IPSOS MORI/Imperial REACT I Antigen study

This agreement is to support Antigen testing study, round 2, (REACT-1-Round 2), one element of the REal-time Assessment of Community Transmission 1 (REACT 1): a study that will provide the basis for estimation of the R value in the community at regional and local authority levels, being conducted by Imperial College London on behalf of the Department of Health and Social Care (DHSC). This was initially thought to be a short study and data was originally requested to support Waves 4 to 6. Following this initial request DHSC have extended the contract to support Waves 15 to 20 due to the pandemic persisting. There have been a number of amendments to this agreement to support the addition of these Waves. This amendment is to support Wave 15 and the "Thank you" trial only a further amendment will be submitted to support Waves 16 to 20.

AGREEMENT HISTORY

The original application was to support three waves of data being supplied to this format and represents waves 4-6 of the antigen (REACT1) testing study. The surveys were completed in August, September and October 2020. In each wave in order to achieve the required sample size of 150,000 names and demographic details of ~750,000 individuals aged 5 years and above were needed. This was based on current response rates, which had fallen over the first three waves. This continued fall resulted in details being required for larger numbers of individuals in September and October in order to reach the target sample size.

An amendment to the agreement (v1.3) was, by extension of Ipsos MORI's contract by DHSC, also requested and had approved a further wave 7 of Demographic data in Late October 2020 at a larger sample size of ~835,000 individuals aged 5 years and above due to a declining response rate. A further amendment (v2.2) requested a further extraction of Demographic data by a further extension of Ipsos MORI's contract by DHSC, for waves 8 to 10 in late November 2020, January and February 2021). A further amendment (v3.3) provided Ipsos MORI with additional drops of PDS data for Waves 11-13, cohort size 835,000. This amendment was also to expand the scope of the purpose. As testing in the community becomes more widely available there is increasing possibility that REACT-1 may underestimate prevalence because potential participants opt to go to Pillar 2 for a test (or Pillar 1 for health care and care home workers). The study additionally aims to overcome this potential source of bias by including test results during the data collection window for positive cases. A further amendment (v4.2) was, by extension of Ipsos MORI's contract with DHSC, to request a further wave 14 of PDS data in mid August 2021 at a sample size of ~835,000 individuals aged 5 years and above. A further amendment (v5.2) was submitted as Ipsos MORI were exploring with the DHSC and Imperial College London a 'thank you' trial at Round 14 to improve swab response. DHSC requested under v5.2 a separate and additional random sample of approximately 45,000 participants aged 13+, half of whom would be offered a 'thank you' if they return a swab.

*** CURRENT AMENDMENT REQUEST

This amendment request v6.1 is to provide Ipsos MORI with additional drops of PDS data for waves 15 plus 1 drop of PDS data for the Wave 15 'thank you' Trial. ***

The main study objectives are:

1. To measure the national prevalence of infection during a window of time in each month.

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

Prevalence refers to the number of disease cases present in a particular population at a given time.

Measures of infection point prevalence will be provided at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response.

This research is being carried out to help the Government establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level. This is based on analysis of a nasopharyngeal self-swab obtained from a random sample of the population ages 5 and above using the NHS register as sampling frame.

The study involves a self-administered nasal and throat swab for adults and children over the age of 12. For children aged 12 or below, a parent or guardian will take the swab. For children aged 5 to 17 years old the invitation will be sent via parents/guardians. There will also be a brief online or telephone questionnaire to complete. Consent will be requested for follow up through routine health records to monitor long-term health outcomes, as has been done for REACT-1 baseline survey. There are no invasive procedures, although the throat and nasal swabs may cause some short-lived discomfort.

The Principal Investigator’s decision about the age criterion for REACT was taken in line with the guidance on the use of the swabs that were available when REACT was set up in April 2020. The age criterion is made clear in all the published reports. The age criterion has been consistent throughout REACT enabling robust estimates about the study population and consistent comparisons over time.

Results of the RT-PCR tests will be fed back to participants. If a positive test is recorded, the participant will be advised to self-isolate for 7 days from onset of any symptoms, or if asymptomatic, for 7 days. Other members of the household will be advised to self-isolate in accordance with Government guidance. If the test is negative but the participant reports symptoms indicative of COVID-19 they and household members will similarly be advised to follow Government guidance. If the participant or a member of the household requires medical advice or attention they will be instructed to dial NHS 111 or if severely ill to dial 999 for an ambulance. With appropriate consent, positive test results could be fed into contact tracing efforts at the local level.

The research aims to contact a representative sample of the population aged 5+ in England in order to meet the study aims:

1. To measure the national prevalence of infection during a window of time in each month.

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

This sample will be selected from the Personal Demographic Service – considered an almost complete sample frame of adults in England, allowing stratification by gender, age and local authority.

REACT-1 aims to provide an accurate estimate of prevalence of infection across England at local authority levels. This sub-study REACT-1 (2), will provide the basis for estimation of the R value in the community at regional and local authority levels. The study is being conducted by Imperial College London (sponsor) on behalf of the Department of Health and Social Care (funder).

Patterns of hospital occupancy, death and 999 calls suggest that the COVID-19 epidemic in England is past its peak overall with a substantial downward trend. The change from increasing to decreasing incidence was almost certainly a result of social distancing and the national lockdown implemented on March 23rd: the net effect of the lockdown and subsequent lockdowns was to reduce the reproduction number R from well above 1 (~3 prior to its implementation) to below 1 currently. However, although it is known that R has likely been below 1 for some time, it cannot accurately infer resulting levels of infectious prevalence in the community nor can it assess its geographical variation, based on available data. As we transition out of lockdown, the relative risk of infection in one area compared to another will be closely related to the number of infectious people in that area, with the absolute risk dependent on the degree of social mixing at that time.

Because contact levels will increase, it is expected that there will be an increase in R which could lead to a subsequent wave of infection if R again becomes greater than 1. Based on experience in other countries, it is likely that R will vary over time and by geography. Also, a variety of policies may be implemented as lockdown is relaxed, some of which – such as contact tracing – may be supported locally and therefore vary geographically.

Although overall and local transmissibility of COVID-19 in the UK will continue to be assessed using mortality and hospital occupancy data, these data lag changes in transmission by a week or more (substantially more for deaths), and reflect transmission in hospitals and care homes as well as the community. Also, healthcare factors including the possibility of levels of transmission varying from one hospital to another, suggest that hospital data would not be useful for inferring local variation in community transmission.

Testing capacity for the UK has increased rapidly, reaching the target of 100,000 tests per day by the end of April, providing sufficient capacity to meet the testing needs of hospital patients, healthcare workers, other key workers; and potentially providing community testing of symptomatic individuals.

The REACT-1 programme is currently being used to establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level.

The data requested from the Personal Demographic Service (PDS) will be used in order to select a nationally representative sample of the population aged 5+ to take part in the testing. The study needs to provide reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

This study is one component of a larger programme and sits alongside the REal-time Assessment of Community Transmission 2 (REACT 2): Usability and feasibility study of widespread home self-testing for SARS-CoV-2 antibodies.

The data required is demographic information from the Personal Demographics Service, to allow individuals aged 5+ to be contacted for the purpose of undertaking a survey collecting nasopharyngeal self-swab tests from an age-stratified representative population sample.

The data is requested in two stages:

- 1 – a record level identifiable population derived from PDS (including NHS Number, patient postcode, gender and age band) in order to allow selection of a representative sample

- 2 – a record level identifiable data request for selected sample only (including NHS Number, patient address, gender, age (month and year), and telephone number – where available)

At wave 12 Ipsos MORI changed the design. Instead of aiming to achieve approximately equal numbers of patients registering to take the test(s) in each Lower Tier Local Authority district, the design aims to achieve a sample where the numbers of patients registering to take the test(s) in each Lower Tier Local Authority district are in their correct proportion relative to the population of England. This change was made to maximise the statistical efficiency of the achieved sample as swab response has been declining. The design now maximises the precision of the prevalence estimates REACT 1 can provide at the national level, has increased the numbers sampled from areas of higher prevalence, the numbers sampled from ethnic minority backgrounds and those living in the key urban areas and the most deprived areas.

ADDITIONAL 'THANK YOU' TRIAL (Introduced in Wave 15)

The purpose of the 'thank you' trial is to gain further insight into the impact of providing both general and targeted 'thank yous' to the population. The intention is to help inform the study and Department of how 'thank yous' would be received, as well as the impact that they could have on the response rate from different groups. In short, it is a proof of concept trial that will help inform the policy and ensure any further usage of the approach is both effective and meets requirements.

Literature suggests not only that 'thank yous' actively improve survey response, but also that it could reduce the worsening of emerging biases as the dynamic of the pandemic changes. As REACT seeks to be as representative as a possible of the community population in order to inform policy that reduces the health burden both on specific sub-groups and the population as a whole, the trial could provide valuable pointers as to how further adjustments would support the aims of the Equalities Act.

It is important that the REACT study achieves a large sample so that it can compare sub-groups and assess change over time, however, the response to the study has fallen. Sampled participants must have a motive for taking part in the study and these motives need to outweigh their reasons for not taking part. Even among those who are motivated to take part, some may encounter other barriers which weigh heavier than their motivation. Some barriers to taking part have also changed over time – for example the risk of potentially testing positive and having to isolate now impacts more people, as society has opened back up. Other barriers are more minor e.g. too much effort to take the extra step from receiving the letter to going online and registering.

The impact of the changes to motivators and barriers is seen in the declining response rate. Therefore, Ipsos MORI need to consider how to maintain the large achieved sample size required for this study for it to provide the most useful data, without just continuously increasing the initial sample drawn (which raises ethical issues around how many people’s data needs to be passed to Ipsos MORI by NHS Digital).

'Thank yous' are commonly used in social research and play a key role in in tipping the balance in favour of participation – something that is usually achieved by well-trained interviewers in telephone and face-to-face surveys. There is plenty of evidence that cash-like 'thank yous' can work to improve response, and are especially effective in the absence of an interviewer. There is also evidence that, in general, 'thank yous' are not coercive in that they do not induce participants to accept higher risks .

Ipsos MORI therefore proposed to trial 'thank yous' of £10 and £20 gift vouchers to assess the impact on response – both overall and by age group. Ipsos MORI will have two experimental groups who are offered either a £10 or £20 gift voucher if they return a swab test, while the rest of the sample will not be offered any 'thank you'. If the impact of offering a 'thank you' is positive, they can then consider rolling a 'thank you' out to the whole sample, or to a targeted sub-group.

Ipsos MORI did not include the 5-12 age group in the 'thank you' trial as the sampled child is not making the decision to take the swab test and they would be incentivising their parent. This might raise significant ethical and reputational concerns.

Being able to conduct the trial specifically in the next Round of REACT is driven by the fact that Ipsos MORI are at a very important pre-Autumn juncture. Ipsos MORI need to make decisions on how to provide the best data we can get in a potentially volatile period for the Ministers who will need it.

DHSC have confirmed that as the data controllers, DHSC and Imperial College London, have determined to conduct the 'thank you' trial and that Ipsos MORI as data processors are processing the data under their instruction.

The 'thank you' trial has ethical approval.

The current study objectives remain as stated above. At wave 15 the national prevalence of infection will be measured in late October 2021, as well as the change in prevalence since September 2021 and the national average R value, with high accuracy. Imperial College London will also look at the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the surveys.

