Since the appearance of Omicron variants in December 2021, the number of people being reinfected with COVID-19 has increased markedly. As Leanne Massie explains, this affects our ability to estimate the total number of people who have had COVID-19 at least once.
Beginning April 2020, the Coronavirus (COVID-19) Infection Survey (CIS) has repeatedly tested a large, representative sample of the UK to track the spread of COVID-19 across the population. By February 2022 we estimated about 70% of the population of England had been infected with COVID-19 at least once. This insight provides an understanding of the proportion of the community that may have some immunity from natural infection helping determine the long-term impact of COVID-19.
Last week, a new technical article showed the number of people infected over each variant dominated period and age group in England. However, determining the total number of people infected throughout the whole pandemic is more challenging as we cannot simply add the total infections together. This is due to the increasing number of people who have had COVID-19 more than once.
Counting Reinfections is Difficult
Before the appearance of the Omicron variants in December 2021, the number of people who had COVID-19 more than once was low. However, reinfections have increased substantially since the Omicron variants emerged but tracking reinfections accurately is difficult. To identify reinfections, we need a positive PCR test every time a person has a COVID-19 infection. People in our survey are tested about once per month so if an infection happens between tests it will be missed, meaning we underestimate the true number of reinfections.
Reinfections Occur with Different Variants
Most reinfections are due to different variants. For example, if you were infected with the Delta variant, a later reinfection is likely to be with another variant such as Omicron. Therefore, we can confidently estimate the number of people who have had a COVID-19 infection in each time period when a particular variant was dominant, since the number of reinfections is low for the same variant.
However, estimating how many people have had COVID-19 in total is not as simple as adding up the number of all infections across the variant dominated periods. This is because even after deducting the number of reinfections we know about, the estimated total would exceed 100% as there are many more reinfections than we can accurately measure, meaning some people are counted more than once. The estimate using this approach reaches 169.2% of the population of England on 11 November 2022. Therefore, to estimate the total number of people infected with COVID-19 at least once, we need to develop alternative methods.
Alternative Method of Estimating How Many People Have Had COVID-19
An alternative approach we are exploring focuses on the people who have never been infected. Based on our survey data we can estimate the number of people that acquire a first infection each day. By subtracting these people from the number of people who had not been infected by that day we can estimate the total that have been infected over time. People cannot be counted more than once using this method as they are removed from the calculations after their first infection. The percentage of the population estimated to have experienced a COVID-19 infection was 81.7% of the population of England on 11 November 2022 using this experimental approach.
The key caveat to this approach is that it assumes everyone is equally likely to become infected, but in reality, infection risk differs between different groups of people. Therefore, the results of these experimental statistics are not official estimates, but for guidance only.
How many people in the UK have had COVID-19 seems a simple question, but since the increasing number of reinfections hinders us, the answer is anything but simple. We are continuing to explore various methods as we learn more about the evolving pandemic and changing patterns of infections.
Author:
Leanne Massie is Senior Statistical Officer at the Office for National Statistics
Discussion about this post