio apparently blindsided and confused by the request from central government for people not to travel in to or out of variant outbreak areas of which it seems they were previously unaware.

1
OP wintertree 25 May 2021
In reply to aksys:

> Therese Coffey did well on GMB this morning on the government’s latest communications failure.

Ouch, that was cringe worthy.  It also hasn't made it any clearer if I can go to Bolton or not to test drive a used 981 Cayman.

1
 Tonker 25 May 2021
In reply to aksys:

> Therese Coffey did well on GMB this morning on the government’s latest communications failure.

This is a joke, surely! 

She was the sacrificial lamb.

 Tonker 25 May 2021
In reply to Si dH:

> If you go to the dashboard for that area and look at the colour chart I copied above, you can click on any part of it to see the case rate for that age band on that date.

Thanks. I was looking for something that you could use to plot by date. 

I downloaded the data from that heatmap but it is in a format that I can't work out how to use with my basic excel skills.

 aksys 25 May 2021
In reply to Tonker:

> Thanks. I was looking for something that you could use to plot by date. 

> I downloaded the data from that heatmap but it is in a format that I can't work out how to use with my basic excel skills.

This might help you to split the data up once you have downloaded the csv file.

https://support.microsoft.com/en-us/office/split-a-cell-f1804d0c-e180-4ed0-...

Post edited at 10:58
 Offwidth 25 May 2021
In reply to Si dH:

"eg the South West where SAGE papers show a higher proportion of infections being thought to be b.1.617.2 than in the North West or London, despite there not yet being any noticeable outbreaks in the area.)"

That's very interesting. Jennie Harris gave the SW as an example of the opposite on Andrew Marr on Sunday.

In reply to captain paranoia:

> I don't see any attempt to have local rollback of the lockdown release in the areas where numbers are now exceeding 1000+ cases/100k

I wasn't expecting them to read my comment and rush into action overnight to impose travel restrictions. No wonder they had no time to tell anyone about them...

 Toerag 25 May 2021
In reply to wintertree:

From the Grauniad this afternoon:-

The number of school children in England with confirmed cases of Covid-19 has jumped by 33% in the space of a week, according to the latest figures released by the Department for Education.

The figures showed that 4,000 pupils were absent from state schools with confirmed coronavirus infections on 20 May last week - an increase of 1,000 compared with the same survey on 12 May.

The DfE figures, based on its weekly school attendance data, showed that the number of pupils self-isolating because of contact within schools has also risen, from 44,000 to 60,000, in the space of a week, and 18,000 pupils were reported to have suspected cases of Covid-19. Some 22,000 were self-isolating because of possible contacts outside of schools.

Absences involving coronavirus cases have now doubled since schools in England returned after the Easter holidays. On 22 April just 2,000 confirmed cases were recorded among pupils.

Secondary school attendance fell from 89% of pupils to 87% nationally, while primary school and special school attendance also dipped. Only national figures are made available by the DfE.

Paul Whiteman, general secretary of National Association of Head Teachers, said:

"There is growing concern about the spread of the India variant in schools. The government must make the data they hold on this public without further delay. Schools need transparency about the levels of infection around the country so they can make sure they have the right measures in place for their local area.  The government must be proactive to ensure that transmission in schools, particularly in relation to the new variant, is not allowed to proceed unchecked."

On 17 May the government dropped its guidance requiring secondary school pupils in England to wear masks in classrooms.

In reply to wintertree:

> Good news - Sunday just gone's total number of doses given is the 3rd highest for a Sunday, and the highest since the March bumper period.  This is typically the low water mark in the weekly cycle, so a bumper week ahead?  Looks like we're gearing up to be ready for the bumper crop of second doses coming due relating back to March/April.

Looks like the biggest Monday we've had in a while (2nd biggest so far??). With the Moderna picking up, are we going to see the magical 1m jab day this week?

OP wintertree 25 May 2021
In reply to Toerag:

As I've said before, the timing is really unfortunate - a relaxation in control measures at the same time as this variant starts rising to prominence.  It takes muddled data and really conflates it a lot more - certainly at the high levels we have access to.

> On 17 May the government dropped its guidance requiring secondary school pupils in England to wear masks in classrooms.

A massive piece of conflation.  I've put the latest demographic rate constant plots in for 3 big outbreak areas and then for the rest of England, excluding those 3, below.  

  • There's some orange appearing in age range 5-10 for the "rest of England" plot, which suggests some of the growth is independent of the outbreaks.  
  • The last couple of days of the outbreaks plot have really taken off in the age ranges 5-15, with the boldest orange for growth.

A differences plot between them is shown - the top right is garbage due to low sample numbers.  The boldest red is in 5-15, suggesting the rise in exponential growth is firming up to be faster in these ages in the outbreak areas compared to the rest of England.

Looking at the cases/day part of the outbreak areas plot, the leading edge is changing fast, and it it now looks less compatible IMO with the vaccines taking a big chunk out of the top right of this graph, shifting the barycentre of age down.   This is still clearly a factor, but I don't think its enough to explain the dramatic shift in structure within those younger age ranges, just the bulk shift to those age ranges.

This demographic data ends ~6 days in to the past, and things look to be developing quickly.  

Muddled, muddled data but it cries out for PHE to be allowed to publish what they know ASAP, and it suggests to me reversing the mask guidance could be beneficial for controlling the rate, certainly in outbreak areas.

Post edited at 16:48

1
OP wintertree 25 May 2021
In reply to Longsufferingropeholder:

> Looks like the biggest Monday we've had in a while (2nd biggest so far??). With the Moderna picking up, are we going to see the magical 1m jab day this week?

Second or third biggest.  Looks like a big week - and it needs to be.  It's almost as if they've planned things spot on.  Good.

The interminable cold and rain is finally set to break as well.

 elsewhere 25 May 2021
In reply to Longsufferingropeholder:

Not seen that - what's the news on Moderna? 

Vaccine uptake incredibly high, 99-100% above age 55 have had first jab in Scotland.

First jabs to ages 18-29 (22%), 30-39 (32%) and 40-49 (70%) on 20/05/21.
First jabs to ages 18-29 (22%), 30-39 (36%) and 40-49 (77%) on 25/05/21.

+0%, +4% and +7% in less than a week for ages 18-29, 30-39 & 40-49 respectively.

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

It looks like high uptake of first jags (got to use the appropriate vocabulary) is percolating down through the ages as is uptake of second jags in older age groups.