Ipsos MORI are acutely conscious of the need for data minimisation. Despite a falling swab response rate and a decline in the number of swabs returned Ipsos MORI has sought to offset this by making the design more statistically efficient and is now exploring the use of 'thank yous' to boost swab response. Ipsos MORI do however continue to need to sample ~835,000 individuals per wave to retain the benefits REACT has provided the Government, academic community studying SARS-CoV-2 infection and society at large.

The justification for the number of records requested remains the requirement for reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

As the National Data Opt Out is not applied patients that have opted out of their confidential patient information being used for research and planning will be included in the sample selected from the Personal Demographic Service. Ipsos MORI delete the data of those who have not registered for the study and requested that there is no further contact from Ipsos MORI in line with the Privacy Notice (typically two months after the end of fieldwork). As no individual can be re-sampled unless six months has elapsed since they were previously sampled and invited to take part, there is no need for Ipsos MORI to retain their contact details longer than that. NHSD do provide data of individuals that have been provided to Ipsos MORI under a previous drop where an individual has been re-sampled for REACT. No individual can be re-sampled unless six months has elapsed since they were previously sampled and invited to take part.

Once participants opt out of the study Ipsos MORI do not contact them again. Ipsos MORI review complaints and suggestions received through its helpline before every wave. The Privacy Notice and other documents have been amended following these reviews.

The study protocol also includes facility to undertake additional testing in specific local areas where an increase in transmission and numbers of cases is suspected. These additional surveys would involve sending out swab test kits to up to ca. 10,000 additional people in a specific area and ensuring rapid delivery and pick up, e.g. both delivery and pick-up by courier. Additional data requests may be needed in response to these local requirements.

The identifiable contact information will be securely destroyed by Ipsos MORI once the analysis has been completed. This has been communicated to participants in the updated privacy notice.

There is no other national register of the population that would allow contact with a named and representative sample of adults.

The data controller determines the purposes for which and the means by which personal data is processed. The DHSC and Imperial College London have designed and funded the study and are both Data Controllers, and Ipsos MORI, who undertake the survey operations, is the primary Data Processor.

Data Controllers:

• Department of Health and Social Care (funder)

• Imperial College London (Sponsor)

Data Processors:

• Imperial College London

• Virtus Holdco Ltd (provision of ICL servers)

• Ipsos MORI - who are working with the following suppliers to deliver the project:

• Formara Limited - Print and despatch

• Adare SEC Ltd - Print and despatch

• Txtlocal Ltd - Text message reminders

• Tivian GmbH (Previously called Questback GmbH) - Online front-end data collection platform accessed by Ipsos Germany.

Please note that Eurofins provide laboratory services to Ipsos MORI for this study, but have not been added as Data Processor as the samples they process are anonymised. Eurofins do not have any access to NHS Digital data.

All suppliers will comply with the relevant provisions of UK GDPR and Regulation 7 of COPI.

The DHSC and Imperial College London as the data controllers satisfy the requirement in Regulation 7(2) COPI. DHSC have a data processing agreement in place with Ipsos MORI who are carrying out the processing under the instruction of DHSC and makes it clear that it is DHSC who has overall control of what happens to the personal data. In the processing agreement it is clear that confidential data must be treated in such a way to meet the Regulation 7(2) requirement (they owe a duty of confidentiality, as per Regulation 7(2)), staff within Ipsos MORI who are handling the data are aware of this fact and are subject to contractual obligations of confidentiality. These provisions in the data processing agreement satisfy the requirement in Regulation 7(2).

LEGAL BASIS for DATA PROCESSING

The lawful basis for processing data under GDPR has been reviewed against the guidance provided by IGARD and been assessed as acceptable. The Department of Health and Social Care and Imperial College London process data under the legal basis of Article 6(1)(e) ‘processing is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the controller’ as the Department of Health and Social Care is a public authority and government department, and Imperial College London is a Public Authority as it has a Royal Charter. The processing of data for this study is a task of public interest as it will provide a measure of the national prevalence of infection with a measure of the national average R value both nationally and across regional and local authority levels.

Both Data Controllers process special category data under Article 9(2)(j) ‘processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes …’ as the data are required for research purposes in the public interest and judged subject to the appropriate safeguards. The survey will involve processing special category data in relation to the health and ethnicity of individuals taking part. The results will be used for processing relevant information to support the current efforts of the UK government authorities for understanding the prevalence and spread of COVID-19, informing the public health response to the ongoing pandemic.

The legal basis for identifiable data to flow is under Regulation 3(4) of the National Health Service (Control of Patient Information Regulations) 2002 (COPI).

While the legal basis is public interest the research has had ethical approval from the South Central Berkshire B Research Ethics Committee (REC) and Health Regulator Authority (HRA).

NON-APPLICATION OF NATIONA DATA OPT OUT

The identifiable data flowing into NHS Digital and out of NHS Digital does not have National Data Opt Out applied because the data flowing into NHS Digital does so under the operation of National Data Opt Out policy guidance which states that the National Data Opt Out does not apply to mandatory flows of data and for the flow of data out of NHS Digital it is flowing under the Control of Patient Information (COPI) Regulations 2012 under a Secretary of State for Public Health Direction. The COPI regulations states that the processing of patient information must align with Regulation 3(1) of COPI particularly in relation to the current outbreak of Covid-19. Regulation 3(1) confirms that confidential patient information may be processed for:

- diagnosing communicable diseases and other risks to public health;

- recognising trends in such diseases and risks;

- controlling and preventing the spread of such diseases and risks;

- monitoring and managing

- outbreaks of communicable disease;

- incidents of exposure to communicable disease;

- the delivery, efficacy and safety of immunisation programmes;

- adverse reactions to vaccines and medicines;

- risks of infection acquired from food or the environment (including water supplies);

- the giving of information to persons about the diagnosis of communicable disease and risks of acquiring such disease.

The COPI Regulations confirm that permission for processing confidential patient information is limited to:

(a) the Public Health Laboratory Service;

(b) persons employed or engaged for the purposes of the health service;

(c) other persons employed or engaged by a Government Department or other public authority in communicable disease surveillance.

The agreement fulfils this criteria and thus under COPI the National Data Opt Out is not applied.

The data requested from NHS Digital will be used to augment the data being collected as part of the REACT-1 programme to evaluate the prevalence of SARS-CoV-2 infection across England based on a random sample of the population.

The Demographic (PDS) data from NHS Digital will provide the contact details for a representative sample of the population aged 5+ to be invited to take part in the testing.

The study aims to provide reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and to feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

Data will not be made available to any third parties, other than those specified in this agreement, except in the form of aggregated outputs with small numbers suppressed in line with the suppression rules stated in the HES analysis guide.

REACT data has provided an accurate and localised view of the spread of COVID-19, feeding directly into the Government’s policy response. The REACT data has been used by SAGE to compute the R (reproduction) number in conjunction with other data sources. The findings from the surveys have directly informed the decisions to bring in regional and national “lockdowns”, the definition of the “tiers” across the country, and the lifting of restrictions in 2021. The study has provided benefits to both patients and the health care system by reducing mortality and morbidity. Public knowledge of the R rate at their local level has provided the information to support the public who are able to make choices and influenced behavioural changes in individuals and how they have personally responded to the pandemic.

The speed and scale of testing have enabled the study to quickly detect rising rates of infections and local hotspots, enabling timely public health responses from government. REACT has secured impact by its regular reports, leading to high profile media coverage and has been used in the Prime Minister’s televised press briefings. A video of how REACT works has been produced, and an infographics highlighting key study findings to help inform the general public and encourage safer behaviours.

In addition, the random sampling design means the study identified several thousand people who tested positive yet displayed no symptoms. By informing participants they were positive and needed to self-isolate, the REACT study has enabled scientists to have a better understanding of how the virus is transmitted and limited the spread of COVID-19.

The REACT study has been critical to ensure that the Government’s decision-making is based on robust, current evidence and that responses are targeted and appropriate.

The REACT study has helped to control the spread of COVID-19 and helped to prevent the spread of the virus across the UK and helped to reduce morbidity and mortality.

The REACT study has provided the public with access to data that has supported individuals to make personal choices in how they respond to the pandemic.

Further to this, the agile and collaborative approach of the REACT model and the wealth of knowledge accumulated has allowed for a very rapid response to changing requests, such as determining the usability and acceptability of different tests.

The study has informed the use of public health control measures such as closing schools, limiting public gatherings, issuing travel restrictions, quarantine rules and introducing the wearing of face coverings.

This application is to support Antigen testing study, round 2, (REACT-1-Round 2), one element of the REal-time Assessment of Community Transmission 1 (REACT 1): a study that will provide the basis for estimation of the R value in the community at regional and local authority levels, being conducted by Imperial College London on behalf of the Department of Health and Social Care (DHSC). This was initially thought to be a short study and data was originally requested to support Waves 4 to 6. Following this initial request DHSC have extended the contract to support Waves 7 to 14 due to the pandemic persisting. There have been a number of amendments to this agreement to support the addition of these Waves.

The main study objectives are:

1. To measure the national prevalence of infection during a window of time in each month.

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

Prevalence refers to the number of disease cases present in a particular population at a given time.

Measures of infection point prevalence will be provided at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response.

This research is being carried out to help the Government establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level. This is based on analysis of a nasopharyngeal self-swab obtained from a random sample of the population ages 5 and above using the NHS register as sampling frame.

The study involves a self-administered nasal and throat swab for adults and children over the age of 12. For children aged 12 or below, a parent or guardian will take the swab. For children aged 5 to 17 years old the invitation will be sent via parents/guardians. There will also be a brief online or telephone questionnaire to complete. Consent will be requested for follow up through routine health records to monitor long-term health outcomes, as has been done for REACT-1 baseline survey. There are no invasive procedures, although the throat and nasal swabs may cause some short-lived discomfort.

The Principal Investigator’s decision about the age criterion for REACT was taken in line with the guidance on the use of the swabs that were available when REACT was set up in April 2020. The age criterion is made clear in all the published reports. The age criterion has been consistent throughout REACT enabling robust estimates about the study population and consistent comparisons over time.

Results of the rt-PCR tests will be fed back to participants. If a positive test is recorded, the participant will be advised to self-isolate for 7 days from onset of any symptoms, or if asymptomatic, for 7 days. Other members of the household will be advised to self-isolate for 14 days in accordance with Government guidance. If the test is negative but the participant reports symptoms indicative of COVID-19 they and household members will similarly be advised to self-isolate in accordance with Government guidance. If the participant or a member of the household requires medical advice or attention they will be instructed to dial NHS 111 or if severely ill to dial 999 for an ambulance. With appropriate consent, positive test results could be fed into contact tracing efforts at the local level.

The research aims to contact a representative sample of the population aged 5+ in England in order to meet the study aims:

1. To measure the national prevalence of infection during a window of time in each month.

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

This sample will be selected from the Personal Demographic Service – considered an almost complete sample frame of adults in England, allowing stratification by gender, age and local authority.

REACT-1 aims to provide an accurate estimate of prevalence of infection across England at local authority levels. This sub-study REACT-1 (2), will provide the basis for estimation of the R value in the community at regional and local authority levels. The study is being conducted by Imperial College London (sponsor) on behalf of the Department of Health and Social Care (funder).

