INTERVAL, COMPARE and STRIDES Bio Resource trial cohorts: Long-term follow up of health outcomes and associations with genetic, biological and lifestyle traits

**** On 1st February 2023 NHS Digital merged with NHS England. Where practicable, all references to NHS Digital have been changed to NHS England ***

***References to “de-identified”, “de-personalised” and “pseudonymised” data/datasets are used interchangeably in this agreement and mean that such data/datasets have had personal identifiable details removed from them.***

The INTERVAL and COMPARE studies are multi-purpose, multi-stage research projects involving blood donors. These efforts are national flagship research projects supported by leading public funders and research charities, including the UK Medical Research Council (MRC), National Institute for Health Research (NIHR), NHS Blood and Transplant (NHSBT), British Heart Foundation (BHF), and the Wellcome Trust. Collectively, these funders have invested more than £15 million into these studies since 2012. It is worth noting that none of these organisations have access to any record level data and will only access aggregate outputs with small numbers suppressed in line with HES analysis guide. Furthermore, none of these organisations have any decision-making powers regarding the use of the data.

The INTERVAL and COMPARE trials’ linkage of data from NHS England (this agreement) is funded by the NIHR Blood and Transplant Research Unit in Donor Health and Genomics (now called NIHR Blood and Transplant Research Unit in Donor Health and Behaviour) under reference number NIHR BTRU-2014-10024. The trials’ coordinating centre at the Department of Public Health and Primary Care at the University of Cambridge, Cambridge, UK, has received core support from the UK Medical Research Council (G0800270), British Heart Foundation (SP/09/002), and the NIHR Cambridge Biomedical Research Centre. DNA sequencing of trial participants was carried out by the Wellcome Sanger Institute using core funding from the Wellcome Trust.

The INTERVAL and COMPARE studies were set up and are managed by the University of Cambridge, who are the data controller for this agreement. The University of Oxford and NHS Blood and Transplant (NHSBT) provide advice on analyses performed by the University of Cambridge across the studies, but do not have access to the data disseminated by NHS England and will only receive aggregate outputs with small numbers suppressed. Originally, the University of Cambridge designed the methodology for processing data within the studies and will decide on any future methods in carrying out study practises, including the processing of NHS England data. No funders mentioned in this agreement make any decisions determining the purposes and means of the processing and are therefore not considered Data Controllers.

These studies have been specifically designed from their inception to have a multi-purpose strategy to be delivered in multiple stages. This strategy has led to generation and uptake of new evidence-based policies by NHSBT as well as rapid completion of major-multiple-purpose studies.

The initial stage of these studies had been related to blood donation research aiming to improve NHSBT’s core services (e.g. safety and efficiency of blood donation):

- The INTERVAL study recruited ~45,000 blood donors in a randomised controlled trial aiming to assess the impact of varying the frequency of blood donation on donor health and the blood supply.

- The COMPARE study was designed to evaluate the optimum method to measure haemoglobin levels in ~30,000 potential whole blood donors in advance of each donation.

Across the two projects, the University originally recruited a total of ~75,000 participants. Under previous versions of this agreement, sub-cohorts of the total population have been confirmed with NHS England for the purposes of receiving hospital data via HES and to compile NHS Numbers of participants using list clean exercises via demographic reports.

Participants for both the INTERVAL and COMPARE studies were recruited across England from NHSBT blood donor centres and mobile teams. Prior to donating blood, donors were invited to join the study where informed consent was gained.

Subsequent stages of both the INTERVAL and COMPARE studies are related to the creation of a resource of healthy volunteers to enable study of associations of genetic, biological, lifestyle, and other exposures with health outcomes.

As outlined in the section above, the INTERVAL and COMPARE studies have been specifically designed to have a multi-purpose strategy to be delivered in multiple stages.

The overall objective is to create a multi-dimensional, multi-purpose resource by linking detailed lifestyle and biological information collected on INTERVAL and COMPARE participants with health-related records. The establishment of such a comprehensive resource of healthy volunteers will enable detailed study of the health of blood donors and, more generally, allow studies of cardiovascular disease and other health-related outcomes. The datasets requested will be used to update the records held about participants and to identify health events to further characterise the study participants. For example, Mortality data will be used to update records held about participants. HES and Demographics data will be used to identify conditions and health events using both existing phenotyping algorithms (such as those developed by the CALIBER initiative or UK Biobank) and study specific code lists. This information from the health records will be combined with genetic and biological characteristics of the participants to address a range of research questions.

For example, this resource will allow researchers at the University of Cambridge to:

- Identify genetic and lifestyle determinants of iron homeostasis and its potential health-related consequences in blood donors, which may in the future be used to predict donation outcomes and personalise the national blood service. The study will initially focus on iron-related biomarkers that have been measured in INTERVAL and COMPARE. Researchers will characterise associations of iron-related biomarkers measured in INTERVAL and COMPARE, quantifying relationships with several health outcomes, Researchers will also define the shapes of dose-response relationships with health outcomes, and explore whether associations with outcomes vary in key subgroups (e.g., by age and sex).

- Study genetic and lifestyle determinants of several metabolic traits and their potential relevance to cardiovascular diseases: participants in INTERVAL and COMPARE have already provided information about their lifestyle and several metabolic traits have been measured, including >1100 metabolites and >1200 lipids. Researchers will study in detail associations of several metabolic traits with incident cardiovascular diseases and other health outcomes. These analyses will: (i) quantify relationships between metabolic traits and health outcomes by correcting associations for potential confounders, and (ii) define shapes of dose-response relationships, explore subgroup effects and heterogeneity, and suggest lifestyle and biological determinants of novel metabolic factors.

- Study of genetic determinants of several soluble plasma proteins and their potential relevance to chronic diseases, such as cardiovascular diseases: analysis of data on >4000 plasma proteins that have been measured in INTERVAL and COMPARE will: (i) quantify relationships between plasma proteins and several health outcomes, and (ii) define shapes of dose-response relationships, explore subgroup effects and heterogeneity, and suggest lifestyle and biological determinants of circulating proteins.

**********************************************************************

REQUEST UNDER CURRENT AMENDMENT (VERSION 9)

**********************************************************************

The University of Cambridge holds a rich set of genetic, biomarker and demographic information and health data for the INTERVAL and COMPARE participants. The University of Cambridge has obtained the following data from NHS England dating back to 2002 (10 years prior to recruitment where datasets allow) and is receiving ongoing quarterly drops of data from NHS England:

- Hospital Episode Statistics (Admitted Patient Care and Outpatients)

- Cancer Registrations

- Deaths Registrations

- COVID-19 UK Non-hospital Antibody Testing Results (Pillar 3) - V9 of this agreement: data provided to July 2022. NHS England discontinued dataset after this date.

- COVID-19 Vaccination Status

- COVID-19 Vaccination Adverse Reactions

- the Sentinel Stroke National Audit Programme (SSNAP) - V9 of this agreement: Quarterly drops cease as no longer required.

- GPES Data for Pandemic Planning and Research (GDPPR)

PLUS National Diabetes Audit on an annual basis and Demographics on a monthly basis

Additionally, Intensive care data was from the Intensive Care National Audit & Research Centre, but this has since ceased and the data destroyed. The University of Cambridge has also obtained the SARS-CoV-2 test results from the UK Health Security Agency (formerly from the organisation known as Public Health England).

The University of Cambridge are requesting under the current amendment of this agreement (Version 9):

1. Addition of the STRIDES BioResource study participants to the agreement (approximately 83,000 further participants) .

2. Continued Quarterly drop of data products to cover the period most recent release to 26/02/2026 for the WHOLE cohort of 156,000 study participants.

3. Addition of annual releases of National Diabetes Core Audit dataset, plus historic data back to June 2002 for the WHOLE cohort of 156,000 study participants.

4. One drop of historic data from all the datasets previously requested, June 2002 to latest only for the additional 83,000 STRIDES participants.

5. Addition of Onward Data Sharing of the Hospital Episode Statistics, Cancer registrations and Death registrations datasets to external researchers (Sub-licencing) to agreement.

> STRIDES BioResource Study

STRategies to Improve Donor ExperienceS (STRIDES) is a study aiming to improve donor experiences within National Health Service Blood and Transplant (NHSBT) by collecting blood samples, questionnaire data and health records (e.g. measured test results, Hospital Episode Statistics etc.) from whole blood donors for research purposes and establishing a BioResource of healthy volunteers who are willing to be contacted and asked if they wish to participate in medical and health-related studies, which may or may not be related with blood donors’ health.

> National Diabetes Core Audit dataset

The National Diabetes Audit data will allow the researchers to cross-check existing information that has been collected on the study participants already, such as their self-reported diabetes status, their use of glucose-lowering medication and their levels of glycaemic markers such as glucose and HbA1c. Secondly, it will allow the researchers to refine definitions of who had diabetes at baseline (i.e. recruitment into the study). Thirdly, it will allow the researchers to identify those participants who were diagnosed with diabetes after baseline. While this information is currently captured by linkage to other routine datasets (e.g. Hospital Episode Statistics), the study team anticipate that the Diabetes Audit information will identify a substantially larger number of patients.

These uses will allow the researchers to conduct several types of study among the cohorts, including:

- Identifying common and rare genetic risk factors for diabetes;

- Testing polygenic risk scores for diabetes derived in other studies in the study's own cohorts;

- Associate dense multi-omic data layers (e.g. proteins, metabolites, lipoproteins, blood cell traits) with risk of developing diabetes;

- Study of multi-morbidity and co-morbidity, to better understand the consequences of having diabetes on other diseases.

> Blood Donors Studies (BDS) BioResource research database

The Blood Donors Studies (BDS) BioResource research database will include de-personalised genetic, biological and lifestyle information linked to health outcomes for up to 156,000 blood donors recruited into the INTERVAL, COMPARE, (TRACK-COVID) and STRIDES BioResource studies. These studies are collectively known as the Blood Donors Studies. The research database provides a framework in which to follow-up and further characterise participants in the Blood Donors Studies using data obtained from various electronic health records. This includes Hospital Episodes Statistics(HES) data and datasets that have been obtained for COVID-19 research including the GPES Data for Pandemic Planning and Research (GDPPR). The University of Cambridge has also obtained SARS-CoV-2 testing data from the UK HSA for COVID-19 research, and plans to request additional data from a range of organisations for example, microbiology data from UK HSA and cardiac audits managed by the National Institute for Cardiovascular Outcomes Research (NICOR).

The BDS BioResource aims to provide fair, consistent and transparent access to the database in order to promote health-related research by bona fide researchers in the public interest. Another central feature of the BDS BioResource, therefore, is for the pseudonymised datasets to be systematically accessible to bona fide researchers in the wider scientific community, including those working in universities, charities, government agencies or commercial companies in the UK and abroad. This includes data obtained from electronic health records where agreements permit. The BDS BioResource hopes to support population health and biomedical research by providing access to data from this large and richly characterised cohort of healthy volunteers. This is expected to lead to a better understanding of the determinants of important diseases and therefore contribute to improvements in risk assessment and treatment of these diseases.