For those outwith Scotland - here you get a letter with an appointment which has worked well so far as no need to make an appointment.

There were some problems over the weekend - lots of missed appointments in Glasgow. Reports of letters to sent to old addresses when GP/NHS doesn't have up to date address for younger people more likely to have moved and rarely go to GP. Maybe the postcodes they were doing were bedsit land.

An online registration system was introduced yesterday.

https://www.nhsinform.scot/covid-19-vaccine/invitations-and-appointments/re...

Post edited at 17:28
In reply to elsewhere:

> Not seen that - what's the news on Moderna? 

There isn't any, but the existing leaks that have been around for a while point to it starting to come in when it did and ramping up around now. Which it seems to be. It's still a big (public) unknown.
If you're interested, Paul Mainwood is twittering about it. Reading his stuff is basically all you could ever want to know, but there are some rabbit holes to follow from there if you feel the need to check his working.

 elsewhere 25 May 2021
In reply to Longsufferingropeholder:

Thanks, I'll look at Paul Mainwood.

A good supply of Moderna would be great.

In reply to aksys:

> Strong evidence that the vaccines work

Or that 70+ age group aren't congregating in classrooms, without protection measures...

OP wintertree 25 May 2021
In reply to Longsufferingropeholder:

> If you're interested, Paul Mainwood is twittering about it. 

It's curious how many ex-academics from a couple of fields are digging away at different parts of the Covid data.

1
In reply to elsewhere:

> Thanks, I'll look at Paul Mainwood.

> A good supply of Moderna would be great.

It's probably best described as a 'meh' supply, but it makes a disproportionate difference to how fast we can go.

In reply to wintertree:

Can't imagine many that wouldn't be. Where they discuss it will be the only variable.

 Bottom Clinger 25 May 2021
In reply to Longsufferingropeholder:

> It's probably best described as a 'meh' supply, but it makes a disproportionate difference to how fast we can go.

Especially as the advice is no Oxford jabs for the under 40s and this is where the majority of spread is happening. 

 Dr.S at work 25 May 2021
In reply to Bottom Clinger:

> Especially as the advice is no Oxford jabs for the under 40s and this is where the majority of spread is happening. 


but the risk balance there changes sharply in the face of increasing case rates...

 Si dH 26 May 2021
In reply to Dr.S at work:

There's a dilemma there isn't there. You're right that the risk balance in Bolton now is certainly towards getting any jab in your 20s or 30s (it's probably still borderline in some of the other publicised outbreak areas as actual infection rates remain low in comparison to recent history.) But given the previous publicity, you'll get higher take-up in those age groups if you offer everyone a Pfizer jab. So if you are trying to drive take up in those areas as fast as you can, then that's what you need to do.

Would be interested to know if AZ is actually being given as a first doses to any under 40s now, or whether it is just being used for second doses with the rest being saved for the booster programme.

Post edited at 07:06
In reply to Si dH:

> Would be interested to know if AZ is actually being given as a first doses to any under 40s now, or whether it is just being used for second doses with the rest being saved for the booster programme.

Drawing on the work of the twitterers, looks like we have enough mRNA not to have to. The wildcard is whether there's enough from moderna to allow for this rate to be maintained without jeopardising the 12 week gap. We'll hit the end of the first pfizer order, just, and are due to get more in September, so we might (or might not) see a slowing down to push some second doses back into sept. That seems to be how we'll find out how much moderna have delivered.

Edit: or maybe we'll be brave and tell a few people sorry, you'll have to wait 14 weeks, your white cells don't have a great wristwatch so it's gonna be ok.

Post edited at 07:35
 Si dH 26 May 2021
In reply to thread:

Results of the Liverpool test-to-party pilot events have been reported.

https://liverpoolexpress.co.uk/liverpool-pilot-events-have-no-impact-on-cov...

Seems a very good outcome. Rates in Liverpool at the time were in the 20/100k range so obviously not appropriate to extend the conclusions to areas/times of relatively high prevalence.

 MG 26 May 2021
In reply to Si dH:

How on earth do risk assessments and ethics approval work for that sort of thing?  "We are going to mix people and see if some die?"  Is it an end justifies the means approach?

 Bottom Clinger 26 May 2021
In reply to Si dH:

> Would be interested to know if AZ is actually being given as a first doses to any under 40s now, or whether it is just being used for second doses with the rest being saved for the booster programme.

To my knowledge, in Wigan AZ is not being given for first jabs for under 40’s and would guess same in Bolton (assuming they have enough Pfizer and hopefully Moderna for for under 40s). 

 Bottom Clinger 26 May 2021
In reply to MG:

I reckon it’s more like: ‘we are going to mix people who have a very, very low chance of dying because at some point we need to get back to a similar normal than before Covid.’  Which I agree with. 

 Misha 26 May 2021
In reply to wintertree:

Over 3k cases today. I know there’s some weekend lag there but Monday’s by specimen date is also over 3k and counting. I suspect your next UTLAs watchlist plot will make for interesting reading as some areas are still relatively low, whereas others are growing and that’s not just in the 8 ‘identified’ focus areas (eg Brum’s cases reported today are up sharply, on top of the last few working days already being significantly up on previous - I know this isn’t proper stats but it’s an example of what’s going on in a number of places).

The impact of the 17th would be just starting to feed in, so it’s hard to say what’s due to B.1.617.2 vs what’s due to the 17th vs what’s due to a combination of both. Hopefully the powers that be have better data.

Hospitalisations seem to be somewhat higher as well but not massively so ‘by eye’.

Still, I’d say it’s fairly clear that we’re going into the third wave now. The only real question is how bad it will be.

Perhaps the fallout from Cummings’s testomony today will lead BoJo to think twice about the 21st but that may be wishful thinking on my part. Suspect it will be too little, too late again. I suspect realistically we need to roll back most of the relaxations which took place on the 17th. I say most as some of them shouldn’t lead to spread in a big way (eg campsite toilet blocks - currently sitting on the campsite at Reiff and feeling fairly safe from Covid; however the pub was full and that’s questionable given there are visitors from much more populated areas there - was chilly sitting outside but the view is much better anyway!). 

In reply to Misha:

> Still, I’d say it’s fairly clear that we’re going into the third wave now. The only real question is how bad it will be.

