Forum menu
Could it make you falsely believe you are president of America?
Harry Potter takes on the Tier system:
Nadhim Zahawi, a minister for business and industry, has been placed in charge of overseeing the deployment of the Covid-19 vaccine, Downing Street has announced.
Is it the vodka talking or was Grayling too busy?
We're doomed now. He'll siphon off the budget for his property portfolio.
Good job we are paying millions for a vaccine Tzar......
Thought this blog posted to Wired on the vaccine trial reporting was very good and worth a read. http://hildabastian.net/index.php/100
Basically, the two mRNA trials were well set out Phase 3 designs with one simple endpoint. Reporting of them in press releases (which is a mandated activity for any trial results material to the share price) was clear and straightforward, as is the interpretation. By contrast, the world-beating Oxford/AZ trial was more convoluted, an amalgam of multiple trials and objectives, and reporting was (ahem) less than transparent. Only after Warp Speed's Moncef Slauoi revealed the randomisation has the true nature of the results come out.
Basically it appears that one contract manufacturer did not transfer the release assay for concentration of virus in the vials. This lower strength vaccine was used unwittingly and the reported reduction in adverse events led to further investigation.
In Biotech Land, the process is the product. If you can't make it reliably then you are absolutely looking for the paddleshop. On a scale of 1 (minor trial issue) to 10 (failed trial) this would rate about an 8-9 maybe a 10. OK the remainder of the data is probably OK (60% efficacy is enough for an emergency license) but the 90% efficacy can be wholly disregarded. The fact that the data is confounded (no >55 received the low dose) and there are no US patients due to an FDA clinical hold during the trial means that the vaccine won't be approved in the US without a further trial. AZ's CEO admitted as much and they will run a further trial.
MHRA may approve the Pfizer/Biontech vaccine this week for emergency use, ahead of the EMA.
If you want some good news, we are presently at the peak of the second wave of deaths and things will continue to go down as we leave lockdown into lockdown lite.
TiRed - if (as they claim) the AZ/Ox vaccine is 60% efficacy, and stops severe illness, possibly transmissibility too, would it be a good, cheap option for the mass population under 50, leaving the Pfizer/Moderna options for the most susceptible?
Yes, I think the OX/AZ vaccine can be approved and used, particularly as all vaccines (and antibodies) prevent serious disease. That is a huge benefit. For the West, cheap is not important - the cost compared to the economic cost of NPIs as a whole is trivial.
Moderna and Pfizer/Biontech have shown how to conduct trials properly rather than the make-do Oxford route - which looks and feels academic at best. Well-defined trials with limited but approvable objectives are what builds public confidence. Plenty of other vaccines coming along next year following the Pharma approval route.
Thanks. My main worry is getting my mum vaccinated (72) so we can see her, and help her move house to my village.
Personal immunity is a lesser worry, but definitely see “no jab, no entry” to some countries being an issue , both to my job and personal travel.
Have we done the effects of Covid-19 on Yankee Candle buyers?
😂😂😂
Week 47 mortality data shows 26% rise above normal means for the year. Or to put it another way, every death in the last eight weeks of 2020 is excess mortality above ten-year mean, and every death in the last five weeks is above the ten-year maximum. That will be about 60k over the past maximum. Things should stabilise by Christmas
My predictions for the coming weeks:
We will see the effects of lockdown for a further week on admissions and two weeks on deaths. Admissions and deaths will then stabilise due to Tier 2/3 going into Christmas. I’m fairly optimistic that over the five days (equivalent to half a Tier 1 doubling time with closed schools) of Christmas mixing we will see a roughly 40% (=sqrt(2)) increase in prevalence which will show in the first ONS survey of the year. This increase in prevalence will feed into (and predict) admissions in 2021-W1-2 and deaths in 2021-W2-3. Peaks will be broadly in line with levels seen now as the Tiers are reintroduced. All-cause mortality will be similar to past 10-year maximum mortality levels which normally peak in Week 2.
60% seems a bit rubbish to a man that doesn't know anything about vaccines (Although similar to flu shot)
Does that mean it has a 60% chance of stopping me contracting the virus? If that's the case then surely when things get back to 'normal' my exposure will be numerous times higher than it is now, and as such my risk of gettimg it will actually be greater.
