The Bigger Picture: Using the COVID-19 Infection Survey to track other infections

a picture of test tubes with chemical symbols in the background

The ONS, the University of Oxford and the UK Health Security Agency are conducting a pilot study to determine whether their COVID-19 Infection Survey could be used to monitor other respiratory infections in the private residential population. The pilot, which has been running since October 2022, tests for Influenza ‘flu’ and another common respiratory virus called Respiratory Syncytial Virus (RSV).

In this post, Lina Lloyd and Charlotte Murphy discuss why the pilot was initiated, explore the focus of the analysis, and share plans for publishing the data.

Background

Respiratory infections or illnesses like colds, flu and COVID-19 involve parts of the body used for breathing, such as the sinuses/nose, throat, airways or lungs.  These illnesses cause a range of symptoms including a cough, sore throat, stuffy or runny nose, high temperature (fever), tiredness, changes to sense of taste or smell, muscle aches or shortness of breath.

Flu A and B viruses are responsible for flu seasons each year. Flu A viruses are the only influenza known to cause flu pandemics as they can be transferred cross-species. Flu B almost exclusively infects humans, is less common than Flu A and mutates about two or three times slower. RSV is also another common respiratory virus that causes coughs and colds in winter.

Flu cases have been historically low since the emergence of COVID-19. This is because public health measures taken to reduce the transmission of COVID-19 also substantially reduce the spread of other viruses that cause respiratory infections like flu. Now these measures have been removed, many people have predicted the 2022-2023 influenza season will be much larger and more challenging than previous years.

Many Southern Hemisphere countries saw large numbers of cases in their winter months (our summer months). Cases started to rise earlier than in pre-pandemic seasons with more severe disease, complications and hospitalisations than the previous year.

Our main protection against flu is from vaccination.  Vaccines chosen each year match the type of flu we expect to see, based on what is circulating in the Southern Hemisphere, when vaccines are made in our spring/summer. However, sometimes these vaccines don’t match the type of flu which ends up circulating, resulting in larger outbreaks of disease. Further, most information about flu comes from people admitted to hospital.  There is currently very little information on how much flu circulates in the community, particularly in people without symptoms, who may spread it without knowing.

The study

The COVID-19 Infection Survey (CIS) has been gathering data on coronavirus infections and antibodies among the population at large since the early months of the pandemic in 2020.

Survey partners, the ONS, the University of Oxford and the UK Health Security Agency have now begun a pilot study to determine whether it could also offer an early warning of levels of flu and RSV in the private residential population.

The pilot, which is set to run for 20 weeks between October 2022 and March 2023, will help us identify early trends in emerging flu and RSV cases in the community. Additionally, it will give us more information on which strains of flu virus are the most prevalent through genomic sequencing.

The study uses two main types of data from CIS; questionnaire data about symptoms and nose and throat swab samples. Each week, around 750 randomly selected swab samples are tested for the presence of SARS-CoV-2 (COVID-19), Flu A or B, and RSV, rather than just SARS-CoV-2 alone. Samples that are identified as having a high concentration of flu virus are tested further to read all the letters of their genetic code (called genomic sequencing), to identify the exact strain of flu virus a person has.

Analysis

Although numbers are small, as well as estimating the percentage of the sample who test positive for flu and RSV overall, we also consider the age distribution, flu vaccination, and symptoms in cases.

We are currently exploring how best to present our early data, as flu and RSV positivity is relatively low, and the sample size is fairly small. Typically, the UK flu season is between the months of December and February. Consistent with this, there are early signs that at the start of December 2022, flu/RSV was circulating in the private residential population already.

It’s important to note that the published data are unweighted and can only estimate the percentage of the sample that is positive for flu/RSV. Although random, the sample is not representative and cannot be generalised to the rest of the population. Nonetheless, it will give us an indication whether CIS data could support the NHS, by giving an early warning of infection levels in the private residential population.

We have released some initial data tables showing the experimental data from this pilot study, which you can view here.

Image of blog author, Lina Lloyd

Lina Lloyd is Head of Statistical Processing for COVID-19 Infection Survey at the ONS

Image of blog author, Chris Jenkins

Charlotte Murphy is Head of Data Processing Pipeline at the ONS