There is absolutely no excuse for it.  All they needed to do was be a little cautious, stop travel from India and take it easy on opening things up any further to make absolutely sure the jags versus Indian strain race was won by the jags.  650k jags a day for one more month is about 20 million jags.

6
 Punter_Pro 27 May 2021
In reply to wintertree:

There are now four cases of the Indian variant in my town in Suffolk, it's spreading and the third wave has unfortunately probably begun...

Got my vaccine booked for June the 7th, I am also off to the grit tonight for a couple of days before summer arrives this weekend. Self contained, no pub visits and taking my own food with me.

Let's hope the warm weather forecasted next week helps keep a lid on things..

 Tonker 27 May 2021
In reply to tom_in_edinburgh:

> There is absolutely no excuse for it.  All they needed to do was be a little cautious, stop travel from India and take it easy on opening things up any further to make absolutely sure the jags versus Indian strain race was won by the jags.  650k jags a day for one more month is about 20 million jags.

Did Nicola put India on the Scottish Red list before we did?

2
 Si dH 27 May 2021
In reply to Misha:

I agree. But to put a positive spin on things, if you had offered us at the beginning of March before schools opened a situation in which the roadmap was followed and by late May, 6 weeks post shops opening and 10 days post pubs, we were still only at 3000 cases per day, I think we'd have all taken it. I think this situation was at the most optimistic end of predictions, some were predicting cases would rapidly turn up the minute schools opened. So in the bigger picture this is a relatively good place to be. The Govt have always made clear since March that the real driver of what they did would be impact on hospitals, not cases alone, and I think we always knew the situation would arise eventually in which cases grew and there was an awkward wait to see the effect. We are now in that time. Obviously a month more of jabs would have helped, but now is much better than if it had happened in March and much better than the situation where countries are opening up abroad. The Indian variant is clearly taking over everywhere now from the S gene data in the last PHE variant report that I linked above, but it doesn't seem to have had a really major effect anywhere outside of Bolton. The spread of cases out from the worst hotspots is quite surprisingly slow in my view and case rates in the oldest groups are still low - so I think there are some grounds for optimism, as well as concern.

Post edited at 06:52
In reply to Tonker:

> Did Nicola put India on the Scottish Red list before we did?

It was a four nation decision based on UK advice.

Scotland had an airport managed quarantine requirement for all flights not just flights from red list countries but it was never effective because it could be dodged by flying through an English airport. 

5
 Si dH 27 May 2021
In reply to tom_in_edinburgh:

Sturgeon has a choice. If she actually believes Scotland would be better off keeping cases low until every adult is vaccinated and wants to do what is best for her country rather than just aiming to create a situation in which she always follows England closely but is perceived to be a bit more cautious and careful, then she should put the country back in to lockdown now for another 1-2 months. She won't do that because that's not what she believes.

2
 Tonker 27 May 2021
In reply to tom_in_edinburgh:

> It was a four nation decision based on UK advice.

> Scotland had an airport managed quarantine requirement for all flights not just flights from red list countries but it was never effective because it could be dodged by flying through an English airport. 

Right so Sturgeon agreed with it. If she really believed it was wrong she could have taken Scotland down a different path. 

4
In reply to Tonker:

> Right so Sturgeon agreed with it. If she really believed it was wrong she could have taken Scotland down a different path. 

The red list date is a red herring.  Scotland has a managed quarantine requirement for non-red list countries.  But it can be circumvented by flying through England.   The UK government refused to help impose Scottish rules on people travelling on to Scotland who landed in England. 

3
 SDM 27 May 2021
In reply to Si dH:

> There's a dilemma there isn't there. You're right that the risk balance in Bolton now is certainly towards getting any jab in your 20s or 30s (it's probably still borderline in some of the other publicised outbreak areas as actual infection rates remain low in comparison to recent history.) But given the previous publicity, you'll get higher take-up in those age groups if you offer everyone a Pfizer jab. So if you are trying to drive take up in those areas as fast as you can, then that's what you need to do.

And it should be based on the future risk tolerance, not the current situation. With cases expected to increase, people have to consider where they are likely to go in the near future, not where they are now.

> Would be interested to know if AZ is actually being given as a first doses to any under 40s now, or whether it is just being used for second doses with the rest being saved for the booster programme.

I don't know the answer to this but I do have some anecdotal  evidence on over 60s: both of my parents had AstraZeneca for their first dose and had their second jabs this week. They were offered a switch to an alternative and were advised by their GP to switch.

They both have underlying health conditions (one has a history of clotting, is on immuno-suppressing medication and had moderately severe side effects from the first dose lasting ~2 weeks, the other has severe asthma) but it would be interesting to see how many older people are being advised away from AZ and whether this will have much affect on supply.

They both decided not to switch; switching would have meant having to book another appointment rather than getting jabbed immediately and you have to get 2 doses of the new vaccine. They decided that increased protection ASAP was more important, especially as side effects for the second AZ dose tend to be milder.

 MG 27 May 2021
In reply to tom_in_edinburgh:

Given that infection rates are higher in Scotland currently, it is clearly the nefarious SNP and their policies infecting England. Shocking.

3
 Si dH 27 May 2021
In reply to SDM:

> And it should be based on the future risk tolerance, not the current situation. With cases expected to increase, people have to consider where they are likely to go in the near future, not where they are now.

It isn't yet clear whether cases will rise to the level that it would make sense for people in their 20s (on a personal risk balance) to get an AZ if there was no alternative (outside of Bolton). That's missing the point though. There will be a lot of people who would rather have the Pfizer and they know that is what they are supposed to be offered - so if only AZ is available they will come back another day and that will slow the overall programme. Even in a hotspot, the question isn't "do I want a jab or not?", it's "do I want an AZ or shall I try again next week?"

I suspect and hope this is academic anyway. Certainly in Bolton, the photo I saw of one vaccine centre had a very prominent sign outside advertising they had pfizers...

Post edited at 08:17
In reply to MG:

> Given that infection rates are higher in Scotland currently, it is clearly the nefarious SNP and their policies infecting England. Shocking.

They aren't infecting England because unlike the Tories, the Scottish Government actually tell people when they have local lockdowns and don't remove them as soon as somebody complains.   Glasgow is in level 3, has been for a while.

7
 Tonker 27 May 2021
In reply to tom_in_edinburgh:

> The red list date is a red herring.  Scotland has a managed quarantine requirement for non-red list countries.  But it can be circumvented by flying through England.   The UK government refused to help impose Scottish rules on people travelling on to Scotland who landed in England. 