The BDS BioResource research database includes a wide range of data including data not available from NHS England. It currently holds the following information for the INTERVAL and COMPARE participants:

● Epidemiology questionnaire and quality of life questionnaire data

● Blood donation records and information on deferrals and adverse events

● Full blood count data including extended parameters

● Molecular assay data including genetic data

● Data from electronic health records:

○ Hospital Episode Statistics obtained from NHS England

○ Cancer and death registrations obtained from NHS England

● Data from electronic health records for COVID-19 research only:

○ GPES Data for Pandemic Planning and Research (GDPPR) obtained from NHS England

○ COVID-19 test results obtained from the UK HSA.

○ COVID-19 UK Non-hospital Antibody Testing Results (Pillar 3) obtained from NHS England - V8 of this agreement: data provided to July 2022. NHS England discontinued dataset after this date.

○ COVID-19 Vaccination Status obtained from NHS England

○ COVID-19 Vaccination Adverse Reactions obtained from NHS England

○ The Sentinel Stroke National Audit Programme (SSNAP) obtained from NHS England - V8 of this agreement: Quarterly drops cease as no longer required.

Similar molecular and electronic health record information will also be obtained for participants in the ongoing STRIDES BioResource study.

In addition, new data generated from molecular assays and from research projects using data from the BioResource may form part of the database. Participants in the INTERVAL, COMPARE and STRIDES BioResource studies also agree to be recontacted with invitations to participate in ethically approved studies. De-personalised data generated by such studies will be integrated into the research database where this provides further information about the participants.

The University of Cambridge will request access to electronic health records (EHR) and administrative data held by a range of organisations. For example, cardiac audits managed by the National Institute for Cardiovascular Outcomes Research (NICOR), to identify and obtain detailed information about cardiovascular disease. Linkage to health records will allow long-term tracking of participants’ health and therefore will enable investigations into the determinants of health outcomes in the general population, e.g. genetic variants and lifestyle. Health records can also provide information on lifestyle factors such as smoking and on important potential confounders such as prescribed drugs. Therefore the participants in the Blood Donors Studies will form a large cohort of healthy volunteers with a richly characterised dataset using information from a range of sources. The value of the data included in the Blood Donors Studies BioResource will increase over time as participants are followed up and health events accrue, making it an important resource to enable broad health-related research. Data from NHS England is one component of this but becomes much more valuable when combined with the other information included in the database.

SUB-LICENCING / ONWARD SHARING OF DATA TO EXTERNAL RESEARCHERS

The Blood Donors Studies (BDS) BioResource research database contains linked data from a number of organisations, but it should be noted that for the purposes of the Onward Sharing of Data (Sublicencing), only the following NHS England datasets will be accessed by researchers beyond the study teams (external to University of Cambridge, including UK, EEA and Worldwide*) applying to use the Blood Donors Studies (BDS) BioResource research database:

- HES Admitted Patient Care and HES Outpatients data

- Cancer Registrations data and

- Civil Registrations Deaths data.

The BDS BioResource aims to provide fair, consistent and transparent access to the database in order to promote health-related research by bona fide researchers in the public interest. Another central feature of the BDS BioResource, therefore, is for the pseudonymised datasets to be systematically accessible to bona fide researchers in the wider scientific community, including those working in universities, charities, government agencies or commercial companies in the UK and abroad. This includes data obtained from electronic health records where agreements permit. External applicants may be asked to pay (on a cost-recovery basis) a variable charge depending on the data that is requested for the research project.

The BDS BioResource will support population health and biomedical research by providing access to data from this large and richly characterised cohort of healthy volunteers. This is expected to lead to a better understanding of the determinants of important diseases and therefore contribute to improvements in risk assessment and treatment. Participants in the Blood Donors Studies gave informed consent for access to their health records and long-term storage and use of their health data for health-related research purposes. The plans for establishing the BDS BioResource as a research database and for access to data from electronic health records were discussed at a workshop in February 2020 with representatives of blood donors, including participants in the studies. The public members supported the research database as a way to maximise use of the participant data for research. Their feedback, for example around communication of the work, has been included in plans. There have been further workshops where data access and sharing have been discussed with public members. The research database has been reviewed by an independent research ethics committee and ethical approval is in place for linkage to health-related datasets and to allow bona fide researchers within the UK and in other countries to access pseudonymised datasets (REC reference number: 20/EE/0115).

The study team currently anticipate 5-10 applications per year for use of the resource. Where data has originally been provided to the BDS BioResource under a data sharing agreement (DSA), no sub-licence will be granted that extends beyond the end of this agreement, with the option to extend if required to complete the research should this agreement term also be extended.

The release of data is regulated by the BDS BioResource Data Access Committee (DAC). This includes senior members from the organisations involved in the studies: the University of Cambridge, NHS Blood and Transplant (NHSBT), the Wellcome Sanger Institute (for requests that include genetic sequencing data) and the University of Oxford, and public members. The DAC has overall responsibility for the data access procedures and decisions, and works according to the guidelines set out in the Data Access Policy.

Researchers requesting access to the data are required to submit a project proposal form which will include details on the scientific rationale of the project and the justification for using the BDS BioResource. The DAC review the proposal form to:

I. assess the applicant/co-applicants and ensure they are bona fide researchers

II. review the specific project aims and justification, and assess whether the data requested is relevant to the project aims

III. determine whether the proposed research use meets the required criteria for access to the data within the database (i.e. it represents high quality research into blood donation or wider health or health-related questions)

IV. ensure that data access requests do not carry a significant potential for participant identification (e.g. through extremely specific data requests)

V. ensure research projects have the relevant scientific approval if required.

The research should aim to benefit healthcare provision, adult social care, or the promotion of health.

For applications that are approved, researchers will be required to sign a Data Transfer Agreement or a Research Collaboration Agreement with the University of Cambridge prior to the transfer of the data sets . The data sets provided are de-personalised and will contain only the subset of participants and data items required for the project. NHS England data will typically be provided to approved proposals linked with other data from the BDS BioResource (e.g. self-report questionnaire data and/or data from molecular assays) since the study team anticipate most proposals will seek to correlate risk factor information with health record outcomes.

The research team at the University of Cambridge will provide only a “bespoke” (ie, customised) and pseudonymised extract of the necessary subset of the study dataset for proposals from bona fide researchers that are approved by the Data Access Committee - only the minimum extract of pseudonymised data needed to allow an approved proposal to achieve its specific scientific goal(s) will be released. For data derived from electronic health records (e.g. NHS England record-level data), the data manager removes any personal identifiable details and replaces them with the unique anonymous study identification number. In addition, the data manager creates derived variables from the data provided by NHS England (e.g. “CALIBER” endpoints which are now commonly used by population health scientists). As a further safeguard, CALIBER endpoints that have a count of <= 5 are “nulled” to prevent the dataset becoming too granular (see also further description in Section 5(b). NHS England's Legal team acknowledge the data to be provided under sub-licence will be derived data.

A database of applications approved to use the BDS BioResource is maintained. This includes the details of the principal applicant and co-applicants, their institution, the dataset requested and the project summary. Summary details of approved applications is published on the Blood and Transplant Research Unit in Donor Health and Behaviour website (www.donorhealth-btru.nihr.ac.uk) for information and transparency. If the datasets include data from NHS England, information about the application will be provided to NHS England for inclusion in their release register within one month of release.

NHS England data released under sub-licencing will be strictly for health-related research purposes, and will not be used for purposes such as marketing, sales or insurance. However, in the interests of full transparency, it is noted here that under the onward sharing of data (sub-licencing) approved researchers in the wider scientific community, including commercial companies (from UK, EEA and Worldwide*), can apply to access data for health-related research. Thus, results of this study may be used in research projects that result in tangible or intangible benefit to commercial companies.

*TERRITORY OF USE

At present, NHS England can only permit sub-licensing to the following countries:

1. The EEA, EU or EEA institutions, bodies, offices or agencies, Andorra, Argentina, Canada (commercial organisations), Faroe Islands, Gibraltar, Guernsey, Isle of Man, Israel, Jersey, New Zealand, Japan (private sector organisations), Switzerland or Uruguay. These territories may be updated by the UK Government from time to time (to add or remove territories from the relevant list). If the researcher is located in a territory covered by an adequacy regulation then the transfer of personal data from the UK to these countries is permitted, pending the University of Cambridge's sub-licence application assessment process.

2. America, Australia, and New Zealand (subject to the University of Cambridge undertaking a DPIA and an Article 46(1) assessment / ICO International Data Transfer Risk Assessment as part of their sub-licence application assessment process).

No other jurisdictions are permitted. Any request to expand this list of countries must be reviewed by the NHS England's Information Law team for approval and an amendment submitted to include the additional country(ies) written in this agreement.

*************************************************

HISTORY OF THE DATA SHARING AGREEMENT WITH NHS ENGLAND (NHS DIGITAL)

This data sharing agreement has undergone several amendments over a number of years to ensure the University of Cambridge receives the health data it requires for the study.

The University of Cambridge requested Medical Research reports data on a monthly basis under the original agreement prior to 2016. Additionally, Hospital Episode Statistics (HES) were requested in later agreements on a bi-annual basis, plus historic HES data going back to 2002 (10 years before the recruitment of participants) in later versions.

By Version 5, the data dissemination frequency increased to monthly from bi-annually due to the urgent need for a better understanding of the risk factors for COVID-19, clinical trajectories after infection with SARS-CoV-2 and the associated health outcomes.

For projects specifically relating to urgent COVID-19 work, access to data provided under the agreement was also permitted for approved visiting academics to the University of Cambridge where the academic's substantive employer had provided assurance that they would consider disciplinary action in the event of a data breach. All visiting academics to the University of Cambridge had a formal visitor agreement that they and their employing institution signed.

University of Cambridge have linked the INTERVAL and COMPARE studies to SARS-CoV-2 testing data from UK HSA (previously Public Health England). Together with the existing genomic, molecular and other data held about participants, this allowed research to:

1) Clarify clinical risk factors associated with COVID-19 status and prognosis in order to inform targeted preventative measures (e.g. prioritisation of vaccinations when available) by linkage of e-health records and COVID 19 status/outcome.

2) Establish molecular factors associated with susceptibility to, and clinical trajectory of, COVID-19.

3) Understand resilience to (and recovery from) severe clinical consequences of SARS-CoV-2 infections.

Version 6 of this agreement saw the University of Cambridge requesting the GDPPR (GPES Data for Pandemic Planning and Research) dataset, with the aim of understanding the risk factors associated with COVID-19 status and prognosis, in order to:

> inform targeted preventative measures (e.g. prioritisation of vaccinations when available);

> establish molecular factors associated with susceptibility to COVID-19 and

> clinical trajectory of the disease, and; understand resilience to (and recovery from) severe clinical consequences of SARS-CoV-2 infections.

