UK data: Impact of vaccines on deaths. Part 3: Recently vaccinated (within 21 days of 1st dose)

This is part 3 in a 4 part blog post series exploring what the data posted by the UK's Office of National Statistics (ONS) on November 1st tell us about the effect of the national vaccination program on death rates.

Part 1 focused on COVID-19 deaths.

Part 2 focused on all-cause deaths, comparing fully vaccinated (2 dose) vs. unvaccinated.

We showed the all-cause death rate is much lower in fully (2 dose) vaccinated individuals than unvaccinated individuals in the decade-based age groups 60-69yr, 70-79yr, and 80-89yr, showing a clear mortality benefit for these groups among those who were fully (2 dose) vaccinated. The higher all-cause death rate observed in fully (2 dose) vaccinated individuals in the 10-59yr age group has been championed by some as clear evidence of vaccine-caused mortality.

However, as I illustrated in part 2 of this series (and the Ghosbuster-themed parody, presented also in this twitter thread), this effect can be fully explained the the major age disparity between the vaccinated and unvaccinated individuals within the 10-59yr age group producing a Simpson's paradox distortion.

After accounting for these age disparities, the ratio of all-cause deaths between vaccinated:unvaccinated observed in this cohort is actually considerably lower than one would expect. Thus, the aggregated 10-59yr data do not provide any evidence of vaccine-caused deaths. If the ONS releases more granular data in this age group, we can further assess, but a reasoned analysis of these aggregated data does not suggest any.

In this blog post comprising part 3 of the series, we will evaluate the all-cause deaths of those recently vaccinated, i.e. within 21 days of receiving their first dose of vaccine, against those not yet vaccinated.

In part 4 to come, we will evaluate those partially vaccinated, more than 21 days after first vaccine dose, against those not yet vaccinated.

It is important to look at the partially vaccinated data, not just because the vaccine benefit takes a few weeks to kick in and the vaccine effectiveness of single dose vaccination is lower, but also to assess whether there is any evidence of a safety signal suggesting excess deaths after vaccination. If an individual has a serious adverse events or major complication after the first dose of vaccine, it is unlikely that they would receive a 2nd dose, making any safety analysis restricted to 2nd dose subject to serious selection bias, as rightly pointed out by many vaccine skeptics.

A complete population-based data source of all cause deaths separated out by vaccination status like the UK ONS data provides is far superior to detect any potential safety signal for death than an open reporting systems like the USA's VAERs, with lack of verification, lack of a control group, reporting bias, and underreporting among its substantial documented limitations.

It is of particular importance to look at the first 21 days after vaccination, given that most serious adverse events, if they are going to occur, will tend to occur very soon after vaccination.

Also, this is the time period of most of the anecdotal reports of deaths following vaccination that have circulated around social media, as well as most of the reported deaths after vaccination in open reporting systems such as the USA's VAERs system that have driven the well-circulated alarmist social media posts claiming that vaccines are killing people, with some even claiming that the vaccines are “killing more people than they save.”

In the USA, the deaths reported to the VAERs system are primarily concentrated in the few days after vaccination, which is pointed out by vaccine skeptics such as Steve Kirsch (who claims the vaccines are killing hundreds of thousands or more) to be a major reason why these deaths must have been caused by vaccines:

Any deaths occurring because of acute serious complications after vaccination should be visible during this time frame, so these data are useful for a careful assessment of whether there is any evidence of such deaths.

In this blog post, I will plot and evaluate the UK death data posted by the ONS on November 1, comparing the weekly all cause death rates from January 2, 2021 through September 24, 2021 between those recently vaccinated (within 21 days of first dose) and unvaccinated, split out separately by the age groups in the ONS report (10-59yr, 60-69yr, 70-79yr, and 80yr+). Here is a link to the Excel spreadsheet containing the data.

To make the plots maximally informative, the data are plotted as follows:

• I will plot all cause deaths, not just COVID-19 deaths, so they are independent of any testing practices or death attribution.

• I will plot the all cause death rates, which are the total deaths in that week for the given age/vaccination status group divided by the population of that group during that week. Plotting the raw death numbers would be misleading since it would not adjust for the number vaccinated, a measure varying substantially over time and by age group. Using these death rates automatically adjust for this factor.

• The population size of each age/vaccine status group changes greatly over time, with the vast majority of the population being unvaccinated at the beginning of the year, and becoming increasingly vaccinated as the year progresses, with the older groups vaccinated earlier and more extensively. To illustrate this effect, I will make the thickness of the line for each time plot proportional to the % of that age group with that vaccination status. By doing this, we can account for the population size for each vaccination status/age group over time, and can also see for a given age group when most received their vaccination, as indicated by the time of the year in which the thickness of the vaccinated lines is increasing rapidly.

• I will plot separately by age groups. Since both risk of death and vaccination rate vary strongly across age groups, failure to stratify by age under these circumstances can produce extremely distorted and misleading results as a result of Simpson's paradox, as previously shown in Israeli data. Stratification by time avoids this effect. As pointed out in part 2 of this series, the 10-59yr age group is far too wide given the wide disparity in vaccination rate and death risk in this age group, so this subset is still subject to distortion by Simpson's paradox and needs to be interpreted with this in mind.