Can you explain what this entails?

1
 jonny taylor 27 May 2021
In reply to SDM:

Re switching away from AstraZeneca: here (Glasgow health board) an early 40s was told they were not allowed to give her Pfizer even when it was requested 

OP wintertree 27 May 2021
In reply to Si dH:

> It isn't yet clear whether cases will rise to the level that it would make sense for people in their 20s (on a personal risk balance)

It's important to look at the demographic distribution of cases as part of assessing that risk; with vaccines having removed a large fraction of older people from the circulation of cases, the barycentre of infection is at a much younger age - and there's perhaps a variant factor involved as well.  

> I suspect and hope this is academic anyway

At the rate things are going, it won't be very long until every adult has been offered a vaccine and their first dose has had some time to take effect, so that will make this doubly academic.

In reply to Misha:

> Over 3k cases today. I know there’s some weekend lag there but Monday’s by specimen date is also over 3k and counting. 

There's an exceptionally high number of LFD positives in the England data for the 24th, currently only very small growth in PCR cases, but we'd expect more PCR cases to land yet.  So I don't think it's as bad as breaking the 3k barrier suggests, but it looks like PCR cases remain in growth - despite the masking effect of Bolton having decreasing cases significant at the national level.

> Hospitalisations seem to be somewhat higher as well but not massively so ‘by eye’.

Rising signals are starting to show for some areas in the rate constant measurements at the regional level.  It should be clearer for the next plotting thread update.  It's becoming very hard to interpret this data though as the longitudinal vaccination status is the critical context we don't have access to with the data, only through figures and tables prepared for reports to SAGE that are released. 

> Still, I’d say it’s fairly clear that we’re going into the third wave now. The only real question is how bad it will be.

I don't think it should be bad; the vaccine uptake figures speak for themselves, and the presumption is that the vaccines are going to still retain high efficacy for illness and hospitalisation against the Indian variant.  Reduced engagement with vaccination seems to be a strong factor in areas ringing alarm bells now.   If the data (undoubtedly being put together now) shows signifiant hospitalisation of vaccinated individuals , there needs to be an emergency stop to the unlocking.  Otherwise, there's only so much those areas can drive top level figures and so only so much they can threaten universal healthcare.  

One big risk of allowing large numbers of sub-clinical infections of a new variant - with reduced immune efficacy in terms of infection - to circulate is that this could lead to a new derived variant, with further clear water between it and the vaccines.  But, this is going to happen at some point, somewhere in the world, and without widespread support for decent travel controls, its coming here.  Perhaps the best protection against that variant is to havre board spectrum immunity against all the viral proteins, not just the spike used by the current vaccines. With the shift in barycentre of infection way down the age scale, quite a few young people in outbreak areas will be getting that kind of immunity now.  At some point, that option stops being the bad option; I think it's too soon.

This all feels like rushing head long to me, and we're still waiting for some of the immune regulating therapeutics to emerge (hopefully successfully) from the trials pipeline that I'd much rather were out first, as an insurance policy against getting it wrong (again).

However, we're not all going to get the unlock we want as different people want very different things.  Back in March 2020 I was appalled at the decisions and actions, but the situation has changed incomparably.  We have wide spread testing of two types and sequencing at scale; it's disappointing that issues are still cropping up with test and trace, but perversely the Bolton outbreak coinciding with T+T breaking in the area suggests it can actually make a big difference when its working.  Vaccines.  Organised, responsive, trusted teams doing doing data analysis and reporting to cabinet.  

> Perhaps the fallout from Cummings’s testomony today will lead BoJo to think twice about the 21st but that may be wishful thinking on my part.

I wish I knew what the testimony was intended to achieve.  

Post edited at 09:37
1
OP wintertree 27 May 2021
In reply to Punter_Pro:

> There are now four cases of the Indian variant in my town in Suffolk, it's spreading and the third wave has unfortunately probably begun...

Seems to be the consensus.  Other than outbreak areas, it's off to a slow start; unless the exponential rates accelerate a lot more in the next few weeks, ongoing first and second doses are going to keep on taking chunks out of the growth.   Big question marks hang over the wisdom of sticking with June 21st.

> Got my vaccine booked for June the 7th, I am also off to the grit tonight for a couple of days before summer arrives this weekend. 

Time to appreciate the A14 upgrade.  

 Si dH 27 May 2021
In reply to wintertree:

> .  Organised, responsive, trusted teams doing doing data analysis and reporting to cabinet.  

A particularly eye opening bit of Cummings' recent output was his tweet of the whiteboard from 13/03/20. I hope the cabinet and other decision-making rooms have moved on a bit since then.

Tweet number 65 for anyone who hasn't seen it...

https://mobile.twitter.com/Dominic2306/status/1397452170249842691?ref_src=t...

 Tonker 27 May 2021
In reply to Si dH:

Interesting comment just now from Hancock ref hospitalisations in the north west.

He has just said 90% of those admitted had not had 2 jabs. You wonder how many have had 1 jab?

 elsewhere 27 May 2021
In reply to Tonker:

> Interesting comment just now from Hancock ref hospitalisations in the north west.

> He has just said 90% of those admitted had not had 2 jabs. You wonder how many have had 1 jab?

They must have that data may not want to admit that if the import of the India variant has made the 12 week second jab delay* a bad idea. Alternatively they don't want ho undermine confidence in first jab (low protection against India variant) as without a first jab there is no second jab with much higher protection.

*I thought it was/is a good idea and I do not blame the government if they change policy when the virus changes. That's what you should do when the facts change.

Post edited at 11:10
 Tonker 27 May 2021
In reply to elsewhere:

> They must have that data may not want to admit that if the import of the India variant has made the 12 week second jab delay* a bad idea. Alternatively they don't want ho undermine confidence in first jab (low protection against India variant) as without a first jab there is no second jab with much higher protection.

> *I thought it was/is a good idea and I do not blame the government if they change policy when the virus changes. That's what you should do when the facts change.

Yes, I noticed at the weekend the line was the vaccines are effective against the Indian variant with virtually no mention of the fact they are less effective, especially after just one jab.

I get it, they need to keep confidence up as a jab is better than no jab.

1
 Offwidth 27 May 2021
In reply to Tonker:

News of a new type of problem variant from down under:

https://www.theguardian.com/australia-news/2021/may/27/record-pace-of-melbo...