Linkage to GP records for INTERVAL and COMPARE participants provided important additional information about pre-existing conditions and co-morbidities, which might have affected susceptibility and resilience to COVID-19, but could not be obtained from other patient records. The University of Cambridge asked for information about a broad range of risk factors, conditions and medications from the dataset, as it was not yet known which were important and relevant to COVID-19. Having access to this range of information allowed rapid investigation of new hypotheses and replication of findings from other cohorts such as UK Biobank.

In version 7 of this agreement changed the dissemination frequency of data to Quarterly rather than monthly, and requested a further suite of COVID-19 tactical products. Researchers at the University of Cambridge are undertaking analyses to identify the risk factors that are associated with SARS-CoV-2 infection and prognosis, and are working as part of national and international COVID-19 consortia to increase biological understanding of the virus and disease.

An example of ongoing research is the TRACK-COVID study; an epidemiological investigation of SARS-CoV-2 virus infection in the population, which is recruiting participants from the INTERVAL, COMPARE and STRIDES BioResource cohorts. This study provides data specific to COVID-19 for these participants, through questionnaires on symptoms and by assaying of SARS-CoV-2 antibodies. Data obtained from health records enables researchers to understand the links between genetic, biological and lifestyle information with disease. The aim of the research is to determine risk factors for infection of the SARS-CoV-2 virus and investigate why some people who carry the virus are symptomatic while others are asymptomatic.

Ongoing linkage to the COVID-19 datasets from NHS England will allow the study team to evaluate the long-term impact of COVID19 infection, vaccination and antibody levels in participants that have been hospitalised due to COVID19 infection or chronic diseases. Given that the main focus of the analysis will be on severe diseases and hospitalisation, the recent policy changes to PCR testing will not affect the study team's analysis and methodology.

Version 8 of this agreement was to implement a short-term extension of 3 months to ensure the Data Sharing Agreement remained active and was consistent with the terms of the Data Sharing Framework Contract, whilst the current (V9) amendment and renewal was negotiated.

DATA MINIMISATION

Data obtained from health records enables researchers to understand the links between genetic, biological and lifestyle information with disease. Researchers at the University of Cambridge are undertaking analyses to identify the risk factors that are associated with SARS-CoV-2 infection and prognosis, and are working as part of national and international COVID-19 consortia to increase biological understanding of the virus and disease.

The University of Cambridge obtains Antibody Testing Results (COVID-19 UK Non-hospital Antibody Testing Results (Pillar 3)) to identify previous exposure to COVID-19. This supplemented the information collected through the TRACK-COVID study (by providing results for participants that are not enrolled into TRACK-COVID) and helped to understand the risk factors for infection with the SARS-CoV-2 virus, defined by the serological test results. These data also enabled investigation of why some people who carry the virus are symptomatic while others are asymptomatic.

Additionally, the University of Cambridge obtains the vaccination datasets (COVID-19 Vaccination Status and COVID-19 Vaccination Adverse Reactions). These will help establish if there are molecular factors or co-morbidities that are associated with SARS-CoV-2 infection or adverse reactions following vaccination, through analysis of this large cohort of demographically and geographically diverse group of participants.

The University of Cambridge also obtained the Sentinel Stroke National Audit Programme (SSNAP) dataset that, in combination with the datasets listed above, enables ana

Additional outputs for urgent COVID-19 work: University of Cambridge will rapidly disseminate ensuing evidence on COVID-19 risk factors to key stakeholders (e.g., policy-makers, healthcare organisations and the scientific community) through preprints and other rapid forms of communication, as well as through traditional publication routes.

The results from the analyses with the SARS-CoV-2 testing data from Public Health England and information on pre-existing conditions and new diagnoses based on Hospital Episode Statistics are already contributing to the international COVID-19 Host Genetics Initiative (https://www.covid19hg.org/), and therefore increasing global knowledge of the biology of SARS-CoV-2 infection and disease.

Research findings will also be disseminated to the public through various routes including the INTERVAL and COMPARE study websites, newsletters and through the active public engagement programmes.

The initial focus of the INTERVAL, COMPARE and STRIDES BioResource has been working to improve NHS Blood and Transplant’s (NHSBT) core services (e.g. safety and efficiency of blood donation) by providing evidence for policy change. Since the initial application, University of Cambridge have published results from a further two years follow up of INTERVAL participants (Kaptoge, Lancet Haematol 2019). This showed that more intensive reminders increased whole blood donation rates and allowed evaluation of the longer-term risks and benefits of varying inter-donation intervals. Outputs are shared with participants through regular updates in the form of web publications, email communications and newsletters. Results are also disseminated to the public and donors through study website, newsletters, national and international press releases and media interviews. The realised and expected benefits of these studies to blood donors and the wider patient population are described below.

The renewal of this agreement will allow for continued follow up of study participants to identify new health events. Findings will be published and disseminated through publications in high impact journals and through dissemination to academic, health service, and general public audiences. Publications arising from these studies will be available on the studies website at http://www.intervalstudy.org.uk/publications/ for INTERVAL and

http://www.comparestudy.org.uk/publications/ for COMPARE and

https://www.strides-study.org.uk/ for STRIDES BioResource.

The STRIDES study aims to demonstrate effectiveness of interventions or a combination of interventions to prevent vasovagal reactions (VVRs) in whole blood donors, a common complication related to blood donation. The study results hope to help shape robust policies for NHSBT and other blood services that will result in improvement of donor health and experience, enhancement of service efficiency, and reduction in medicolegal liability. The STRIDES BioResource study hopes to enable detailed study of the health of a subset of blood donors and provide additional data to address the main aim of the STRIDES study. Recruitment into the STRIDES BioResource study has now completed, with study results to be published when the analysis of the study data has been completed.

Study results are disseminated to policy makers, academic and health service audiences through publications in high impact journals, webinars and talks.

Outputs are shared with participants through regular updates in the form of web publications, email communications and newsletters. Results are also disseminated to the public and donors through the study websites, newsletters, social media, national and international press releases and media interviews. Publications arising from these studies aim to be available on the study websites at http://www.intervalstudy.org.uk/publications/ for INTERVAL,

http://www.comparestudy.org.uk/publications/ for COMPARE, and

https://www.strides-study.org.uk/publications/ for STRIDES BioResource.

Members of the public are also informed about results, for example at public engagement events such as the Cambridge Festival.

The University of Cambridge will continue to work closely with the Patient and Public Involvement and Engagement panel managed by the NIHR Blood and Transplant Research Unit in Donor Health and Genomics, who provide both advice on plans and on the dissemination of results and the unit has an active public engagement programme (see http://www.donorhealth-btru.nihr.ac.uk/btru_events/).

COVID-19 RESEARCH

The University of Cambridge aims to rapidly disseminate ensuing evidence on COVID-19 risk factors to key stakeholders (e.g., policy-makers, healthcare organisations and the scientific community) through preprints and other rapid forms of communication, as well as through traditional publication routes. The results from the analyses with the SARS-CoV-2 testing data from UKHSA (previously Public Health England) and information on pre-existing conditions and new diagnoses based on Hospital Episode Statistics are contributing to the international COVID-19 Host Genetics Initiative (https://www.covid19hg.org/), and therefore increasing global knowledge of the biology of SARS-CoV-2 infection and disease. The results from the TRACK-COVID study (an epidemiological investigation of SARS-CoV-2 virus infection in the population) are expected to be published at the end of 2022.

THE BLOOD DONORS STUDIES BIORESOURCE RESEARCH DATABASE

Subsequent stages of the INTERVAL, COMPARE and STRIDES BioResource studies are related to the creation of a comprehensive resource that will enable detailed studies of health-related questions by linking health outcomes data to genetic, biological and lifestyle information. For example, researchers from the University of Cambridge have published a novel analysis of the human plasma proteome (Sun, Nature 2018) combining genomic and proteomic measurements from the INTERVAL study to identify potential therapeutic targets, opportunities for matching existing drugs with new disease indications, and potential safety concerns for drugs under development.

Use of the linked data from NHS England is at an earlier stage but, as examples of the types of output to be expected, manuscripts describing the shared underpinnings of five distinct cardiometabolic conditions (coronary disease, atrial fibrillation, stroke, type 2 diabetes, chronic kidney disease) are at advanced stages. These use novel methods that combine polygenic risk scores, molecular ‘omics, and disease databases and are expected to result in multiple high impact scientific publications. As such, it is expected that findings from this resource will extensively advance biomedical research and inform public health policy. Given that health outcomes will accrue over time and that the University of Cambridge intends to track participants’ health over many years, the University of Cambridge anticipates that the outputs of this research will be realised for a considerable time into the future. The renewal of this agreement aims to allow the University of Cambridge to continue and extend this work. Onward sharing increases the number of users of the linked NHS England data and therefore the outputs generated.

Applicants to the Blood Donors Studies BioResource are required to use their best endeavours to publish the findings of any research deriving from the data in the BioResource in an academic journal or on an open source publication site within the amount of time stated in the agreement. Summaries and/or links to publications that are derived from use of the data aim to be be provided on the study websites and the NIHR Blood and Transplant Research Unit (BTRU) in Donor Health and Behaviour website. Research findings hope to be highlighted in study newsletters, which are sent to participants of the blood donor studies.

When sharing research findings, results will be displayed as aggregate data only (with small numbers suppressed, in line with the HES analysis guide), therefore individual data cannot be recognised.

Results from the INTERVAL trial published in The Lancet (Di Angelantonio et al, 2017), are already shaping policy nationally and internationally, and leading to the improvements in the health of donors. In particular, results from INTERVAL have shown that more frequent blood donations from donors are possible without causing harm to donor health. It has provided policy-makers with evidence that more frequent collection from donors than is now standard can be done over two years without causing harm to donor health, allowing better management of the supply to the NHS of units of blood with in-demand blood groups. Furthermore, INTERVAL has led to NHSBT’s adoption of comprehensive multi-modal reminders (e.g. SMS messages) to help donors make and keep appointments and therefore will help to meet the demand for blood required by patients in the NHS.

Since the initial agreement, researchers from the University of Cambridge have published results from a further two years’ follow up of INTERVAL participants (Kaptoge, Lancet Haematol 2019). The findings have provided policy-makers with two key evidence-based options to meet blood supply needs; the use of frequent reminders to help donors keep appointments and shorter inter-donation intervals than are now standard. This study has also quantified the extent of iron depletion within four years of repeated donation, thus informing safety guidelines. Results have been disseminated to the public and donors through the study website, newsletters, national and international press release and media interviews.

The COMPARE study was designed to evaluate the optimum method to measure haemoglobin levels in ~30,000 potential whole blood donors in advance of each donation. Results from COMPARE have led to an evidence-based decision by NHSBT to replace the copper sulphate-based haemoglobin screening with finger-prick haemoglobin testing. It is estimated that this will prevent about 30,000 donors annually from experiencing anaemia and potential iron deficiency, due to being inappropriately bled at a blood donation session (Bell, Sweeting, Transfusion Medicine 2020) and will therefore help to protect the health of blood donors.

