HSTEd
Veteran Member
- Joined
- 14 Jul 2011
- Messages
- 20,282
Many thousands of people would still be hospitalised, just that it is unlikely that the wave could grossly overmatch healthcare capacity.
-
Our new ticketing site is now live! Using either this or the original site (both powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!
Vaccine Progress, Approval, and Deployment
- Thread starter Yew
- Start date
- Status
Many thousands of people would still be hospitalised, just that it is unlikely that the wave could grossly overmatch healthcare capacity.
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R
RailUK Forums
brad465
Veteran Member
Yes, we just have to make sure it doesn't cause a major panic and regardless sort out healthcare capacity and resilience long term.
No it isn't. Once the most vulnerable are vaccinated, the risk of hospitalisation drops dramatically. And that's because for the vast majority of people covid will be no more than a heavy cold or flu, and for many more even less. I thought we had established this some time ago?
HSTEd
Veteran Member
- Joined
- 14 Jul 2011
- Messages
- 20,282
It drops dramatically, but not to zero.
Ah! I see, you want zero covid. The very thing scientists tell us is impossible?
HSTEd
Veteran Member
- Joined
- 14 Jul 2011
- Messages
- 20,282
Uh what?
No?
I can accept that allowing another wave will lead to substantial numbers of hospitalisations and still believe this is the correct choice.
@Bald Rick has partially answered; my other observation would be that the present assumption is that immunity is not lasting and therefore vaccination (including boosters) will be required as immunity acquired through illness fades.
By the time it gets any steam, the vulnerable will have very effective immunity from vaccination.
The models are worst case scenarios, and many of the input parameters are hopelessly pessimistic, to the point where they have been shown to be wrong. Ultimately, this incessant wheeling out of worst-case scenarios is counterproductive, we need to look at the reasonably probable scenarios, and make decisions based on those.
Vaccination of groups 1-9 will eliminate 80% of hospital admissions, and severely slow down transmission.
== Doublepost prevention - post automatically merged: ==
That's very unlikely, there is still an active and powerful immune response in SARS patients over a decade after the outbreak.
They are absolutely not worst case scenarios, they are based on a range of assumptions, with a range of answers, naturally. Some of those assumptions are cautious, yes, but I wouldn’t say they are ‘hopelessly pessimistic’.
== Doublepost prevention - post automatically merged: ==
Indeed. But 20% of a big number (approx half the adult population) is still a big number.
And I’d also suggest that some of the assumptions about how the speed and durability of the impact of vaccination should inform policy have erred on the optimistic side.
Interesting article at https://www.quantamagazine.org/the-hard-lessons-of-modeling-the-coronavirus-pandemic-20210128/ on the subject of Covid-19 modelling, which echoes this. The University of Illinois came up with an excellent model but "what we didn't anticipate was that they would break the law" referring to the behaviour of the students they were attempting to model.
But it won't necessarily lead to that because of the vaccine. I mean that's the whole point of it right, to significantly reduce the risk of serious illness requiring hospitalisation.
Why would that assumption be correct? Our immune systems don't simply forget infections after antibodies fade, they often retain that genetic knowledge in B-cells. And this is shown by the fact that there are people with pre-immunity to covid, i.e. their immune systems were capable of tackling the virus even before we knew it existed. So the presumption of zero immunity at the start as modelled by Ferguson and his team, along with the presumption that naturally induced immunity is short lived are clearly wrong.
That’s a lot of reliance on modelling being right as a guide to policy. With other information, I’d be more cautious.
It'll be less than 20% once all the other factors (in particular seasonality) are taken into account. It will be down at the level of a bad flu year (at most).
You can’t have the argument both ways on modelling, denying one course of action because the model is “wrong” (actually less clear cut than often alleged), yet insisting on the use of predictions as the basis for policy.
So it's fine to rely on models from someone who has demonstrably over-estimated by huge amounts, but not to rely on models which are realistic based on current understanding?
Why don't you just come out and say that you think lockdowns should continue indefinitely, and that you couldn't care less about the impact this has on some sectors of society? Because that is clearly the standpoint you are taking on this!