Your point on 'effective' after one dose is based on percentage chance of catching the Indian variant. We don't have reliable data yet on hospitalisations and serious illness/death. The indications of real risk seems to be with respect to potential spread to those not jabbed (or jabbed but very vulnerable) rather than risks to the typical single jabbed individuals.

Post edited at 11:29
1
 Tonker 27 May 2021
In reply to Offwidth:

> News of a new type of problem variant from down under:

> Your point on 'effective' after one dose is based on percentage chance of catching the Indian variant. We don't have reliable data yet on hospitalisations and serious illness/death. The indications of real risk seems to be with respect to potential spread to those not jabbed (or jabbed but very vulnerable) rather than risks to the typical single jabbed individuals.

Yes but if we assume it is just as effective against protecting against hospitalisation if infected then any drop in efficacy against infection will reduce the protection against hospitalisation.

OP wintertree 27 May 2021
In reply to Tonker:

> Yes but if we assume it is just as effective against protecting against hospitalisation if infected then any drop in efficacy against infection will reduce the protection against hospitalisation.

To my limited understanding: I don’t think the two are trivially correlated; protection against infection is a more specific kind of immunity with more specific targets - so the selective pressure it provides results in changes around those specific sites perhaps, where as the protection from serious illness if infected is a broader immune response.

1
 Offwidth 27 May 2021
In reply to Tonker:

Sure but that only matters if the percentage effect is the same. All the evidence so far indicates the vaccines protect much better against hospitalisations than infection after a single jab. We just don't know reliable statistics on by how much. 

This is important for public messaging as the 33% for AZ can worry people needlessly and feed vaccine reluctance.

 Tonker 27 May 2021
In reply to wintertree:

> > Yes but if we assume it is just as effective against protecting against hospitalisation if infected then any drop in efficacy against infection will reduce the protection against hospitalisation.

> To my limited understanding: I don’t think the two are trivially correlated; protection against infection is a more specific kind of immunity with more specific targets - so the selective pressure it provides results in changes around those specific sites perhaps, where as the protection from serious illness if infected is a broader immune response.

As I understand it the two % are used to work out the overall effectiveness against hospitalisation or death.

I sort of worked this out from a PHE document that said protection against death from one dose of i think AZ was 58% if infected but overall was 81%. The protection against symptomatic infection was given as 55%.

The calculation below gives 81% which can't just be a coincidence:

1-(1-0.58)*(1-0.55) then x by 100

OP wintertree 27 May 2021
In reply to Tonker:

> I sort of worked this out from a PHE document that said protection against death from one dose of i think AZ was 58% if infected but overall was 81%. The protection against symptomatic infection was given as 55%.

I think you may have got the wrong end of a stick somewhere?  Easily done with all the different source of information floating about and "efficacy" being used interchangeably to mean several different types of prevention.  The large phase III clinical trials of the AZ vaccine found no hospitalisations or severe cases from 3 weeks after the first dose, so I'm not sure where your 58% came from?  There's data floating around in a PHE document out there showing that the numbers being hospitalised drop off to zero roughly linearly over the ~3 week period after a first dose.

Things might be different with the most vulnerable to have received a vaccine, but the efficacy against severe illness is so high, that any statistics on serious adverse outcomes for people who are infected > 3 weeks after their first dose would be quite suspect, I think.

1
 Tonker 27 May 2021
OP wintertree 27 May 2021
In reply to Tonker:

To quote your link:

Vaccine effectiveness against mortality was based on PHE estimates of effectiveness of vaccination against symptomatic infection (58%) and of death given infection (54%) which combined gives 81% protection against death

That seems like an astoundingly bad methodology to me, one not supported at all by the results of various clinical trials or an understanding of immunology.

It looks to me like some GCSE level solution to a homework problem set by a minister to produce a headline number of lives saved by the vaccine program.  It doesn't look like science in any meaningful capacity.

I'm appalled.  It could serve as a baseline estimate for the minimum bound for the fall in deaths one might expect from vaccination, but that's it.  

Edit:  In the context of what this report seeks to do - produce a number for lives saved by vaccination - it also misses the massive second order effect that the vaccines have taken a big chunk out of R, leading to a faster decay of cases and hospitalisations than would otherwise have been the case.  The number it produces really is a hard minimum bound for lives saved.  I saw this number of the news a while back and thought it sounded like a lowball - I didn't dig in to the source then but I understand why now...

Perhaps I'm missing something, bit this seems like appalling methodology to me.

I absolutely do not think the probability of death after becoming unvaccinated is independent of vaccination status.  That's just indefensible, and flies in the face of data PHE have and have published, and in the face of clinical trial results and in the face of what immunology would suggest.

I think this document was put together for a very specific purpose and I would not suggest using its methods to infer anything about changes to efficacy against serious illness or death of the vaccines against the new India variant.  

I still think you've got the wrong end of the stick, but so has whoever put that PHE document together.  

Minimike, if you're reading, what's your take on this?

Post edited at 12:46
2
 Tonker 27 May 2021
In reply to wintertree:

I'm only going off what they wrote in that document. I'd have expected they knew what they are on about.

OP wintertree 27 May 2021
In reply to Tonker:

> I'm only going off what they wrote in that document.

It's a BS methodology used to produce a BS number for a press release IMO.  I strongly think the method is of absolutely no relevance to understanding the intersection of vaccination, new variants and hospitalisation/death.  Perhaps you could use it to produce an upper bound, but it would be one so far removed from reality as to be pointless.

> I'd have expected they knew what they are on about.

Although I semi regularly get labelled as a conspiracy type, I find it best to take very little at face value and to do some basic cross checking.   You could for example compare the numbers that PHE document gives with those AZ reported in their trial results, for example:

https://www.astrazeneca.com/media-centre/press-releases/2021/covid-19-vacci...

Quote - my emphasis: The primary analysis of the Phase III clinical trials from the UK, Brazil and South Africa, [...] confirmed [the vaccine] is safe and effective at preventing COVID-19, with no severe cases and no hospitalisations, more than 22 days after the first dose.

The two documents clearly can't be squared off.  The later is a press release, but its content falls under medical and financial licensing authorities with severe repercussions for the company if they mislead.  The former document has no accountability.  

Post edited at 13:29
2
 Si dH 27 May 2021
In reply to Tonker:

> Interesting comment just now from Hancock ref hospitalisations in the north west.

