A new week and a new update from the UKHSA on the performance of the vaccines. Table 12, the rates data for vaccinated and unvaccinated, once again shows data for those triple-jabbed rather than two-or-more doses as was given for the data for 2021 since September up to Week 2’s report, however, we can estimate the rates for those with only two jabs using data in tables 9-11 and vaccination data available from the NHS.
Using these data we can estimate infection rates by vaccination status (above); from these data we can then estimate the unadjusted vaccine effectiveness for infection for those having two or three doses (below).
The vaccine effectiveness data for those receiving two doses are particularly concerning. Because we do not have full data available to us there is a risk of some biases creeping into the data and making things look worse than they are in reality. However, the very very negative estimates of vaccine effectiveness for the double-jabbed shown in the graph above (lower than minus 300% in those aged 40-59, meaning four times as many infections than in the unvaccinated) are likely to be much greater than what could easily be accounted for. What’s more, the data show significantly negative vaccine effectiveness for those aged between 18 and 50, a group which has not received the booster in particularly high rates, reducing the impact of some potential biases that occur with small numbers in each cohort, meaning we can be more confident the figures are closer to reality.
The new estimates of vaccine effectiveness by dose for this week allow us to continue the graphs of the change in vaccine effectiveness with time.
Note that data for 2022 is separated out into those with two doses only and those with three doses only. Note also the scale on the y-axis: almost all the points are highly negative; those for two doses are hundreds of percent below zero, equating to infection rates many times higher than in the unvaccinated.
It is interesting to see in the graph that for those aged over 60 the trend in the data from last year most closely tracks into those having had a booster dose for the 2022 data, while for those aged under 50 the data most closely follows those only having had two doses for the 2022 data. This reflects the earlier boosting and much higher booster rates in those aged over 60 producing the ‘hump’ in the estimate of vaccine effectiveness going into the last month or two of last year.
The negative estimate of vaccine effectiveness shown in the graph above will be affected to some extent by the biases inherent in the raw, uncorrected data that we have available to us. However, it reflects the negative effectiveness seen elsewhere. Of course, the UKHSA would be keen to point out that its own data still shows at worst zero vaccine effectiveness, but their estimates are influenced by biases of their own, mainly associated with their choice of method (test-negative case-control (TNCC) methods are significantly affected by testing biases in the vaccinated versus unvaccinated populations – a bias warned against in scientific papers that describe how the TNCC method should be used). It is likely that the true vaccine effectiveness lies somewhere between the raw analysis shown above and the ‘zero or slightly negative’ estimate given by the TNCC method. I note that the cohort studies published last year suggested that two doses of vaccines had a negative effectiveness of between approximately minus-20% and minus-150% (for example, see S12 in a study of infections in Qatar, a study of infections in Israel National Airport, or cohort study undertaken by Imperial College, data given in Table 2) – given that Omicron has further escaped the protection offered by the vaccines it is likely that the true vaccine effectiveness is now significantly negative and that the estimates shown in the graph above aren’t too far out.
It is likely that the small uptick in apparent vaccine effectiveness in this week’s data indicated in the graph above is an artefact related to the biases that remain in using raw data – if only the UKHSA would gather and release better data…
Of course, the approach we’re taking to estimate vaccine effectiveness should be used with caution; the question is how much bias it introduces. While there will be some bias introduced by using raw data, the impact should not be too great and certainly the trend in the data shows a very worrying situation. The authorities could assist by releasing more useful data or by undertaking vaccine effectiveness studies that are more robust (what’s needed are large scale matched cohort longitudinal studies). However, they’re extraordinarily keen not to do this and merely rely on the highly biased TNCC method (presumably because it significantly overestimates vaccine effectiveness). Of course, more robust estimates would cost a little more money to undertake – but it isn’t as though governments worldwide aren’t throwing huge amounts of cash at other aspects associated with the Covid pandemic (for example, the nearly useless Test and Trace programme). As it stands, the data that are available to us paint a very worrying picture that really should be rigorously investigated by our health ‘security’ authority.