The key point for the vaccine is the absence of ANY severe disease for those vaccinated. This will keep people out of healthcare, and that is the reason for restricting transmission with Tiers and Lockdowns. It's a good result. Think of the vaccine as giving you that lifetime of colds you would have had. A reduction in transmission is also possible but has not been tested in the studies.
If people around you have the vaccine then over time the likelihood of you coming into contact with it dramatically decreases from, oh I dunno probably <5% now, to <<0.1% in the summer
The key point for the vaccine is the absence of ANY severe disease for those vaccinated
So do we know how effective it is at reducing severe disease? Is that 60% (ie it effectively cuts the risk of becoming severely ill in half)
Apologies if this has recently been covered, but what are the current estimates for % population to have been infected? We were @ ~6% in June; where are we thought to be now?
I don't think there is any evidence that any(?) of the vaccines are either better or worse at stopping severe cases, than they are at stopping mild cases. Reasonable assumption may be that they are similarly effective.
60% wouldn't be fantastic on an individual basis but would certainly aid control and reduce the load on healthcare. 90% would be really good though and depending on uptake and persistence could fix this problem quite conclusively.
So do we know how effective it is at reducing severe disease
100% - nobody in the vaccinated arm had a serious COVID19 infection. In any of the three trials.
but what are the current estimates for % population to have been infected?
Based on most up-to-date ONS survey - note lower rates of infection mean that people lose their seropositivity, so these numbers may not reflect total numbers. Less than 1/8 is a robust answer.
Analysis of antibodies evidence from individuals who have had the infection in the past show: in England, an estimated 6.9% (95% confidence interval: 6.3% to 7.4%) of people would have tested positive for antibodies against SARS-CoV-2 on a blood test in October, suggesting they had the infection in the past; there is substantial variation in antibody positivity between regions, from 10.8% (95% confidence interval: 9.3% to 12.5%) in London compared with 3.1% (95% confidence interval: 2.1% to 4.4%) in the South West.
Anyone seen the latest anti-vax bollocks doing the rounds on Facebook?
Apparently "forced vaccination is against the Nurumbrg [sic] Rules."
That " " taken directly off a comment on my FB from someone who clearly can't spell Nuremberg nor has any idea about the Trials. But hey it sounds good and also it sort of mentions The War so one for all the proud Brexiteers to get behind - there seems to be quite an overlap on the Venn Diagram of Brexiteer and Anti-Vaxxer.
There will not be compulsory vaccination. There may be a nudge along the lines that other countries won't admit you if you have not been vaccinated. Come Brexit we won't be going anywhere anyway 🙁
TiReD, that's a silly claim. Only a handful were infected, the proportion of serious cases is expected to be small. We certainly don't know any of the vaccines to be 100% effective against serious cases any more than if you toss a coin a couple of times and it comes up heads, you could claim to have invented a foolproof way of always getting heads.
@thecaptain - although you're "correct" in absolute terms, do you realistically expect the trial to test the total population?
Assuming an AZ breakdown of infections (they haven't given the breakdown) of Active/Placebo:
3/30 low dose trial (90% effective)
28/70 normal dose trial (60% effective)
31/100 (reported 69% effective)
That makes no serious infections in 28 treated on normal dose gives a 95% credibility interval of (0,0.098145). So I'm 95% certain that the vaccine is at least 90% effective in preventing serious COVID19 infections that lead to hospitalisation. Including the low dose (0/31) doesn't change this much.
I never mentioned prevention of infection - only serious disease. One could take the prior of, say 2-3% as the normal complication rate and update with a posterior. But this is now across three vaccines with no hospitalisations.
So I’m 95% certain that the vaccine is at least 90% effective in preventing serious COVID19 infections that lead to hospitalisation. Including the low dose (0/31) doesn’t change this much.