> He has just said 90% of those admitted had not had 2 jabs. You wonder how many have had 1 jab?

I don't know, but I would expect a few.

It's worth remembering that even if there is no drop in efficacy with this variant, you would probably still expect to see more double-jabbed 80 yos and more single-jabbed 50 yos going to hospital than you would un-jabbed 30 yos. As good as the efficacy of these vaccines is, being young is still even better protection with this disease. So seeing high proportions of hospitalised people who have had a jab shouldn't be alarming, as long as the absolute numbers don't get too high.

Wintertree - are you able to produce side-by-side comparisons for now and September 2020, of case and hospitalisation rate constants for the areas with lots Indian variant? That would hopefully show us really clearly if cases are accelerating ahead of hospitalisations faster now than they did then, as we would hope. I appreciate it might need another week or two of data but would be really useful.

Post edited at 13:34
 Tonker 27 May 2021
In reply to wintertree:

> It's a BS methodology used to produce a BS number for a press release IMO.  I strongly think the method is of absolutely no relevance to understanding the intersection of vaccination, new variants and hospitalisation/death.  Perhaps you could use it to produce an upper bound, but it would be one so far removed from reality as to be pointless.

> Although I semi regularly get labelled as a conspiracy type, I find it best to take very little at face value and to do some basic cross checking.   You could for example compare the numbers that PHE document gives with those AZ reported in their trial results, for example:

> Quote - my emphasis: The primary analysis of the Phase III clinical trials from the UK, Brazil and South Africa, [...] confirmed [the vaccine] is safe and effective at preventing COVID-19, with no severe cases and no hospitalisations, more than 22 days after the first dose.

Fully aware of that trial data but you don''t think that means the vaccine is 100% effective do you? 

They musty have got that 58% protection against hospitialisation if infected from somewhere though.

TBH if it's bollox so be it. It's not that important.

 Tonker 27 May 2021
In reply to wintertree:

https://khub.net/documents/135939561/430986542/Effectiveness+of+BNT162b2+mR...

I have no intention of reading this (beyond my pay grade ) but it might be of interest to you.

 TomD89 27 May 2021
In reply to wintertree:

> Big question marks hang over the wisdom of sticking with June 21st.

I'd be interested to hear how long people are suggesting we hold off of ending restrictions and to exactly what end. I feel more concrete justification than generalised concern about new variants is required.

 Toerag 27 May 2021
In reply to Si dH:

> Wintertree - are you able to produce side-by-side comparisons for now and September 2020, of case and hospitalisation rate constants for the areas with lots Indian variant? That would hopefully show us really clearly if cases are accelerating ahead of hospitalisations faster now than they did then, as we would hope. I appreciate it might need another week or two of data but would be really useful.

^^ This. Comparisons with the autumn rise in cases will be super interesting, even better if we can correlate levels of restrictions in force at the two times.

OP wintertree 27 May 2021
In reply to Si dH:

I missed your edit until Toerag quoted it.  Probably - it would be useful to have a list of NHS trusts to help me localise admissions figures to the hotspot areas...  

 Offwidth 27 May 2021
In reply to Tonker:

You're totally forgiven. I'm also appalled if that's from PHE (not the first time sadly in this pandemic). Terrible science. We just don't have the data yet but early indications are fairly optimistic (variant information is two weeks old so deaths should be higher if you look at the typical lag time to deaths and apply those numbers).

Post edited at 14:46
OP wintertree 27 May 2021
In reply to Tonker:

> Fully aware of that trial data but you don''t think that means the vaccine is 100% effective do you? 

I didn't claim that, did I?

There is a vast gulf between the trial results and the figure in your PHE link however.  One that sets alarm bells flashing.

> I have no intention of reading this (beyond my pay grade  ) but it might be of interest to you.

This document cited by your PHE link is one you should at least skim read.  An excerpt from each document below, with my emphasis added:

PHE document: Vaccine effectiveness against mortality was based on PHE estimates of effectiveness of vaccination against symptomatic infection (58%) and of death given infection (54%) which combined gives 81% protection against death (3).

Reference (3), the pre-print cited by the PHE document: 19,495 admissions were reported by participating acute hospital trusts, of which 13,907 were included in the analysis. In those aged 80+ years, VE against hospitalisation was 80% (95% CI 74-85%) 28 days after one dose of any vaccine, rising to 92% (87-95%) 14 days after the second dose. 

  • The PHE document is for a fixed time after the first dose (31 days, vs 28 days in the paper).
  • The PHE document is giving a protection rate from  death across all ages as about the same as the protection from hospitalisation from the data driven scientific pre-print is giving for those aged 80+.  They can cite it, but that doesn't explain the gaping discrepancy of giving an all-ages mortality rate the same as the hospitalisation rate for 80+...  

Perhaps I'm missing something really obvious and making a tit out of myself, but reading through the reference hasn't illuminated this for me at all.  

As near as I can tell, the PHE document is a garbage method with data of unclear provenance serving as its inputs.  Where-as the pre-print you then link looks like high quality, reputable science.

(Edit: If it's not clear, my horror here is 100% at the PHE document and 0% at anyone else).

Post edited at 14:54
2
OP wintertree 27 May 2021
In reply to Offwidth:

>  Terrible science.

At least it passed dimensional analysis...

It's always helpful if a student's work fails dimensional analysis, as it's a powerful short cut to outright rejecting it without having to delve in to the details to figure out exactly where the big mistakes are...

1
 Tonker 27 May 2021
In reply to wintertree:

Ok mate, not looking for any kind of argument.

If the PHE document is crap, it's crap.

 Si dH 27 May 2021
In reply to wintertree:

> I missed your edit until Toerag quoted it.  Probably - it would be useful to have a list of NHS trusts to help me localise admissions figures to the hotspot areas...  

Bolton has it's own trust so that will be easy to compare to changes in cases.

https://coronavirus.data.gov.uk/details/healthcare?areaType=nhstrust&ar...

Blackburn is in East Lancashire health trust. That also covers Burnley so you'd need to look at case rate constants for the two together. (Edit - I assume those hospitals also pull people in from Hyndburn, Rossendale and Pendle.)

https://coronavirus.data.gov.uk/details/healthcare?areaType=nhstrust&ar...

Bedford is in the Bedfordshire trust so it'd need comparing with total cases for Bedford and the rest of Bedfordshire.

https://coronavirus.data.gov.uk/details/healthcare?areaType=nhstrust&ar...