• Further, we will plot the entire time curve, not aggregated summaries over time. Key pandemic factors including the overall SARS-CoV-2 infection levels, the predominant variant, the season of the year, and the mitigation strategies in place all change over time and could have a strong impact on death rates, and of course the vaccination rate has also changed substantially over time. Summaries that aggregate over time under these conditions can also produce a distorted and misleading picture of reality via a Simpson's paradox effect, as shown in a previous blog post. Plotting the entire time curves using weekly summaries avoids this effects.

For maximal transparency, here I share the script I used to download the data and produce these plots, so anyone can see the data for themselves and adapt the code to change the figures in any way they want. The plotting is done in the freely available statistical package R,

UK All Cause mortality recently vaccinated vs. unvaccinated.R

.txt

Download TXT • 5KB

Results:

I will perform the comparison for the older groups first, followed by the younger 10-59yr age group that is affected by differential age distributions between the vaccinated and unvaccinated groups that we previously observed in the fully vaccinated data.

Here is the plot for the oldest 80yr+ group:

We see that the all cause deaths were much lower in the recently vaccinated (within 21 days of first dose) than unvaccinated from weeks 0-10, a time during which 95.2% of this cohort was vaccinated. This was during the massive winter Alpha variant surge, and by week 10 onward was on the same level as the unvaccinated groups.

As of week 10, only 4.8% of this age cohort was unvaccinated, with the 95.2% vaccinated including 73.3% receiving their second dose, 20.7% receiving their first dose more than 21 days ago, and just 1.2% receiving their first dose within 21 days.

Those within 21 days of first dose after week 10 are an extremely small and select group, with only 6942, 2632, 1814, 913, at 501 individuals in this group during weeks 15, 20, 25, 30, and 35, respectively, out of a population of 2,518,496 in this age group during week 1.

We do not see any evidence of excess deaths in those recently vaccinated in this age group.

Here is the data for the 70-79yr old group:

For this group as well, the all cause deaths in the first 10 weeks during which the vast majority were vaccinated were considerably higher in the unvaccinated group than those within 21 days of first dose. If there were many vaccine-caused deaths, as some have claimed, we should have seen a signal during this time period. As of week 10, only 7.8% of this age cohort remained unvaccinated, with the 92.2% vaccinated including 1.1% receiving their second dose and 89.6% more than 21 days since their first dose, and just 1.5% having received their first dose within 21 days. Again, all the data for those within 21 days of first dose are sparse and selective after week 10, with only 11,503, 4605, 3232, 1855, and 978 individuals during weeks 15, 20, 25, 30, and 35, respectively, out of a population of 4,266,838 in this age group during week 1.

These small sample sizes lead to high volatility in the curves. For example, the visually evident volatility in weeks 18-22 corresponds to observed death rates of 5/4798, 13/4427, 4/4785, 1/4605, 4/4785, and 7/4175, respectively. So, while dramatic, there is no reason to interpret these spikes as indicating substantial vaccine-caused deaths.

We do not see any evidence of excess deaths in those recently vaccinated in this age group.

Here is the data for the 60-69yr old group:

In this age group, we see the all cause deaths in the first 15 weeks during which the vast majority were vaccinated were considerably higher in the unvaccinated group than those recently vaccinated (within 21 days of first dose). The lagged timing of the spike of unvaccinated deaths caused by the winter Alpha surge may be the result of longer time from infection to death in the 60-69yr than the older cohorts. As of week 15, only 6.6% of this age cohort remained unvaccinated, with the 93.4% vaccinated including 14.6% receiving their second dose, 77.5% more than 21 days after their first dose, and just 1.4% having their first dose within 21 days. The sparsity of data after week 15 for those within 21 days of first dose is also evident, with just 14,749, 8940, 5457, and 3096 individuals in this group during weeks 20, 25, 30, and 35, respectively. This sparsity produces the apparent volatility in the curves, for example with the data from weeks 35-38 resulting from observed death rates of 0/3096, 0/2814, 2/2559, and 1/2553, respectively.

We do not see any evidence of excess deaths in those recently vaccinated in this age group.

Finally, here is the youngest cohort 10-59yr:

From this we see the all-cause death rates of the newly vaccinated is lower than unvaccinated in weeks 0-5, similar in weeks 12-18, and much lower in weeks 19+.

However, the dominant feature in this cohort is a spike centered at week 7 in which the recently vaccinated in this age group have higher all-cause death rate than the unvaccinated. From weeks 5-11, the ratio of all cause deaths from recently vaccinated:unvaccinated are 1.0x, 1.4x, 2.0x, 1.9x, 1.5x, 1.4x, and 1.0x, respectively.

This difference could easily be an artifact of the differential age distribution between vaccinated and unvaccinated cohorts in the 10-59yr old age group.