Technically I am 19/20 certain that the frequency of serious infections is less than 10%. But my best estimate is zero if I am a frequentist and less than 3% if I am a Bayesian. If I strongly believed before the trial that serious infections occur in 5%, then I am now very confident that the vaccine does something. If I have no prior belief before the trial, then my best estimate now is 0%, but larger values are possible. How's that? 🙂
Reporting of the OX/AZ trial has been an exercise in how NOT to report a clinical trial. And I've co-written those press releases for some of our trials in the past.
There will not be compulsory vaccination. There may be a nudge along the lines that other countries won’t admit you if you have not been vaccinated. Come Brexit we won’t be going anywhere anyway
It was hinted on BBC Breakfast that you might be allowed into pubs unless you have been vaccinated. I think that'll get plenty of volunteers ready for the jab.
TiReD, we already know that the probability of serious infection is around 10% (ballpark) as a proportion of cases. If your calculation doesn't take that into account it *cannot* be correct as a matter of principle. Saying that the frequency of serious infections (as a proportion of all infections) is no more than 10% in the treatment arm doesn't really say very much at all since it would only be 10% in the case of no treatment, and certainly your calculation does not support the claim that the vaccine is 100% effective in preventing these serious infections.
It was hinted on BBC Breakfast that you might be allowed into pubs unless you have been vaccinated. I think that’ll get plenty of volunteers ready for the jab.
Seems to be a good level of support for it in the licence trade, if Radio Scotia today is anything to go by.
Zahawi - says yes
Gove - says no
Same old Govt mixed messages
TiReD, we already know that the probability of serious infection is around 10% (ballpark)
Probably 3-10x lower based on the ONS survey of prevalence. In the Lilly antibody trial, hospital admission rate for placebo AFTER symptom onset was 9/156. When dosed with antibody it went down to 6/421. This ignores mild and asymptomatic infections who did not report.
The point estimate for none-in-N is always zero. Hence 100% effective. The true interpretation depends on your statistical belief structure. Giving exogenous antibodies AFTER infection (with symptoms) reduces hospitalisations by 75%. Having them before infection, even polyclonal vaccine-home-made at varying potency and titer, will give protection to a lower viral load challenge.
I'm less worried about the effects on transmission.
Yes, I think the OX/AZ vaccine can be approved and used, particularly as all vaccines (and antibodies) prevent serious disease. That is a huge benefit. For the West, cheap is not important – the cost compared to the economic cost of NPIs as a whole is trivial.
but what are the current estimates for % population to have been infected?
Based on most up-to-date ONS survey – note lower rates of infection mean that people lose their seropositivity, so these numbers may not reflect total numbers. Less than 1/8 is a robust answer.
I'm still wondering what is the reason to give the vaccine to people with active antibodies ?
Does a vaccine do something different or last longer to acquired antibodies or is it just cheaper than testing and we have surplus vaccine available? (non of which I thought to be the case but I stopped following)
The home where my mum resides contacted my dad today to seek permission to administer her a vaccine next week. He’s agreed and it appears that it’s being lined up that with a negative test, home visits are back on soon.
Let's hope so rockhopper, fingers crossed.
Yes, he’s already said it’s the best present he could ever receive. At 90 and 91 respectively, married for 70 years, certainly my dad isn’t thinking long term consequences of the vaccine. For clarity, if it wasn’t obvious, my mum hasn’t the capacity to make the decision herself.
Does that mean it has a 60% chance of stopping me contracting the virus? If that’s the case then surely when things get back to ‘normal’ my exposure will be numerous times higher than it is now, and as such my risk of gettimg it will actually be greater.
You can only catch it if someone else has got it. So if everyone was magically vaccinated on the same day, and piled into their favourite pub/tubetrain/orgy the week after, then what you said would be correct.
Slowly building up the number of immune people and slowly increasing societal contact will cause it to (not completely) die down.
I’m still wondering what is the reason to give the vaccine to people with active antibodies ?
How are you identifying people with active antibodies?
Yes, he’s already said it’s the best present he could ever receive. At 90 and 91 respectively, married for 70 years
That's awesome mate 💙
Basically it is faster and easier to give the vaccine than test for antibodies to covid. So it becomes a simple calculation.