If you want to do something simple due to time, you could just look at the data for the entire North West region?

Post edited at 17:01
 Bottom Clinger 27 May 2021
In reply to Si dH:

Quick snippet of info: Bolton’s cases have gone up 1.3%. Neighbouring Wigan has gone up 128.9%. 

This will be our fourth wave (not third as many folk say), hopefully not as big, or of deaths etc. Suits the govt to ignore the fact that vast parts of the UK had a spring, then an autumn, then a winter wave. (I know you know this btw). 

OP wintertree 27 May 2021
In reply to Si dH:

Thanks; it won’t be much work to make up an NHS pseudo-region to go with outbreak areas.

Notable that both acute and ITU occupancy are rising in these; the rise is just poking through at national level.

1
 Tonker 27 May 2021
In reply to Bottom Clinger:

Patients in Bolton hospital up to 49 today. Has doubled in the last 8 days.

In reply to Bottom Clinger:

> This will be our fourth wave

That's my view, too. March rise, September rise, December rise, May rise...?

I notice the BBC page has added a new map showing prevalence of the Indian variant:

https://www.bbc.co.uk/news/uk-51768274

 Si dH 27 May 2021
In reply to captain paranoia:

> That's my view, too. March rise, September rise, December rise, May rise...?

> I notice the BBC page has added a new map showing prevalence of the Indian variant:

Because the sequencing data has variable coverage around the country and a significant lag, you get a better idea of prevalence of the Indian variant in most areas by looking at s gene positivity. It's covered in the latest PHE variant report. The variant is more widespread already than that map suggests.

Post edited at 19:26
In reply to Si dH:

> The variant is more widespread already than that map suggests.

Yeah; I just noticed it's already 12 days out of date...

I just thought it was interesting that the BBC were actually trying to present some data about the Indian variant for mass audience.

Post edited at 19:30
OP wintertree 27 May 2021
In reply to Si dH:

Speaking of which, the latest weekly report has landed.

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/...

  • Figure 1 - The grey shading on the RHS of this figure gives a good indication on how much latency there is in the reporting of sequencing data; perhaps an average of 7 days with a ~uniform distribution over 0 to 14 days.  Also, 60% sequenced - impressive.
  • Figure 2 - Even allowing for the reporting lag, the cumulative cases for B.1.617.2 looks to be dropping below its initial fast exponential.  Hopefully that means the tabulated data on the variants webpage if/when that gets updated later today is going to drop below its 5.1 day doubling time...
  • Figure 3 - it's taking over pretty damned fast.
  • Figure 5 - this is why NZ quarantines travellers - look how one curve leads the other.
  • Figure 17 - ability to sample the SGTF is poor in the North East.  
  • "Live virus and pseudovirus neutralization" and "Monitoring of vaccine effectiveness" - basically the same news as last week on reduced vaccine efficacy against transmission but with tighter CIs.  Well in to the "warning shot - we should not have let this variant get loose but it could have been worse" territory IMO.  

Edit:  Shout out to whoever is making the effort to dislike all my posts.  Haven't seen you in a couple of months.

Post edited at 19:43
1
OP wintertree 27 May 2021
In reply to wintertree:

> Hopefully that means the tabulated data on the variants webpage if/when that gets updated later today is going to drop below its 5.1 day doubling time...

Updated variants plot below.  The latest update for the Indian variant has fallen well below a 5.1 day doubling time trendline; the doubling time measured over the whole dataset is now 6.1 days; the change over the last week represents a doubling time closer to 7 days.  Even through we're pretty sure there's a lot of lag between the pillar 1 and 2 cases and these numbers, it's good to see the initial fast rise starting to slack off here too.  7 days isn't great, but I hope there's going to be another slackening off next week.

Phew.

On the good news front, the Manaus variant is really dropping off it's already much slower exponential curve.

https://www.gov.uk/government/publications/covid-19-variants-genomically-co...

Post edited at 19:56

1
 Misha 27 May 2021
In reply to tom_in_edinburgh:

Agree. I think an uptick was inevitable with further opening up, although the 12 April reopening didn’t seem to have any significant impact (broadly equivalent to level 3 in Scotland I think?). A ripple is ok but we’re going to get a wave I suspect but hopefully not a tsunami given the vaccines drive.

In reply to wintertree:

Tables 10 & 14 are interesting with the %age of cases who have 'travelled' (I'm not sure how far they have travelled)

Figure 10 is a bit worrying with the sudden rise in East England figures.

Figure 11 is disappointing in the number of cases still 'under investigation'; track & trace is still pretty poor, only managing to trace ~50% of cases within two weeks.

Post edited at 22:06
OP wintertree 27 May 2021
In reply to Tonker:

> Ok mate, not looking for any kind of argument.

Nor am I looking for an argument.  

I didn't think your estimate was a sensible method, and so I said so.   

I didn't think the source for your estimate (the PHE document) was a sensible method, and I explained why.   I noted that the phase 3 trial results were so different from that source that it sets alarm bells ringing.

You then gave me a link to the primary source referenced by your source - although you didn't explain why - and I explained why I did not think the content of that link related to the PHE document in any meaningful way, and how it further showed the PHE document to be frankly bizarre.

> If the PHE document is crap, it's crap.

That's my take on it.  Data of unknown providence being fed in to a grossly inappropriate method.  

The only thing that has any validity is the longitudinal data that will eventually emerge from studies of people infected with the particular variant.  Knowing the change in efficacy of the vaccine against infection by the variant clearly offers a bound on that, but it can't be used beyond that IMO.

3
OP wintertree 27 May 2021
In reply to captain paranoia:

> Figure 10 is a bit worrying with the sudden rise in East England figures.

Yowsers.

> Figure 11 is disappointing in the number of cases still 'under investigation'; track & trace is still pretty poor, only managing to trace ~50% of cases within two weeks.

I’ve pogod on interpreting that.  This is “by specimen date”  and we know the sequencing data lags that a fair bit, so many of the grey cases may only just have been assigned some sort of priority status.

But, you’d hope the initiation side of tracing - speaking with the initially detected case - would be happening far sooner than two weeks later, regardless of priority status.  I mean, if it hasn’t happened after two weeks...

I have this vague notion that it should be possible to determine travel status immediately from an NHS number cross referenced to the persons identity and so to state travel records.   I don’t know what to infer from us not doing that; felt to be a step too far against civil liberties, or is big brother a bit less organised than we thought?

