UKC

Worth a read for those interested - a lot of detail in here.

https://cmmid.github.io/topics/covid19/reports/uk-novel-variant/2020_12_23_...

Table 1 is the key summary of predictions - basically it’s looking every bit as bad as various people on here have wondered, and the only real answer appears to be a massively accelerated vaccine program.

Its early days and I expect their modelling will be updated as the next week or two of data rolls out.

A phenomenal amount of work pulled together so quickly.  

At least there are the vaccines (assuming, as is believed, they work).

Can you imagine how much more frightening it would have been to be there in March?

(This of course holds lessons for future planning - we were in essence lucky).

Figure 2 in the paper gives strong support to this variant having increased transmissibility rather than reduced immunity.  That is a big relief.  This comes from analysing the progress of the different strains to date.  Porton Down will hopefully soon be reporting up the chain the results of their lab based tests of immunity.  Hopefully these align with the findings in this preprint.

As the preprint notes, increased transmissivity is in agreement with the reduced “Ct” value, something a couple of posters here also noted from the NERVTAG minutes last week.  (Lower ct meaning samples are detected in PCR with less amplification, provisionally suggesting a higher shedding of virus particles by an infected person).

I’d rather not.  I’ve found the last week distressing enough, and that’s after 9 months of adapting to everything.  Let us hope permanent lessons are learnt.  Some of the lessons learnt after SARS were lost before this outbreak.

Those graphs are pretty grim unless the vaccine roll out gets a shot in the arm...

Ah, yes: I'm sure this government will learn those lessons, seeing how much they learned from the first wave about the need to lockdown early and hard...

Thanks very much for posting the link. This was the main take-home for me:

"The most stringent intervention scenario with Tier 4 England-wide and schools closed during January, and 2 million individuals vaccinated per week, is the only scenario we considered which reduces peak ICU burden below the levels seen during the first wave."

Also given that increased transmissibility fits the data the best, I'll be wearing a mask in all contact situations from now on, including outside. Brother in Australia commented on this recently - apparently mask wearing everywhere outside the home is de rigeur.

I actually feel slightly reassured looking at that....but this all hinges on a mass roll out of the Oxford vaccine in the New Year.

Clearly London, the South East and East are in trouble whatever happens but the rest of the country looks like it can avoid a disaster in the next month or so.

Obviously we need the vaccine programme to be ramped up. I wish they had modelled 1m a week rather than 200k as if the Oxford vaccine gets approved we will clearly be doing more than 200k but 2m a week seems very ambitious.

I have two questions if anyone can give their view:

1. It states it assumes the vaccine will protect 95% against disease. Are they assumung this means it will stop 95% of people going to hospital i.e. 5% will be admitted. I ask because the efficacy of the Oxford was ony 62% on the double dose but I believe no one was hospitalised. Clearly the main thing we need in the immediate term is stopping people going to hospital.

2. On Table 1 it give weeks in Tier 3 and 2 after coming out of Tier 4. On the 2m/week vax programmme this suggests we come out of Tier 2 in Mid March.....to what? Surely it could not be no restrictions could it, this would seem far too optimistic to me.

Thanks, this is grim reading but only what we anticipated I think. It seems the debate has moved on and the increased transmissibility is now becoming accepted. Impressive that you predicted this a full 2 weeks ago from the raw data.

From my position in the health service in Yorkshire it feels like we are in an island of false calm with fires burning all around right now.

What strikes me is that the modelling requires the novel variant to fit mortality in the SE already, indicating that the proportion of novel variant has been high for a while. It all fits together I suppose but first time I’ve seen it translating into deaths.

Is cigarette smoke a useful/realistic indicator of the spread of exhaled breath? For example on the bike this morning I could smell smoke from someone walking a good 25m away, carried on a gentle breeze on a frosty morning. Yesterday, in the car in traffic, I could smell e-cig vapour from a car four cars in front. 