COVID-19 RESEARCH

As described above, the results from the analyses with the SARS-CoV-2 testing data from Public Health England and information on pre-existing conditions and new diagnoses based on Hospital Episode Statistics are already contributing to the international COVID-19 Host Genetics Initiative (https://www.covid19hg.org/), and this is increasing global knowledge of the biology of SARS-CoV-2 infection and disease. Work by this consortium has identified areas in the genome that are associated with SARS-CoV-2 infection or severe COVID-19 disease and this has the potential to inform the development of therapeutics for patients (The COVID-19 Host Genetics Initiative, Andrea Ganna, medRxiv [Preprint], 2021).

Additionally, University of Cambridge researchers and collaborators have identified a gene (IFNAR2) that is more likely to play a role in COVID-19 hospitalisation, with findings from this work prioritising trials of drugs targeting specific proteins (IFNAR2 and ACE2) for early management of COVID-19 (Gaziano, Nature Medicine 2021).

*****

This version of the agreement (Version 8) is to implement a short-term extension of 3 months to ensure this Data Sharing Agreement remains active and is consistent with the terms of the Data Sharing Framework Contract, whilst a further amendment and renewal is negotiated.

*****

A sub study of the INTERVAL AND COMPARE cohorts is currently underway. The Track-COVID study is an epidemiological investigation of COVID-19 virus infection in the population, which is recruiting participants from the INTERVAL and COMPARE cohorts. This study provides data specific to COVID-19 for these participants, through questionnaires on symptoms and by assaying of SARS-CoV-2 antibodies.

Data obtained from health records enables research to understand the links between genetic, biological and lifestyle information with disease. Researchers at the University of Cambridge are undertaking analyses to identify the risk factors that are associated with SARS-CoV-2 infection and prognosis, and are working as part of national and international COVID-19 consortia to increase biological understanding of the virus and disease.

The University of Cambridge are requesting Antibody Testing Results to identify previous exposure to COVID-19. This will supplement the information collected through the Track-COVID study (by providing results for participants that are not enrolled into Track-COVID) and will help to understand the risk factors for infection with the SARS-CoV-2 virus, defined by the serological test results. These data will also enable investigation of why some people who carry the virus are symptomatic while others are asymptomatic.

Additionally The University of Cambridge are requesting the vaccination datasets. These will help establish if there are molecular factors or co-morbidities that are associated with SARS-CoV-2 infection or adverse reactions following vaccination, through analysis of this large cohort of demographically and geographically diverse group of participants.

The University of Cambridge are also requesting the SSNAP dataset that, in combination with the datasets listed above, will enable analyses to be performed to help establish if there is an association between SARS-CoV-2 infection or severe COVID-19 symptoms, and stroke prevalence and severity.

The data has been minimised by selecting the specific variables that are required for analysis from each dataset.

The previous amendment to this agreement (version 6) was to request the GPES data for Pandemic Planning and Research (GDPPR) for participants in the INTERVAL and COMPARE studies.

The INTERVAL and COMPARE studies are large, richly characterised cohorts with approximately 75,000 participants in total. A range of genomic, molecular and lifestyle information is already available for the participants as well as Hospital Episodes Statistics (HES), mortality data and cancer registration data provided by NHS Digital under this Data Sharing Agreement. SARS-CoV-2 test results are provided by Public Health England (PHE) and records of intensive care admissions from the Intensive Care National Audit & Research Centre.

Further information related to COVID-19 is being collected through the TRACK-COVID study, which is recruiting participants from the INTERVAL and COMPARE studies. The aim of this research is to determine risk factors for infection of the SARS-CoV-2 virus and investigate why some people who carry the virus are symptomatic while others are asymptomatic. This study includes data collection from participants relating to symptoms and targeted bio-sampling for viral assays.

The University of Cambridge are requesting the GDPPR (GPES Data for Pandemic Planning and Research) dataset as an aim to understand the risk factors associated with COVID-19 status and prognosis in order to inform targeted preventative measures (e.g. prioritisation of vaccinations when available), establish molecular factors associated with susceptibility to COVID-19 and clinical trajectory of the disease, and understand resilience to (and recovery from) severe clinical consequences of SARS-CoV-2 infections.

Linkage to GP records for INTERVAL and COMPARE participants will provide important additional information about pre-existing conditions and co-morbidities, which may affect susceptibility and resilience to COVID-19, but cannot be obtained from other patient records. The University of Cambridge are asking for information about a broad range of risk factors, conditions and medications from the dataset, as it is not yet known which are important and relevant to COVID-19. Having access to this range of information will allow rapid investigation of new hypotheses and replicate findings from other cohorts such as UK Biobank.

The fields from the GDPPR data set have been selected and minimised in order to

- provide updates on participant medical history relating to COVID-19 diagnosis or risk.

- allow investigation of how pre-existing conditions (e.g. specific underlying conditions such as Diabetes and multi-morbidities) may affect susceptibility and resilience to COVID-19.

- provide information on risk factors (e.g. BMI, smoking status) and recent interactions with primary care that may be relevant to COVID-19 status and severity.

- identify prescriptions, treatments and vaccinations that may be associated with COVID-19 susceptibility and resilience (e.g. immunosuppressive agents).

Data such as Year of Birth and Ethnicity has not been requested as details are already held by the University of Cambridge or are not required for research.

Patient and Public Involvement and Engagement

NHSX’s ‘Data Access Request Form E’ and a lay summary of this data request were circulated to the Health Data Research UK Cambridge Advisory Group, which includes blood donors. The public members were asked if they thought that accessing GP data for participants in the INTERVAL and COMPARE studies to enable research into COVID-19 is acceptable, or if they had any concerns. The public members supported the data access request. They thought that accessing the GP records for these participants was acceptable and important, though they emphasised that it is essential that participant confidentiality is maintained.

---

The last amendment to this agreement (Version 7) was to renew the current data sharing agreement for the INTERVAL and COMPARE cohorts, to amend the drops of data to quarterly rather than monthly, and to request the following extra datasets be added to this agreement:

COVID-19 UK Non-hospital Antibody Testing Results (Pillar 3),

COVID-19 Vaccination Status,

COVID-19 Vaccination Adverse Reactions and

the Sentinel Stroke National Audit Programme (SSNAP).

It has been confirmed that the cohort number has been reduced to approximately 73,000 (but may reduce due to withdrawals)

The University of Cambridge already holds a rich set of data for the INTERVAL and COMPARE participants. This includes genetic and biomarker data, demographic information and the following data from electronic health records (EHR):

- Hospital Episode Statistics obtained from NHS Digital

- Cancer Registrations and death registrations obtained from NHS Digital

- Data from electronic health records for COVID-19 research only:

- GPES Data for Pandemic Planning and Research (GDPPR) obtained from NHS Digital

- Intensive care data obtained from the Intensive Care National Audit & Research Centre

- COVID-19 test results obtained from Public Health England.

A previous iteration of this agreement (Version 5) was amended to support COVID-19 research. Under this version of the agreement - it was approved that NHS Digital would increase the data dissemination frequency from bi-annual to monthly. There is urgent need for a better understanding of the risk factors for COVID-19, clinical trajectories after infection with SARS-CoV-2 and the associated health outcomes. This increased frequency is being continued under v7 of this agreement also.

The University of Cambridge are the named data processor in this agreement. For projects specifically relating to urgent COVID-19 work, access to data provided under this agreement will also be permitted for approved visiting academics to the University of Cambridge where the academic's substantive employer has provided assurance that they would consider disciplinary action in the event of a data breach. All visiting academics to the University of Cambridge have a formal visitor agreement that they and their employing institution sign.

University of Cambridge have linked the INTERVAL and COMPARE studies to SARS-CoV-2 testing data from Public Health England. Together with the existing genomic, molecular and other data held about participants, this will allow research to:

1) Clarify clinical risk factors associated with COVID-19 status and prognosis in order to inform targeted preventative measures (e.g. prioritisation of vaccinations when available) by linkage of e-health records and COVID 19 status/outcome.

2) Establish molecular factors associated with susceptibility to, and clinical trajectory of, COVID-19.

3) Understand resilience to (and recovery from) severe clinical consequences of SARS-CoV-2 infections.

To maximise the value of this research, University of Cambridge need timely information about study participants’ health status, both prior to diagnosis with COVID-19 and, as the outbreak progresses, during and after treatment. Moving from biannual to monthly HES updates will enable this.

BACKGROUND:

INTERVAL and COMPARE studies are multi-purpose, multi-stage research projects involving blood donors. These efforts are national flagship research projects supported by leading public funders and research charities, including the UK Medical Research Council (MRC), National Institute for Health Research (NIHR), NHS Blood and Transplant (NHSBT), British Heart Foundation (BHF), and Wellcome Trust. Collectively, these funders have invested more than £15 million into these studies since 2012. It is worth noting that none of these organisations have access to any record level data and will only access aggregate outputs with small numbers suppressed in line with HES analysis guide. Furthermore, none of these organisations have any decision-making powers regarding the use of the data.

INTERVAL and COMPARE study has received support from the organisations mentioned above in different areas. The INTERVAL and COMPARE trials were funded by NHSBT, the NIHR Blood and Transplant Research Unit in Donor Health and Genomics with the linkage of data from NHS Digital funded by the NIHR Blood and Transplant Research Unit in Donor Health and Genomics (NIHR BTRU-2014-10024). The trials’ coordinating centre at the Department of Public Health and Primary Care at the University of Cambridge, Cambridge, UK, has received core support from the UK Medical Research Council (G0800270), British Heart Foundation (SP/09/002), and the NIHR Cambridge Biomedical Research Centre. DNA sequencing of trial participants was carried out by the Wellcome Sanger Institute using core funding from the Wellcome Trust.

The INTERVAL and COMPARE studies were set up and are managed by the University of Cambridge, who are the data controller for this agreement. The University of Oxford and NHS Blood and Transplant (NHSBT) provide advice on analyses performed by University of Cambridge across the studies, but do not have access to the data disseminated by NHS Digital and will only receive aggregate outputs with small numbers suppressed. Originally, the University of Cambridge designed the methodology for processing data within the studies and will decide on any future methods in carrying out study practises, including the processing of NHS Digital data.

These studies have been specifically designed from their inception to have a multi-purpose strategy to be delivered in multiple stages. This strategy has led to generation and uptake of new evidence-based policies by NHSBT (see below) as well as rapid completion of major-multiple purpose studies.

The initial stage of these studies had been related to blood donation research aiming to improve NHSBT’s core services (e.g. safety and efficiency of blood donation):

- The INTERVAL study recruited ~50,000 blood donors in a randomised controlled trial aiming to assess the impact of varying the frequency of blood donation on donor health and the blood supply.

- The COMPARE study was designed to evaluate the optimum method to measure haemoglobin levels in ~30,000 potential whole blood donors in advance of each donation.

COHORT

Across the two projects, the University originally recruited a total of ~75,000 participants. Under previous versions of this agreement, sub-cohorts of the total population have been confirmed with NHS Digital for the purposes of receiving hospital data via HES and to compile NHS Numbers of participants using list clean exercises via demographic reports.

