Should Oxford Be Trusted to Assess the Safety of its Own Vaccine?

In December 2021, I received the Pfizer COVID-19 booster while pre-symptomatic with Covid. These events culminated in chronic heart issues consistent with inflammation of the heart (non-ischaemic myocarditis). So I was interested to see that Baroness Hallet, Chair of the Covid Inquiry, has “directed that an expert team of pharmaco-epidemiologists led by Professor Prieto-Alhambra… to consider the main serious adverse events that were observed during the vaccine rollout”. Prieto-Alhambra, a pharmacoepidemiologist at the University of Oxford, presented his findings to the inquiry on January 22nd, alongside Professor Evans, a pharmacoepidemiologist at the London School of Hygiene and Tropical Medicine.

Only approved questions were allowed and Baroness Hallet got very upset when Anna Morris KC “asked questions for which I didn’t give permission” on behalf of the vaccine injured. Prieto-Alhambra was not asked about the limitations of the methods used to measure vaccine effectiveness and harm. Perhaps these were not allowed? The risk-benefit analysis of vaccination was briefly discussed when lead consul to the inquiry, Hugo Keith KC, observed that regulators must weigh up the number of deaths prevented against the “very rare” possibility of side effects. Prieto-Alhambra’s response was emphatic: “There is very strong data that the vaccines still worked in severe outcomes like hospitalisation or death.” It is, therefore, perhaps useful to consider one of Prieto-Alhambra studies on vaccine effectiveness against death.

On August 18th 2021, Prieto-Alhambra was a co-author of a cohort study published in the BMJ that claimed that two doses of the COVID-19 vaccine were 97% effective (Hazard Ratio (HR) 0.03, 95% confidence interval (CI) 0.02-0.04) at preventing COVID-19 deaths in nursing home residents in Catalonia.

Four months later, on December 8th 2021, a cohort study was published in the New England Journal of Medicine by Arbel et al. The paper claimed that the Pfizer booster was 90% effective (HR 0.10, CI 0.07-0.14) at preventing COVID-19 deaths in older age groups in Israel. On July 19th 2023 the Arbel et al. were forced to disclose that the rate of non-Covid mortality in the boosted group was 77% lower than the unboosted group (HR 0.23, CI 0.20-0.26) – a common issue in vaccine studies known as ‘healthy vaccinee’ bias.

It is notable that Arbel et al. did not disclose the non-Covid mortality rates in their original paper which would, inevitably, have cast doubt on the 90% effectiveness claim attributed to the booster. The difference in non-Covid mortality rates is indicative that the vaccinated and unvaccinated groups differ in underlying health and that this difference may have a major effect on the reported vaccine effectiveness.

Like Arbel et al., Prieto-Alhambra’s Catalan study, which used a similar cohort methodology, did not disclose the non-Covid mortality rates, thus leaving out the same crucial context that could indicate the presence of healthy vaccinee bias.

During the COVID-19 vaccination rollout in England, all-cause mortality rates, according to the Office for National Statistics (ONS), “were lower in the first three weeks after a vaccine dose than in subsequent weeks after that dose”. The ONS acknowledges that this could be due to the healthy vaccinee effect which, as per the ONS, is “where people who are ill (either due to COVID-19 or another relevant illness) are likely to delay vaccination. Therefore, the people who have been recently vaccinated are, in the short term, in better health than the general population”. The incredible results reported by Arbel et al. were probably due to the healthy vaccinee effect.

The interpretation of Prieto-Alhambra’s Catalan study, which claimed that vaccination was 97% effective against COVID-19 deaths in nursing home residents, depends on how much consideration is given to unmeasured biases such as the healthy vaccinee effect. If potential study limitations are ignored, vaccination was 97% effective against death. If potential study limitations are considered, particularly in the context of Arbel et al., did the study overestimate vaccine effectiveness?

In his evidence to the COVID-19 inquiry and as evidenced by the results of his Catalan study, it seems that Prieto-Alhambra places a high degree of trust in cohort studies. However, Arbel et al. serves as an important reminder of the pitfalls of accepting the results of cohort studies at face value. As Evans made clear during his evidence to the inquiry, observational studies are less reliable than randomised clinical trials. The results of cohort studies are less certain because the groups being studied may differ in ways which are not measured by the study. Therefore, a single study can be interpreted in different ways depending on how much consideration is given to unmeasured differences between the groups.

In this context, perhaps Baroness Hallet should not have chosen an epidemiologist from a university that received £143 million in royalties from its COVID-19 vaccine research – more than all other British Universities earned from intellectual property in the same period.

Particularly, when the royalties were generated from the sale of one of the vaccines being investigated.