​​Those smells emerged from their lungs, so I would assume at the very least I'm getting their exhaust CO2 etc. Would the virus be transmitted a similar distance? An interesting report lately from Australia shows bushfire smoke columns are transporters of microbes/fungi. And the e-cig 'smoke' is actually a vapour and in my mind more similar to normally exhaled humid air, which is what we wear masks to reduce the spread of. 

Ignoring for the moment if the virus is actually carried in smoke and/or vapour, is the spread/dissipation of these media a realistic visual/olfactory indicator of the spread/dissipation of normally exhaled breath? 

Trying to take some comfort from that:

- at least there is a potential way through that would contain it. And I still feel like March's lockdown was harder than Tier 4 with schools out, which means there is potential to turn the screw tighter (i.e. the scenario they tested isn't the most stringent we could potentially do)

- as far as I recall none (or very few) of the Nightingales got used in wave 1. That means "first wave ICU burden" isn't quite the cliff edge. It may not be far from it, but there is some room, however small, for the government to screw things up again.

This is not in any sense to try to underplay the difficulties ahead, but whilst it is a fairly narrow and hard path it is some comfort to know that there is one.

Worth saying that looking at their projections without vaccination tier 4 plus schools does not give a big benefit over tier 4 alone - and they have not modelled ( or reported) tier 4 alone with vaccination.

Apologies, I didn't see this before starting my other thread.

Apart from what people have already said, what struck me the most was that the effect of the most helpful measures they modelled (Tier 4 everywhere + 2m vaccine per week + schools closed in January) actually made the most difference to deathsn in the rest of the country outside London/SE/East. Essentially the model predicts in Figure 4 and 5 that it is already too late for further measures or accelerated vaccinations to really change the dial much in London, they could make a small difference in the SE or East, but make a huge difference in the rest of the country. Yet current policy is to leave the rest of the country outside Tier 4 which is likely to lead to this opportunity being lost. This constitutes the most powerful demonstration I have seen of the reason for the entire country to enter Tier 4 straight away.

Does anyone disagree with this interpretation? It's not a conclusion that the paper explicitly draws itself.

This was the most obvious conclusion I drew from it......that and if we can jab up 2m people a week we can get out of this mess quite quickly.

I don't really get that from figure 4 (edit: figure 4 does set the scene in that vaccination is the only solution short on an un-modelled much harder lockdown and near elimination), but I do from figure 5 by eyeballing the area under the various curves.  It looks like the 2m/week vaccine roll out roughly halves the death toll in the "red" areas from my plots and has a much larger effect for the other regions.  I think perhaps the new strain has risen to prominence faster than expected in the South West - inferred from the recent change in doubling times for cases there -  so that's one part to watch if they update their preprint with more data.

Given that the effect of a progressive vaccine roll out accrues over time, and that this strain is already dominant in London and the South East, there's a simple and unhappy interpretation for why it's too late to make a significant difference there compared to regions where the new strain is still in earlier stages of exponential growth.  

My best estimate is on a Boxing Day announcement of full lockdown, approval of the AZ vaccine and immediate roll out assisted by the army and army medics doing inoculations.  

I would much rather things had gone in to full lockdown the moment this came together, and I think that allowing the mass exodus from London on the night of the press conference may shorten the timelines for the "blue" regions from my plot to fall; I don't think the effects of this exodus will have percolated into the LSHTM analysis yet.  Let's hope I'm overly pessimistic on this.

, and the only real answer appears to be a massively accelerated vaccine program.

In which case I'd be worried.  Very worried. 

Am I misreading this - it suggests that with Tier 4 and school open and no vaccination the estimated peak ICU is 113% of 1st wave (with a 90-151%CI), that really doesn't seem too bad considering what I am currently seeing in the numbers.  

Thanks.  I tried to get a write up of it to places it could make a difference at the time; now it's moot as the genomic and healthcare data unambiguously speaks for itself as shown by actual medical scientists in the LSHTM pre-print.

I still think your plot 18 rate constant methodology is particularly good and allowed you to find the turnover point and regional variation faster than approaches based on the absolute numbers.

It's an awful lot worse than what I was expecting when case numbers started to plunge in November.  