Post edited at 22:31
1
 Misha 27 May 2021
In reply to Si dH:

That’s a fair point about us being in a good place in the scheme of things. The concern is that it will get significantly worse, especially if the Indian variant proves to be significantly more transmissible. I wouldn’t be as concerned if it weren’t for this variant. 

 Misha 27 May 2021
In reply to wintertree:

Re LFDs, they’ve been in the stats for a while now and I wonder whether it’s more appropriate to look at the total numbers now. I suppose the big question is whether there’s much transmission happening from school aged children to their families. Someone must have a decent idea of that. It seems that since March this transmission has been limited, in that cases in the under 20s have gone up more than in the 30s to 50s.

A vaccine resistant variant would be a significant concern if cases increase to the tens of thousands say (appreciate that’s pretty vague). As you say, it might be imported one day anyway but no need to fuel the fire at home. We’ve already given the world one nasty variant.

As I’ve been saying for a while, I’d have managed the unlock on the 17th differently to try to figure out what’s going on first… I just hope they see sense and rethink the 21st. 

OP wintertree 28 May 2021
In reply to Misha:

> Re LFDs, they’ve been in the stats for a while now and I wonder whether it’s more appropriate to look at the total numbers now.

The number of LFDs done varies a lot over time, both in relation to schools and in general.  They’re in a downwards trend recently I think despite cases rising.  Their contribution to controlling cases is a lot more useful than their inclusion in most of the data IMO; I think the PCR data is clearer and more translatable to other measures; but it’s only available at the England level AFAIK.

> I suppose the big question is whether there’s much transmission happening from school aged children to their families. Someone must have a decent idea of that.

Well, perhaps PHE will be allowed soon to publish what they know...

> As I’ve been saying for a while, I’d have managed the unlock on the 17th differently to try to figure out what’s going on first… I just hope they see sense and rethink the 21st. 

There was some softening up stories in the news yesterday for the 21st.  We’ll see - there’ll be a much better idea of the situation by then; not perhaps for us, that depends on what longitudinally derived data including vaccine status gets published by SAGE.

 Tonker 28 May 2021
In reply to wintertree:

PHE saying 3% of those infected with the Indian variant have been double jabbed, 24% had one jab and 73% unvaccinated.

 Si dH 28 May 2021
In reply to wintertree:

I think you can download separate PCR and LFT data for every local authority (at least in England.) They just don't present it by default on the dashboard, but I think it's there (it's certainly available in the drop-down list anyway...)

OP wintertree 28 May 2021
In reply to Si dH:

> I think you can download separate PCR and LFT data for every local authority (at least in England.) They just don't present it by default on the dashboard, but I think it's there (it's certainly available in the drop-down list anyway...)

I got (very) excited there for a moment until I tried it.  It's possible to create every permutation of options in the downloads page, but many of them return blank files - and as far as I can tell, the [new/cum]CasesPCROnlyBySpecimenDate are still returning no data at the regional and UTLA level.

Shame...

 Si dH 28 May 2021
In reply to wintertree:

Doh. Ok well worth a try!

OP wintertree 28 May 2021
In reply to wintertree:

> "Live virus and pseudovirus neutralization" and "Monitoring of vaccine effectiveness" - basically the same news as last week on reduced vaccine efficacy against transmission but with tighter CIs.  Well in to the "warning shot - we should not have let this variant get loose but it could have been worse" territory IMO.  

A noddy plot of the data from table 12 in the report.  The blue squares represent, by area, the fraction of people protected from either variant by each of 1 and 2 doses, and the grey boxes are the confidence intervals (range unspecified) from the table.

Given where we are with first and second doses, this is definitely going to put a chunk back on R for at least the next two months, as many people have not had their second dose or had time for it to take full effect.  The people with one dose at least 3 weeks should reasonably be protected from hospitalisation and quite protected from the worst effects, and given the published vaccine uptake rates there aren't that many people vulnerable by age who  unvaccinated to cause serious hospitalisation problems in big picture terms amidst what increasingly feels like an inevitable - but hopefully slow - next wave.  There's still lots of time for people in older age ranges to get vaccinated, and the sense I get from the last few weeks of news reports is that recent events are proving a useful encouragement; I've not seen hard data on this though.

I do wonder how forwards looking this data is though, as much of it must derive from the outbreak areas, where there are likely a couple of confounding factors.

Post edited at 14:48

In reply to wintertree:

I did chuckle ironically this morning, when the BBC headlines were going "shock horror: Indian variant cases doubling in a week!". Err... yeah, it's been doubling faster than that for the last four weeks; have you only just noticed...?. It's actually slowing down now...

Just shows how spoiled we have been with wintertree's analyses...

 Si dH 29 May 2021
In reply to captain paranoia:

Impact of local rises breaking in to rises in the national level figures = people down south thinking "oh shit, it's not just Bolton" = newspaper headlines.

Rises are starting to get more noticeable over most of Lancs and GM now, so I think the national level figures are in to a sustained rise. The leading edge data looks bad in a lot of places.

Post edited at 06:43
 Tonker 29 May 2021
In reply to Si dH:

> Impact of local rises breaking in to rises in the national level figures = people down south thinking "oh shit, it's not just Bolton" = newspaper headlines.

> Rises are starting to get more noticeable over most of Lancs and GM now, so I think the national level figures are in to a sustained rise. The leading edge data looks bad in a lot of places.

The significant rises have bled into western West Yorkshire now. Both Bradford and Kirklees have seen increases >50% over the 7 days to the 23rd May. 

 Toerag 29 May 2021
In reply to Si dH:

> Impact of local rises breaking in to rises in the national level figures = people down south thinking "oh shit, it's not just Bolton" = newspaper headlines.

> Rises are starting to get more noticeable over most of Lancs and GM now, so I think the national level figures are in to a sustained rise. The leading edge data looks bad in a lot of places.


Hardly surprising, if you let it the virus have an opportunity it does it's stuff. It's just a matter of how quickly now.

 Šljiva 29 May 2021
In reply to Toerag:

The bastard child? Vietnam health minister Nguyen Thanh Long said on Saturday the country has detected a new variant of the coronavirus, which is a mix of the India and UK Covid-19 variants and spreads quickly by air.


New Topic
This topic has been archived, and won't accept reply postings.
Loading Notifications...