Vaccine Effectiveness Heads Towards Minus-400% as Death Rate Among Vaccinated Approaches Same as Unvaccinated, UKHSA Data Show

The clock might be ticking down to the final UKHSA Vaccine Surveillance report that includes data by vaccination status, but at least this week there are more data to explore.


Anyone managing to avoid the 24/7 coverage of the disaster in Ukraine might have had the opportunity to be told about the ‘disaster’ that is the ever increasing number of Covid cases in the country. The UKHSA data are a week or so behind the times, but nevertheless they are now showing a substantial increase in cases compared with last week’s release (details of methodology and limitations can be found here).

A comparison with last week’s data shows that in the 18-80 age range cases are up about 30% in the triple vaccinated, whereas they’re only up approximately 15% in the unvaccinated and those vaccinated with only one or two doses. Thus it appears that this is a Covid wave of the boosted. This trend is most visible in the rate of change of case rate over the last few weeks (data shown for those aged 40-80 only).

Not only are the case rates in the unvaccinated far lower than the triple jabbed, but in recent weeks they’ve also not been rising at the same rate either. Hopefully we’ll get at least one more data point from the UKHSA to see how this changes as we progress further into the current Covid wave.

It isn’t clear whether the rise in cases in the unvaccinated and one or two dosed is because of an intrinsic sensitivity to what is probably Omicron BA.2 variant, or whether it is because of the immense infectious pressure caused by the high levels of infection in the boosted.

The press is full of stories about how the BA.2 variant is as infectious as measles and thus it shouldn’t be surprising that we’ve got ever increasing case numbers at present. However, it seems that this isn’t the case everywhere. For example, compare Covid case rates in the U.K. with those in South Africa.

Note how both the U.K. and South Africa had a pronounced Omicron wave at the end of 2021, but South Africa just hasn’t had any BA.2 variant wave at all, let alone a pronounced one. Omicron BA.2 is known to be present in South Africa, so it isn’t simply because it hasn’t got there yet. Readers are welcome to suggest any differences in the populations of the U.K. and South Africa that might explain their lack of a BA.2 wave.

The relative increase in infections has of course led to changes in the estimate of vaccine effectiveness against infection this week.

The raw UKHSA data is once again suggesting that vaccination increases the risk of infection, and the more doses you receive the higher the increase in risk and thus the lower the vaccine effectiveness estimate. The unadjusted vaccine effectiveness in the triple-jabbed in the 60-69 year-old age bracket is pushing on for minus-400%, corresponding to an infection rate five times that of the unvaccinated.

The drop in the estimate of vaccine effectiveness against infection since the start of the year has been rather remarkable.

In those aged under 50, for those that have received only one or two doses of vaccine there appears to be a slow recovery back to the baseline levels (i.e., unvaccinated). However, for the boosted we see the year starting with low vaccine effectiveness (possibly reflecting an increased susceptibility to Covid immediately after vaccination), then a sharp rise in effectiveness (yet still deeply negative) for a few weeks as the vaccines went through a short period when they at least tried to offer protection, and then a week-by-week reduction in effectiveness that shows no signs of slowing down. Whether this is a simple time effect following a vaccine dose (and the triple vaccinated will eventually recover also) or whether it is a dose effect (and the triple vaccinated will remain more susceptible to Covid) remains to be seen.

A similar trend is seen in the data for those aged over 50, albeit without the initial with very low vaccine effectiveness (this may reflect those aged over 50 getting boosted earlier than those aged under 50).


Hospitalisation rates remain low; this is likely because Omicron variants have intrinsically lower pathogenicity. Hospitalisation rates in the younger age groups are particularly low.

The data appear to show little effect of vaccine against hospitalisation for those under the age of 60. The outlier is for those aged under 18 – this is probably an artefact related to the specifics of which individuals were vaccinated and when, but it is possible that Omicron BA.2 has mutated to become more infectious in this age group.

The change in hospitalisation rates by vaccination status since the start of the year shows some interesting effects:

In general the effectiveness of one or two doses of the vaccines at preventing hospitalisation has reached a steady state at about minus-50% and the protection offered by three doses of vaccine has been dropping each week, though remains positive – albeit marginally so for those over 80.

However, there appears to be an improvement in vaccine effectiveness for some age groups, most obviously seen in the data for those aged under 30. This is an interesting effect and may be related to differences seen in the time course of infection discussed in previous posts. A similar effect was seen at the start of previous Covid waves. It was first noted back in June 2021, when at the start of that Covid wave there were many stories in the press about how the unvaccinated made up the vast majority of hospitalisations; there were no updates in the press a few weeks later when the majority of hospitalisations were in the vaccinated. Thus the apparent increase in vaccination effectiveness seen in the triple-jabbed young may be an artefact related to the unvaccinated becoming hospitalised sooner, and, if so, it should reverse over the coming weeks.


The data for deaths by vaccine status show a similar progression as seen for hospitalisation – those that have been given only one or two doses of vaccine show an increased risk of death compared with the unvaccinated, whereas the triple-vaccinated appear still to have some protection.

Of note in the above graph are the data for those aged 40-50; in this age group the protection offered by three doses of vaccine is now close to zero.

As I have mentioned in previous weeks, it is possible that the mortality data has been corrupted by those closest to death not being vaccinated with the latest dose (a form of the healthy vaccinee effect) – this would have the effect of raising the apparent effectiveness of the latest dose of vaccine but would substantially lower the apparent effectiveness of one or two doses. Because of this effect I believe that it is more instructive to analyse the deaths data by ‘any dose of vaccine’ versus the unvaccinated. In our exemplar group (those aged 40 to 50) the mortality rates in the unvaccinated are now statistically identical to those that have taken at least one dose of vaccine.

While the data suggest that the mortality rate within 60 days of a positive test is now greater in the vaccinated, this is not yet statistically significant – data for other age groups suggest that the vaccines simply offer no protection against death after infection with Omicron variant in those aged under 70.

It is important to recognise that the death rate from Covid in the younger age groups is very low, and even if there were a slight increased risk of death in the vaccinated aged under 70, this would not have much real world impact (although the vaccinated might find it very annoying). Also, note that these are only the deaths within 60 days of a positive test – there is no information on non-Covid death rates and vaccination status in the UKHSA data.

Last week I noted that death rates in the triple-vaccinated between 28 and 60 days after a positive test are now proportionately greater than those in the unvaccinated.  This effect has only increased in magnitude in the latest data:

It is unclear why there should now be so many deaths in the period between 28 days and 60 days following a positive test, but these latest data offer more evidence that any Covid statistics that only consider deaths within 28 (or 30) days of infection will be significantly under-reporting the true death rates from (or with) Covid.

Amanuensis is an ex-academic and senior Government scientist. He blogs at Bartram’s Folly.

Notify of

Profanity and abuse will be removed and may lead to a permanent ban.

Newest Most Voted
Inline Feedbacks
View all comments
November 2022
Free Speech Union

Welcome Back!

Login to your account below

Create New Account!

Please note: To be able to comment on our articles you'll need to be a registered donor

Retrieve your password

Please enter your username or email address to reset your password.