The total deaths given for this scenario are far worse than those to date, and a quite set of estimates (e.g. eyeballing graphs, dividing total deaths by peak deaths) suggests hospitals will be running beyond the 1st wave capacity for at least 3 months in that scenario. 

Did you see how asymmetric the confidence interval is?  They're - by necessity - projecting the current early situation forwards in a largely exponential situation, and the uncertainties on that are significant.  I'm worried that the South West may be ahead of their current curves, with developments presumably since this data came together.  

I agree though that if we look at the growth in case numbers for e.g. London right now the table could be worse.  

Edit:  Also, the new strain is everywhere including T1/2/3 regions and is almost certainly growing exponentially in all of those regions right now.   

Yes, I'm still trying to get that approach to some people.  I think that normally people turn this value in to an R number which introduces large errorbars.  The origin of those errorbars is in the difficulties of relating R to case growth, not in the measurement data itself, and their presences weakens the interpretation of the data.  I also think that as a direct measurable of infections (random sampling), cases, hospitalisations and deaths it provides a unique way of directly relating the effects of policy and behaviour through the data.  In many ways the direct measurables are of more immediate concern as well.  On the down side, it is harder to clearly explain to, and to interpret by, people from broader backgrounds.

I'm always hesitant to wade in to an established field and say "You should try this instead" because it's a sure-fire way to get yourself labelled as a know-it-all idiot and ignored...  My general experience is that it takes a couple of years of collaboration and working together to get to that point.  

At the moment it looks like England is vaccinating about 250000 people a week.

https://www.england.nhs.uk/statistics/wp-content/uploads/sites/2/2020/12/CO...

Broadly speaking we need to increase that by an order of magnitude. *IF* the vaccines can be produced in sufficient quantities then that sounds difficult but not impossible to me. I also wonder if efforts at mass vaccination should give the SE of England priority.

I've been pretty unhappy for the last couple of weeks.  I'm less worried now because as AJM said up-thread "but whilst it is a fairly narrow and hard path it is some comfort to know that there is one."  The other alternatives were bleaker.  

Yup, but maybe it’ll be useful for the next time.. but I must say it looks like a physicists approach to me (which is probably why I like it) and I’ve had ‘fun’ trying to introduce similar things to medical problems in the past. If you struggle at peer review just stick a (meaningless) p-value on it.. but make sure it’s between 0.05 and 0.005!

edit: I once got a paper rejected for saying p<10-6 (it was valid, but the reviewer didn’t ‘believe in p values less than 0.005’)

It seems the UK production of the Oxford/AstroZeneca vaccine has suffered a setback; where we were set to have 30m locally made doses by year-end, this article suggests it's more like 4m imported doses. 

https://www.ft.com/content/651be7e7-2a4e-410f-8089-b4b7e887f6e8

Hopefully money and resources are now being thrown at the production problems in as large a quantity as is useful.

That doesn't seem unreasonable with the Oxford/AZ vaccine if it's available in sufficient quantity.  It needs a fridge not a -80 freezer.  The Royal Army Medical Crops have about 30 regular and reserve brigades.  (Edit: about 30 units, not brigades....!!)

One take of the pre-print is that more lives may be saved by focusing on areas were the new variant is not yet prevalent, and that may also be a better way to preserve net hospital capacity across the country.  I think there will be at least another week to gather more data and to make that decision, and I am very grateful I do not have to make it.

Yes, I've been making 6-sigma detections on living systems and I wish I'd had your advice to stick with p < 0.005.  It would have made life easier.  p<1e-6 certainly raises questions about the role of unidentified variables and the relevance of a p-value to interpreting the result.

That seemed like a lot, and a quick perusal of https://en.wikipedia.org/wiki/Royal_Army_Medical_Corps#Brigades suggests its one brigade with two or three regular field hospitals and ten reserve field hospitals. No idea what the medically-trained personnel strength is.

Sorry, I meant units and not brigades!  30 brigades would be totally bonkers.  I can’t find the official page I counted from; but this FOIA response is closer to 25 - https://assets.publishing.service.gov.uk/government/uploads/system/uploads/...