Patterns of hospital occupancy, death and 999 calls suggest that the COVID-19 epidemic in England is past its peak overall with a substantial downward trend. The change from increasing to decreasing incidence was almost certainly a result of social distancing and the national lockdown implemented on March 23rd: the net effect of the lockdown and subsequent lockdowns was to reduce the reproduction number R from well above 1 (~3 prior to its implementation) to below 1 currently. However, although it is known that R has likely been below 1 for some time, it cannot accurately infer resulting levels of infectious prevalence in the community nor can it assess its geographical variation, based on available data. As we transition out of lockdown, the relative risk of infection in one area compared to another will be closely related to the number of infectious people in that area, with the absolute risk dependent on the degree of social mixing at that time.

Because contact levels will increase, it is expected that there will be an increase in R which could lead to a subsequent wave of infection if R again becomes greater than 1. Based on experience in other countries, it is likely that R will vary over time and by geography. Also, a variety of policies may be implemented as lockdown is relaxed, some of which – such as contact tracing – may be supported locally and therefore vary geographically.

Although overall and local transmissibility of COVID-19 in the UK will continue to be assessed using mortality and hospital occupancy data, these data lag changes in transmission by a week or more (substantially more for deaths), and reflect transmission in hospitals and care homes as well as the community. Also, healthcare factors including the possibility of levels of transmission varying from one hospital to another, suggest that hospital data would not be useful for inferring local variation in community transmission.

Testing capacity for the UK has increased rapidly, reaching the target of 100,000 tests per day by the end of April, providing sufficient capacity to meet the testing needs of hospital patients, healthcare workers, other key workers; and potentially providing community testing of symptomatic individuals.

The REACT-1 programme is currently being used to establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level.

The data requested from the Personalt Demographic Service (PDS) will be used in order to select a nationally representative sample of the population aged 5+ to take part in the testing. The study needs to provide reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

This study is one component of a larger programme and sits alongside the REal-time Assessment of Community Transmission 2 (REACT 2): Usability and feasibility study of widespread home self-testing for SARS-CoV-2 antibodies.

The data required is demographic information from the Personal Demographics Service, to allow individuals aged 5+ to be contacted for the purpose of undertaking a survey collecting nasopharyngeal self-swab tests from an age-stratified representative population sample.

The data is requested in two stages:

- 1 – a record level identifiable population derived from PDS (including NHS Number, patient postcode, gender and age band) in order to allow selection of a representative sample

- 2 – a record level identifiable data request for selected sample only (including NHS Number, patient address, gender, age (month and year), telephone number and email – where available)

The original application was to support three waves of data being supplied to this format and represents rounds 4-6 of the antigen testing study. The surveys were completed in August, September and October 2020. In each wave in order to achieve the required sample size of 150,000 names and demographic details of ~750,000 individuals aged 5 years and above were needed. This was based on current response rates, which had fallen over the first three waves. This continued fall resulted in details being required for larger numbers of individuals in September and October in order to reach the target sample size.

An amendment to the agreement (v1.3) was, by extension of Ipsos MORI's contract by DHSC, also requested and had approved a further wave 7 of Demographic data in Late October 2020 at a larger sample size of ~835,000 individuals aged 5 years and above due to a poor response rate. A further amendment (v2.2) requested a further extraction of Demographic data by a further extension of Ipsos MORI's contract by DHSC, for waves 8 to 10 in late November 2020, January and February 2021). A further amendment (v3.3) provided Ipsos MORI with additional drops of PDS data for Waves 11-13, cohort size 835,000. This amendment was also to expand the scope of the purpose. As testing in the community becomes more widely available there is increasing possibility that REACT-1 may underestimate prevalence because potential participants opt to go to Pillar 2 for a test (or Pillar 1 for health care and care home workers). The study additionally aims to overcome this potential source of bias by including test results during the data collection window for positive cases. A further amendment (v4.2) was, by extension of Ipsos MORI's contract with DHSC, to request a further wave 14 of PDS data in mid August 2021 at a sample size of ~835,000 individuals aged 5 years and above.

****

In this amendment (5.1), Ipsos MORI have been exploring with the DHSC and Imperial College London an incentive trial at Round 14 to improve swab response. DHSC are requesting a separate and additional random sample of approximately 45,000 participants aged 13+, half of whom would be offered an incentive if they return a swab under this amendment to this agreement.

The purpose of the incentive trial is to gain further insight into the impact of providing both general and targeted incentives to the population. The intention is to help inform the study and Department of how incentives would be received, as well as the impact that they could have on the response rate from different groups. In short, it is a proof of concept trail that will help inform the policy and ensure any further usage of the approach is both effective and meets requirements.

Literature suggests not only that incentives actively improve survey response, but also that it could reduce the worsening of emerging biases as the dynamic of the pandemic changes. As REACT seeks to be as representative as a possible of the community population in order to inform policy that reduces the health burden both on specific sub-groups and the population as a whole, the trail could provide valuable pointers as to how further adjustments would support the aims of the Equalities Act.

It is important that the REACT study achieves a large sample so that it can compare sub-groups and assess change over time, however, the response to the study has fallen. Sampled participants must have a motive for taking part in the study and these motives need to outweigh their reasons for not taking part. Even among those who are motivated to take part, some may encounter other barriers which weigh heavier than their motivation. Some barriers to taking part have also changed over time – for example the risk of potentially testing positive and having to isolate now impacts more people, as society has opened back up. Other barriers are more minor e.g. too much effort to take the extra step from receiving the letter to going online and registering.

The impact of the changes to motivators and barriers is seen in the declining response rate. Therefore, Ipsos MORI need to consider how to maintain the large achieved sample size required for this study for it to provide the most useful data, without just continuously increasing the initial sample drawn (which raises ethical issues around how many people’s data needs to be passed to Ipsos MORI by NHS Digital).

Incentives are commonly used in social research and play a key role in in tipping the balance in favour of participation – something that is usually achieved by well-trained interviewers in telephone and face-to-face surveys. There is plenty of evidence that cash-like incentives can work to improve response, and are especially effective in the absence of an interviewer. There is also evidence that, in general, incentives are not coercive in that they do not induce participants to accept higher risks .

Ipsos MORI are therefore proposing to trial incentives of £10 and £20 gift vouchers to assess the impact on response – both overall and by age group. Ipsos MORI will have two experimental groups who are offered either a £10 or £20 gift voucher if they return a swab test, while the rest of the sample will not be offered any incentive. If the impact of offering an incentive is positive, they can then consider rolling an incentive out to the whole sample, or to a targeted sub-group.

Ipsos MORI have not included the 5-12 age group in the incentive trial as the sampled child is not making the decision to take the swab test and they would be incentivising their parent. This might raise significant ethical and reputational concerns.

Being able to conduct the trial specifically in the next Round of REACT is driven by the fact that Ipsos MORI are at a very important pre-Autumn juncture. Ipsos MORI need to make decisions on how to provide the best data we can get in a potentially volatile period for the Ministers who will need it.

DHSC have confirmed that as the data controllers, DHSC and Imperial College London, have determined to conduct the incentives trial and that Ipsos MORI as data processors are processing the data under their instruction.

The incentive trial has ethical approval.

The current study objectives remain as stated above. At wave 14 the national prevalence of infection will be measured at an important time – one week after schools return in September 2021, as well as the change in prevalence since late June/early July and the national average R value, with high accuracy. Imperial College London will also look at the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

****

At wave 12 Ipsos MORI changed the design. Instead of aiming to achieve approximately equal numbers of patients registering to take the test(s) in each Lower Tier Local Authority district, the design aims to achieve a sample where the numbers of patients registering to take the test(s) in each Lower Tier Local Authority district are in their correct proportion relative to the population of England. This change was made to maximise the statistical efficiency of the achieved sample as swab response has been declining. The design now maximises the precision of the prevalence estimates REACT 1 can provide at the national level, has increased the numbers sampled from areas of higher prevalence, the numbers sampled from ethnic minority backgrounds and those living in the key urban areas and the most deprived areas.

Ipsos MORI are acutely conscious of the need for data minimisation. Despite a falling swab response rate and a decline in the number of swabs returned Ipsos MORI has sought to offset this by making the design more statistically efficient and is now exploring the use of incentives to boost swab response. Ipsos MORI do however continue to need to sample ~835,000 individuals per wave to retain the benefits REACT has provided the Government, academic community studying SARS-CoV-2 infection and society at large.

The justification for the number of records requested remains the requirement for reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

As the National Data Opt Out is not applied patients that have opted out of their confidential patient information being used for research and planning will be included in the sample selected from the Personal Demographic Service. Ipsos MORI delete the data of those who have not registered for the study and requested that there is no further contact from Ipsos MORI in line with the Privacy Notice (typically two months after the end of fieldwork). As no individual can be re-sampled unless six months has elapsed since they were previously sampled and invited to take part, there is no need for Ipsos MORI to retain their contact details longer than that. NHSD do provide data of individuals that have been provided to Ipsos MORI under a previous drop where an individual has been re-sampled for REACT. No individual can be re-sampled unless six months has elapsed since they were previously sampled and invited to take part.

Once participants opt out of the study Ipsos MORI do not contact them again. Ipsos MORI review complaints and suggestions received through its helpline before every wave. The Privacy Notice and other documents have been amended following these reviews.

The study protocol also includes facility to undertake additional testing in specific local areas where an increase in transmission and numbers of cases is suspected. These additional surveys would involve sending out swab test kits to up to ca. 10,000 additional people in a specific area and ensuring rapid delivery and pick up, e.g. both delivery and pick-up by courier. Additional data requests may be needed in response to these local requirements.

The identifiable contact information will be securely destroyed by Ipsos MORI once the analysis has been completed. This has been communicated to participants in the updated privacy notice.

There is no other national register of the population that would allow contact with a named and representative sample of adults.

The data controller determines the purposes for which and the means by which personal data is processed. The DHSC and Imperial College London have designed and funded the study and are joint Data Controllers, and Ipsos MORI, who undertake the survey operations, is the primary Data Processor.

Data Controllers:

• Department of Health and Social Care (funder)

• Imperial College London (Sponsor)

Data Processors:

• Imperial College London

• Virtus Holdco Ltd (provision of ICL servers)

• Ipsos MORI - who are working with the following suppliers to deliver the project:

• Formara Limited - Print and despatch

• Adare SEC Ltd - Print and despatch

• Txtlocal Ltd - Text message reminders

• Tivian GmbH (Previously called Questback GmbH) - Online front-end data collection platform accessed by Ipsos Germany.

Please note that Eurofins provide laboratory services to Ipsos MORI for this study, but have not been added as Data Processor as the samples they process are anonymised. Eurofins do not have any access to NHS Digital data.

All suppliers will comply with the relevant provisions of UK GDPR and Regulation 7 of COPI.

The DHSC and Imperial College London as the data controllers satisfy the requirement in Regulation 7(2) COPI. DHSC have a data processing agreement in place with Ipsos MORI who are carrying out the processing under the instruction of DHSC and makes it clear that it is DHSC who has overall control of what happens to the personal data. In the processing agreement it is clear that confidential data must be treated in such a way to meet the Regulation 7(2) requirement (they owe a duty of confidentiality, as per Regulation 7(2)), staff within Ipsos MORI who are handling the data are aware of this fact and are subject to contractual obligations of confidentiality. These provisions in the data processing agreement satisfy the requirement in Regulation 7(2).

