The latest ONS data on deaths by vaccination status show that of the 5,678 Covid deaths in April and May, 93% or 5,276 were of vaccinated people. Given that the Government’s Coronavirus Dashboard reports that 93% of the over-12 population had a first dose of the vaccine, 87% had a second dose and 70% had a third dose, these real-world data suggest that the vaccinated did no better than the unvaccinated in being protected against death.
Breaking this down by age does not improve the outlook. In the 50-59 age cohort, in April 86% and in May 87% of the deaths were of vaccinated people, compared to 87% of the cohort vaccinated with at least one dose. In the 60-69 cohort, in April 88% and in May 91% of the deaths were of vaccinated people, compared to 91% of the cohort. In the 70-79 cohort, in April 94% and in May 90% of the deaths were of vaccinated people, compared to 92% of the cohort. In the 80-plus cohort, in April and May 94% of the deaths were of vaccinated people, compared to 94% of the cohort. Note that ONS’s definition of a Covid death includes deaths where Covid is mentioned anywhere on the death certificate, not necessarily as the underlying cause.
Many readers will be familiar with the ongoing debate about vaccination statistics and the size of the unvaccinated population. We know how many people have been vaccinated with the first, second, third and fourth doses of vaccine but we don’t know what the population of the U.K. is so we can’t know exactly what percentage of the population the vaccinated represent. However, we can have a pretty good guess. Figure 1 is taken from the Covid dashboard and shows that just over 90% of the population had a first dose, just under 90% had a second dose and just under 70% had a third dose.
Of course, vaccine take-up varies greatly with age and vaccine efficacy (or lack of efficacy) depends on how many doses you’ve had and when you had your last dose.
Figure 2 shows vaccine uptake for the first three rounds of vaccine, though not the spring second booster campaign among over-75s. On the right hand side of the chart I’ve drawn the 90%, 80% and 70% lines in red to make reading easier. I’ve then written in red the age cohorts that fall in various percentage tranches. Broadly speaking, 90% of people over the age of 70 have had all three doses whereas, if you look at the 50-59 year-olds you’ll see that only about 75% have gone on to have the booster dose.
I’ve then inserted a table showing the approximate uptake of the various vaccine doses by age cohort that will then be used to compare vaccine uptake to the death data.
I should add that this level of accuracy is somewhat spurious and overstates the percentage of the various age cohorts that have taken up the vaccines. There’s been quite a lot of controversy about the percentage take up of vaccine due to the mystery surrounding the U.K.’s population. The problem was first addressed by Professor Sir David Spiegelhalter, who brought up the complications around the NIMS and ONS datasets; I wrote a piece taking issue with some of his findings here. Professor Norman Fenton also covered this in his recent critical review of the BBC’s “Unvaccinated”. I won’t revisit the debate here except to say that I think most would agree that the estimates for the proportion of people vaccinated by age group tend to be a bit overstated, so let’s simply regard them as a ceiling.
It’s perhaps also worth noting that Professor Fenton has also criticised the ONS data for undercounting Covid deaths, but, as with the percentage of vaccine uptake, I’m going to use the ONS’s figures as that is what the Government has published.
The ONS published “Deaths by vaccination status, England” on July 6th. I, along with many others had been complaining that since the end of March there had been no raw data released showing deaths by vaccination status. It was a pleasant surprise to me that this new report made good this shortfall and included deaths by vaccination status for April and May.
Figure 3 shows the count of Covid deaths by vaccination status for April and May 2022. In April there were 4,119 Covid deaths, of which 3,837 were of people vaccinated with at least one dose, that’s 93% of the total. In May, 1,559 Covid deaths were recorded, 1,439 of which were of vaccinated people, that’s 92% of the total. These figures are in line with the proportion of deaths amongst the vaccinated being shown in the UKHSA data up to the point when the data stopped being published at the end of March. The final report can be found here, with the real-world data found in tables 12 and 13, which show that the proportion of deaths by vaccination status were much the same then as in April and May.