That and how quickly they can train more inncoualators are the key questions.

The medically trained reserve  personnel may well be NHS staff already - best not to double count!

That's a very good point - thanks.  Hopefully there's some former paramedics, former nurses and vets in there too.  I haven't found a good breakdown of the staffing.

This [1] PHE document suggests training takes two days. As an indication of cost, Immunisations UK do a two day New Immunisers’ Foundation course for £250+vat [2]. Seems like it might scale as part of a NHS+military programme.

Also, many pharmacies already have people trained to administer flu inoculations. And there are vets too.

[1] https://www.gov.uk/government/publications/national-minimum-standards-and-c...

[2] https://immunisationsuk.co.uk/training/

Thanks.  That’s reassuring. I’ll feel a lot happier when it’s not just speculation and some news is announced.  It does seem like this is possible - once the AZ production is sorted out (the news story I posted above is a few weeks or so perhaps it already is).

Legally, March basically was Tier 4 with schools out.  However, a very large number of businesses closed voluntarily, most notably the likes of McDonald's, because they felt they couldn't keep their staff safe.  We could certainly legislate to create the same conditions.  One way would be that only outlets that sell predominantly non-VATable items would be able to open - basically supermarkets and other non-prepared food shops.

For those who don't pay for FT access:

https://www.bloomberg.com/news/articles/2020-11-04/delays-hit-oxford-vaccin...

An observation. Hospital numbers are now similar to wave1 peak and have been for a while. Deaths are about 40% lower than w1.

Why?

Primarily, dexamethasone.

The slight "downside" of that is that it means hospitals will be fuller because people aren't dying as quickly.

Um.. hopefully the people in question leave for happier reasons. But yes average occupancy time is probably higher. I struggle to see this as a downside!

More good news on immunity..

https://www.medrxiv.org/content/10.1101/2020.11.18.20234369v1

I know some places maybe went further through lack of preparedness/interpretation/pressure/etc, but I have to confess I also thought it was legally tighter (was working in homes still allowed in March? I didn't think it was although I'm quite prepared to be wrong). Anyway, as you say, in practice you could legislate to recreate those conditions - it doesn't really matter how they arrived in some respects as long as they are replicable...

It's a "downside" in quotes - it causes overwhelming to be reached sooner, though for good reasons as you say.

I assume that dexamethasone and learning when it’s better to use CPAP style oxygen delivery than ventilation are the main reasons why number of patients in ventilation beds is about half the number it was when hospitals were similarly full in March?

I don’t know how well “Peak ICU” in the preprint corresponds to ventilation beds, but if it was a 1:1 correspondence, their “116% Peak ITU” noted by ABlackett up thread would correspond to about a “230% peak hospital occupancy”.  If my understanding is anywhere near reality, that’s not good because we don’t have 230% of the hospital capacity we did in March.  

Projecting the recent short term forwards and looking at their most optimistic scenario’s peak deaths, I don’t look forwards to the next update to their modelling.

I’m hoping I’m quite thoroughly wrong.

Yes, it’s not just dex, but a lot of it is.

you’re right re: the projections, the nightingales help in principle but thanks to (politics warning) years of cuts there’s no one to staff them.

True on staffing - but if you plonk the folk who need moderate care levels of care rather than the genuine ICU candidates then the staffing cost is not as bad, and can be done by a wider range of folk. 
 

It was being done on a trust by trust basis rather than nationally, but some prep was made in the spring/summer to recruit and train a fair number of “respiratory care assistants” who were going to be deployed in Nightingales to support Normal staff.

It took about a couple of hours to teach me and several others when I was in the RN.  

However:- there are many thousands of farmers who regularily inject their own stock.  A local dairy and beef farmer I've worked for has probably injected far more cattle than some GPs have injected humans.  Put them in s suit and they'd look just the part and no mucking about either.

Thanks for sharing. As you say, the paper paints a bleak picture. I’m not familiar with the Bayesian modelling used but how are these types of models validated? The authors give the deviance  Information Criterion but the values don’t really mean a lot to me. As mentioned on the Reddit sub for the paper the data they fit to looks very noisy.

