Just as a clarification of the significance of these reports: Phase III clinical data is normally the last data in, so these two vaccines should have the last components of their product licence applications submitted to regulatory bodies imminently. In most territories rolling review of the available early data (ie things like chemistry and preclinical) will have begun some time ago. In the current climate, I would have thought that, for these new vaccines, authorisation/emergency supply etc should be possible very quickly, perhaps in a matter of weeks.
Still no new record daily highs in reported cases in any state (although a couple of them show sharply rising trends). It’s been quite a while now.
Even internationally, things are trending down. Portugal’s numbers are starting to decline, although Spain has hit a new record high and is going sharply up.
So the Johnson & Johnson results are a little disappointing and a little odd.
Only 72% effective in the US trials, only 66% in Latin America and only 57% in S. Africa. There are some fears that it might not do great with newer variants.
The good news, 85% effective in preventive severe forms that would require hospitalization.
So remember, this is a single shot vaccine that doesn’t require special freezers.
It will probably get approved by next Friday from what I can glean.
So there are 16,513,896 cases that have been resolved one way or another. Of those, 443,769 cases have been resolved by death. 443,769 ÷ 16,513,896 = 0.02687. So am I calculating correctly if I say the mortality rate at the time the data were posted is 2.7%?
Yes and no. The mortality rate for people who were tested is 2.7%, but this includes all the people who experienced symptoms severe enough to go get tested. There’s almost certainly a large number of people who actually caught the disease but had a very mild case or were asymptomatic who didn’t get tested, and therefore didn’t get diagnosed or recorded. So the true mortality rate is probably substantially lower.
The trials on the J&J vaccine were on a single shot - 60-odd percent effective overall. It’s an adenovirus vector vaccine, as is the A-Z vaccine - which after a single shot was also 60-odd percent effective, IIRC. However, the A-Z vaccine had a second shot in the clinical trial dosage regimen, after which the efficacy was in the 80 - 90% range, depending when the second shot was delivered (again, IIRC). Consider:
Crucially, the Janssen trial looked at giving just one dose of the vaccine, which makes it significantly easier to roll out than those requiring two. It is also investigating whether giving two doses will give either stronger or longer-lasting protection.
So 1 shot was comparable with 1 shot of A-Z. Maybe 2 shots will also be comparable with 2 of A-Z. We shall see.
The regional variations are concerning though, I agree.
Yeah. A while back I mentioned that a 60-70% effective vaccine was pretty good and a self styled internet disease expert told me how silly I was and people only think that because flu vaccines aren’t very good.
The number you quoted is known as the “case fatality rate.” That number includes only confirmed cases. However, there are far more cases than are officially recorded - somewhere in the neighborhood of 10x. The number that includes all cases is called the “infection fatality rate.” That number is currently believed to be 0.26% with huge variation based on age:
You know, I have a theory that some folks react that way to such news because, as a matter of literal life and death, they feel judged for not doing it, ie “not wearing three layers on your face is irresponsible and stupid. Enjoy killing yourself and everyone you come into contact with, asshole.”
10x is likely too high at this point, I agree. A few months ago the best estimate was 8x-10x. CDC’s current best estimates are 4x-5x confirmed cases, so in the neighborhood of 85,000,000-100,000,000 cases in the US, or approximately 1/4 people.
No, that wasn’t the best estimate. That was one estimate of many, all of which can’t be right.
The CDC’s most recent estimates are between 72 and 97 million actual infections at the end of December, with a 95% confidence interval. I can find other sources that say 3x, others still that say 20x. The real problem is that almost every seroprevalence study out there is not a random sample, but instead lab sampled. We can’t currently extrapolate the general population from the population who are making lab visits and giving blood.
We don’t know.
Either way, I’m happy to drop this as you’ve backed off your 10x claim.
I know this is a bit of a hobby horse for you, so not to continue the hijack, but I think Belgium’s numbers and methods might interest you. They seem to have a high covid case count compared to other western countries but they don’t have the “extra deaths”, which might suggest their accounting method is more accurate.
I started looking around at this and I can’t tell what is true. It appears that they have some combination of excellent COVID death reporting in concert with vastly underreported case counts, or if they just suck at keeping people alive (I’m doubtful of this, but still uncertain), or something else entirely. They have a CFR of almost 3% while the US is around 1.6%.
So, while I think there’s a good chance that you are correct with them being somewhat accurate with their COVID death reporting, that would mean that either their testing was underwhelming (furthering the gap between case count and infection count) or that the US is missing half of its COVID deaths. While I do believe we have an issue with underreporting, I’ve seen no evidence that it could be that far off.
I continue to feel that we are not only seeing people dealing with imperfect data, we’re also seeing them extrapolate things from that imperfect data incorrectly. It’s not quite as bad as that early survey that stopped people on a street corner for blood testing in a busy area around Boston back in April, but we still suck at it mostly. I get the desire to publish so I understand it, but it doesn’t make the science better. It’s also why we have shit like Hydroxychloroquine stockpiled in Oklahoma.
Again, probably not the place for this rant, but I do miss good science.
