The Coronavirus Infection Survey study

This research used data made available via the Office for National Statistics (ONS) Secure Research Service (SRS), which is being expanded and improved with ADR UK funding.

Authors: Office for National Statistics (ONS), University of Oxford, University of Manchester, Public Health England and the Wellcome Trust.

Date: May 2020

Research summary

Since May 2020, data made available by the ONS has been used to provide a weekly snapshot of coronavirus (Covid-19) infections within the UK. The Coronavirus Infection Survey (CIS) provides vital insights into the number of positive cases, incidence rates and antibody levels and continues to be crucial in informing the government’s response to the Covid-19 pandemic.

Following the onset of the Covid-19 outbreak in the UK, understanding how the coronavirus spread through the population was essential in controlling the pandemic. The ONS and its partners were asked by the UK Government to track infections for a sample of participants of all ages across England to estimate infection rates. From an initial six-month pilot of 20,000 participants in private residential households, the CIS has increased to 150,000 weekly samples across all UK nations.

Data used

The CIS is a nationally representative survey that provides important insights into the underlying characteristics of those infected, helping to uncover potential health inequalities within the population. It is the first longitudinal community survey of Covid-19 infection rates within the UK and provides the following statistics:

• percentage of the population testing positive at a given time
• incidence rates (the number of new infections per day)
• percentage of the population testing positive for antibodies.

Data for the CIS is collected through nose and throat swabs from UK households for participants over two-years old. Additional analysis for a sub-sample of participants aged 15 years and over identifies Covid-19 antibodies, providing evidence as to whether an individual has had an immune response due to either a Covid-19 infection or vaccination.

An accompanying household questionnaire provides socio-demographic data along with contextual information such as whether individuals are self-isolating, working from home or have been in contact with a suspected carrier of Covid-19. More recently, additional questions have collected data on participant experiences of the pandemic, looking at whether a participant has been vaccinated, is travelling to work or has experienced symptoms of long Covid. CIS data is provided twice a week for researchers to analyse within the ONS Secure Research Service (SRS).

Methods used

Several modelling techniques, such as dynamic Bayesian multi-level regression with post-stratification, are used to account for socio-demographic and contextual data variability and bias. Once processed, data for groupings of participants testing positive for Covid-19 are compared using pairwise statistical testing. This establishes any significance in infection rates between the groupings, helping to identify potential risk factors.

Research findings

The CIS continues to provide weekly bulletins detailing the levels of coronavirus within the UK population along with the percentage testing positive for antibodies. At the peak of the pandemic in England (27 December 2020 to 2 January 2021), the study estimated that 1 in 50 people (2.06%) were infected with Covid-19. Respective infection rates during this period for Wales was 1 in 70 people (1.45%). At the peak of the pandemic in Scotland (6 December to 12 December 2020), 1 in 100 people (1.00%) were estimated to be infected. While at the peak in Northern Ireland (17 January to 23 January 2021), this was 1 in 50 (2.01%).