I note that the authors “assumed that vaccine protection was conferred immediately upon receipt of one vaccine dose” when 2 doses are required for the BioNTech vaccine. Also, is the assumption that of the paper that the vaccine will be administered indiscriminately? How does the fact that the vaccine is targeted at high risk people first affect the results?

All vey good questions and I'm afraid I'm not qualified to answer most of them well. I'm definitely not a Bayesian person.

The model they use looks to be a very standard, previously published model of two competing infections.

Me neither.  The CIs look surprisingly small to me, I wondered if the analysis was over-powered, but that's a first take and not a robust academic criticism!  Or even a poor academic criticism.

The noise doesn't bother me - they are fitting to multiple (many) independent data sets each with uncorrelated noise, and there are many more data points any one of the datasets than free parameters in their model.  This looks like a very robust approach to me.  My noddy analysis of just cases data shows the effect of this new strain; they are using something like 8 separate datasets.  

I think it's likely to be the Oxford/AZ vaccine shortly rolled out not BioNTech if this paper pans out.  I think this is also 2-dose but both will start protecting people some time after the first dose.  But it does look like an optimistic assumption on their behalf. 

Aged ranked decreasing order up to a 85% threshold in each cohort - about herd immunity threshold for the new variant based on their estimate of increased transmissibility I assume.

Vaccines are given first to 70+ year olds until 85% coverage is reached in this age group, then to 60+ year olds until 85% coverage is reached in this age group, continuing into younger age groups in 10-year decrements

Thanks for your responses wintertree

Isn't the efficacy of the Oxford/AZ vaccine 70%? I think the authors assume 95%. Is there any good data on the efficacy against infection of this and other vaccines? I'm wondering how accurate the author's 60% assumption is.

That'll teach me for not reading the paper fully  

Regardless, things dont look good. Like others I find it odd that the government hasn't locked the entire country down yet. I'm also surprised that the impact this will have on the economy hasn't been priced into the FTSE yet. Hopefully I'm wrong. 

Boxing Day announcement - possibly. I reckon near certain by the end of the year. One thing Hancock said yesterday which didn’t really get picked up on in the media is he hoped this would be the last time he’d be doing a press conference before Christmas. He didn’t say New Year. 

There were no hospitalisations in the vaccine arm of the Oxford stage 3 trial which is the main thing when looking at the impact on health care pressure.....however the data was not sufficient in those aged >55 (only 12% of the arm were in this age group) ......new data was submitted to the MHRA on Monday and the noises coming out from those close to the project suggest it is good. 

To be fair they seem to have managed 500k in a week with a limited roll out. Seems that 1m might be manageable with the AZ one on stream in the new year. Don’t know about 2m.

But.  500k in one week?  We have approx,  60,000,000 population, but for the sake of practicality we say only 30,000,000 (The oldest, most vulnerable etc),  get the vaccine.  Thats half the population done in just over a  year. But we/they need two injections a month apart -  So 15,000.000 done and dusted in the first year?     Perhaps my maths isn't so hot.  Fire away!

Our local  vaccination centre is at a building on a pretty steep coastal cliff.  The arrivals arrive how ever they want at the top car park.  They then walk down some 60ft of steep stairs.  To return to their car or mobility scooter, they must leave via the lower entrance and walk the 60ft back to the entrance via a pretty steep drive way.  Oh, and there's no lift access.  I can't see how many of the older, less mobile will manage getting down the steps, let alone back to the top.  At the moment its blowing a gale, directly off the North Sea and its only a degree or two above freezing.

I hope you are right in things will get better, and that we'll not be discussing this all like Brexit, 4 years after the event and wondering why people still haven't been vaccinated and why people are still dying.

Its going to be next Christmas before we know it!!!

Forgive my negativity, but its a pretty depressing Xmas for me this year.

I think the final efficacy of the Oxford vaccine isn’t settled - at least not yet publicly.  The initial findings had two different values but without clear blue water - statistically speaking - between them.

