select search filters
briefings
roundups & rapid reactions
before the headlines
Fiona fox's blog
The Office for National Statistics (ONS) have released new data for the number of deaths from COVID-19 in England in relation to vaccination status.
Dr James Doidge, Senior Statistician, Intensive Care National Audit & Research Centre (ICNARC); and Honorary Associate Professor, London School of Hygiene and Tropical Medicine, said:
“There are some interesting and important findings in this release, but also some very important limitations.
“Throughout the pandemic, people have voiced concerns about the classification of deaths; in particular that deaths due to COVID-19 may be overcounted and that deaths due to vaccine-related adverse events may be undercounted. Identifying the causal factors contributing to any one person’s death is an almost impossibly complex problem that is of more interest to the courts than to scientists. In medical research, we compare all-cause mortality between different groups of people (such as different treatment arms in a clinical trial) to ascribe differences in overall mortality to differences between the groups (such as which treatment each received), without relying on subjective decisions about what caused any one death. This release from ONS is therefore very important in providing data on all-cause mortality. It builds on the previous release of this data by providing a breakdown by age groups (age-specific mortality) in addition to the overall analysis (age-standardised mortality) that was previously provided, and by extending the analysis to include deaths occurring up to September 24, thereby covering the bulk of the Delta wave to date.
“The data show that, overall (age-standardised) and specifically among people aged 60+ (age-specific – see data tables), all-cause mortality was lower among people who had received two doses of the COVID-19 vaccines than among people who were unvaccinated, at all points in time. However, the results for people aged 10-59 and for those who have received only one dose of the vaccines appear inconsistent and, at first glance, alarming. A closer look at the methods and consideration of who is included in each vaccination group at different points in time provides insight as to the likely explanation. The “10-59” age group covers an obviously large range. This was done to avoid statistical problems and privacy concerns associated with reporting the very small numbers of deaths that occur in people of these ages in any one week. But at any given point in time, there are therefore very large differences in the age distribution between unvaccinated and vaccinated 10-59 year olds, with vaccinated individuals being much older on average. This makes it appear as though mortality rates are higher among the vaccinated group and what is urgently needed to correct for this is analyses that are both age-standardised and age-stratified, or stratified by smaller age groups.
“The issue with the single-dose vaccination groups is more subtle. While the age composition of the single-dose groups also changes over time, they are also likely to change in other ways: in particular, people whose health is too poor to consider a second dose, and perhaps those who experienced the worst side effects after a first dose, become the only people left in the single-dose group over time. This is a group of patients who are obviously at higher risk of death than the population in general and we must be extremely cautious when interpreting their data. Only more sophisticated analyses that factor in information on people’s health status will be able to tease out what is happening here with any degree of certainty.
“Lastly, Table 7 of the data download highlights the known problem in UKHSA vaccine surveillance reports, whereby the number of unvaccinated people was grossly overestimated, and event rates among unvaccinated individuals grossly underestimated by UKHSA using NIMS data to define the population of unvaccinated individuals. In Table 7, you can see that the population of people in England aged 18+ has been overestimated by more than 5 million individuals, the entirety of which were classified in UKHSA reports as “unvaccinated”, more than doubling the apparent number of unvaccinated individuals. These extra 5 million people represent those who have emigrated from England or were in England only briefly (but appear in the NIMS database) and as such should never have been counted. We can only hope that the ONS are able to rectify their issues with apparent mortality by vaccination status in a more effective and timely manner than the UKHSA have (not) with theirs. The ONS have been a pillar of transparency throughout the pandemic and I expect that an update will be issued forthwith.”
Dr Peter English, Retired Consultant in Communicable Disease Control, Former Editor of Vaccines in Practice, Immediate past Chair of the BMA Public Health Medicine Committee, said:
“These are among the best available data on the effectiveness of vaccines at preventing the most serious adverse outcome of Covid-19.
“They emphasise that vaccines are very effective.
“Two factors have to be borne in mind:
“This can lead to confusion, as an increasing proportion of the population who gets ill or dies has been vaccinated. It may even appear at first glance (and antivaccine activists who know better sometimes use this deliberately) that vaccination makes you more likely to die.
“The important message here is that vaccination is highly effective against death from Covid-19. You are 32 times less likely to die if you have been fully vaccinated than if you are unvaccinated; and that a single dose also provides some (although less than two doses) protection.
“The ‘Age standardised mortality rates’ (ASMR) were calculated as a means for taking into the effect of the age structure of a population, so you can compare populations with different age structures. It is a very long-standing, widely-used and uncontroversial epidemiological technique.”
All our previous output on this subject can be seen at this weblink:
www.sciencemediacentre.org/tag/covid-19
Declared interests
Dr James Doidge: “None.”
Dr Peter English: “Dr English is on the editorial board of Vaccines Today: an unpaid, voluntary, position. While he is also a member of the BMA’s Public Health Medicine Committee, this comment is made in a personal capacity. Dr English sometimes receives honoraria for acting as a consultant to various vaccine manufacturers, most recently to Seqirus.”