Oxford vaccine is £10, moderna is £40.. The private antibody tests are £65 so assuming some uplift they are probably about comparable with the expensive vaccine but considering run times, reporting etc it will be much faster to just give people the vaccine (15mins including waiting time) and you then don't need all the machine run time and additional staff to run the tests.
Financially the costs are probably about the same but we don't have the capacity to test everyone for antibodies.
give the vaccine to people with active antibodies ?
There is no data as these people were excluded from the first trials (I’m sure they’ll be studied in smaller trials). I imagine it will act as a boosting effect similar to other coronavirus infections where antibodies wane over time. Think of the shingles vaccine that does the same. Repriming your defences.
I view this virus like the other four of its type. More pathogenic perhaps, but likely to have waning immunity, reinfection and boosting. The vaccine has been shown to be therapeutic in that it protects from severe infection. It may reduce infectiousness too, but this was not the subject of the trials.
A protective vaccine is great news. If it gives you the equivalent of the first ten years of past infections, we’ll be in a good place. I expect seasonal vaccination, possibly annually or biannually with a flu jab. Then in ten years time this will be a seasonal cold. It may still pose a problem for some abs treatments will be available. That’s not always a given for the vulnerable who are susceptible to other viruses.
Thanks for sharing rockhopper, made my day.
Interesting Opinion article in the Guardian on today's vote
Green light for the Pfizer vaccine in UK. Now how to actually distribute it.
You would assume that it will have to be hospital focused. So this vaccine should be for health workers, and clinically vulnerable including care homes as they should be relatively easy to get.
it would be the general public that would be a challenge.
Let's hope they do it that way
It's all happening, my wife and her colleagues have volunteered to begin vaccinating (she's an opthalmic nurse), Her team to train in January to vaccinate the public then.
When do we think the men in grey suits will be taking the ceremonial revolver to 10 Downing St ?
Not until after January next year.
they need a scapegoat ready to use.
How are the Vaccine success rates calculated?
Is it simply a matter of giving say 1000 people the vaccine and then seeing how many of them contract Covid in the next n weeks/months?
Those people will carry on using masks, washing hands and social distancing. What is to say the success of the vaccine is not dowm to those factors?
I guess they are comparing the figures against another sample group that has had a placebo.
Are there any results from testing where the vaccinated group were subsequently deliberately infected with the virus? I know this option was mooted several months ago as being the most accurate way of determining efficacy.
It is exactly as you describe. The numbers infected in the vaccine group measured against the numbers infected in the control group.
live virus challenge studies have not been started yet as far as I know. I suspect they are recruiting though
Efficacy is calculated based on events not total population treated. So you dose 40,000 people, half on each arm, and you look when you’ve acquired 100 infections. The null hypothesis is that there will be about 50 infections in a each. The alternative is that the split is not even because the vaccinated arm has done something.
The mRNA vaccines split about 5/95. Think heads and tails. The AZ/OX vaccine split about 29/70 (but they haven’t released the real numbers). Both are very unlikely by chance.
There are other analyses that correct for months of exposure too, because one has monitored the early recipients for longer. Or you cut the data at 28 days post last dose. The analysis is about as robust and assumption-free as is possible.
I remain unconvinced by challenge studies for sars-cov-2. The pathogenesis is still not well characterised. We have rescue treatments but they have to be given early at the onset of symptoms to work. So we shall see about the ethics. I wouldn’t volunteer, and I would for influenza challenge. I’ve had both.
[tl:dr] vaccine efficacy is measured similar to the toss of a coin. Flipping only five heads in 100 is very unlikely.
Just returning to the passport question interesting article here that enters the moral maze over ethics.
https://www.hl.co.uk/news/2020/12/2/should-pubs-cinemas-and-airlines-require-covid-vaccine-passports
The argument reminds me of Vroomfondel and Majikthise discussing Deepthought in Hitchhikers Guide.
My own view is vaccinate for the greater good.
If ant-vaxers die out I don't care as long as they don't take me with them.
Dulce et Decorum est and all that.
More Cash
How are you identifying people with active antibodies?
I'm not as I'm not PHE. France seem to be doing it so I don't understand why we can't?