It is important to note that the impact of negative vaccine effectiveness on case numbers would be greater than suggested simply by looking at the proportional risk of infection indicated by the estimates of vaccine effectiveness – this is because infectious diseases have an exponential growth pattern. The impact of highly negative vaccine effectiveness on case numbers would be for there to be a very rapid increase in cases to a much higher eventual peak, than would have been the case with only unvaccinated – note the speed and scale of this winter’s Omicron wave…
We can also analyse the new data to get hospitalisation and death rates by vaccination status.
Hospitalisation data suggest that only having two doses of the vaccines now have negative protection against hospitalisation for all age groups aged 18 or over. The data suggest that three doses are still offering some protection, although even this protection appears to be waning.
The trend lines in the data suggest that there has been a significant drop in vaccine effectiveness against hospitalisation with the Omicron variant. However, there was probably a more gradual drop over the last few months of 2021, hidden in the data above by the insistence of UKHSA in not releasing data by vaccine dose for two and three doses separately. Perhaps the very poor performance of the vaccines at protecting against hospitalisation is the reason why they didn’t want to release these data?
If anything, the data suggest that the protection against death (below) offered by two doses of vaccine is even worse – however, it is possible that the particularly negative data for those aged over 70, and to a lesser extent those aged 60-70, reflects a ‘healthy vaccinee’ effect. This would occur if those most close to death were spared from the booster vaccine, resulting in relatively higher death rates in the small two-dose-only group. However, the booster take-up in those aged under 50 is much lower, and would mean that a healthy vaccinee effect would not significantly affect the data – thus it is very likely that the loss in protection against death offered by two doses of vaccine has fallen to zero, as indicated by the data, and might well have fallen below zero for those aged over 60 (i.e., those whose vaccine protection has waned the most).
If we are seeing a ‘healthy vaccinee’ effect then we should expect to see the very negative effectiveness figures for those aged over 60 rebound a little further and then flatten out over the next few months.
Finally, in all of the data above it is clear that the booster doses are now having little positive protection against infection (and probably make infection more likely), but still have some worthwhile protective value against hospitalisation and death. However, it is likely that even the booster will repeat the spectacular drop in the effectiveness of two doses of vaccine seen in recent months. While our authorities will be keen to promise that the booster can just be repeated at regular intervals, immunology is not that simple. The boosters appear to work by increasing antibody levels even further, rather than boosting antibody levels back to ‘where they should be’. It is likely that further boosters will have a small positive impact and for a shorter duration of effect. Omicron specific boosters might well improve on this,. However, it is likely that these will merely introduce new evolutionary pressures in the virus to evolve escape to those vaccines, and might introduce new viral behaviours that won’t be as benign as those apparently seen with Omicron.
In summary, the latest data suggest that:
- Infection rates are soaring in the vaccinated, with even those boosted seeing increased case rates than the unvaccinated. The apparent lack of any ability of the vaccines to protect against infection (and onwards infection) suggests that the current mania in governments worldwide to vaccinate everyone to keep Covid under control is counterproductive. The UKHSA really should comment on the purpose of vaccinating the non-vulnerable in the population.
- There now appears to be an increased risk of hospitalisation for those with only two vaccine doses. This risk appears to be rather high for those aged over 70, but it must be considered that this might be an artefact due to a healthy vaccinee effect. The data for younger individuals should be more robust. Urgent work should be undertaken by our authorities to fully understand what is going on with hospitalisation rates. It may be necessary to inform the vaccinated that they need to keep getting booster doses to keep their risk of hospitalisation below that of the unvaccinated.
- The increased risk of death in the double-vaccinated compared with the unvaccinated is troubling. I hope that this is an artefact due to a healthy vaccinee effect and that the real situation is only that vaccine protection against death falls to zero, as suggested by the under 50 data. I also hope that things don’t get any worse.
Hope is not a strategy.
Amanuensis is an ex-academic and senior Government scientist. Find him on his Substack page, Bartram’s Folly.