I’m starting to wonder if I’ve gone a bit batty and am imagining it all.  It’s hard to tell anything from today’s data update as it’s mainly reporting on a weekend with sampling lull.  Tomorrow’s may shed some light for England, other nations aren’t reporting over Christmas.  I think there’s going to be an unusually large amount of sample lag from postal samples as well with bank holidays bracketing the coming weekend.  Today’s ONS update was the third in a row not to include an estimate of the daily infection rate.  So a frustrating stagnation ahead in the data perhaps.

If we can do 200-300k COVID tests a day, I can’t see 2m day or many more vaccinations really being a problem with the will.  I suggested a while ago, DIY injections would work only to be shot down by those claiming giving an injection a requires being a paramedic. I wonder if they might re-think this now.

Availability of vaccine may be an issue of course.

Seeing as my plotting thread filled up and got archived, I'll post this here.

It's a plot of the demographic breakdown of cases for all London UTLAs to have risen above their level on Nov 17th.  The left side is cases/day for each age bin, and the right side is the doubling time for each age bin.

The more positive the exponential constant, the faster cases are rising.  I make it rising fastest in the 25-30 age bin, with a doubling time of 3.8 days.  The second plot shows the exponential rates and doubling times for the final day in each age bin in the first plot.

As always this is the leading edge of the data where things are most sensitive to noise.

I know what you mean. Time will tell I guess. I feel similar to how I felt when I first read the imperial paper back in March where I went overnight from "yeah, yeah whatever" to "this is obviously extremely serious. Why isn't it being treated as more of a big deal". 

I agree, 500k a week isn’t actually that much but it’s a good start. Question how quickly they can scale this up. You’re right that walk / drive in centres aren’t for everyone. The very old and disabled will need home visits, the not very mobile (old and/or without access to a car) will need to go to local GPs / pharmacies.

Case numbers seemed to level off around 35k last week before the weekend. Whether that’s temporary is another question. As you say, Xmas will cause havoc as far as the stats are concerned. It might look ok-ish and then really bad - perfect excuse to bring in a nationwide lockdown...

How do you vaccinate 2 million people a week without having queues of people.  It isn't just the shot itself its also waiting 10 minutes after you get it in case there's a reaction.  If the vaccine takes a few weeks to be effective there needs to be precautions against catching it when you go.

You have queues of socially distanced people then....make use of sport stadiums etc where appropriate. 

We sent a man to the moon 51 years ago....i am sure jabbing 300k people a day is possible. 

It's about the right time for the effects of Tier 4 to start kicking in, and the effects of the schools being out will show up soon too.

As long as they are queueing 2m apart outdoors with masks on that isn't a major problem.

You won't see Tier 4 (19th Dec) in the weekly-average data until new year. Putting aside the unknown effect of Christmas reporting, you might see any effect of Tier 3 (,16th Dec I think?) before that, but not much. ,(Based on experience of timescales from the beginning of lockdown 2.)

Happy Christmas everyone!

Ah, sorry, I thought you were referring to daily data (the incubation period being about a week meaning that measures take about a week to show there).

That alone is a near guarantee that it is exactly what is not going to happen, judging on past performance...

In practice it has always previously taken slightly longer than that to see any noticeable change in the (complete) daily data (I wrote about this on another thread but I think it was 10-12 days.) ThisFurthermore though, at that point it still looks like noise. Unless Wintertree thinks his algorithm can improve the understanding of the leading edge then you really need to see the weekly average drop to have any confidence in a downturn. This means just over 2 weeks from a change in policy. I don't remember any instances of that duration being lower.

Queueing outside in the cold isn't great for many people's health,  people aren't disciplined enough to always keep 2m apart if they are in a queue for ages, 2m is a tradeoff, it isn't magic and it was based on experiments with the older, less infectious, strain.  If you multiply a small risk by a very large number of people you get a substantial risk, it won't all be outside, presumably these queues are going to end up with going into a building where the vaccination happens and they need people to sit down somewhere for 10 minutes after to see if they get an adverse reaction.

