advertisement


Coronavirus - the new strain XXII

Status
Not open for further replies.
A lot depends on the quality of the statistical modelling. I have to admit to having expected universities to work together with the NHS data and modelling groups to develop better statistical models, whereby the likely distributions of certain characteristics can be determined (e.g. effectiveness of masks in a country, effectiveness of WFH, effectiveness of being able to take time of work when sick etc) but this has not been borne out.

I still feel that delaying the surge until spring next spring (maybe parts of Europe will achieve that), will have allowed more time for people to be vaccinated, possibly with improved vaccines and for further development of treatments, with improved and less intrusive NPIS in place, whilst not keeping the NHS at saturation. By saturation I mean that the NHS has capacity to reduce its waiting list in many other areas, which it has struggled to do due to Covid (staff off sick, staff treating Covid patients, staff leaving the profession due to the pressure, extra procedures to deal with Covid).

But I would not rule out that the modelling predictions may have been in the ball park and the decision ultimately for the government was to have the excess deaths now or later.



His delays in 2020 (at the start and from September onwards) did result in a lot of excess deaths (could be up to 50k). In 2021, I would agree that his decisions were not batshit or evil. However, I still feel that my comments above are valid in terms of delaying the surge.


My mistake re "infantile leftism" on PFM. I can list and have listed areas where I think iSage are at fault, but I still think that they have informed by bringing many areas to the attention of people that have been watching their presentations. The data analysis is good and supports the wider use of data in the pandemic. Their recommendations have generally been quite soft and not dealt clearly with some of the areas that I think are important. But I still think that they are an asset, so we will have to continue to disagree on that!
I can’t really comment on the quality of the modelling: I have to assume that it’s as good as one might reasonably expect, and a useful tool for policymakers in considering the possible consequences of their actions. I don’t believe those campaigning for ongoing restrictions criticised the modelling, although it was often dismissed or misrepresented.

I’m not sure maintaining restrictions into spring 22 would ever have been a goer.
 
In explosive revelations, one source told the Mirror there were “many social gatherings” in Downing Street last year while the public faced restrictions. They even suggested there were “always parties” in the flat Mr Johnson shares with wife, adding: “Carrie’s addicted to them”.

There were also claims of a third, smaller gathering on November 13, the night Dominic Cummings walked out of No 10, "where they were all getting totally plastered".


https://www.mirror.co.uk/news/politics/boris-johnson-broke-covid-lockdown-25585238

I can kind of forgive that last one. I think I'd be cracking open the Red Stripe too....
 
I can’t really comment on the quality of the modelling: I have to assume that it’s as good as one might reasonably expect, and a useful tool for policymakers in considering the possible consequences of their actions. I don’t believe those campaigning for ongoing restrictions criticised the modelling, although it was often dismissed or misrepresented.
My view is based on 35 years of working in mathematical modelling. My domain knowledge is limited and therefore do not understand in detail the limitations that they would have come up against, but I understand the processes that they would need to have followed etc. Having said all that your view is more reasonable and more likely to be correct.

I’m not sure maintaining restrictions into spring 22 would ever have been a goer.
That is why I said less intrusive e.g. masks in shops and public transport but not everywhere, WFH where practical.

And or non intrusive, e,g. ventilation and or air filtering in schools and workplaces, funding for people to not be at work if sick with Covid and their jobs kept for them, better and more consistent messaging

The experts (including behavioural science) could of course come up with a really well targeted list.
 
Ministers may allow GPs in England to halt regular monitoring of millions of patients with underlying health problems as part of the urgent new blitz on delivering Covid booster jabs.

Sajid Javid and NHS bosses are locked in talks with GP representatives at the British Medical Association (BMA) about relaxing rules which mean family doctors undertake checks on people with diabetes, high blood pressure and other conditions that mean they are at higher risk of having a heart attack or stroke.

https://www.theguardian.com/world/2021/dec/01/gps-in-england-may-stop-monitoring-vulnerable-patients
 
My view is based on 35 years of working in mathematical modelling. My domain knowledge is limited and therefore do not understand in detail the limitations that they would have come up against, but I understand the processes that they would need to have followed etc.

The models (and there were three feeding into SAGE i.e. Warwick, IC and Bristol - I don't believe that the Government was sharing the full datasets more widely than SAGE) were as good as they could have been in my opinion, but they're only models - experiment and observation always trumps modelling. The uncertainties were enormous in all of them, especially the further out in time they were being pushed. They couldn't included waning or indeed boosters, changes in public behaviour and the influence of Government messaging, decisions of the JCVI and, most importantly, they couldn't account for the fact that no single strain of the virus had been dominant for more than about 5 or 6 months. People were reading far too much into them, especially setting the whole ideological roadmap in stone around them, as first the dominance of Delta and the new emergence of Omicron shows. In reality they're living models and are being updated, refined and adjusted as the situation changes but we only get occasional snapshots. My view was always to take the future with a pinch of salt with the need to protect lives always being in the present.

