Urban75 Home About Offline BrixtonBuzz Contact

Omicron news

I use hand santiser when I am out & about, and have touched surfaces such as door handles.
Also, I wash my hands when I get home, as well as cleaning touch points.
 
I use cotton gloves and hand sanitiser when out, wash hands with soap on arriving home - but more for general infections not particularly for covid. Don’t think any fomite transmission of covid has definitely been proved (doesn’t mean it doesn’t happen?) whereas the evidence on aerosol transmission has been clear for over well a year.

Plenty of other unpleasant illnesses do transmit via surface contact tho so I’m still doing these things. Adopted them initially cos of covid.
 
It's my theory now that maybe in the days before handwashing, I may have got innoculated against minor viruses by inhaling or otherwise imbibing broken bits from public keyboards ...
 
Last edited:
Hand sanitiser everywhere annoys me a little now as its main purpose seems to be to allow people to feel as if they are doing something meaningful whilst completely ignoring ventilation.
I'm also concerned that a legacy of this pandemic will be a residual germophobia. We used to view it as a sign of concern for people to constantly sanitise their hands, that this indicated an obsessive concern about an excessively magnified risk, making it difficult for people to live their lives. I wonder about the long-term effect of normalising this.
 
I'm also concerned that a legacy of this pandemic will be a residual germophobia. We used to view it as a sign of concern for people to constantly sanitise their hands, that this indicated an obsessive concern about an excessively magnified risk, making it difficult for people to live their lives. I wonder about the long-term effect of normalising this.

I have one colleague who was pretty openly delighted that the hand sanitiser thing as she saw it vindicated her previous germ obsession.
 
LynnDoyleCooper I've not gone looking for the modelling myself yet but I see its now in the news:


The work by the London School of Hygiene and Tropical Medicine is not a crystal ball. It does not say what will happen with Omicron but gives a range of possible outcomes.

In the most optimistic scenario the numbers being admitted to hospitals every day would be 40% lower than the peak last winter. In the most pessimistic scenario it would be nearly twice as high.

But the report said "the majority of scenarios" concluded that with current measures there would be more admissions than last winter.

Dr Davies said: "I think our projections are worrying, it doesn't paint an optimistic picture."
 
Last edited:
In the most optimistic scenario the numbers being admitted to hospitals every day would be 40% lower than the peak last winter. In the most pessimistic scenario it would be nearly twice as high.

But the report said "the majority of scenarios" concluded that with current measures there would be more admissions than last winter.

Dr Davies said: "I think our projections are worrying, it doesn't paint an optimistic picture."

Previous such scenarios and projections for alpha turned out to be wildly inaccurate, with the more transmissible delta wave that came along delivering fewer hospitalizations than even the most optimistic of scenarios modeled for alpha. So when someone says "even in the most optimistic scenario" it should really be read as "in our most optimistic scenario, from our pessimistic model"
 
What would you have them do differently? I recall from the past that you didnt like the vaccine assumptions used, and its fair enough to debate what the realistic parameters for those should be. But when it comes to whether the model itself is pessimistic, I note that they do try to test and adjust their model via things like fitting their model to real data observed in the past (and there are graphs of them doing that in this latest document, using real data from the entire pandemic so far).

And its rare for me to expect the reality to match any of the modelling projections because they have very little chance of making the correct assumptions about how peoples behaviours/the result of restrictions will change at different moments in time. Plus the modelling usually involves dates for new restrictions being imposed that dont end up perfectly mirroring the dates government actually do those things.

I am broadly content with pessimistic or realistic modelling because they end up affecting behaviour in ways that reduces the pandemic burden. 'self defeating prophecy' stuff which allows some people to indulge in stupid narratives but the modelling and its effects ultimately end up serving the cause of public health in a manner I highly approve of.
 
Last edited:
What would you have them do differently? I recall from the past that you didnt like the vaccine assumptions used, and its fair enough to debate what the realistic parameters for those should be. But when it comes to whether the model itself is pessimistic, I note that they do try to test and adjust their model via things like fitting their model to real data observed in the past (and there are graphs of them doing that in this latest document, using real data from the entire pandemic so far).

I was going to mention their vaccine efficacy figures for hospitalisation and death (falling to 66% for the unboosted in the most optimistic of the two scenerios) in Table 1. These seem to be plucked from thin air, but I wasn't sure it worth arguing on here.
 
Previous such scenarios and projections for alpha turned out to be wildly inaccurate, with the more transmissible delta wave that came along delivering fewer hospitalizations than even the most optimistic of scenarios modeled for alpha. So when someone says "even in the most optimistic scenario" it should really be read as "in our most optimistic scenario, from our pessimistic model"
Is it not more the case that prrdictions were made for what would happen without restrictions, restrictions were then put in in a half baked way, and we never got to see what would have actually happened without them.
 
I was going to mention their vaccine efficacy figures for hospitalisation and death (falling to 66% for the unboosted in the most optimistic of the two scenerios) in Table 1. These seem to be plucked from thin air, but I wasn't sure it worth arguing on here.