Using official data on daily cumulative vaccination rates by age in England and age-specific mortality data for the UK in 2019, we can calculated the exact age disparity on a given date and assess the expected ratio of mortality risk in recently vaccinated:unvaccinated on that day using the methods laid out step by step in the previous blog post. We do this for 2/13/21, the middle of week 7 which is the center of this spike, computing the age distribution of those vaccinated with first dose within 21 days of 2/13/21 (between 1/24/21 and 2/13/21) with those yet unvaccinated as of 2/13/21.

Following is the respective age distribution, as of 2/13/21 of recently vaccinated and unvaccinated, along with the expected mortality rate of the two cohorts using a weighted average of the respective 5-year age groups computed from the official UK ONS 2019 age-specific mortality rates.

The calculations are detailed in the following spreadsheet:

Computing expected background deaths for vaccinated and unvaccinated subgroups of 10-59yr

.

Download • 22KB

To clarify and avoid confusion, nothing in this table has any data about actual deaths from vaccinated or unvaccinated cohorts.

It is presented to derive the expected death rates in the recently vaccinated and unvaccinated cohorts based solely on their differential age distributions.

• The bar plot in the figure shows the age distribution in vaccinated (blue) and unvaccinated (orange) cohorts, that shows the recently vaccinated group is systematically older than the unvaccinated.

• The black numbers at the top are the mortality rates by age from 2019, showing older people, unsurprisingly, inherently have a much higher risk of death.

• The bottom numbers are obtained by multiplying the mortality rate (black at top) and % for that age group (bar plot), which when summed across all age groups yields the expected death rate for vaccinated and unvaccinated cohorts based on their differential age distribution.

From these calculations, we see that based on the systematic disparity in age distribution between those recently vaccinated (within 21 days of 1st dose) and unvaccinated, we would expect the ratio of all-cause deaths between vaccinated:unvaccinated to be 233.7/145.1=1.6x as of February 13th (middle of week 5-11 spike), which is the same order of magnitude as what is observed during this week 5-11 spike. The ratios seen during this prominent spike in the plot are not out of line with the inherent mortality ratios based on disparate age distributions of the two groups.

Thus, there is no evidence of excess deaths in the recently vaccinated subset of this 10-59yr cohort either.

If the ONS releases more granular data in this age group, we can reassess these results, but these aggregated data do not provide any evidence of a measurable number of vaccine-caused deaths.

Conclusions:

In this blog post, we have compared the weekly profiles of all-cause mortality between January 2, 2021 and September 24, 2021 in the UK between those recently vaccinated (within 21 days of 1st dose) and those yet unvaccinated. We see no evidence of any excess mortality during this time period that would indicate a substantial number of all-cause deaths caused by acute vaccine-induced complications. For the 60+ decade age groups, we see lower death rates during the winter Alpha surge in the just-vaccinated group relative to unvaccinated, and then roughly equivalent death rates afterwards, during a time with very sparse data in the newly vaccinated cohort.

We see no evidence in the 10-59yr old cohort of excess deaths in the newly vaccinated cohort either, with lower all-cause deaths in the newly vaccinated group in the first 5 weeks, similar in weeks 12-18, and much lower in weeks 19+. An initial (naive) view of the aggregated data for weeks 5-12 might suggest excess deaths in the newly vaccinated subset. However, a more careful look that accounts for the age disparities between the groups shows that this difference is the magnitude expected based on their actual respective age distributions, so does not provide evidence of excess deaths. This age group can be re-evaluated when/if the UK ONS releases more granular data for this age group

Given many of the anecdotal reports and claims of vaccine safety alarmists (e.g. Steve Kirsch) focus on the overabundance of deaths within 21 days of vaccination based on VAERs reports, we can safely say that these UK ONS data show no evidence of any noticeable number of such deaths.

Kirsch claims that the higher number of deaths reported to VAERs in the days immediately following vaccination provide evidence they are not background deaths, but vaccine-caused, as shown in his recent slide deck:

From the UK ONS data presented here, there is no evidence of excess deaths in the time frame immediately following vaccination as he suggests should be the case.

Kirsch never considers that deaths occurring within a few days of vaccination are much more likely to be reported to VAERs, and thus the fact "these bars are not the same height" might be as a result of reporting bias, with deaths occurring within days of vaccination much more likely to be reported to VAERs, and not “proof” of vaccine-caused deaths as he tries to argue. This is one example of the myriad scientific, logical, and data interpretation errors made by Kirsch and other vaccine safety alarmists in formulating their narratives that "vaccines are killing more than they save.”

While we have now shown that the UK ONS all-cause death data show no evidence of excess deaths for 2-dose vaccinated and recently (within 21 days of 1st dose) vaccinated cohorts, there is still one more cohort in the UK ONS data set: those given first dose more than 21 days ago, but not yet receiving second dose.

In principle, this cohort could contain individuals who experienced vaccine-induced serious adverse events after first dose that did not result in death within 21 days, but prevented them from receiving 2nd dose, and in some could have later led to their death. These deaths would be recorded in this remaining cohort, so it is important for us to carefully evaluate this cohort as well before completely concluding this UK ONS data does not suggest evidence of vaccine-caused deaths.

Part 4 of this series will perform a detailed analysis of this cohort.