The total cohort is split across the two projects as below:

INTERVAL: 48448

COMPARE: 29029

Participants for both the INTERVAL and COMPARE studies were recruited across England from blood donor centres and mobile teams. Prior to donating blood, donors were invited to join the study where informed consent was gained.

Subsequent stages of both INTERVAL and COMPARE studies are related to the creation of a resource of healthy volunteers to enable study of associations of genetic, biological, lifestyle, and other exposures with health outcomes.

For the purposes of this agreement, only researchers from the University of Cambridge/visiting academics investigating COVID-19 with the appropriate honorary contracts will access the data products requested in this agreement.

As outlined in the section above, INTERVAL and COMPARE studies have been specifically designed to have a multi-purpose strategy to be delivered in multiple stages .

The overall objective is to create a multi-dimensional, multi-purpose resource by linking detailed lifestyle and biological information collected on INTERVAL and COMPARE participants with health-related records. The establishment of such a comprehensive resource of healthy volunteers will enable detailed study of the health of blood donors and, more generally, allow studies of cardiovascular disease and other health-related outcome at the University of Cambridge (ie, only researchers from the University of Cambridge/ or visiting academics investigating COVID-19 with appropriate honorary contracts and visiting arrangements will access the data requested in this application). The datasets requested will be used to update the records held about participants and to identify health events to further characterise the study participants. Mortality data will be used to update records held about participants. HES and Demographics data will be used to identify conditions and health events using both existing phenotyping algorithms (such as those developed by the CALIBER initiative or UK Biobank) and study specific code lists. This information will be combined with genetic and biological characteristics of the participants to address a range of research questions.

For example, this resource will allow researchers at the University of Cambridge/visiting academics with relevant honorary contracts to:

- Identify genetic and lifestyle determinants of iron homeostasis and its potential health-related consequences in blood donors, which may in the future be used to predict donation outcomes and personalise the national blood service. The study will initially focus on iron-related biomarkers that have been measured in INTERVAL and COMPARE. Researchers will characterise associations of iron-related biomarkers measured in INTERVAL and COMPARE and quantifying relationships with several health outcomes and will define the shapes of dose-response relationships with health outcomes, and explore whether associations with outcomes vary in key subgroups (eg, by age and sex).

- Study genetic and lifestyle determinants of several metabolic traits and their potential relevance to cardiovascular diseases: Participants in INTERVAL and COMPARE have already provided information about their lifestyle and several metabolic traits have been measured, including >1100 metabolites and >1200 lipids. Using developed methods, researchers will study in detail associations of several metabolic traits with incident cardiovascular diseases and other health outcomes. These analyses will: (i) quantify relationships by correcting associations for potential confounders, and (ii) define shapes of dose-response relationships, explore subgroup effects and heterogeneity, and suggest lifestyle and biological determinants of novel metabolic factors.

- Study of genetic determinants of several soluble plasma proteins and their potential relevance to chronic diseases, such as cardiovascular diseases: >4000 plasma proteins have been measured in INTERVAL and COMPARE. These analyses will: (i) quantify relationships with several health outcomes, and (ii) define shapes of dose-response relationships, explore subgroup effects and heterogeneity, and suggest lifestyle and biological determinants of circulating proteins.

The University of Cambridge require data on a monthly basis (Quarterly from Version 7 of this agreement) for the term of the agreement. The University has requested further data within the following datasets:

• HES – Admitted Patient Care

• HES – Outpatients

• Cancer Registration Data

• Civil Registration Mortality Data

• GPES Data for Pandemic Planning and Research.

• COVID-19 UK Non-hospital Antibody Testing Results (Pillar 3)

• COVID-19 Vaccination Status

• COVID-19 Vaccination Adverse Reactions

• The Sentinel Stroke National Audit Programme (SSNAP).

The University has requested identifiable record level data. Historic HES data is included within the data already held by the University going as far back as 2002 (10 years before the recruitment of participants). For HES data, the University of Cambridge has received historical data (going back up to 10 years from the recruitment of the first INTERVAL participant, 2002-2018/19) and is requesting future disseminations every month. Linkage to HES records will be used to enhance information already recorded at baseline in the INTERVAL and COMPARE studies about the donors’ prior medical history. Given the age of participants enrolled in these cohorts it is believed that 10 years will provide a reliable timeframe to capture pre-existing medical conditions.

The data controller for the INTERVAL and COMPARE studies is the University of Cambridge, which processes the data as part of its public task under GDPR Article 6(1)(e). Processing special category personal data is necessary for scientific research purposes under GDPR Article 9(2)(j). Appropriate safeguards to the processing of data are in place, as are appropriate technical and organisational measures, described elsewhere in the application. The public interest nature of the processing is assessed as part of the research ethics committee review of the studies and through peer review from the public bodies funding the research.

Additional outputs for urgent COVID-19 work: University of Cambridge will rapidly disseminate ensuing evidence on COVID-19 risk factors to key stakeholders (e.g., policy-makers, healthcare organisations and the scientific community) through preprints and other rapid forms of communication, as well as through traditional publication routes.

The results from the analyses with the SARS-CoV-2 testing data from Public Health England and information on pre-existing conditions and new diagnoses based on Hospital Episode Statistics are already contributing to the international COVID-19 Host Genetics Initiative (https://www.covid19hg.org/), and therefore increasing global knowledge of the biology of SARS-CoV-2 infection and disease.

Research findings will also be disseminated to the public through various routes including the INTERVAL and COMPARE study websites, newsletters and through the active public engagement programmes.

---

The initial focus for INTERVAL and COMPARE has been working to improve NHSBT’s core services (e.g. safety and efficiency of blood donation). Since our initial application, University of Cambridge have published results from a further two years follow up of INTERVAL participants (Kaptoge, Lancet Haematol 2019). This showed that more intensive reminders increased whole blood donation rates and allowed evaluation of the longer-term risks and benefits of varying inter-donation intervals. Outputs are shared with participants through regular updates in the form of web publications, email communications and newsletters. Results are also disseminated to the public and donors through study website, newsletters, national and international press releases and media interviews. The realised and expected benefits of these studies to blood donors and the wider patient population are described below.

The renewal of this agreement will allow for continued follow up of study participants to identify new health events. Findings will be published and disseminated through publications in high impact journals and through dissemination to academic, health service, and general public audiences. Publications arising from these studies will be available on the studies website at http://www.intervalstudy.org.uk/publications/ for INTERVAL and http://www.comparestudy.org.uk/publications/ for COMPARE.

Subsequent stages of both the INTERVAL and COMPARE studies are related to the creation of a comprehensive resource that will enable detailed studies of health-related questions by linking health outcomes data to genetic, biological and lifestyle information. For example, University of Cambridge have published a novel analysis of the human plasma proteome (Sun, Nature 2018) combining genomic and proteomic measurements from the INTERVAL study to identify potential therapeutic targets, opportunities for matching existing drugs with new disease indications, and potential safety concerns for drugs under development.

Use of the linked data from NHS Digital is at an earlier stage but, as examples of the types of output to be expected, manuscripts describing the shared underpinnings of five distinct cardiometabolic conditions (coronary disease, atrial fibrillation, stroke, type 2 diabetes, chronic kidney disease) are at advanced stages. These use novel methods that combine polygenic risk scores, molecular ‘omics, and disease databases and expect this to result in multiple high impact scientific publications. As such, it is expected that findings from this resource will extensively advance biomedical research and inform public health policy. Given that health outcomes will accrue over time and University of Cambridge intend to track participants’ health over many years University of Cambridge anticipates that the outputs of this research will be realised for a considerable time into the future. The renewal of this agreement will allow the university to continue and extend this work.

Datasets that have been collected under a specific COVID-19 related direction (e.g. GPES Data for Pandemic Planning and Research - GDPPR) can only be used for COVID-19 research related purposes. In the absence of a continuing legal basis for NHS Digital to provide this data - the long term follow up of the cohort will not apply to these datasets.

As described above, researchers will communicate results through both traditional academic routes (papers, seminars, etc.) but also through the study websites, newsletters, press releases and media interviews. University of Cambridge will continue to work closely with the Patient and Public Involvement and Engagement panel managed by the NIHR Blood and Transplant Research Unit in Donor Health and Genomics, who provide both advice on plans and on the dissemination of results and the unit has an active public engagement programme (see http://www.donorhealth-btru.nihr.ac.uk/btru_events/).

When sharing research findings, results will be displayed as aggregate data only (with small numbers suppressed, in line with the HES analysis guide), therefore individual data cannot be recognised.

Results from the INTERVAL trial published in The Lancet (Di Angelantonio et al, 2017), are already shaping policy nationally and internationally, and leading to the improvements in the health of donors. In particular, results from INTERVAL have shown that more frequent blood donations from donors are possible without causing harm to donor health. It has provided policy-makers with evidence that more frequent collection from donors than is now standard can be done over two years without causing harm to donor health, allowing better management of the supply to the NHS of units of blood with in-demand blood groups. Furthermore, INTERVAL has led to NHSBT’s adoption of comprehensive multi-modal reminders (e.g. SMS messages) to help donors make and keep appointments and therefore will help to meet the demand for blood required by patients in the NHS.

Since the initial agreement, researchers from the University of Cambridge have published results from a further two years’ follow up of INTERVAL participants (Kaptoge, Lancet Haematol 2019). The findings have provided policy-makers with two key evidence-based options to meet blood supply needs; the use of frequent reminders to help donors keep appointments and shorter inter-donation intervals than are now standard. This study has also quantified the extent of iron depletion within four years of repeated donation, thus informing safety guidelines. Results have been disseminated to the public and donors through the study website, newsletters, national and international press release and media interviews.

The COMPARE study was designed to evaluate the optimum method to measure haemoglobin levels in ~30,000 potential whole blood donors in advance of each donation. Results from COMPARE have led to an evidence-based decision by NHSBT to replace the copper sulphate-based haemoglobin screening with finger-prick haemoglobin testing. It is estimated that this will prevent about 30,000 donors annually from experiencing anaemia and potential iron deficiency, due to being inappropriately bled at a blood donation session (Bell, Sweeting, Transfusion Medicine 2020) and will therefore help to protect the health of blood donors.

COVID-19 RESEARCH

As described above, the results from the analyses with the SARS-CoV-2 testing data from Public Health England and information on pre-existing conditions and new diagnoses based on Hospital Episode Statistics are already contributing to the international COVID-19 Host Genetics Initiative (https://www.covid19hg.org/), and this is increasing global knowledge of the biology of SARS-CoV-2 infection and disease. Work by this consortium has identified areas in the genome that are associated with SARS-CoV-2 infection or severe COVID-19 disease and this has the potential to inform the development of therapeutics for patients (The COVID-19 Host Genetics Initiative, Andrea Ganna, medRxiv [Preprint], 2021).