LEGAL BASIS for DATA PROCESSING

The lawful basis for processing data under GDPR has been reviewed against the guidance provided by IGARD and been assessed as acceptable. The Department of Health and Social Care and Imperial College London process data under the legal basis of Article 6(1)(e) ‘processing is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the controller’ as the Department of Health and Social Care is a public authority and government department, and Imperial College London is a Public Authority as it has a Royal Charter. The processing of data for this study is a task of public interest as it will provide a measure of the national prevalence of infection with a measure of the national average R value both nationally and across regional and local authority levels.

Both Data Controllers process special category data under Article 9(2)(j) ‘processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes …’ as the data are required for research purposes in the public interest and judged subject to the appropriate safeguards. The survey will involve processing special category data in relation to the health and ethnicity of individuals taking part. The results will be used for processing relevant information to support the current efforts of the UK government authorities for understanding the prevalence and spread of COVID-19, informing the public health response to the ongoing pandemic.

The legal basis for identifiable data to flow is under Regulation 3(4) of the National Health Service (Control of Patient Information Regulations) 2002 (COPI).

While the legal basis is public interest the research has had ethical approval from the South Central Berkshire B Research Ethics Committee (REC) and Health Regulator Authority (HRA).

NON-APPLICAITON OF NATIONA DATA OPT OUT

The identifiable data flowing into NHS Digital and out of NHS Digital does not have National Data Opt Out applied because the data flowing into NHS Digital does so under the operation of National Data Opt Out policy guidance which states that the National Data Opt Out does not apply to mandatory flows of data and for the flow of data out of NHS Digital it is flowing under the Control of Patient Information (COPI) Regulations 2012 under a Secretary of State for Public Health Direction. The COPI regulations states that the processing of patient information must align with Regulation 3(1) of COPI particularly in relation to the current outbreak of Covid-19. Regulation 3(1) confirms that confidential patient information may be processed for:

- diagnosing communicable diseases and other risks to public health;

- recognising trends in such diseases and risks;

- controlling and preventing the spread of such diseases and risks;

- monitoring and managing

- outbreaks of communicable disease;

- incidents of exposure to communicable disease;

- the delivery, efficacy and safety of immunisation programmes;

- adverse reactions to vaccines and medicines;

- risks of infection acquired from food or the environment (including water supplies);

- the giving of information to persons about the diagnosis of communicable disease and risks of acquiring such disease.

The COPI Regulations confirm that permission for processing confidential patient information is limited to:

(a) the Public Health Laboratory Service;

(b) persons employed or engaged for the purposes of the health service;

(c) other persons employed or engaged by a Government Department or other public authority in communicable disease surveillance.

The agreement fulfils this criteria and thus under COPI the National Data Opt Out is not applied.

The data requested from NHS Digital will be used to augment the data being collected as part of the REACT-1 programme to evaluate the prevalence of SARS-CoV-2 infection across England based on a random sample of the population.

The Demographic (PDS) data from NHS Digital will provide the contact details for a representative sample of the population aged 5+ to be invited to take part in the testing.

The study aims to provide reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and to feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

Data will not be made available to any third parties, other than those specified in this agreement, except in the form of aggregated outputs with small numbers suppressed in line with the suppression rules stated in the HES analysis guide.

REACT data has provided an accurate and localised view of the spread of COVID-19, feeding directly into the Government’s policy response. The REACT data has been used by SAGE to compute the R (reproduction) number in conjunction with other data sources. The findings from the surveys have directly informed the decisions to bring in regional and national “lockdowns”, the definition of the “tiers” across the country, and the lifting of restrictions in 2021. The study has provided benefits to both patients and the health care system by reducing mortality and morbidity. Public knowledge of the R rate at their local level has provided the information to support the public who are able to make choices and influenced behavioral changes in individuals and how they have personally responded to the pandemic.

The speed and scale of testing have enabled the study to quickly detect rising rates of infections and local hotspots, enabling timely public health responses from government. REACT has secured impact by its regular reports, leading to high profile media coverage and has been used in the Prime Minister’s televised press briefings. A video of how REACT works has been produced, and an infographics highlighting key study findings to help inform the general public and encourage safer behaviours.

In addition, the random sampling design means the study identified several thousand people who tested positive yet displayed no symptoms. By informing participants they were positive and needed to self-isolate, the REACT study has enabled scientists to have a better understanding of how the virus is transmitted and limited the spread of COVID-19.

The REACT study has been critical to ensure that the Government’s decision-making is based on robust, current evidence and that responses are targeted and appropriate.

The REACT study has helped to control the spread of COVID-19 and helped to prevent the spread of the virus across the UK and helped to reduce morbidity and mortality.

The REACT study has provided the public with access to data that has supported individuals to make personal choices in how they respond to the pandemic.

Further to this, the agile and collaborative approach of the REACT model and the wealth of knowledge accumulated has allowed for a very rapid response to changing requests, such as determining the usability and acceptability of different tests.

The study has informed the use of public health control measures such as closing schools, limiting public gatherings, issuing travel restrictions, quarantine rules and introducing the wearing of face coverings.

This application is to support Antigen testing study, round 2, (REACT-1-Round 2), one element of the REal-time Assessment of Community Transmission 1 (REACT 1): a study that will provide the basis for estimation of the R value in the community at regional and local authority levels, being conducted by Imperial College London on behalf of the Department of Health and Social Care (DHSC). This was initially thought to be a short study and data was originally requested to support Waves 4 to 6. Following this initial request DHSC have extended the contract to support Waves 7 to 14 due to the pandemic persisting. There have been a number of amendments to this agreement to support the addition of these Waves.

The main study objectives are:

1. To measure the national prevalence of infection during a window of time in each month.

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

Prevalence refers to the number of disease cases present in a particular population at a given time.

Measures of infection point prevalence will be provided at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response.

This research is being carried out to help the Government establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level. This is based on analysis of a nasopharyngeal self-swab obtained from a random sample of the population ages 5 and above using the NHS register as sampling frame.

The study involves a self-administered nasal and throat swab for adults and children over the age of 12. For children aged 12 or below, a parent or guardian will take the swab. For children aged 5 to 17 years old the invitation will be sent via parents/guardians. There will also be a brief online or telephone questionnaire to complete. Consent will be requested for follow up through routine health records to monitor long-term health outcomes, as has been done for REACT-1 baseline survey. There are no invasive procedures, although the throat and nasal swabs may cause some short-lived discomfort.

The Principal Investigator’s decision about the age criterion for REACT was taken in line with the guidance on the use of the swabs that were available when REACT was set up in April 2020. The age criterion is made clear in all the published reports. The age criterion has been consistent throughout REACT enabling robust estimates about the study population and consistent comparisons over time.

Results of the rt-PCR tests will be fed back to participants. If a positive test is recorded, the participant will be advised to self-isolate for 7 days from onset of any symptoms, or if asymptomatic, for 7 days. Other members of the household will be advised to self-isolate for 14 days in accordance with Government guidance. If the test is negative but the participant reports symptoms indicative of COVID-19 they and household members will similarly be advised to self-isolate in accordance with Government guidance. If the participant or a member of the household requires medical advice or attention they will be instructed to dial NHS 111 or if severely ill to dial 999 for an ambulance. With appropriate consent, positive test results could be fed into contact tracing efforts at the local level.

The research aims to contact a representative sample of the population aged 5+ in England in order to meet the study aims:

1. To measure the national prevalence of infection during a window of time in each month.

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

This sample will be selected from the Personal Demographic Service – considered an almost complete sample frame of adults in England, allowing stratification by gender, age and local authority.

REACT-1 aims to provide an accurate estimate of prevalence of infection across England at local authority levels. This sub-study REACT-1 (2), will provide the basis for estimation of the R value in the community at regional and local authority levels. The study is being conducted by Imperial College London (sponsor) on behalf of the Department of Health and Social Care (funder).

Patterns of hospital occupancy, death and 999 calls suggest that the COVID-19 epidemic in England is past its peak overall with a substantial downward trend. The change from increasing to decreasing incidence was almost certainly a result of social distancing and the national lockdown implemented on March 23rd: the net effect of the lockdown and subsequent lockdowns was to reduce the reproduction number R from well above 1 (~3 prior to its implementation) to below 1 currently. However, although it is known that R has likely been below 1 for some time, it cannot accurately infer resulting levels of infectious prevalence in the community nor can it assess its geographical variation, based on available data. As we transition out of lockdown, the relative risk of infection in one area compared to another will be closely related to the number of infectious people in that area, with the absolute risk dependent on the degree of social mixing at that time.

Because contact levels will increase, it is expected that there will be an increase in R which could lead to a subsequent wave of infection if R again becomes greater than 1. Based on experience in other countries, it is likely that R will vary over time and by geography. Also, a variety of policies may be implemented as lockdown is relaxed, some of which – such as contact tracing – may be supported locally and therefore vary geographically.

Although overall and local transmissibility of COVID-19 in the UK will continue to be assessed using mortality and hospital occupancy data, these data lag changes in transmission by a week or more (substantially more for deaths), and reflect transmission in hospitals and care homes as well as the community. Also, healthcare factors including the possibility of levels of transmission varying from one hospital to another, suggest that hospital data would not be useful for inferring local variation in community transmission.

Testing capacity for the UK has increased rapidly, reaching the target of 100,000 tests per day by the end of April, providing sufficient capacity to meet the testing needs of hospital patients, healthcare workers, other key workers; and potentially providing community testing of symptomatic individuals.

The REACT-1 programme is currently being used to establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level.

The data requested from the Personalt Demographic Service (PDS) will be used in order to select a nationally representative sample of the population aged 5+ to take part in the testing. The study needs to provide reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

This study is one component of a larger programme and sits alongside the REal-time Assessment of Community Transmission 2 (REACT 2): Usability and feasibility study of widespread home self-testing for SARS-CoV-2 antibodies.

The data required is demographic information from the Personal Demographics Service, to allow individuals aged 5+ to be contacted for the purpose of undertaking a survey collecting nasopharyngeal self-swab tests from an age-stratified representative population sample.

The data is requested in two stages:

- 1 – a record level identifiable population derived from PDS (including NHS Number, patient postcode, gender and age band) in order to allow selection of a representative sample

- 2 – a record level identifiable data request for selected sample only (including NHS Number, patient address, gender, age (month and year), telephone number and email – where available)

The original application was to support three waves of data being supplied to this format and represents rounds 4-6 of the antigen testing study. The surveys were completed in August, September and October 2020. In each wave in order to achieve the required sample size of 150,000 names and demographic details of ~750,000 individuals aged 5 years and above were needed. This was based on current response rates, which had fallen over the first three waves. This continued fall resulted in details being required for larger numbers of individuals in September and October in order to reach the target sample size.

An amendment to the agreement (v1.3) was, by extension of Ipsos MORI's contract by DHSC, also requested and had approved a further wave 7 of Demographic data in Late October 2020 at a larger sample size of ~835,000 individuals aged 5 years and above due to a poor response rate. A further amendment (v2.2) requested a further extraction of Demographic data by a further extension of Ipsos MORI's contract by DHSC, for waves 8 to 10 in late November 2020, January and February 2021). A further amendment (v3.3) provided Ipsos MORI with additional drops of PDS data for Waves 11-13, cohort size 835,000. This amendment was also to expand the scope of the purpose. As testing in the community becomes more widely available there is increasing possibility that REACT-1 may underestimate prevalence because potential participants opt to go to Pillar 2 for a test (or Pillar 1 for health care and care home workers). The study additionally aims to overcome this potential source of bias by including test results during the data collection window for positive cases.