Maybe drive through is an option.

Drive through is a good idea (you could close roads for the queue if needs be) but an awful lot of the "target market" (to start with) doesn't drive.

You can never contrive to catch everyone. If you can get the majority (who probably can, if we're honest, be put in a car and driven) you're winning.

What I'm not understanding is why we aren't going ahead at risk now with the AZD1222 deployment. Safety data isn't going to get any better is it?

And efficacy/effectiveness is 'good enough to make a difference', yet it's still worth pondering while the next day adds another today's x e^R/t cases and guarantees today's x e^R/t deaths? What gives?

I'm sure the MHRA, JCVI, COBR (and, praise be, the great prophet Tony Blair, who conceiveth of concepts so blindingly simple that no mortal could come up with before he), have had a good think on it and concluded best not to pull the trigger just yet, but why? Rules is rules but couldn't we ask for volunteers for a 5m+ participant clinical trial?

You are assuming that the vaccine is indeed safe and is indeed at least 60% effective (I assume, as that is what AZ have claimed), but that is what the MHRA are checking.

I don't know anything about how the MHRA works specifically but as a regulator in another high risk industry I could perhaps pass some comment. As a regulator they are accountable to the public for ensuring the vaccine has been developed and tested in a way that adequately proves it is safe and effective. This is critical both to minimise the risks of adverse events but also to provide public confidence. They do not have accountability for delivery of the vaccine - this is a key principle of being an independent regulator. Because the actual provider of the vaccine is responsible for safety and efficacy but also for delivery, costs and making a profit, an independent regulator is required to provide assurance that the safety and efficacy aspects are dealt with adequately and not de-prioritised. As a result of all this, it is an important principle of regulation that the regulator is not rushed to meet the programme of a manufacturer. I'm sure their usual modus operandi will be to do what they feel needs to be done to meet their objectives and to resist any programme pressure. Of course, with the timescales of the pandemic and public health benefits of getting the vaccine out quickly, there is a huge amount of pressure on them. Their jobs must feel extremely difficult right now if they feel there are any difficult judgements to make about the data provided on efficacy or safety, and are feeling like they need to consider public health benefits of a faster decision into that equation. To be fair the Government have on multiple occasions made clear that the MHRA should not be rushed, which will no doubt give them some reassurance, but I am sure it will still play into their decisions.

It is also possible that they might reconsider recommendations they make for how the vaccine is used in light of this new strain appearing; if so that will also take extra time.

Obviously as a member of the public I'm hoping this vaccine will be available ASAP. But imagine the consequences if it comes out, the Govt predicate their pandemic response on its delivery and then within a few weeks it turns out not to be safe or effective after all...it would derail vaccine efforts in the country for years to come. So they have to be very confident.

For what it's worth, I went back and noted that the period from final data submittal to approval for Pfizer was 9 days. If the AZ vaccine takes the same period of time then that would be approval on New Year's Day. Perhaps add a couple of days for bank holidays? I'd be surprised if anything came quicker than that as it sounds like this is a more complex case.

Fully appreciate all that. Wasn't suggesting the MHRA would be rushed, more that we could get started at risk, even if only with a massively expanded and properly consented clinical trial. 

Stage 2 trials are supposed to be the safety ones, and I thought they concluded ages ago? Seems like the worst outcome from forging ahead would be if it turns out it's less effective than it was efficacious, in which case we wouldn't have lost much. 

I do agree with that - if it's already considered safe, it would seem worth trying it very widely.  At worst it's a bit of a waste of money, and money has certainly been wasted on far worse things during the pandemic.

I was quite shocked when I read it wouldn't be until April/May when it would be more widely available (even as a wider trial or emergency authorisation) despite its potential to save a lot of lives if it does work.

When lockdowns were working then MHRA should not be rushed.  When lockdown is not working and there are hundreds of thousands of cases and a doubling time of 4 days then provided the safety part of the analysis has been done and is OK and there's a reasonable expectation of efficacy they should start hand out whatever they've got before dotting i's and crossing t's. at MHRA,  9 days is two doubling times.