I cant say how close to reality they will be, but they arent plucked from thin air:

Neutralisation studies on Omicron are ongoing, so we rely on a combination of estimates of the fold reduction in neutralisation titre for previous variants of concern (VOCs) as well as early studies explicitly considering Omicron neutralisation (1–5) to inform our assumptions about the level of immune escape Omicron might possess. The largest drop in neutralisation titre (8.8 fold) was estimated for the Beta VOC (20), with a 3.9 fold reduction for the Delta variant, compared to the ancestral SARS-CoV-2 virus. We consider two scenarios for the immune escape of Omicron relative to Delta: 5.1-fold (escape low, EL) and 12.8-fold (escape high, EH) reductions compared to our existing assumptions for Delta (19). Since the Delta variant was estimated to have a 3.9 fold reduction in neutralisation compared to the ancestral SARS-CoV-2 virus, these assumptions correspond to overall fold reductions of approximately 20 and 50-fold between the ancestral SARS-CoV-2 virus and the Omicron variant.

We use the relationship between mean neutralisation titre and protective efficacy from Khoury et al. (7) to arrive at assumptions for vaccine efficacy against infection with Omicron, given each drop in neutralisation. We then use Khoury et al.’s modelled relationship between efficacy against any infection and efficacy against severe infection to generate vaccine effectiveness estimates against severe outcomes (Fig. 1c). For the effectiveness of booster vaccinations against Omicron, we base two main scenarios on two studies which measured increases in neutralisation titres following the second dose of the primary vaccination course and after booster vaccinations with the Moderna (2.5-fold) and Sinovac (4.9-fold) vaccines (20). We assume that protection against infection for individuals who have received a primary course of the AstraZeneca COVID-19 vaccine before being boosted with either full-dose Pfizer or half-dose Moderna (the current policy in England) is initially increased to the same levels as Pfizer/Moderna, before using the relationship in Fig. 1c to scale protection against infection to protection against severe outcomes (hospitalisation and death). We assume that individuals in the recovered disease state who have previously been infected with SARS-CoV-2 have the same level of protection against Omicron as individuals who have received two doses of Pfizer/Moderna. Our estimates for vaccine protection against different outcomes for the Delta variant and for the various Omicron scenarios are shown in Table 1.

These estimates are broadly in line with early vaccine effectiveness estimates against Omicron and Delta from the UK’s Health Security Agency (21), with the exception that our assumptions for dose 2 AstraZeneca vaccine protection against Omicron disease are high in comparison (we assume 38.2% and 23.3% protection against disease, whereas the UKHSA study’s highest estimate for protection against Omicron disease with two doses of AstraZeneca is 5.9%).
I have mentioned in other posts some acknowledged limitations with the UKHSA AZ estimates that they mention in that last bit, so I'm not surprised they have used higher ones for that.
 
Is it not more the case that prrdictions were made for what would happen without restrictions, restrictions were then put in in a half baked way, and we never got to see what would have actually happened without them.

The models included restrictions, the reality of transmissibility and restrictions was worse than modeled, but the outcome in terms of cases and deaths was better than their most optimistic scenario. I’d dig it out and match it up to what actually happened again, but I don’t think there’s much appetite for criticism of SAGE on here.
 
Its what uncertain science looks like, have you got a better methodology they could use instead? Would you rather they pluck optimistic stuff from thin air instead? And this isnt SAGE, although obviously they are one of the sources SAGE relies on and there is overlap between people in the academic institutions and members of SAGE.

Please do dig the old stuff out, so we can evaluate your criticisms of past modelling.
 
Its what uncertain science looks like, have you got a better methodology they could use instead? Would you rather they pluck optimistic stuff from thin air instead? And this isnt SAGE, although obviously they are one of the sources SAGE relies on and there is overlap between people in the academic institutions and members of SAGE.

A good starting point would be to use the protection against hospitalisation and death figures for delta in the main model. There's currently no evidence to suggest the figures for omicron will be lower. They could always produce a secondary set of scenarios using stepwise reductions in those figures.
 
Please do dig the old stuff out, so we can evaluate your criticisms of past modelling.

No. Last time I did something like that you said "I am not interested in playing this stupid game with you" so I'll try and stick to reddit for that stuff.
 
I'd agree that evidence isnt there yet but I also think its reasonable for them to assume it could be lower, especially if its only a little lower. Especially given that even a small reduction can make quite a difference to the numbers.

But yes, I'd have nothing against a larger number of scenarios being presented in modelling reports. Including more optimistic ones. But then it comes down to what some people will want build off the back of more optimistic scenarios being presented at this time. An excuse to do less at this stage? That carries its own risk, better safe than sorry is a reasonable approach given the stakes.
 
No. Last time I did something like that you said "I am not interested in playing this stupid game with you" so I'll try and stick to reddit for that stuff.
I'll do it myself then when I find time. I expect I'll be able to build a case that you are misrepresenting the limitations of past modelling in order to serve your dubious agenda.
 
But yes, I'd have nothing against a larger number of scenarios being presented in modelling reports. Including more optimistic ones. But then it comes down to what some people will want build off the back of more optimistic scenarios being presented at this time. An excuse to do less at this stage? That carries its own risk, better safe than sorry is a reasonable approach given the stakes.

This is the problem. Their output is presented as covering a range of scenarios, and the impression given is that that this is the range of likely scenarios, whereas it's actually a subset of pessimistic scenarios.
 
Back
Top Bottom