This amendment (v4.1) is, by extension of Ipsos MORI's contract with DHSC, to also request a further wave 14 of PDS data in mid August 2021 at a sample size of ~835,000 individuals aged 5 years and above.

In addition, Ipsos MORI are exploring with the DHSC and Imperial College London an incentive trial at Round 14 to improve swab response. If this is agreed and also approved by the Research Ethics Committee in time, Ipsos MORI will request a separate and additional random sample of approximately 45,000 participants aged 13+, half of whom would be offered an incentive if they return a swab via a further amendment to this agreement.

The current study objectives remain as stated above. At wave 14 the national prevalence of infection will be measured at an important time – one week after schools return in September 2021, as well as the change in prevalence since late June/early July and the national average R value, with high accuracy. Imperial College London will also look at the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

At wave 12 Ipsos MORI changed the design. Instead of aiming to achieve approximately equal numbers of patients registering to take the test(s) in each Lower Tier Local Authority district, the design aims to achieve a sample where the numbers of patients registering to take the test(s) in each Lower Tier Local Authority district are in their correct proportion relative to the population of England. This change was made to maximise the statistical efficiency of the achieved sample as swab response has been declining. The design now maximises the precision of the prevalence estimates REACT 1 can provide at the national level, has increased the numbers sampled from areas of higher prevalence, the numbers sampled from ethnic minority backgrounds and those living in the key urban areas and the most deprived areas.

Ipsos MORI are acutely conscious of the need for data minimisation. Despite a falling swab response rate and a decline in the number of swabs returned Ipsos MORI has sought to offset this by making the design more statistically efficient and is now exploring the use of incentives to boost swab response. Ipsos MORI do however continue to need to sample ~835,000 individuals per wave to retain the benefits REACT has provided the Government, academic community studying SARS-CoV-2 infection and society at large.

The justification for the number of records requested remains the requirement for reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

As the National Data Opt Out is not applied patients that have opted out of their confidential patient information being used for research and planning will be included in the sample selected from the Personal Demographic Service. Ipsos MORI delete the data of those who have not registered for the study and requested that there is no further contact from Ipsos MORI in line with the Privacy Notice (typically two months after the end of fieldwork). As no individual can be re-sampled unless six months has elapsed since they were previously sampled and invited to take part, there is no need for Ipsos MORI to retain their contact details longer than that. NHSD do provide data of individuals that have been provided to Ipsos MORI under a previous drop where an individual has been re-sampled for REACT. No individual can be re-sampled unless six months has elapsed since they were previously sampled and invited to take part.

Once participants opt out of the study Ipsos MORI do not contact them again. Ipsos MORI review complaints and suggestions received through its helpline before every wave. The Privacy Notice and other documents have been amended following these reviews.

The study protocol also includes facility to undertake additional testing in specific local areas where an increase in transmission and numbers of cases is suspected. These additional surveys would involve sending out swab test kits to up to ca. 10,000 additional people in a specific area and ensuring rapid delivery and pick up, e.g. both delivery and pick-up by courier. Additional data requests may be needed in response to these local requirements.

The identifiable contact information will be securely destroyed by Ipsos MORI once the analysis has been completed. This has been communicated to participants in the updated privacy notice.

There is no other national register of the population that would allow contact with a named and representative sample of adults.

The data controller determines the purposes for which and the means by which personal data is processed. The DHSC and Imperial College London have designed and funded the study and are joint Data Controllers, and Ipsos MORI, who undertake the survey operations, is the primary Data Processor.

Data Controllers:

• Department of Health and Social Care (funder)

• Imperial College London (Sponsor)

Data Processors:

• Imperial College London

• Virtus Holdco Ltd (provision of ICL servers)

• Ipsos MORI - who are working with the following suppliers to deliver the project:

• Formara Limited - Print and despatch

• Adare SEC Ltd - Print and despatch

• Txtlocal Ltd - Text message reminders

• Tivian GmbH (Previously called Questback GmbH) - Online front-end data collection platform accessed by Ipsos Germany.

Please note that Eurofins provide laboratory services to Ipsos MORI for this study, but have not been added as Data Processor as the samples they process are anonymised. Eurofins do not have any access to NHS Digital data.

All suppliers will comply with the relevant provisions of UK GDPR and Regulation 7 of COPI.

The DHSC and Imperial College London as the data controllers satisfy the requirement in Regulation 7(2) COPI. DHSC have a data processing agreement in place with Ipsos MORI who are carrying out the processing under the instruction of DHSC and makes it clear that it is DHSC who has overall control of what happens to the personal data. In the processing agreement it is clear that confidential data must be treated in such a way to meet the Regulation 7(2) requirement (they owe a duty of confidentiality, as per Regulation 7(2)), staff within Ipsos MORI who are handling the data are aware of this fact and are subject to contractual obligations of confidentiality. These provisions in the data processing agreement satisfy the requirement in Regulation 7(2).

LEGAL BASIS for DATA PROCESSING

The lawful basis for processing data under GDPR has been reviewed against the guidance provided by IGARD and been assessed as acceptable. The Department of Health and Social Care and Imperial College London process data under the legal basis of Article 6(1)(e) ‘processing is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the controller’ as the Department of Health and Social Care is a public authority and government department, and Imperial College London is a Public Authority as it has a Royal Charter. The processing of data for this study is a task of public interest as it will provide a measure of the national prevalence of infection with a measure of the national average R value both nationally and across regional and local authority levels.

Both Data Controllers process special category data under Article 9(2)(j) ‘processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes …’ as the data are required for research purposes in the public interest and judged subject to the appropriate safeguards. The survey will involve processing special category data in relation to the health and ethnicity of individuals taking part. The results will be used for processing relevant information to support the current efforts of the UK government authorities for understanding the prevalence and spread of COVID-19, informing the public health response to the ongoing pandemic.

The legal basis for identifiable data to flow is under Regulation 3(4) of the National Health Service (Control of Patient Information Regulations) 2002 (COPI).

While the legal basis is public interest the research has had ethical approval from the South Central Berkshire B Research Ethics Committee (REC) and Health Regulator Authority (HRA).

NON-APPLICAITON OF NATIONA DATA OPT OUT

The identifiable data flowing into NHS Digital and out of NHS Digital does not have National Data Opt Out applied because the data flowing into NHS Digital does so under the operation of National Data Opt Out policy guidance which states that the National Data Opt Out does not apply to mandatory flows of data and for the flow of data out of NHS Digital it is flowing under the Control of Patient Information (COPI) Regulations 2012 under a Secretary of State for Public Health Direction. The COPI regulations states that the processing of patient information must align with Regulation 3(1) of COPI particularly in relation to the current outbreak of Covid-19. Regulation 3(1) confirms that confidential patient information may be processed for:

- diagnosing communicable diseases and other risks to public health;

- recognising trends in such diseases and risks;

- controlling and preventing the spread of such diseases and risks;

- monitoring and managing

- outbreaks of communicable disease;

- incidents of exposure to communicable disease;

- the delivery, efficacy and safety of immunisation programmes;

- adverse reactions to vaccines and medicines;

- risks of infection acquired from food or the environment (including water supplies);

- the giving of information to persons about the diagnosis of communicable disease and risks of acquiring such disease.

The COPI Regulations confirm that permission for processing confidential patient information is limited to:

(a) the Public Health Laboratory Service;

(b) persons employed or engaged for the purposes of the health service;

(c) other persons employed or engaged by a Government Department or other public authority in communicable disease surveillance.

The agreement fulfils this criteria and thus under COPI the National Data Opt Out is not applied.

The data requested from NHS Digital will be used to augment the data being collected as part of the REACT-1 programme to evaluate the prevalence of SARS-CoV-2 infection across England based on a random sample of the population.

The Demographic (PDS) data from NHS Digital will provide the contact details for a representative sample of the population aged 5+ to be invited to take part in the testing.

The study aims to provide reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and to feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

Data will not be made available to any third parties, other than those specified in this agreement, except in the form of aggregated outputs with small numbers suppressed in line with the suppression rules stated in the HES analysis guide.

REACT data has provided an accurate and localised view of the spread of COVID-19, feeding directly into the Government’s policy response. The REACT data has been used by SAGE to compute the R (reproduction) number in conjunction with other data sources. The findings from the surveys have directly informed the decisions to bring in regional and national “lockdowns”, the definition of the “tiers” across the country, and the lifting of restrictions in 2021. The study has provided benefits to both patients and the health care system by reducing mortality and morbidity. Public knowledge of the R rate at their local level has provided the information to support the public who are able to make choices and influenced behavioral changes in individuals and how they have personally responded to the pandemic.

The speed and scale of testing have enabled the study to quickly detect rising rates of infections and local hotspots, enabling timely public health responses from government. REACT has secured impact by its regular reports, leading to high profile media coverage and has been used in the Prime Minister’s televised press briefings. A video of how REACT works has been produced, and an infographics highlighting key study findings to help inform the general public and encourage safer behaviours.

In addition, the random sampling design means the study identified several thousand people who tested positive yet displayed no symptoms. By informing participants they were positive and needed to self-isolate, the REACT study has enabled scientists to have a better understanding of how the virus is transmitted and limited the spread of COVID-19.

The REACT study has been critical to ensure that the Government’s decision-making is based on robust, current evidence and that responses are targeted and appropriate.

The REACT study has helped to control the spread of COVID-19 and helped to prevent the spread of the virus across the UK and helped to reduce morbidity and mortality.

The REACT study has provided the public with access to data that has supported individuals to make personal choices in how they respond to the pandemic.

Further to this, the agile and collaborative approach of the REACT model and the wealth of knowledge accumulated has allowed for a very rapid response to changing requests, such as determining the usability and acceptability of different tests.

The study has informed the use of public health control measures such as closing schools, limiting public gatherings, issuing travel restrictions, quarantine rules and introducing the wearing of face coverings

This application is to support Antigen testing study, round 2, (REACT-1-Round 2), one element of the REal-time Assessment of Community Transmission 1 (REACT 1): a study that will provide the basis for estimation of the R value in the community at regional and local authority levels, being conducted by Imperial College London on behalf of the Department of Health and Social Care (DHSC).

The main study objectives are:

1. To measure the national prevalence of infection during a window of time in each month (July 2020 through to March 2021).

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

Measures of infection point prevalence will be provided at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response.

This research is being carried out to help the Government establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level. This is based on analysis of a nasopharyngeal self-swab obtained from a random sample of the population ages 5 and above using the NHS register as sampling frame.

The study involves a self-administered nasal and throat swab for adults and children over the age of 12. For children aged 12 or below, a parent or guardian will take the swab. For children aged 5 to 17 years old the invitation will be sent via parents/guardians. There will also be a brief online or telephone questionnaire to complete. Consent will be requested for follow up through routine health records to monitor long-term health outcomes, as has been done for REACT-1 baseline survey. There are no invasive procedures, although the throat and nasal swabs may cause some short-lived discomfort.