Additionally, University of Cambridge researchers and collaborators have identified a gene (IFNAR2 ) that is more likely to play a role in COVID-19 hospitalisation, with findings from this work prioritising trials of drugs targeting specific proteins (IFNAR2 and ACE2) for early management of COVID-19 (Gaziano, Nature Medicine 2021).

****** The current amendment to this agreement (Version 7) is to renew the current data sharing agreement for the INTERVAL and COMPARE cohorts, to amend the drops of data to quarterly rather than monthly, and to request the following extra datasets be added to this agreement:

COVID-19 UK Non-hospital Antibody Testing Results (Pillar 3),

COVID-19 Vaccination Status,

COVID-19 Vaccination Adverse Reactions and

the Sentinel Stroke National Audit Programme (SSNAP).

It has been confirmed that the cohort number has been reduced to approximately 73,000 (but may reduce due to withdrawals)

*************

The University of Cambridge already holds a rich set of data for the INTERVAL and COMPARE participants. This includes genetic and biomarker data, demographic information and the following data from electronic health records (EHR):

- Hospital Episode Statistics obtained from NHS Digital

- Cancer Registrations and death registrations obtained from NHS Digital

- Data from electronic health records for COVID-19 research only:

- GPES Data for Pandemic Planning and Research (GDPPR) obtained from NHS Digital

- Intensive care data obtained from the Intensive Care National Audit & Research Centre

- COVID-19 test results obtained from Public Health England.

A sub study of the INTERVAL AND COMPARE cohorts is currently underway. The Track-COVID study is an epidemiological investigation of COVID-19 virus infection in the population, which is recruiting participants from the INTERVAL and COMPARE cohorts. This study provides data specific to COVID-19 for these participants, through questionnaires on symptoms and by assaying of SARS-CoV-2 antibodies.

Data obtained from health records enables research to understand the links between genetic, biological and lifestyle information with disease. Researchers at the University of Cambridge are undertaking analyses to identify the risk factors that are associated with SARS-CoV-2 infection and prognosis, and are working as part of national and international COVID-19 consortia to increase biological understanding of the virus and disease.

The University of Cambridge are requesting Antibody Testing Results to identify previous exposure to COVID-19. This will supplement the information collected through the Track-COVID study (by providing results for participants that are not enrolled into Track-COVID) and will help to understand the risk factors for infection with the SARS-CoV-2 virus, defined by the serological test results. These data will also enable investigation of why some people who carry the virus are symptomatic while others are asymptomatic.

Additionally The University of Cambridge are requesting the vaccination datasets. These will help establish if there are molecular factors or co-morbidities that are associated with SARS-CoV-2 infection or adverse reactions following vaccination, through analysis of this large cohort of demographically and geographically diverse group of participants.

The University of Cambridge are also requesting the SSNAP dataset that, in combination with the datasets listed above, will enable analyses to be performed to help establish if there is an association between SARS-CoV-2 infection or severe COVID-19 symptoms, and stroke prevalence and severity.

The data has been minimised by selecting the specific variables that are required for analysis from each dataset.

The previous amendment to this agreement (version 6) was to request the GPES data for Pandemic Planning and Research (GDPPR) for participants in the INTERVAL and COMPARE studies.

The INTERVAL and COMPARE studies are large, richly characterised cohorts with approximately 75,000 participants in total. A range of genomic, molecular and lifestyle information is already available for the participants as well as Hospital Episodes Statistics (HES), mortality data and cancer registration data provided by NHS Digital under this Data Sharing Agreement. SARS-CoV-2 test results are provided by Public Health England (PHE) and records of intensive care admissions from the Intensive Care National Audit & Research Centre.

Further information related to COVID-19 is being collected through the TRACK-COVID study, which is recruiting participants from the INTERVAL and COMPARE studies. The aim of this research is to determine risk factors for infection of the SARS-CoV-2 virus and investigate why some people who carry the virus are symptomatic while others are asymptomatic. This study includes data collection from participants relating to symptoms and targeted bio-sampling for viral assays.

The University of Cambridge are requesting the GDPPR (GPES Data for Pandemic Planning and Research) dataset as an aim to understand the risk factors associated with COVID-19 status and prognosis in order to inform targeted preventative measures (e.g. prioritisation of vaccinations when available), establish molecular factors associated with susceptibility to COVID-19 and clinical trajectory of the disease, and understand resilience to (and recovery from) severe clinical consequences of SARS-CoV-2 infections.

Linkage to GP records for INTERVAL and COMPARE participants will provide important additional information about pre-existing conditions and co-morbidities, which may affect susceptibility and resilience to COVID-19, but cannot be obtained from other patient records. The University of Cambridge are asking for information about a broad range of risk factors, conditions and medications from the dataset, as it is not yet known which are important and relevant to COVID-19. Having access to this range of information will allow rapid investigation of new hypotheses and replicate findings from other cohorts such as UK Biobank.

The fields from the GDPPR data set have been selected and minimised in order to

- provide updates on participant medical history relating to COVID-19 diagnosis or risk.

- allow investigation of how pre-existing conditions (e.g. specific underlying conditions such as Diabetes and multi-morbidities) may affect susceptibility and resilience to COVID-19.

- provide information on risk factors (e.g. BMI, smoking status) and recent interactions with primary care that may be relevant to COVID-19 status and severity.

- identify prescriptions, treatments and vaccinations that may be associated with COVID-19 susceptibility and resilience (e.g. immunosuppressive agents).

Data such as Year of Birth and Ethnicity has not been requested as details are already held by the University of Cambridge or are not required for research.

Patient and Public Involvement and Engagement

NHSX’s ‘Data Access Request Form E’ and a lay summary of this data request were circulated to the Health Data Research UK Cambridge Advisory Group, which includes blood donors. The public members were asked if they thought that accessing GP data for participants in the INTERVAL and COMPARE studies to enable research into COVID-19 is acceptable, or if they had any concerns. The public members supported the data access request. They thought that accessing the GP records for these participants was acceptable and important, though they emphasised that it is essential that participant confidentiality is maintained.

---

A previous iteration of this agreement (v5) was amended to support COVID-19 research. Under this version of the agreement - it was approved that NHS Digital would increase the data dissemination frequency from bi-annual to monthly. There is urgent need for a better understanding of the risk factors for COVID-19, clinical trajectories after infection with SARS-CoV-2 and the associated health outcomes. This increased frequency is being continued under v7 of this agreement also.

The University of Cambridge are the named data processor in this agreement. For projects specifically relating to urgent COVID-19 work, access to data provided under this agreement will also be permitted for approved visiting academics to the University of Cambridge where the academic's substantive employer has provided assurance that they would consider disciplinary action in the event of a data breach. All visiting academics to the University of Cambridge have a formal visitor agreement that they and their employing institution sign.

University of Cambridge have linked the INTERVAL and COMPARE studies to SARS-CoV-2 testing data from Public Health England. Together with the existing genomic, molecular and other data held about participants, this will allow research to:

1) Clarify clinical risk factors associated with COVID-19 status and prognosis in order to inform targeted preventative measures (e.g. prioritisation of vaccinations when available) by linkage of e-health records and COVID 19 status/outcome.

2) Establish molecular factors associated with susceptibility to, and clinical trajectory of, COVID-19.

3) Understand resilience to (and recovery from) severe clinical consequences of SARS-CoV-2 infections.

To maximise the value of this research, University of Cambridge need timely information about study participants’ health status, both prior to diagnosis with COVID-19 and, as the outbreak progresses, during and after treatment. Moving from biannual to monthly HES updates will enable this.

BACKGROUND:

INTERVAL and COMPARE studies are multi-purpose, multi-stage research projects involving blood donors. These efforts are national flagship research projects supported by leading public funders and research charities, including the UK Medical Research Council (MRC), National Institute for Health Research (NIHR), NHS Blood and Transplant (NHSBT), British Heart Foundation (BHF), and Wellcome Trust. Collectively, these funders have invested more than £15 million into these studies since 2012. It is worth noting that none of these organisations have access to any record level data and will only access aggregate outputs with small numbers suppressed in line with HES analysis guide. Furthermore, none of these organisations have any decision-making powers regarding the use of the data.

INTERVAL and COMPARE study has received support from the organisations mentioned above in different areas. The INTERVAL and COMPARE trials were funded by NHSBT, the NIHR Blood and Transplant Research Unit in Donor Health and Genomics with the linkage of data from NHS Digital funded by the NIHR Blood and Transplant Research Unit in Donor Health and Genomics (NIHR BTRU-2014-10024). The trials’ coordinating centre at the Department of Public Health and Primary Care at the University of Cambridge, Cambridge, UK, has received core support from the UK Medical Research Council (G0800270), British Heart Foundation (SP/09/002), and the NIHR Cambridge Biomedical Research Centre. DNA sequencing of trial participants was carried out by the Wellcome Sanger Institute using core funding from the Wellcome Trust.

The INTERVAL and COMPARE studies were set up and are managed by the University of Cambridge, who are the data controller for this agreement. The University of Oxford and NHS Blood and Transplant (NHSBT) provide advice on analyses performed by University of Cambridge across the studies, but do not have access to the data disseminated by NHS Digital and will only receive aggregate outputs with small numbers suppressed. Originally, the University of Cambridge designed the methodology for processing data within the studies and will decide on any future methods in carrying out study practises, including the processing of NHS Digital data.

These studies have been specifically designed from their inception to have a multi-purpose strategy to be delivered in multiple stages. This strategy has led to generation and uptake of new evidence-based policies by NHSBT (see below) as well as rapid completion of major-multiple purpose studies.

The initial stage of these studies had been related to blood donation research aiming to improve NHSBT’s core services (e.g. safety and efficiency of blood donation):

- The INTERVAL study recruited ~50,000 blood donors in a randomised controlled trial aiming to assess the impact of varying the frequency of blood donation on donor health and the blood supply.

- The COMPARE study was designed to evaluate the optimum method to measure haemoglobin levels in ~30,000 potential whole blood donors in advance of each donation.

Cohort:

Across the two projects, the University originally recruited a total of ~75,000 participants. Under previous versions of this agreement, sub-cohorts of the total population have been confirmed with NHS Digital for the purposes of receiving hospital data via HES and to compile NHS Numbers of participants using list clean exercises via demographic reports.

The total cohort is split across the two projects as below:

INTERVAL: 48448

COMPARE: 29029

Participants for both the INTERVAL and COMPARE studies were recruited across England from blood donor centres and mobile teams. Prior to donating blood, donors were invited to join the study where informed consent was gained.

Subsequent stages of both INTERVAL and COMPARE studies are related to the creation of a resource of healthy volunteers to enable study of associations of genetic, biological, lifestyle, and other exposures with health outcomes.

For the purposes of this agreement, only researchers from the University of Cambridge/visiting academics investigating COVID-19 with the appropriate honorary contracts will access the data products requested in this agreement.

As outlined in the section above, INTERVAL and COMPARE studies have been specifically designed to have a multi-purpose strategy to be delivered in multiple stages .