What I'm trying to understand is why PHE are actively preventing it? Is it science or politics?
The same for active testing why are Govt/PHE not acknowledging negative tests for people in track and trace?
Is it because they are unreliable or too expensive or not available or is it just political?
A personal example is Jnr is self isolating ... the instructions are NOT to get tested unless he has symptoms but aged 11 his chance of symptoms seems pretty low (or certainly not high) - don't know what the exact figures are.
3 kids in his class later tested positive ...
but of these 1 broke the rules and got tested when they were asymptomatic... so who knows who has and hasn't had the virus.
The other kids got tested against the rules still can't go back to school regardless of how many negative tests so we didn't even bother getting a test as it makes no difference if its negative or a subsequent one a week later is still negative.
Meanwhile his mother is teaching and I'm out and about... and he's not even allowed to leave the house for a bike ride.
His isolation is over tomorrow but I can't understand how that means anything when his mother is in school and I as a minimum had an operation the day before.
If track and trace worked I wouldn't even have been able to have the operation... as had there not been delays in the testing of the 3 initial kids he would have been told to self isolate and my operation would have been cancelled. So basically his isolation actually started before my operation but as nooone had results or whatever I was able to go through ....
The other side is people who are pretty sure they already had it can't get tested (and if they do privately the results are ignored)
TiReD
There is no data as these people were excluded from the first trials (I’m sure they’ll be studied in smaller trials). I imagine it will act as a boosting effect similar to other coronavirus infections where antibodies wane over time. Think of the shingles vaccine that does the same. Repriming your defences.
I don't disagree based on my very limited knowledge but I understood and may be wrong that we have very limited supply of the SARS-COV19 vaccines but as much shingles vaccine as we can use and if we don't use the shingles vaccine before it expires it will need to be destroyed anyway ???
I view this virus like the other four of its type. More pathogenic perhaps, but likely to have waning immunity, reinfection and boosting. The vaccine has been shown to be therapeutic in that it protects from severe infection. It may reduce infectiousness too, but this was not the subject of the trials.
Again from my limited knowledge surely this would have been a prudent assumption to start out with?
The government led response seems to have been the entire opposite from day 1?
The absence of evidence seems to have been taken as evidence... it absolutely couldn't be spread by airborne means and could only be spread by touch. This seems counter to the common 4 ... surely a working assumption in the absence of evidence to the contrary would be to "view this virus like the other four of its type. More pathogenic perhaps.." We were told kids cannot spread the virus... again it seems based on lack of evidence constituting evidence. 7-8 months later NHE are still doing their best to prevent children being tested to skew the results.
A protective vaccine is great news. If it gives you the equivalent of the first ten years of past infections, we’ll be in a good place. I expect seasonal vaccination, possibly annually or biannually with a flu jab. Then in ten years time this will be a seasonal cold. It may still pose a problem for some abs treatments will be available. That’s not always a given for the vulnerable who are susceptible to other viruses.
From a personal perspective my only real motivation to get vaccinated is to prevent spreading the virus to the vulnerable.
I've got 3-4 of the yearly nag letters from the GP about flu virus vaccines...but they don't say anything about protecting others or why this years expected mutation would affect me. Perhaps naïve but I've never managed to develop flu (as in disease symptoms) regardless of exposure so why would this year be any different? I know it could be different but in the 40 odd years of having a full immune system the odds of this year being different for me don't seem very high.
Just to be clear I've nothing against getting the vaccine, it just seems like a waste of time and NHS resources and this year a chance to pick on SARS-Cov19.
I think it's fair to say, that as a nation, we've lagged behind the curve every step of the way and I personally have little faith or trust in the current government.
I hear on the news today, that we're first to approve the new vaccine but that Europe claim their checks are safer than ours.
I'm not against vaccines. After all, I've had the live version of it, I can't see what harm the dead version would do. I am however, curious as to why we're now suddenly first to something.
There are a lot of people in this thread more learned that I. Perhaps they could shed some light on this sudden change in competence.
The MHRA ain't the government.
True, but it seems harder than it should be to remove political influence from a lot if places it shouldn't be.
Is there any truth in the claim that the European version is more thorough?