Results of the rt-PCR tests will be fed back to participants. If a positive test is recorded, the participant will be advised to self-isolate for 7 days from onset of any symptoms, or if asymptomatic, for 7 days. Other members of the household will be advised to self-isolate for 14 days in accordance with Government guidance. If the test is negative but the participant reports symptoms indicative of COVID-19 they and household members will similarly be advised to self-isolate in accordance with Government guidance. If the participant or a member of the household requires medical advice or attention they will be instructed to dial NHS 111 or if severely ill to dial 999 for an ambulance. With appropriate consent, positive test results could be fed into contact tracing efforts at the local level.

The research aims to contact a representative sample of the population aged 5+ in England in order to meet the study aims:

1. To measure the national prevalence of infection during a window of time in each month (August, September and October).

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

This sample will be selected from the PDS – considered an almost complete sample frame of adults in England, allowing stratification by gender, age and local authority.

REACT-1 aims to provide an accurate estimate of prevalence of infection across England at local authority levels. This sub-study REACT-1 (2), will provide the basis for estimation of the R value in the community at regional and local authority levels. The study is being conducted by Imperial College London (sponsor) on behalf of the Department of Health and Social Care (funder).

Patterns of hospital occupancy, death and 999 calls suggest that the COVID-19 epidemic in England is past its peak overall with a substantial downward trend. The change from increasing to decreasing incidence was almost certainly a result of social distancing and the national lockdown implemented on March 23rd: the net effect of the lockdown was to reduce the reproduction number R from well above 1 (~3 prior to its implementation) to below 1 currently. However, although it is known that R has likely been below 1 for some time, it cannot accurately infer resulting levels of infectious prevalence in the community nor can it assess its geographical variation, based on available data. As we transition out of lockdown, the relative risk of infection in one area compared to another will be closely related to the number of infectious people in that area, with the absolute risk dependent on the degree of social mixing at that time.

Because contact levels will increase, it is expected that there will be an increase in R which could lead to a second wave of infection if R again becomes greater than 1. Based on experience in other countries, it is likely that R will vary over time and by geography. Also, a variety of policies may be implemented as lockdown is relaxed, some of which – such as contact tracing – may be supported locally and therefore vary geographically.

Although overall and local transmissibility of COVID-19 in the UK will continue to be assess using mortality and hospital occupancy data, these data lag changes in transmission by a week or more (substantially more for deaths), and reflect transmission in hospitals and care homes as well as the community. Also, healthcare factors including the possibility of levels of transmission varying from one hospital to another, suggest that hospital data would not be useful for inferring local variation in community transmission.

Testing capacity for the UK has increased rapidly, reaching the target of 100,000 tests per day by the end of April, providing sufficient capacity to meet the testing needs of hospital patients, healthcare workers, other key workers; and potentially providing community testing of symptomatic individuals.

The REACT-1 programme is currently being used to establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level.

The data requested from the Patient Demographic Service (PDS) will be used in order to select a nationally representative sample of the population aged 5+ to take part in the testing. The study needs to provide reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

This study is one component of a larger programme and sits alongside the REal-time Assessment of Community Transmission 2 (REACT 2): Usability and feasibility study of widespread home self-testing for SARS-CoV-2 antibodies.

The data required is demographic information from the Personal Demographics Service, to allow individuals aged 5+ to be contacted for the purpose of undertaking a survey collecting nasopharyngeal self-swab tests from an age-stratified representative population sample.

The data is requested in two stages:

- 1 – a record level identifiable population derived from PDS (including NHS Number, patient postcode, gender and age band) in order to allow selection of a representative sample

- 2 – a record level identifiable data request for selected sample only (including NHS Number, patient address, gender, age (month and year), telephone number and email – where available)

The original application is to support three waves of data being supplied to this format and represents rounds 4-6 of the antigen testing study. The surveys were completed in August, September and October 2020. In each wave in order to achieve the required sample size of 150,000 names and demographic details of ~750,000 individuals aged 5 years and above are needed. This is based on current response rates, which have fallen over the first three waves. If this continues, details may be required for larger numbers of individuals in September and October in order to reach the target sample size.

An amendment to this agreement (v1.3) was, by extension of Ipsos MORI's contract by DHSC, also requested and had approved a further wave 7 of Demographic data in Late October 2020 at a larger sample size of ~835,000 individuals aged 5 years and above due to a poor response rate. A further amendment amendment (v2.2) requested a further extraction of Demographic data by a further extension of Ipsos MORI's contract by DHSC, for waves 8 to 10 in late November 2021, January and February 2021)

*** This amendment (v3.3) is to provide Ipsos MORI with additional drops of Demographics GPES data for Waves 11-13. This amendment is also to expand the scope of the purpose. As testing in the community becomes more widely available there is increasing possibility that REACT-1 may underestimate prevalence because potential participants opt to go to Pillar 2 for a test (or Pillar 1 for health care and care home workers). The study additionally aims to overcome this potential source of bias by including test results during the data collection window for positive cases. Imperial College London are a Data Processor who will receive record-level Pillar 1 and Pillar 2 testing data and Civil Registration (Deaths)data extract linked to cohorts provided by Ipsos MORI for Waves 1 through 13. ****

The study protocol also includes facility to undertake additional testing in specific local areas where an increase in transmission and numbers of cases is suspected. These additional surveys would involve sending out swab test kits to up to ca. 10,000 additional people in a specific area and ensuring rapid delivery and pick up, e.g. both delivery and pick-up by courier. Additional data requests may be needed in response to these local requirements.

The identifiable contact information will be securely destroyed by Ipsos MORI once the analysis has been completed. This has been communicated to participants in the updated privacy notice as end March 2021.

There is no other national register of the population that would allow contact with a named and representative sample of adults.

Data controllers:

• Department of Health and Social Care (funder)

• Imperial College London (Sponsor)

Data processors:

• Imperial College London

• Virtus Holdco Ltd (provision of ICL servers)

• Ipsos MORI - who are working with the following suppliers to deliver the project:

• Formara Limited - Print and despatch

• Adare SEC Ltd - Print and despatch

• Txtlocal Ltd - Text message reminders

• Questback GmbH - Online front-end data collection platform accessed by Ipsos Germany.

Please note that Eurofins provide laboratory services to Ipsos MORI for this study, but have not been added as Data Processor as the samples they process are anonymised. Eurofins do not have any access to NHS Digital data.***

LEGAL BASIS for DATA PROCESSING

The lawful basis for processing data under GDPR has been reviewed against the guidance provided by IGARD and been assessed as acceptable. The Department of Health and Social Care and Imperial College London process data under the legal basis of Article 6(1)(e) ‘processing is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the controller’ as the Department of Health and Social Care is a public authority and government department, and Imperial College London is a Public Authority as it has a Royal Charter. The processing of data for this study is a task of public interest as it will provide a measure of the national prevalence of infection with a measure of the national average R value both nationally and across regional and local authority levels.

Both Data Controllers process special category data under Article 9(2)(j) ‘processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes …’ as the data are required for research purposes in the public interest and judged subject to the appropriate safeguards. The survey will involve processing special category data in relation to the health and ethnicity of individuals taking part. The results will be used for processing relevant information to support the current efforts of the UK government authorities for understanding the prevalence and spread of COVID-19, informing the public health response to the ongoing pandemic.

The legal basis for identifiable data to flow is under Regulation 3(4) of the National Health Service (Control of Patient Information Regulations) 2002 (COPI).

While the legal basis is public interest the research has had ethical approval from the South Central Berkshire B Research Ethics Committee (REC) and Health Regulator Authority (HRA).

The data requested from NHS Digital will be used to augment the data being collected as part of the REACT-1 programme to evaluate the prevalence of SARS-CoV-2 infection across England based on a random sample of the population.

The Demographic (PDS) data from NHS Digital will provide the contact details for a representative sample of the population aged 5+ to be invited to take part in the testing.

The study aims to provide reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and to feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

Data will not be made available to any third parties, other than those specified in this agreement, except in the form of aggregated outputs with small numbers suppressed in line with the suppression rules stated in the HES analysis guide.

**** ADDITIONAL INFORMATION UNDER THIS AMENDMENT (v3.3)

REACT 1 has become one of the most important surveillance studies. Data from Rounds 1 to 9, all of which have used Demographic (PDS) data from NHS Digital. It has provided real-time assessment of virus prevalence which has provided very early information about the epidemiological trajectory at regional and national levels: informing officials, Ministers and others and continues to support COVID-19 policy development and implementation across Government which affects everyone in England.

Further to this, the agile and collaborative approach of the REACT model and the wealth of knowledge accumulated has allowed for a very rapid response to changing requests, such as determining the usability and acceptability of different tests. ****

****This amendment (v2.2) is to provide Ipsos MORI with four additional drops of Demographics data for three rounds of data (rounds 8-10). ****

This application is to support Antigen testing study, round 2, (REACT-1-Round 2), one element of the REal-time Assessment of Community Transmission 1 (REACT 1): a study that will provide the basis for estimation of the R value in the community at regional and local authority levels, being conducted by Imperial College London on behalf of the Department of Health and Social Care (DHSC).

The main study objectives are:

1. To measure the national prevalence of infection during a window of time in each month (July 2020 through to March 2021).

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

Measures of infection point prevalence will be provided at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response.

This research is being carried out to help the Government establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level. This is based on analysis of a nasopharyngeal self-swab obtained from a random sample of the population ages 5 and above using the NHS register as sampling frame. The lawful basis for processing is Article 6(1) (e) - “processing is necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller and Article 9(2) (j) ‘processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes’. The processing of data for this study is a task of public interest as it will provide a measure of the national prevalence of infection with a measure of the national average R value both nationally and across regional and local authority levels.

The legal basis for identifiable data to flow is under Regulation 3(4) of the National Health Service (Control of Patient Information Regulations) 2002 (COPI).

The survey will involve processing special category data in relation to the health and ethnicity of individuals taking part. The results will be used for processing relevant information to support the current efforts of the UK government authorities for understanding the prevalence and spread of COVID-19, informing the public health response to the ongoing pandemic.

While the legal basis is public interest the research has had ethical approval from the South Central Berkshire B Research Ethics Committee (REC) and Health Regulator Authority (HRA).

The study involves a self-administered nasal and throat swab for adults and children over the age of 12. For children aged 12 or below, a parent or guardian will take the swab. For children aged 5 to 17 years old the invitation will be sent via parents/guardians. There will also be a brief online or telephone questionnaire to complete. Consent will be requested for follow up through routine health records to monitor long-term health outcomes, as has been done for REACT-1 baseline survey. There are no invasive procedures, although the throat and nasal swabs may cause some short-lived discomfort.

Results of the rt-PCR tests will be fed back to participants. If a positive test is recorded, the participant will be advised to self-isolate for 7 days from onset of any symptoms, or if asymptomatic, for 7 days. Other members of the household will be advised to self-isolate for 14 days in accordance with Government guidance. If the test is negative but the participant reports symptoms indicative of COVID-19 they and household members will similarly be advised to self-isolate in accordance with Government guidance. If the participant or a member of the household requires medical advice or attention they will be instructed to dial NHS 111 or if severely ill to dial 999 for an ambulance. With appropriate consent, positive test results could be fed into contact tracing efforts at the local level.