The maximum percentage of the population vaccinated is just over 90%, yet the percentage of deaths in April and May was also over 90%. Why is this important? Because, what’s always needed in any experiment is a ‘control group’, a group of people who didn’t participate in the experimental drug, vaccine or whatever was the subject of the study. If the vaccines are effective in preventing deaths you’d expect to see a higher rate of fatalities among the people who didn’t take the vaccines. But we don’t.
Let’s go into a little more detail. Around 99% of the deaths in April and May occurred in the over-50s so I’ll only look at these older age groups.
In the 50-59 age cohort there were 105 deaths in April and 47 in May. In April 86% and in May 87% of the deaths were of vaccinated people. The third column in grey shows the estimated percentage of people in this age cohort who had either no vaccine or the various doses. The unvaccinated do not appear to have suffered disproportionately to the vaccinated. While at least 13% of this cohort have never had any vaccine dose at all, they accounted for less than 14% of the deaths in April and May.
Figure 5 shows that it’s the same story in the 60-69 year-old cohort. In April there were 284 deaths of which 250 or 88% were of vaccinated people. In May 106 of the 117 deaths or 91% were vaccinated. The proportion of deaths was in line with the proportion of the vaccinated in that age group. If the vaccines had no effect this is exactly what you’d expect to see.
Figure 6 tells the same story for the over 70s: deaths from Covid are proportionately the same for the vaccinated and the unvaccinated. About 91% are vaccinated and over 90% of deaths were among the vaccinated.
And finally, figure 7 shows the situation for the 80+ cohort. In this age cohort, just about everyone who had the first dose turned up for their second and third dose and quite probably their fourth dose too, but it doesn’t seem to have made much difference to the outcome. Again, it’s perfectly clear that with 94% of the deaths occurring amongst the vaccinated, the 6% unvaccinated are proportionately no more at risk that the vaccinated.
Similar findings were reported in the Daily Sceptic recently for Canada and the Netherlands so this isn’t British exceptionalism.
Why might this be? Waning efficacy perhaps. However, they’ve been at this level for some months, suggesting that isn’t the whole story. Maybe the unvaccinated are a self-selecting group who are disproportionately healthy or have been previously infected – though if so that would need to be shown not assumed. Whatever the reason, the ONS data clearly raise important questions that need addressing. They seem to confirm that in the real world, regardless of how many doses of vaccines someone may have, it makes absolutely no difference to the likelihood that they may die from Covid.
Postscript: It has been pointed out that some experts have warned about using raw counts, like the number of COVID-19 infections and deaths among the vaccinated versus the unvaccinated, to draw conclusions about the effectiveness of the vaccines. According to a November 2021 FactCheck article, “If the large majority of a population is vaccinated, it’s not surprising if most deaths are among the vaccinated… It’s simply math.” Moreover, the U.K. Health Security Agency said in its July 7th, 2022 report that for people aged 50 and over, COVID-19 vaccines were 47% effective against COVID-19 deaths 25 weeks or more after the second dose, and 87& effective 10 weeks or more after a third dose.
In Nick Rendell’s defence, his article was expressly looking at the death rate compared to the vaccination rate in the population, so the FactCheck article is not relevant. The point being made in his article is that the proportion of deaths in the vaccinated is higher than the proportion of the population that is vaccinated (and this applies to each age group so is not just a result of older people being more likely to be vaccinated). Notwithstanding the UKHSA’s data on the efficacy of the COVID-19 vaccines, this doesn’t change the fact that the vaccinated groups are not dying with Covid at a lower rate (per population) than the unvaccinated, according to the ONS data. True, the ONS warns against interpreting its data to measure the effectiveness of COVID-19 vaccines, stating: “The [age-standardised mortality rates] are not equivalent to vaccine effectiveness and both the COVID-19 and non-COVID-19 ASMRs can be affected by other various factors, such as health status and changes in mortality rates over the year.” Nick Rendell mentions these potentially confounding factors in his article, but nevertheless we believe that looking at differential rates of outcomes by vaccination status is a valid way to measure vaccine effectiveness.