The MHRA ain’t the government.
That's a matter of perspective... we have been through schools, councils, NHS etc. (even BBC) elsewhere and earlier in this thread and the distinction isn't black and white when "the government" can appoint an overseer or czar at it's whim if it doesn't get the answers it wants.
the distinction isn’t black and white when “the government” can appoint an overseer or czar at it’s whim if it doesn’t get the answers it wants.
Where is the evidence that the regulatory body has been populated by government stooge yes men (or women) ?
That’s a matter of perspective
No - that's the FDA, where the FDA Commissioner is a political appointment. MHRA is wholly independent statutory body. Just like the CAA for aviation.
The MHRA have traditionally done a lot of the EMA "heavy lifting" for drug reviews. The reward for this, of course, was location in London and I have friends that work there. Not any more. All country's licensing authorities can approve for their own country independently if an application is submitted. Pfizer/Biontech will have submitted to individual countries for emergency approval - on a rolling basis as soon as data was available. The UK was first - by a few days.
as much shingles vaccine as we can use and if we don’t use the shingles vaccine before it expires it will need to be destroyed anyway ???
GSK was unable to meet shingrix demand actually. Recovered for the wrong reasons as vaccination declined!
Thanks TiRed, that's reassuring. Should I take that to mean that EMA isn't any better and on this occasion, the coercion of facts is on their part rather than ours?
Any EMA delay is probably down to their project managers trying to schedule zoom meetings. There's quite a bit of cat-herding and a lot of academics involvement 😉 . The FDA, by contrast, is a slick professional organisation of full-time employees.
I read nothing into it. Agencies like to assert their independence. If you want to upset the FDA in a meeting, just say "The EMA said..." 😀
Where is the evidence that the regulatory body has been populated by government stooge yes men (or women) ?
This is the black and white point....
The government don't need to appoint someone if they simply make an implicit threat that they will.
They can either just appoint someone where they can or just invent a new kwango if they can't.
BBC? Test and Trace? NHS ?
BBC? Test and Trace? NHS ? SPACE LIZARDS?
This is the black and white point….
Just look at the US and FDA. We are not the US.
https://www.google.com/search?client=firefox-b-d&q=fda+commissioner+summoned
That’s a matter of perspective
It's really not.
BBC? Test and Trace? NHS ? SPACE LIZARDS?
BBC a direct threat to report what they are told and carried through
Test and Trace - failed jockey and miserably failed CEO married to a conservative MP and Tory whip in the Lords then appointed Chair of NHS improvement then votes in a Tory Lord as Chair of NHS?
Not sure where the space lizards come in ?
The government don’t need to appoint someone if they simply make an implicit threat that they will.
They can either just appoint someone where they can or just invent a new kwango if they can’t.BBC? Test and Trace? NHS ?
Is paranoia a symptom of Covid itself, or a sign that people have had way too much time home alone with the internet?
Is paranoia a symptom of Covid itself, or a sign that people have had way too much time home alone with the internet?
He's never been spotted in the same place as chewkw. Makes ye think.
Bit unfair chaps. Consider his position.
Steve, as you're no doubt convalescing, re-register for BBC sounds and listen to 'how to vaccinate the world'. There are a few things explained there I think you'll find useful.
Yeah, sorry, too late to edit. I was a bit trigger happy there.
It looks like i'm getting it in the next couple of weeks so will report back if I grow scaly green skin
Efficacy is calculated based on events not total population treated. So you dose 40,000 people, half on each arm, and you look when you’ve acquired 100 infections. The null hypothesis is that there will be about 50 infections in a each. The alternative is that the split is not even because the vaccinated arm has done something.
The mRNA vaccines split about 5/95. Think heads and tails. The AZ/OX vaccine split about 29/70 (but they haven’t released the real numbers). Both are very unlikely by chance.
@Tired are you mistaken with your figures/calculation above or have I been misunderstanding the figures quoted for vaccine effectiveness ~95% for mRNA and ~70% for OX/AZ.