The research aims to contact a representative sample of the population aged 5+ in England in order to meet the study aims:

1. To measure the national prevalence of infection during a window of time in each month (August, September and October).

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

This sample will be selected from the PDS – considered an almost complete sample frame of adults in England, allowing stratification by gender, age and local authority.

REACT-1 aims to provide an accurate estimate of prevalence of infection across England at local authority levels. This sub-study REACT-1 (2), will provide the basis for estimation of the R value in the community at regional and local authority levels. The study is being conducted by Imperial College London (sponsor) on behalf of the Department of Health and Social Care (funder).

Patterns of hospital occupancy, death and 999 calls suggest that the COVID-19 epidemic in England is past its peak overall with a substantial downward trend. The change from increasing to decreasing incidence was almost certainly a result of social distancing and the national lockdown implemented on March 23rd: the net effect of the lockdown was to reduce the reproduction number R from well above 1 (~3 prior to its implementation) to below 1 currently. However, although it is known that R has likely been below 1 for some time, it cannot accurately infer resulting levels of infectious prevalence in the community nor can it assess its geographical variation, based on available data. As we transition out of lockdown, the relative risk of infection in one area compared to another will be closely related to the number of infectious people in that area, with the absolute risk dependent on the degree of social mixing at that time.

Because contact levels will increase, it is expected that there will be an increase in R which could lead to a second wave of infection if R again becomes greater than 1. Based on experience in other countries, it is likely that R will vary over time and by geography. Also, a variety of policies may be implemented as lockdown is relaxed, some of which – such as contact tracing – may be supported locally and therefore vary geographically.

Although overall and local transmissibility of COVID-19 in the UK will continue to be assess using mortality and hospital occupancy data, these data lag changes in transmission by a week or more (substantially more for deaths), and reflect transmission in hospitals and care homes as well as the community. Also, healthcare factors including the possibility of levels of transmission varying from one hospital to another, suggest that hospital data would not be useful for inferring local variation in community transmission.

Testing capacity for the UK has increased rapidly, reaching the target of 100,000 tests per day by the end of April, providing sufficient capacity to meet the testing needs of hospital patients, healthcare workers, other key workers; and potentially providing community testing of symptomatic individuals.

The REACT-1 programme is currently being used to establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level.

The data requested from the Patient Demographic Service (PDS) will be used in order to select a nationally representative sample of the population aged 5+ to take part in the testing. The study needs to provide reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

This study is one component of a larger programme and sits alongside the REal-time Assessment of Community Transmission 2 (REACT 2): Usability and feasibility study of widespread home self-testing for SARS-CoV-2 antibodies.

The data required is demographic information from the Personal Demographics Service, to allow individuals aged 5+ to be contacted for the purpose of undertaking a survey collecting nasopharyngeal self-swab tests from an age-stratified representative population sample.

The data is requested in two stages

- 1 – a record level identifiable population derived from PDS (including NHS Number, patient postcode, gender and age band) in order to allow selection of a representative sample

- 2 – a record level identifiable data request for selected sample only (including NHS Number, patient address, gender, age (month and year), telephone number and email – where available)

The original application is to support three waves of data being supplied to this format and represents rounds 4-6 of the antigen testing study. The surveys were completed in August, September and October 2020. In each wave in order to achieve the required sample size of 150,000 names and demographic details of ~750,000 individuals aged 5 years and above are needed. This is based on current response rates, which have fallen over the first three waves. If this continues, details may be required for larger numbers of individuals in September and October in order to reach the target sample size.

Further amendments to this agreement (v1.3 and v2.2), by extension of Ipsos MORI's contract by DHSC, also requested and had approved a further wave 7 of Demographic data in Late October 2020 at a larger sample size of ~835,000 individuals aged 5 years and above due to a poor response rate. Then a further extraction of Demographic data by a further extension of Ipsos MORI's contract by DHSC, for waves 8 to 10 in late November, January and February 2021.

The study protocol also includes facility to undertake additional testing in specific local areas where an increase in transmission and numbers of cases is suspected. These additional surveys would involve sending out swab test kits to up to ca. 10,000 additional people in a specific area and ensuring rapid delivery and pick up, e.g. both delivery and pick-up by courier. Additional data requests may be needed in response to these local requirements.

The identifiable contact information will be securely destroyed by Ipsos MORI once the analysis has been completed. This has been communicated to participants in the updated privacy notice as end March 2021.

There is no other national register of the population that would allow contact with a named and representative sample of adults.

Data controllers:

• Department of Health and Social Care (funder)

• Imperial College London (Sponsor)

Data processors:

*** Imperial College London ****

• Ipsos MORI - who are working with the following suppliers to deliver the project:

• Formara Limited - Print and despatch

• Adare SEC Ltd - Print and despatch

• Txtlocal Ltd - Text message reminders

• Questback GmbH - Online front-end data collection platform accessed by Ipsos Germany.

Please note that Eurofins provide laboratory services to Ipsos MORI for this study, but have not been added as Data Processor as the samples they process are anonymised. Eurofins do not have any access to NHS Digital data.

LEGAL BASIS for DATA PROCESSING

The lawful basis for processing data under GDPR has been reviewed against the guidance provided by IGARD and been assessed as acceptable. The Department of Health and Social Care and Imperial College London process data under the legal basis of Article 6(1)(e) ‘processing is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the controller’ as the Department of Health and Social Care is a public authority and government department, and Imperial College London is a Public Authority as it has a Royal Charter. Both Data Controllers process special category data under Article 9(2)(j) ‘processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes …’ as the data are required for research purposes in the public interest and judged subject to the appropriate safeguards.

The Demographic (PDS) data from NHS Digital will provide the contact details for a representative sample of the population aged 5+ to be invited to take part in the testing.

The study aims to provide reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and to feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

Data will not be made available to any third parties, other than those specified in this agreement, except in the form of aggregated outputs with small numbers suppressed in line with the suppression rules stated in the HES analysis guide.

This amendment is to provide Ipsos MORI with an additional drop of Demographics data for an additional round of data. Due to a lower than expected response rate at Round 6, Ipsos MORI are requesting an additional drop of data to support Round 7.

This application is to support Antigen testing study, round 2, (REACT-1-Round 2), one element of the REal-time Assessment of Community Transmission 1 (REACT 1): a study that will provide the basis for estimation of the R value in the community at regional and local authority levels, being conducted by Imperial College London on behalf of the Department of Health and Social Care (DHSC).

The main study objectives are:

1. To measure the national prevalence of infection during a window of time in each month (August, September and October).

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

Measures of infection point prevalence will be provided at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response.

This research is being carried out to help the Government establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level. This is based on analysis of a nasopharyngeal self-swab obtained from a random sample of the population ages 5 and above using the NHS register as sampling frame. The lawful basis for processing is Article 6(1) (e) - “processing is necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller and Article 9(2) (j) ‘processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes’. The processing of data for this study is a task of public interest as it will provide a measure of the national prevalence of infection with a measure of the national average R value both nationally and across regional and local authority levels.

The legal basis for identifiable data to flow is under Regulation 3(4) of the National Health Service (Control of Patient Information Regulations) 2002 (COPI).

The survey will involve processing special category data in relation to the health and ethnicity of individuals taking part. The results will be used for processing relevant information to support the current efforts of the UK government authorities for understanding the prevalence and spread of COVID-19, informing the public health response to the ongoing pandemic.

While the legal basis is public interest the research has had ethical approval from the South Central Berkshire B Research Ethics Committee (REC) and Health Regulator Authority (HRA).

The study involves a self-administered nasal and throat swab for adults and children over the age of 12. For children aged 12 or below, a parent or guardian will take the swab. For children aged 5 to 17 years old the invitation will be sent via parents/guardians. There will also be a brief online or telephone questionnaire to complete. Consent will be requested for follow up through routine health records to monitor long-term health outcomes, as has been done for REACT-1 baseline survey. There are no invasive procedures, although the throat and nasal swabs may cause some short-lived discomfort.

Results of the rt-PCR tests will be fed back to participants. If a positive test is recorded, the participant will be advised to self-isolate for 7 days from onset of any symptoms, or if asymptomatic, for 7 days. Other members of the household will be advised to self-isolate for 14 days in accordance with Government guidance. If the test is negative but the participant reports symptoms indicative of COVID-19 they and household members will similarly be advised to self-isolate in accordance with Government guidance. If the participant or a member of the household requires medical advice or attention they will be instructed to dial NHS 111 or if severely ill to dial 999 for an ambulance. With appropriate consent, positive test results could be fed into contact tracing efforts at the local level.

The research aims to contact a representative sample of the population aged 5+ in England in order to meet the study aims:

1. To measure the national prevalence of infection during a window of time in each month (August, September and October).

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

This sample will be selected from the PDS – considered an almost complete sample frame of adults in England, allowing stratification by gender, age and local authority.

REACT-1 aims to provide an accurate estimate of prevalence of infection across England at local authority levels. This sub-study REACT-1 (2), will provide the basis for estimation of the R value in the community at regional and local authority levels. The study is being conducted by Imperial College London (sponsor) on behalf of the Department of Health and Social Care (funder).

Patterns of hospital occupancy, death and 999 calls suggest that the COVID-19 epidemic in England is past its peak overall with a substantial downward trend. The change from increasing to decreasing incidence was almost certainly a result of social distancing and the national lockdown implemented on March 23rd: the net effect of the lockdown was to reduce the reproduction number R from well above 1 (~3 prior to its implementation) to below 1 currently. However, although it is known that R has likely been below 1 for some time, it cannot accurately infer resulting levels of infectious prevalence in the community nor can it assess its geographical variation, based on available data. As we transition out of lockdown, the relative risk of infection in one area compared to another will be closely related to the number of infectious people in that area, with the absolute risk dependent on the degree of social mixing at that time.

Because contact levels will increase, it is expected that there will be an increase in R which could lead to a second wave of infection if R again becomes greater than 1. Based on experience in other countries, it is likely that R will vary over time and by geography. Also, a variety of policies may be implemented as lockdown is relaxed, some of which – such as contact tracing – may be supported locally and therefore vary geographically.

Although overall and local transmissibility of COVID-19 in the UK will continue to be assess using mortality and hospital occupancy data, these data lag changes in transmission by a week or more (substantially more for deaths), and reflect transmission in hospitals and care homes as well as the community. Also, healthcare factors including the possibility of levels of transmission varying from one hospital to another, suggest that hospital data would not be useful for inferring local variation in community transmission.

Testing capacity for the UK has increased rapidly, reaching the target of 100,000 tests per day by the end of April, providing sufficient capacity to meet the testing needs of hospital patients, healthcare workers, other key workers; and potentially providing community testing of symptomatic individuals.

The REACT-1 programme is currently being used to establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level.

The data requested from the Patient Demographic Service (PDS) will be used in order to select a nationally representative sample of the population aged 5+ to take part in the testing. The study needs to provide reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

This study is one component of a larger programme and sits alongside the REal-time Assessment of Community Transmission 2 (REACT 2): Usability and feasibility study of widespread home self-testing for SARS-CoV-2 antibodies.

The data required is demographic information from the Personal Demographics Service, to allow individuals aged 5+ to be contacted for the purpose of undertaking a survey collecting nasopharyngeal self-swab tests from an age-stratified representative population sample.