The overall objective is to create a multi-dimensional, multi-purpose resource by linking detailed lifestyle and biological information collected on INTERVAL and COMPARE participants with health-related records. The establishment of such a comprehensive resource of healthy volunteers will enable detailed study of the health of blood donors and, more generally, allow studies of cardiovascular disease and other health-related outcome at the University of Cambridge (ie, only researchers from the University of Cambridge/ or visiting academics investigating COVID-19 with appropriate honorary contracts and visiting arrangements will access the data requested in this application). The datasets requested will be used to update the records held about participants and to identify health events to further characterise the study participants. Mortality data will be used to update records held about participants. HES and Demographics data will be used to identify conditions and health events using both existing phenotyping algorithms (such as those developed by the CALIBER initiative or UK Biobank) and study specific code lists. This information will be combined with genetic and biological characteristics of the participants to address a range of research questions.

For example, this resource will allow researchers at the University of Cambridge/visiting academics with relevant honorary contracts to:

- Identify genetic and lifestyle determinants of iron homeostasis and its potential health-related consequences in blood donors, which may in the future be used to predict donation outcomes and personalise the national blood service. The study will initially focus on iron-related biomarkers that have been measured in INTERVAL and COMPARE. Researchers will characterise associations of iron-related biomarkers measured in INTERVAL and COMPARE and quantifying relationships with several health outcomes and will define the shapes of dose-response relationships with health outcomes, and explore whether associations with outcomes vary in key subgroups (eg, by age and sex).

- Study genetic and lifestyle determinants of several metabolic traits and their potential relevance to cardiovascular diseases: Participants in INTERVAL and COMPARE have already provided information about their lifestyle and several metabolic traits have been measured, including >1100 metabolites and >1200 lipids. Using developed methods, researchers will study in detail associations of several metabolic traits with incident cardiovascular diseases and other health outcomes. These analyses will: (i) quantify relationships by correcting associations for potential confounders, and (ii) define shapes of dose-response relationships, explore subgroup effects and heterogeneity, and suggest lifestyle and biological determinants of novel metabolic factors.

- Study of genetic determinants of several soluble plasma proteins and their potential relevance to chronic diseases, such as cardiovascular diseases: >4000 plasma proteins have been measured in INTERVAL and COMPARE. These analyses will: (i) quantify relationships with several health outcomes, and (ii) define shapes of dose-response relationships, explore subgroup effects and heterogeneity, and suggest lifestyle and biological determinants of circulating proteins.

The University of Cambridge require data on a monthly basis (Quarterly from Version 7 of this agreement) for the term of the agreement. The University has requested further data within the following datasets:

• HES – Admitted Patient Care

• HES – Outpatients

• Cancer Registration Data

• Civil Registration Mortality Data

• GPES Data for Pandemic Planning and Research.

• COVID-19 UK Non-hospital Antibody Testing Results (Pillar 3)

• COVID-19 Vaccination Status

• COVID-19 Vaccination Adverse Reactions

• The Sentinel Stroke National Audit Programme (SSNAP).

The University has requested identifiable record level data. Historic HES data is included within the data already held by the University going as far back as 2002 (10 years before the recruitment of participants). For HES data, the University of Cambridge has received historical data (going back up to 10 years from the recruitment of the first INTERVAL participant, 2002-2018/19) and is requesting future disseminations every month. Linkage to HES records will be used to enhance information already recorded at baseline in the INTERVAL and COMPARE studies about the donors’ prior medical history. Given the age of participants enrolled in these cohorts it is believed that 10 years will provide a reliable timeframe to capture pre-existing medical conditions.

The data controller for the INTERVAL and COMPARE studies is the University of Cambridge, which processes the data as part of its public task under GDPR Article 6(1)(e). Processing special category personal data is necessary for scientific research purposes under GDPR Article 9(2)(j). Appropriate safeguards to the processing of data are in place, as are appropriate technical and organisational measures, described elsewhere in the application. The public interest nature of the processing is assessed as part of the research ethics committee review of the studies and through peer review from the public bodies funding the research.

Additional outputs for urgent COVID-19 work: University of Cambridge will rapidly disseminate ensuing evidence on COVID-19 risk factors to key stakeholders (e.g., policy-makers, healthcare organisations and the scientific community) through preprints and other rapid forms of communication, as well as through traditional publication routes.

The results from the analyses with the SARS-CoV-2 testing data from Public Health England and information on pre-existing conditions and new diagnoses based on Hospital Episode Statistics are already contributing to the international COVID-19 Host Genetics Initiative (https://www.covid19hg.org/), and therefore increasing global knowledge of the biology of SARS-CoV-2 infection and disease.

Research findings will also be disseminated to the public through various routes including the INTERVAL and COMPARE study websites, newsletters and through the active public engagement programmes.

---

The initial focus for INTERVAL and COMPARE has been working to improve NHSBT’s core services (e.g. safety and efficiency of blood donation). Since our initial application, University of Cambridge have published results from a further two years follow up of INTERVAL participants (Kaptoge, Lancet Haematol 2019). This showed that more intensive reminders increased whole blood donation rates and allowed evaluation of the longer-term risks and benefits of varying inter-donation intervals. Outputs are shared with participants through regular updates in the form of web publications, email communications and newsletters. Results are also disseminated to the public and donors through study website, newsletters, national and international press releases and media interviews. The realised and expected benefits of these studies to blood donors and the wider patient population are described below.

The renewal of this agreement will allow for continued follow up of study participants to identify new health events. Findings will be published and disseminated through publications in high impact journals and through dissemination to academic, health service, and general public audiences. Publications arising from these studies will be available on the studies website at http://www.intervalstudy.org.uk/publications/ for INTERVAL and http://www.comparestudy.org.uk/publications/ for COMPARE.

Subsequent stages of both the INTERVAL and COMPARE studies are related to the creation of a comprehensive resource that will enable detailed studies of health-related questions by linking health outcomes data to genetic, biological and lifestyle information. For example, University of Cambridge have published a novel analysis of the human plasma proteome (Sun, Nature 2018) combining genomic and proteomic measurements from the INTERVAL study to identify potential therapeutic targets, opportunities for matching existing drugs with new disease indications, and potential safety concerns for drugs under development.

Use of the linked data from NHS Digital is at an earlier stage but, as examples of the types of output to be expected, manuscripts describing the shared underpinnings of five distinct cardiometabolic conditions (coronary disease, atrial fibrillation, stroke, type 2 diabetes, chronic kidney disease) are at advanced stages. These use novel methods that combine polygenic risk scores, molecular ‘omics, and disease databases and expect this to result in multiple high impact scientific publications. As such, it is expected that findings from this resource will extensively advance biomedical research and inform public health policy. Given that health outcomes will accrue over time and University of Cambridge intend to track participants’ health over many years University of Cambridge anticipates that the outputs of this research will be realised for a considerable time into the future. The renewal of this agreement will allow the university to continue and extend this work.

Datasets that have been collected under a specific COVID-19 related direction (e.g. GPES Data for Pandemic Planning and Research - GDPPR) can only be used for COVID-19 research related purposes. In the absence of a continuing legal basis for NHS Digital to provide this data - the long term follow up of the cohort will not apply to these datasets.

As described above, researchers will communicate results through both traditional academic routes (papers, seminars, etc.) but also through the study websites, newsletters, press releases and media interviews. University of Cambridge will continue to work closely with the Patient and Public Involvement and Engagement panel managed by the NIHR Blood and Transplant Research Unit in Donor Health and Genomics, who provide both advice on plans and on the dissemination of results and the unit has an active public engagement programme (see http://www.donorhealth-btru.nihr.ac.uk/btru_events/).

When sharing research findings, results will be displayed as aggregate data only (with small numbers suppressed, in line with the HES analysis guide), therefore individual data cannot be recognised.

As described above, the results from the analyses with the SARS-CoV-2 testing data from Public Health England and information on pre-existing conditions and new diagnoses based on Hospital Episode Statistics are already contributing to the international COVID-19 Host Genetics Initiative (https://www.covid19hg.org/), and it is therefore hoped this is increasing global knowledge of the biology of SARS-CoV-2 infection and disease.

Work by this consortium has identified areas in the genome that are associated with SARS-CoV-2 infection or severe COVID-19 disease and this has the potential to inform the development of therapeutics for patients. Other analysis by University of Cambridge researchers and collaborators has identified a gene (IFNAR2 ) that is more likely to play a role in COVID-19 hospitalisation, with findings from this work prioritising trials of drugs targeting specific proteins (IFNAR2 and ACE2) for early management of COVID-19 (Gaziano, Nature Medicine 2021).

---

Results from the INTERVAL trial published in The Lancet (Di Angelantonio et al, 2017), are already shaping policy nationally and internationally, and leading to the improvements in the health of donors. In particular, results from INTERVAL have shown that more frequent blood donations from donors are possible without causing harm to donor health. It has provided policy-makers with evidence that more frequent collection from donors than is now standard can be done over two years without causing harm to donor health, allowing better management of the supply to the NHS of units of blood with in-demand blood groups. Furthermore, INTERVAL has led to NHSBT’s adoption of comprehensive multi-modal reminders (eg, SMS messages) to help donors make and keep appointments and therefore will help to meet the demand for blood required by patients in the NHS.

Since the initial agreement the University of Cambridge have published results from a further two years follow up of INTERVAL participants (Kaptoge, Lancet Haematol 2019). The findings have provided policy-makers with two key evidence-based options to meet blood supply needs; the use of frequent reminders to help donors keep appointments and shorter inter-donation intervals than are now standard. This study has also quantified the extent of iron depletion within four years of repeated donation, thus informing safety guidelines. Results have been disseminated to the public and donors through the study website, newsletters, national and international press release and media interviews.

The COMPARE study was designed to evaluate the optimum method to measure haemoglobin levels in ~30,000 potential whole blood donors in advance of each donation. Results from COMPARE have led to an evidence-based decision by NHSBT to replace the copper sulphate-based haemoglobin screening with finger-prick haemoglobin testing. It is estimated that this will prevent about 30,000 donors annually from experiencing anaemia and potential iron deficiency, due to being inappropriately bled at a blood donation session (Bell, Sweeting, Transfusion Medicine 2020) and will therefore help to protect the health of blood donors.

The amendment to this agreement (v6) is to request the GPES data for Pandemic Planning and Research (GDPPR) for participants in the INTERVAL and COMPARE studies.

The INTERVAL and COMPARE studies are large, richly characterised cohorts with approximately 75,000 participants in total. A range of genomic, molecular and lifestyle information is already available for the participants as well as Hospital Episodes Statistics (HES), mortality data and cancer registration data provided by NHS Digital under this Data Sharing Agreement. SARS-CoV-2 test results are provided by Public Health England (PHE) and records of intensive care admissions from the Intensive Care National Audit & Research Centre.

Further information related to COVID-19 is being collected through the TRACK-COVID study, which is recruiting participants from the INTERVAL and COMPARE studies. The aim of this research is to determine risk factors for infection of the SARS-CoV-2 virus and investigate why some people who carry the virus are symptomatic while others are asymptomatic. This study includes data collection from participants relating to symptoms and targeted bio-sampling for viral assays.