If someone told me a vaccine was 50% effective then I’d assume it would cut the number of infections in half, but the inference from your numbers above is it would do absolutely nothing, 50/50 infections vaccine/placebo
For the ~70% effective oxford vaccine where you quote 29/70, then zero effect would be 49.5/49.5 (99 samples). Therefore number of infections prevented based on 29/70 is 49.5-29=20.5, surely then the effectiveness is 20.5/49.5 = 41%
but the inference from your numbers above is it would do absolutely nothing, 50/50 infections vaccine/placebo
No, Tired said:
The null hypothesis is that there will be about 50 infections in a each. The alternative is that the split is not even because the vaccinated arm has done something.
The null hypothesis is essentially what would happen if there was no difference, in this case of 100 infections, you would expect about 50 from each if the drug does nothing vs placebo.
Therefore, if there IS a statistically significant difference between the arms it's attributable to the drug.
That is, assuming you get your randomization (who goes into what treatment arm) correct of course, and don't unconsciously bias one of the groups towards better/worse outcomes. Thats pretty basic and I wouldn't usually bother saying it - but you know, we seem to have given a load of people the wrong dose so.....
The null hypothesis is essentially what would happen if there was no difference, in this case of 100 infections, you would expect about 50 from each if the drug does nothing vs placebo.
Therefore, if there IS a statistically significant difference between the arms it’s attributable to the drug.That is, assuming you get your randomization (who goes into what treatment arm) correct of course, and don’t unconsciously bias one of the groups towards better/worse outcomes. Thats pretty basic and I wouldn’t usually bother saying it – but you know, we seem to have given a load of people the wrong dose so…..
Yes I know and that is what I stated - 50/50 is zero effect, based on the numbers tired gave the 95% mRNA is only 90% effective and the ~70% ox/az is only 41%. My point was using tired’s rationale for his numbers above then a completely useless vaccine would be rated at 50% effective (50/100) but the actual equation should be (50-50)/100 = 0% effective
Edit :- What I’m trying to say is if you have a 70% effective vaccine then in my mind the split should be 15/85 not ~30/70 as tired was suggesting.
I’ve just listened to the BioNTech boss on news night - the 90% efficacy means 90% reduction in symptoms.
I have it in my head the Ox/Az 70% is actually reduction in infection rather than Symptoms. They also have had no serious cases.
but the actual equation should be (50-50)/100 = 0% effective
Meant (50-50)/50 = 0%
I’ve just listened to the BioNTech boss on news night – the 90% efficacy means 90% reduction in symptoms.
I have it in my head the Ox/Az 70% is actually reduction in infection rather than Symptoms. They also have had no serious cases.
Didn’t they routinely screen the Oxford subjects irrespective of symptoms but only tested people who had symptoms for the biontech vaccine, so asymptomatic cases in the mRNA trial would be missed, so the Oxford vaccine might actually have a similar efficacy when that is factored in.
Yes, that was what I was trying to get at!!
Yes I know and that is what I stated – 50/50 is zero effect
Sorry Mudmuncher - I misunderstood your point. I think I'm probably conditioned to answering super-basic questions from elderly relatives!
@mudmuncher: no
Two key points:
1. The unblinding criterion. They roughly said "we'll stop and take a look when we get 100 infections detected". It's true that if the vaccine does nothing then when you look then you'd expect close to 50-50, but really what you're saying in that scenario is "in <however long it takes to get 50 infections in unvaccinated people> we got 50 infections in vaccinated people". However, imagine if the vaccine were perfect, then your first 100 detected infections would all be in unvaccinated people, and it would have taken twice as long to get there. Those unvaccinated people weren't affected by how good the vaccine is, so the effectiveness of the vaccine has a big effect (up to 2x) on the time at which you stop and look at the data!
2. Because of 1, your denominator is wrong. Easy way to think if it is "when we stopped and took a look, 70 unvaccinated people had been infected. If the vaccine did nothing then we'd have expected 70 vaccinated people to have been infected by then too". (It doesn't matter that if it *did* do nothing then you'd have got to the magic 100 sooner.) The sum then becomes "out of 70 people in that group who we'd expect to be infected if the vaccine did nothing, we actually only got 30 infections. So 40/70 were protected by the vaccine."