The data is requested in two stages

- 1 – a record level identifiable population derived from PDS (including NHS Number, patient postcode, gender and age band) in order to allow selection of a representative sample

- 2 – a record level identifiable data request for selected sample only (including NHS Number, patient address, gender, age (month and year), telephone number and email – where available)

This application is to support three waves of data being supplied to this format and represents rounds 4-6 of the antigen testing study. The surveys will be completed in August, September and October 2020. In each wave order to achieve the required sample size of 150,000 names and demographic details of ~750,000 individuals aged 5 years and above are needed. This is based on current response rates, which have fallen over the first three waves. If this continues, details may be required for larger numbers of individuals in September and October in order to reach the target sample size.

The study protocol also includes facility to undertake additional testing in specific local areas where an increase in transmission and numbers of cases is suspected. These additional surveys would involve sending out swab test kits to up to ca. 10,000 additional people in a specific area and ensuring rapid delivery and pick up, e.g. both delivery and pick-up by courier. Additional data requests may be needed in response to these local requirements.

The identifiable contact information will be securely destroyed by Ipsos MORI once the analysis has been completed. This has been communicated to participants in the privacy notice as December 2020.

There is no other national register of the population that would allow contact with a named and representative sample of adults.

Month and year of birth is sufficient for processing, rather than full date of birth.

Data controllers:

• Department of Health and Social Care (funder)

• Imperial College London (Sponsor)

Data processor:

• Ipsos MORI

Ipsos MORI are working with the following suppliers to deliver the project:

• Print and despatch (Formara and Adare)

• Mail service (Royal Mail)

• Text message reminders (TextLocal)

• Online data collection (Ipsos Germany)

• Courier services (co-ordinated by The Delivery Group)

• Laboratory Services (Eurofins)

These suppliers are approved and compliant with the General Data Protection Regulations.

LEGAL BASIS for DATA PROCESSING

The lawful basis for processing data under GDPR has been reviewed against the guidance provided by IGARD and been assessed as acceptable. The Department of Health and Social Care and Imperial College London process data under the legal basis of Article 6(1)(e) ‘processing is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the controller’ as the Department of Health and Social Care is a public authority and government department, and Imperial College London is a Public Authority as it has a Royal Charter. Both Data Controllers process special category data under Article 9(2)(j) ‘processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes …’ as the data are required for research purposes in the public interest and judged subject to the appropriate safeguards.

The data from NHS Digital will provide the contact details for a representative sample of the population aged 5+ to be invited to take part in the testing.

Data file shared with Imperial College London, for the purposes of further analysis. This file will include all survey responses, some sample information (postcode, gender, age) with additional variables mapped (including IMD, local authority name).

This application is to support Antigen testing study, round 2, (REACT-1-Round 2), one element of the REal-time Assessment of Community Transmission 1 (REACT 1): a study that will provide the basis for estimation of the R value in the community at regional and local authority levels, being conducted by Imperial College London on behalf of the Department of Health and Social Care (DHSC).

The main study objectives are:

1. To measure the national prevalence of infection during a window of time in each month (August, September and October).

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

Measures of infection point prevalence will be provided at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response.

This research is being carried out to help the Government establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level. This is based on analysis of a nasopharyngeal self-swab obtained from a random sample of the population ages 5 and above using the NHS register as sampling frame. The lawful basis for processing is Article 6(1) (e) - “processing is necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller and Article 9(2) (j) ‘processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes’. The processing of data for this study is a task of public interest as it will provide a measure of the national prevalence of infection with a measure of the national average R value both nationally and across regional and local authority levels.

The legal basis for identifiable data to flow is under Regulation 3(4) of the National Health Service (Control of Patient Information Regulations) 2002 (COPI).

The survey will involve processing special category data in relation to the health and ethnicity of individuals taking part. The results will be used for processing relevant information to support the current efforts of the UK government authorities for understanding the prevalence and spread of COVID-19, informing the public health response to the ongoing pandemic.

While the legal basis is public interest the research has had ethical approval from the South Central Berkshire B Research Ethics Committee (REC) and Health Regulator Authority (HRA).

The study involves a self-administered nasal and throat swab for adults and children over the age of 12. For children aged 12 or below, a parent or guardian will take the swab. For children aged 5 to 17 years old the invitation will be sent via parents/guardians. There will also be a brief online or telephone questionnaire to complete. Consent will be requested for follow up through routine health records to monitor long-term health outcomes, as has been done for REACT-1 baseline survey. There are no invasive procedures, although the throat and nasal swabs may cause some short-lived discomfort.

Results of the rt-PCR tests will be fed back to participants. If a positive test is recorded, the participant will be advised to self-isolate for 7 days from onset of any symptoms, or if asymptomatic, for 7 days. Other members of the household will be advised to self-isolate for 14 days in accordance with Government guidance. If the test is negative but the participant reports symptoms indicative of COVID-19 they and household members will similarly be advised to self-isolate in accordance with Government guidance. If the participant or a member of the household requires medical advice or attention they will be instructed to dial NHS 111 or if severely ill to dial 999 for an ambulance. With appropriate consent, positive test results could be fed into contact tracing efforts at the local level.

The research aims to contact a representative sample of the population aged 5+ in England in order to meet the study aims:

1. To measure the national prevalence of infection during a window of time in each month (August, September and October).

2. To measure the change in prevalence between each survey and hence measure the national average R value with high accuracy for each period.

3. To characterise the geographical variation in SARS-CoV-2 infection prevalence and R across regional and local authority levels.

4. To investigate the prevalence of infection by sociodemographic characteristics including age, sex, ethnicity, area deprivation and socio-economic status (SES), and potential changes in those patterns between the baseline and second survey.

5. To compare prevalence of SARS-CoV-2 infection as obtained by nasopharyngeal swab with prevalence from a saliva sample collected on the same participants on the same occasion (REACT-1-2b).

This sample will be selected from the PDS – considered an almost complete sample frame of adults in England, allowing stratification by gender, age and local authority.

REACT-1 aims to provide an accurate estimate of prevalence of infection across England at local authority levels. This sub-study REACT-1 (2), will provide the basis for estimation of the R value in the community at regional and local authority levels. The study is being conducted by Imperial College London (sponsor) on behalf of the Department of Health and Social Care (funder).

Patterns of hospital occupancy, death and 999 calls suggest that the COVID-19 epidemic in England is past its peak overall with a substantial downward trend. The change from increasing to decreasing incidence was almost certainly a result of social distancing and the national lockdown implemented on March 23rd: the net effect of the lockdown was to reduce the reproduction number R from well above 1 (~3 prior to its implementation) to below 1 currently. However, although it is known that R has likely been below 1 for some time, it cannot accurately infer resulting levels of infectious prevalence in the community nor can it assess its geographical variation, based on available data. As we transition out of lockdown, the relative risk of infection in one area compared to another will be closely related to the number of infectious people in that area, with the absolute risk dependent on the degree of social mixing at that time.

Because contact levels will increase, it is expected that there will be an increase in R which could lead to a second wave of infection if R again becomes greater than 1. Based on experience in other countries, it is likely that R will vary over time and by geography. Also, a variety of policies may be implemented as lockdown is relaxed, some of which – such as contact tracing – may be supported locally and therefore vary geographically.

Although overall and local transmissibility of COVID-19 in the UK will continue to be assess using mortality and hospital occupancy data, these data lag changes in transmission by a week or more (substantially more for deaths), and reflect transmission in hospitals and care homes as well as the community. Also, healthcare factors including the possibility of levels of transmission varying from one hospital to another, suggest that hospital data would not be useful for inferring local variation in community transmission.

Testing capacity for the UK has increased rapidly, reaching the target of 100,000 tests per day by the end of April, providing sufficient capacity to meet the testing needs of hospital patients, healthcare workers, other key workers; and potentially providing community testing of symptomatic individuals.

The REACT-1 programme is currently being used to establish the point prevalence of SARS-CoV-2 infection across England at Local Authority level.

The data requested from the Patient Demographic Service (PDS) will be used in order to select a nationally representative sample of the population aged 5+ to take part in the testing. The study needs to provide reliable estimates of infection point prevalence at the level of local authority, as this is the administrative level responsible for local government and will feed into the local public health response. It is also powered to explore differences by key sociodemographic variables (age, gender, ethnicity and socio-economic status (SES)).

This study is one component of a larger programme and sits alongside the REal-time Assessment of Community Transmission 2 (REACT 2): Usability and feasibility study of widespread home self-testing for SARS-CoV-2 antibodies.

The data required is demographic information from the Personal Demographics Service, to allow individuals aged 5+ to be contacted for the purpose of undertaking a survey collecting nasopharyngeal self-swab tests from an age-stratified representative population sample.

The data is requested in two stages

- 1 – a record level identifiable population derived from PDS (including NHS Number, patient postcode, gender and age band) in order to allow selection of a representative sample

- 2 – a record level identifiable data request for selected sample only (including NHS Number, patient address, gender, age (month and year), telephone number and email – where available)

This application is to support three waves of data being supplied to this format and represents rounds 4-6 of the antigen testing study. The surveys will be completed in August, September and October 2020. In each wave order to achieve the required sample size of 150,000 names and demographic details of ~750,000 individuals aged 5 years and above are needed. This is based on current response rates, which have fallen over the first three waves. If this continues, details may be required for larger numbers of individuals in September and October in order to reach the target sample size.

The study protocol also includes facility to undertake additional testing in specific local areas where an increase in transmission and numbers of cases is suspected. These additional surveys would involve sending out swab test kits to up to ca. 10,000 additional people in a specific area and ensuring rapid delivery and pick up, e.g. both delivery and pick-up by courier. Additional data requests may be needed in response to these local requirements.

The identifiable contact information will be securely destroyed by Ipsos MORI once the analysis has been completed. This has been communicated to participants in the privacy notice as December 2020.

There is no other national register of the population that would allow contact with a named and representative sample of adults.

Month and year of birth is sufficient for processing, rather than full date of birth.

Data controllers:

• Department of Health and Social Care (funder)

• Imperial College London (Sponsor)

Data processor:

• Ipsos MORI

Ipsos MORI are working with the following suppliers to deliver the project:

• Print and despatch (Formara and Adare)

• Mail service (Royal Mail)

• Text message reminders (TextLocal)

• Online data collection (Ipsos Germany)

• Courier services (co-ordinated by The Delivery Group)

• Laboratory Services (Eurofins)

These suppliers are approved and compliant with the General Data Protection Regulations.

LEGAL BASIS for DATA PROCESSING

The lawful basis for processing data under GDPR has been reviewed against the guidance provided by IGARD and been assessed as acceptable. The Department of Health and Social Care and Imperial College London process data under the legal basis of Article 6(1)(e) ‘processing is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the controller’ as the Department of Health and Social Care is a public authority and government department, and Imperial College London is a Public Authority as it has a Royal Charter. Both Data Controllers process special category data under Article 9(2)(j) ‘processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes …’ as the data are required for research purposes in the public interest and judged subject to the appropriate safeguards.

The data from NHS Digital will provide the contact details for a representative sample of the population aged 5+ to be invited to take part in the testing.

Data file shared with Imperial College London, for the purposes of further analysis. This file will include all survey responses, some sample information (postcode, gender, age) with additional variables mapped (including IMD, local authority name).

Yielded Benefits is not a requirement for new applications.