The Office for National Statistics (ONS) have released the latest report from the COVID-19 infection survey, looking at antibody and vaccination data for the UK.
Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:
“In terms of broad trends, the latest data from the ONS coronavirus infection survey (CIS) on antibody prevalence and the percentages of people vaccinated against the virus that can cause Covid-19 continues what we’ve seen in previous bulletins, and indeed (in relation to vaccination) on the data on numbers of people vaccinated that are published on the dashboard at coronavirus.data.gov.uk. In round numbers, for the latest available week (5-11 April), about 7 in 10 in the community population of England would test positive for antibodies, and about 6 in 10 in the populations of each of the other UK countries would test positive. Those are quite substantial increases on the last ONS bulletin a fortnight ago, which said that about half the people in the community population would test positive.
“On the test used in these data, someone can be positive for antibodies because of a previous infection or because they have been vaccinated. There are testing methods that distinguish between people whose antibodies came from an infection and people whose antibodies came only from being vaccinated, and ONS have said that they have started using methods that allow for this distinction to be made, but it is not made yet in this data release. Another point to be aware of is that the data are only for the community population ages 16 and over, so it excludes people who live in institutional settings such as care homes or prisons. In most age groups, the percentage of the population living in such institutions is very low, but it’s relatively high in the oldest age groups. Even for those aged 80 and over, though, ONS estimate that 9 in 10 live in the community and only 1 in 10 in an institutional setting such as a care home.
“There are differences in the percentages who would test positive between the different UK countries, and the regions of England, but generally they are not very large. There are clear differences between different age groups, with the older groups having much larger estimated percentages testing positive. Since most people testing positive, in those age groups anyway, get their antibodies from being vaccinated, that’s in line with the roll-out of vaccines by age. The numbers who would test positive is generally over 8 in every 10 for age groups from 60 years up, an indeed still over 8 in 10 for those in their 50s in England. It’s a little lower, though still over 7 in 10, in those in their 50s in the other UK countries. That might be because the proportion of people in that age group who were infected is higher in England, or it might have something to do with the timing of vaccinations. The estimated numbers testing positive in younger age groups are smaller, but still quite substantial.
“The numbers of people who are positive for antibodies will vary over time for a variety of reasons, and what’s seen in these data is a balance between different trends. Antibody levels can decrease naturally over time – this is called ‘waning’ – and at some point it’s possible that a person would no longer test positive. That doesn’t mean they have no immunity – they might still have antibodies in their blood, but the level has fallen too low to be detected by the test. In any case, the test measures antibodies, not immunity, and cannot take into account other aspects of immunity (such as through T cells). When someone has an infection, or is vaccinated, the antibody level takes some time to become detectable – maybe two or three weeks after vaccination, according to ONS. Then evidence from studies of vaccine effectiveness indicate that antibody levels can fall after the first vaccine dose, at different average rates in people of different ages, and is likely to pick up again after a second dose. There’s evidence of some increases after the second dose in people in the older age groups in these latest data.
“The bulletin also gives data on what proportion of people have been vaccinated. These go up to a week later than the antibody data, so up to the week 12-16 April this time, because the vaccination data are quicker to collect and process. Very broadly speaking, the data on antibodies track the data on vaccinations fairly closely, particularly in the older age groups where most people with antibodies will have got them from being vaccinated. Generally, though, the proportion who would test positive for antibodies is smaller than the proportion who have been vaccinated, across those aged 50 and over. That’s because a small proportion of people do not develop detectable antibodies after vaccination – but again that doesn’t necessarily mean they have no immunity. They may have antibody levels below the level detectable by the test, and/or they may have other kinds of immunity. The vaccination levels in the older age groups are very high, with 98, 99 or more in every 100 having received at least one vaccine dose in age groups above 60, and over 95 in every hundred in people in their 50s. Even in some younger age groups, some of the vaccination levels are already quite high, for example half those aged 35-49 have already been vaccinated in England.”
“You might wonder what the point is of including data on people vaccinated, since data on vaccination numbers appear regularly elsewhere (for example on coronavirus.data.gov.uk). I think there is a good reason for having these ONS estimates as well, and it’s because they may well give more accurate measures of the percentage of people vaccinated, across the UK countries as a whole and also within specific age groups. In the vaccination data on the dashboard, the counts of people vaccinated, by age, are likely to be pretty accurate. But if you want to know what percentage of, say, those in their 70s have been vaccinated, you also need to know how many people in their 70s there are in the country in total. ONS do provide estimates of that number, and there are other sources such as NHS records of various kinds. But all of these will be inaccurate to a certain extent. Data from the recent Census have not been processed yet, and there was no Census this year in Scotland anyway. It’s 10 years since the last Census, and though ONS revise their population estimates to take into account trends and other data, they can’t be exactly correct. The estimates of percentages of people vaccinated in this bulletin don’t use those population estimates, but are instead based on data from the ONS infection survey itself, processed and modelled in appropriate ways. The statistical analysis here also takes into account the administrative data on numbers of people vaccinated. Because of this wider use of data, I believe that the ONS estimates in this series of bulletins is likely to be more accurate on the percentages of people vaccinated than are figures from other sources. Against that, though, these figures don’t include people who live in institutional settings (either in the count of people vaccinated or in the count of people in the relevant age group), and that is an issue for the oldest age groups.”
All our previous output on this subject can be seen at this weblink:
Prof Kevin McConway: “I am a Trustee of the SMC and a member of its Advisory Committee. I am also a member of the Public Data Advisory Group, which provides expert advice to the Cabinet Office on aspects of public understanding of data during the pandemic. My quote above is in my capacity as an independent professional statistician.”