Cage Match: Vinay Prasad's "staggering efficacy in adults" vs. Summer 2021 data
In one of my 1st posts, I questioned how the claims of vaccine efficacy against death could be reconciled with what we saw in the summer of 2021, by which time a significant portion of the population (and a large majority of the elderly population) had been vaccinated.
This topic has re-entered my mind as I was just listening to Vinay Prasad’s recent appearance on EconTalk with Russ Roberts. Dr. Prasad has had very good things to say with regards to disaster of school closures, the lack of evidence for masking, and the potential harms of Covid vaccines, particularly in young. In this discussion and his recent paper, however, he starts with the premise that these vaccines showed remarkable efficacy against serious illness and death , and saved countless lives, particularly among the elderly. In fact, his recent paper began “The COVID-19 vaccine has been a miraculous, life-saving advance, offering staggering efficacy in adults, and was developed with astonishing speed.” There are other folks whose opinion I respect who would support that claim. Hearing this, my mind drifted back to the thought that I can’t square what we saw in the summer of 2021 with that supposition, so I revisited data from that period again with a focus on the older populations and the six months of May- October 2021.
To start, let’s recognize that per the CDC data, 84% of seniors had at 1 dose, and 74% had competed the series of the vaccine by the end of April of that year. The vaccines, therefore, should have been at maximum efficacy for the summer months.
Source: CDC Vaccination History
Next, we will look at the impact on deaths (COVID & All Cause) during these summer months for the older groups (data source: CDC Wonder). Starting with COVID labeled deaths, we have the following:
From above charts, of particular interest is the comparison between 2020 (pre-vaccine) and 2021 (74% of seniors fully vaccinated by end of April). From the above charts, one could certainly find support for the notion that the vaccines prevented significant COVID deaths in May, June, and July. However, what we see in August, September, and October is quite remarkable. For the groups between 65 & 79, significantly more deaths attributed to COVID happened in 2021 vs. 2020. In fact, September of 2021 was one of the worst months, in terms of COVID attributed deaths, of the entire pandemic for these age groups, despite an 84% fully vaccinated rate of seniors heading into September. For the 80-84 and 85-89 year olds, the jumps in late summer are not as significant, but we also don’t see evidence of reduced COVID deaths.
In terms of All Cause deaths, we see a similar story:
By looking at all cause deaths, we can have confidence that what we saw above is not just an artifact of different attribution of COVID to deaths over time. Consisted with above, we see the 2021 Aug-Oct is particularly bad for 65-79 year olds & while not as bad, compared to 2020, for the >80 group, we can also see that 2021 was very bad compared to 2018 & 2019. Further, other than the 85-89 group, as of 2022 we saw little evidence of all cause deaths returning to their 18/19 levels.
I continue to be open to the notion that the COVID vaccines saved many elderly lives, but, as I said, I can’t square that with the summer of 2021. If anyone can help me make sense of this seeming contradictory evidence, I’m all ears.
Contradictions do not exist. Whenever you think you are facing a contradiction, check your premises. You will find that one of them is wrong. -Ayn Rand
The question to pose is "Ok, up to ~15% of seniors experienced a first time infection with no vaccine beforehand in summer 2021 - is that in the ballpark of how many experienced an infection before that point?" If it is in the ballpark then you have enough fuel to generate an AUC of senior deaths in the Delta wave that is higher than the AUC of 2020-May2021: Essentially, you can get a peak twice as high as the average in 2020 because you have all these first time infections squeezed into fewer months during the Delta wave.
Here is how that math works out. In blood donors, the 65+ were in the low 10s% seropositive in May 2021, though donors might introduce a bias for more risk-averse and pro-vax, it still should be ballpark for the future unvaxxed, https://jamanetwork.com/journals/jama/fullarticle/2784013 fig 3.
So now we have this idea, 10-20% of seniors had an infection before May 2021, this generated an AUC of deaths of "1 unit." Next, 16% of seniors are unvaxxed in May 2021, 10 - 20% of them have natural immunity, so if all of the ones without natural immunity are fuel for Delta, and let's say nine out of ten of those wind up infected with Delta, maybe 13% of all seniors are "unvaxxed with infection in Delta wave." So right away that's close to how many were unvaxxed and infected before May. Given that Delta is more severe, can you get AUC of "~1 unit" again in the Delta wave?
Yes. Easily.
Without the vaccines, but supposing that Delta infects nine out of ten of the 80-90% of seniors with no prior infection as of May 2021, what AUC is expected? ******Up to "7 units."******
Now, for me, this doesn't say that seniors should get vaccinated. It's a big number but plenty of seniors would have been fine - after all, there are plenty of unvaccinated people of all ages still around after Delta and up to today. But it's clear you can start from the expectation that seniors still had "7 units" of AUC deaths to generate in May 2021 and that amount never happened, either because the vaccines prevent death or they put off a lot of those still-to-occur first time infections until Omicron which was milder.
A lot of this comes down to a very common false premise when analyzing 2021 deaths which is to benchmark 2020 as "the worst the virus could do" when in most of the West (aside from NYC, Italy) it was just 1/5 to 1/9 of what it could do. Again, still not a big deal in the grand scheme of things, but more than what we saw afterward.
Oh dear, that data is not dispositive in favor of the vaccines... quite surprising.