[BTW I'm an experimentalist not a modeller but I've always done some modelling. My first ever research project for my BSc was a modelling one!]
 
The team - in Cardiff and the US - have shown in exquisite detail how a protein in the blood is attracted to a key component of the vaccine

Oi! I'll do links from the BBC - you stick to The Guardian, thank you very much:D

On a serious note - if they can confirm this, they can stop it happening. Also in the article is the vanishingly low numbers of clot fatailties compared to vaccines used etc...
 
The models (and there were three feeding into SAGE i.e. Warwick, IC and Bristol - I don't believe that the Government was sharing the full datasets more widely than SAGE) were as good as they could have been in my opinion, but they're only models - experiment and observation always trumps modelling. The uncertainties were enormous in all of them, especially the further out in time they were being pushed. They couldn't included waning or indeed boosters, changes in public behaviour and the influence of Government messaging, decisions of the JCVI and, most importantly, they couldn't account for the fact that no single strain of the virus had been dominant for more than about 5 or 6 months. People were reading far too much into them, especially setting the whole ideological roadmap in stone around them, as first the dominance of Delta and the new emergence of Omicron shows. In reality they're living models and are being updated, refined and adjusted as the situation changes but we only get occasional snapshots. My view was always to take the future with a pinch of salt with the need to protect lives always being in the present.

[BTW I'm an experimentalist not a modeller but I've always done some modelling. My first ever research project for my BSc was a modelling one!]
I have been changing my view as to how this probably should be viewed over time, once it was clear that deterministic modelling was not applicable (Debs explained that to me pretty clearly). That the left the non deterministic models and the requirement to determine statistical distributions for the relevant configuration parameters. Due to so many confounding issues, data quality and very little chance to configure data experiments, that was also very challenging at best. So that then left the models to be used to look at directions and rough scale for interventions, and the results to be used to help inform as to policy.

Your more practical view is I think more appropriate. My expectation is that the modellers, over time, will learn a lot from this and will improve their techniques and make better sense of the data over time and be ready to make more useful predictions in the future.

And yes I still support the view, that I have mentioned recently, of reducing deaths now, as how we can deal will the virus will improve in the future but do accept that the government have used the rough modelling results and that these may be roughly correct. And that has to be balanced with @Seanm comment that keeping full restrictions in place to 2022 was not a goer. Which leaves me with my view having evolved to what I have put here and in #2327.
 
Ministers may allow GPs in England to halt regular monitoring of millions of patients with underlying health problems as part of the urgent new blitz on delivering Covid booster jabs.

Sajid Javid and NHS bosses are locked in talks with GP representatives at the British Medical Association (BMA) about relaxing rules which mean family doctors undertake checks on people with diabetes, high blood pressure and other conditions that mean they are at higher risk of having a heart attack or stroke.

https://www.theguardian.com/world/2021/dec/01/gps-in-england-may-stop-monitoring-vulnerable-patients

That's an interesting one as the Quality Outcome Framework they're talking about suspending seems to have been a bone of contention with GPs for a while and it's the BMA calling for it to be removed, not the government.
 
And that has to be balanced with @Seanm comment that keeping full restrictions in place to 2022 was not a goer.

I didn't think anyone was proposing full restrictions into 2022 were they? I favoured a more cautious reopening in the summer, masks, distancing, restrictions on large public events e.g., Euros, festivals etc untill all adults had received at least one dose and a full reassessment in August. Unfortunately we went into the school year with very high prevalence and unvaccinated children with esssentially no mitigations in place. There's no doubt that a more considered approach would have saved thousands of lives.
 
That's an interesting one as the Quality Outcome Framework they're talking about suspending seems to have been a bone of contention with GPs for a while and it's the BMA calling for it to be removed, not the government.

Indeed. I don't know what goes on behind the scenes but I don't get anymore than an annual blood test in normal times.
 
The Torygraph made me actually LOL this morning :Do_O

JHsTNNp.png
 
Received my first notification through the COVID App yesterday whilst at work. So came home to isolate and have done a PCR this morning. We do LFTs everyday and they have all been clear so I shall await the PCR result.
 
Status
Not open for further replies.


advertisement


Back
Top