Indeed, about 5 million in England alone that are susceptible to hospitalisation, although clearly only some fraction of that will actually be hospitalised if they get Covid
But not, I feel, so significant to be gravely concerning.
HSTEd
Veteran Member
- Joined
- 14 Jul 2011
- Messages
- 20,282
I'm not sure it would be that mild, but I am fairly certain that healthcare capacity would not be grossly overmatched by such a wave, especially with the vaccination campaign continuing to advance in the meantime.
Because the politicians don't understand one bit of the science, so they end up either completely trusting or completely distrusting scientists like Ferguson. Both approaches are wrong.
The SIR model which is a vastly simpler model which underpins most of the models used today is just one differential equation, but even that gives answers. With R=3, the SIR model predicts that 94% of the population will be infected before the virus runs out of hosts, and 94% of the UK population with a 0.35% mortality rate leads to a "prediction" of 200,000 deaths. That was easy. But is it correct?
The problem with modelling is that it's incredibly sensitive to the assumptions you make.
I was reading an analysis which an independent statistician had done - he had replicated their model but then tweaked some of the assumptions, and that had huge impacts on the results. For example, he initially assumed that the vaccine blocks 50% of transmission after one dose, and 65% after two. Under that, we'd see a fairly large (but manageable) peak in August, about half the size of the current one. But if you change those figures to 40% and 55%, you see a much larger peak than the current one in July instead. If those figures were actually 60% and 75%, we might barely see another peak at all, and it would be pushed all the way to the winter. We still don't exactly know what the true levels of protection are, yet even tiny differences in this would have massive future ramifications.
I was reading an analysis which an independent statistician had done - he had replicated their model but then tweaked some of the assumptions, and that had huge impacts on the results. For example, he initially assumed that the vaccine blocks 50% of transmission after one dose, and 65% after two. Under that, we'd see a fairly large (but manageable) peak in August, about half the size of the current one. But if you change those figures to 40% and 55%, you see a much larger peak than the current one in July instead. If those figures were actually 60% and 75%, we might barely see another peak at all, and it would be pushed all the way to the winter. We still don't exactly know what the true levels of protection are, yet even tiny differences in this would have massive future ramifications.
If that model is predicting large peaks in July or August, it's clearly not taking seasonality into account - we know from last year that this has a significant impact.
He acknowledges that, but he doesn't actually think seasonality will play that big a role in the size of the peak, just when it happens. The stronger the seasonal effect, the later the peak.
Modelling has nothing to do with it, we know that our immune systems adapt to new threats and retain that information beyond infection & initial antibody response. SARS-CoV-2 may be a new virus, but it uses proteins that other coronaviruses do to build it's RNA and it's envelope / spikes. Hence the reason that blood samples taken before the pandemic show B-cells that can trigger a response to this current virus. Based on this evidence it is fair to say that exposure to SARS-CoV-2 will likely lead to a long term genetic memory in many people, resulting in minimal effect on re-infection.
Work out in the last fortnight showed that antibody levels produced in response to vaccination over time were far higher and longer lasting that due to infection (three separate studies for Pfizer, Moderna and Oxford/AZ).
== Doublepost prevention - post automatically merged: ==
Presumably you are aware of the Hoskins effect? the body will try to use a pattern of antibody it already has in response to an infection if it thinks that antibody might work as this mean the body's antibody response is far quicker than if it starts from scratch with new pattern of antibody. It doesn't mean that than existing antibody pattern (presumably against OC43 as this uses the same mechanism as SARS-Cov2 to enter cells) is actually effective against a particular Covid strain, it just means the body is trying it first as it thinks it might work.
The apparent comparative reduced / in-effectiveness of vaccines to reduce mild covid cases with 501 and 484 mutations is probably down to lower antibody bonding and existing antibodies being ineffective leading to a several day lag till new pattern antibodies are made. There is still a very good specific T-cell response (87% in Ox/AZ South Africa trial) in those cases which prevents them becoming serious.
Last edited:
A peak in September won't be an issue, as we'll have have vaccinated everyone by then.
- Status