The University of Cambridge are requesting the GDPPR (GPES Data for Pandemic Planning and Research) dataset as an aim to understand the risk factors associated with COVID-19 status and prognosis in order to inform targeted preventative measures (e.g. prioritisation of vaccinations when available), establish molecular factors associated with susceptibility to COVID-19 and clinical trajectory of the disease, and understand resilience to (and recovery from) severe clinical consequences of SARS-CoV-2 infections.

Linkage to GP records for INTERVAL and COMPARE participants will provide important additional information about pre-existing conditions and co-morbidities, which may affect susceptibility and resilience to COVID-19, but cannot be obtained from other patient records. The University of Cambridge are asking for information about a broad range of risk factors, conditions and medications from the dataset, as it is not yet known which are important and relevant to COVID-19. Having access to this range of information will allow rapid investigation of new hypotheses and replicate findings from other cohorts such as UK Biobank.

The fields from the GDPPR data set have been selected and minimised in order to

- provide updates on participant medical history relating to COVID-19 diagnosis or risk.

- allow investigation of how pre-existing conditions (e.g. specific underlying conditions such as Diabetes and multi-morbidities) may affect susceptibility and resilience to COVID-19.

- provide information on risk factors (e.g. BMI, smoking status) and recent interactions with primary care that may be relevant to COVID-19 status and severity.

- identify prescriptions, treatments and vaccinations that may be associated with COVID-19 susceptibility and resilience (e.g. immunosuppressive agents).

Data such as Year of Birth and Ethnicity has not been requested as details are already held by the University of Cambridge or are not required for research.

Patient and Public Involvement and Engagement

NHSX’s ‘Data Access Request Form E’ and a lay summary of this data request were circulated to the Health Data Research UK Cambridge Advisory Group, which includes blood donors. The public members were asked if they thought that accessing GP data for participants in the INTERVAL and COMPARE studies to enable research into COVID-19 is acceptable, or if they had any concerns. The public members supported the data access request. They thought that accessing the GP records for these participants was acceptable and important, though they emphasised that it is essential that participant confidentiality is maintained.

---

A previous iteration of this agreement (v5) was amended to support COVID-19 research. Under this version of the agreement - it was approved that NHS Digital would increase the data dissemination frequency from bi-annual to monthly. There is urgent need for a better understanding of the risk factors for COVID-19, clinical trajectories after infection with SARS-CoV-2 and the associated health outcomes. This increased frequency is being continued under v6 of this agreement also.

For projects specifically relating to urgent COVID-19 work, access will also be permitted for approved visiting academics where the academic's substantive employer has provided assurance that they would consider disciplinary action in the event of a data breach. All visiting academics have a formal visitor agreement that they and their employing institution sign.

University of Cambridge have linked the INTERVAL and COMPARE studies to SARS-CoV-2 testing data from Public Health England. Together with the existing genomic, molecular and other data held about participants, this will allow research to:

1) Clarify clinical risk factors associated with COVID-19 status and prognosis in order to inform targeted preventative measures (e.g. prioritisation of vaccinations when available) by linkage of e-health records and COVID 19 status/outcome.

2) Establish molecular factors associated with susceptibility to, and clinical trajectory of, COVID-19.

3) Understand resilience to (and recovery from) severe clinical consequences of SARS-CoV-2 infections.

To maximise the value of this research, University of Cambridge need timely information about study participants’ health status, both prior to diagnosis with COVID-19 and, as the outbreak progresses, during and after treatment. Moving from biannual to monthly HES updates will enable this.

BACKGROUND:

INTERVAL and COMPARE studies are multi-purpose, multi-stage research projects involving blood donors. These efforts are national flagship research projects supported by leading public funders and research charities, including the UK Medical Research Council (MRC), National Institute for Health Research (NIHR), NHS Blood and Transplant (NHSBT), British Heart Foundation (BHF), and Wellcome Trust. Collectively, these funders have invested more than £15 million into these studies since 2012. It is worth noting that none of these organisations have access to any record level data and will only access aggregate outputs with small numbers suppressed in line with HES analysis guide. Furthermore, none of these organisations have any decision-making powers regarding the use of the data.

INTERVAL and COMPARE study has received support from the organisations mentioned above in different areas. The INTERVAL and COMPARE trials were funded by NHSBT, the NIHR Blood and Transplant Research Unit in Donor Health and Genomics with the linkage of data from NHS Digital funded by the NIHR Blood and Transplant Research Unit in Donor Health and Genomics (NIHR BTRU-2014-10024). The trials’ coordinating centre at the Department of Public Health and Primary Care at the University of Cambridge, Cambridge, UK, has received core support from the UK Medical Research Council (G0800270), British Heart Foundation (SP/09/002), and the NIHR Cambridge Biomedical Research Centre. DNA sequencing of trial participants was carried out by the Wellcome Sanger Institute using core funding from the Wellcome Trust.

The INTERVAL and COMPARE studies were set up and are managed by the University of Cambridge, who are the data controller for this agreement. The University of Oxford and NHS Blood and Transplant (NHSBT) provide advice on analyses performed by University of Cambridge across the studies, but do not have access to the data disseminated by NHS Digital and will only receive aggregate outputs with small numbers suppressed. Originally, the University of Cambridge designed the methodology for processing data within the studies and will decide on any future methods in carrying out study practises, including the processing of NHS Digital data.

These studies have been specifically designed from their inception to have a multi-purpose strategy to be delivered in multiple stages. This strategy has led to generation and uptake of new evidence-based policies by NHSBT (see below) as well as rapid completion of major-multiple purpose studies.

The initial stage of these studies had been related to blood donation research aiming to improve NHSBT’s core services (e.g. safety and efficiency of blood donation):

- The INTERVAL study recruited ~50,000 blood donors in a randomised controlled trial aiming to assess the impact of varying the frequency of blood donation on donor health and the blood supply.

- The COMPARE study was designed to evaluate the optimum method to measure haemoglobin levels in ~30,000 potential whole blood donors in advance of each donation.

Cohort:

Across the two projects, the University originally recruited a total of ~80,000 participants. Under previous versions of this agreement, sub-cohorts of the total population have been confirmed with NHS Digital for the purposes of receiving hospital data via HES and to compile NHS Numbers of participants using list clean exercises via demographic reports.

The total cohort is split across the two projects as below:

INTERVAL: 48448

COMPARE: 29029

Participants for both the INTERVAL and COMPARE studies were recruited across England from blood donor centres and mobile teams. Prior to donating blood, donors were invited to join the study where informed consent was gained.

Subsequent stages of both INTERVAL and COMPARE studies are related to the creation of a resource of healthy volunteers to enable study of associations of genetic, biological, lifestyle, and other exposures with health outcomes. For the purposes of this agreement, only researchers from the University of Cambridge will access the data products requested in this application.

As outlined in the section above, INTERVAL and COMPARE studies have been specifically designed to have a multi-purpose strategy to be delivered in multiple stages .

The overall objective is to create a multi-dimensional, multi-purpose resource by linking detailed lifestyle and biological information collected on INTERVAL and COMPARE participants with health-related records. The establishment of such a comprehensive resource of healthy volunteers will enable detailed study of the health of blood donors and, more generally, allow studies of cardiovascular disease and other health-related outcome at the University of Cambridge (ie, only researchers from the University of Cambridge/ or visiting academics investigating COVID-19 with appropriate honorary contracts and visiting arrangements will access the data requested in this application). The datasets requested will be used to update the records held about participants and to identify health events to further characterise the study participants. Mortality data will be used to update records held about participants. HES and Demographics data will be used to identify conditions and health events using both existing phenotyping algorithms (such as those developed by the CALIBER initiative or UK Biobank) and study specific code lists. This information will be combined with genetic and biological characteristics of the participants to address a range of research questions.

For example, this resource will allow researchers at the University of Cambridge/visiting academics with relevant honorary contracts to:

- Identify genetic and lifestyle determinants of iron homeostasis and its potential health-related consequences in blood donors, which may in the future be used to predict donation outcomes and personalise the national blood service. The study will initially focus on iron-related biomarkers that have been measured in INTERVAL and COMPARE. Researchers will characterise associations of iron-related biomarkers measured in INTERVAL and COMPARE and quantifying relationships with several health outcomes and will define the shapes of dose-response relationships with health outcomes, and explore whether associations with outcomes vary in key subgroups (eg, by age and sex).

- Study genetic and lifestyle determinants of several metabolic traits and their potential relevance to cardiovascular diseases: Participants in INTERVAL and COMPARE have already provided information about their lifestyle and several metabolic traits have been measured, including >1100 metabolites and >1200 lipids. Using developed methods, researchers will study in detail associations of several metabolic traits with incident cardiovascular diseases and other health outcomes. These analyses will: (i) quantify relationships by correcting associations for potential confounders, and (ii) define shapes of dose-response relationships, explore subgroup effects and heterogeneity, and suggest lifestyle and biological determinants of novel metabolic factors.

- Study of genetic determinants of several soluble plasma proteins and their potential relevance to chronic diseases, such as cardiovascular diseases: >4000 plasma proteins have been measured in INTERVAL and COMPARE. These analyses will: (i) quantify relationships with several health outcomes, and (ii) define shapes of dose-response relationships, explore subgroup effects and heterogeneity, and suggest lifestyle and biological determinants of circulating proteins.

The University of Cambridge require data on a monthly basis for the term of the agreement. The University has requested further data within the following datasets:

• HES – Admitted Patient Care

• HES – Outpatients

• Cancer Registration Data

• Demographic Data

• Civl Registration Mortality Data

• GPES Data for Pandemic Planning and Research.

The University has requested record level data with identifiable mortality & embarkation data as well as record level HES pseudonymised at NHS Number. Historic HES data is included within the data already held by the University going as far back as 2002 (10 years before the recruitment of participants). For HES data, the University of Cambridge has received historical data (going back up to 10 years from the recruitment of the first INTERVAL participant, 2002-2018/19) and is requesting future disseminations every month. Linkage to HES records will be used to enhance information already recorded at baseline in the INTERVAL and COMPARE studies about the donors’ prior medical history. Given the age of participants enrolled in these cohorts it is believed that 10 years will provide a reliable timeframe to capture pre-existing medical conditions.

The data controller for the INTERVAL and COMPARE studies is the University of Cambridge, which processes the data as part of its public task under GDPR Article 6(1)(e). Processing special category personal data is necessary for scientific research purposes under GDPR Article 9(2)(j). Appropriate safeguards to the processing of data are in place, as are appropriate technical and organisational measures, described elsewhere in the application. The public interest nature of the processing is assessed as part of the research ethics committee review of the studies and through peer review from the public bodies funding the research.

Additional outputs for urgent COVID-19 work: University of Cambridge will rapidly disseminate ensuing evidence on COVID-19 risk factors to key stakeholders (e.g., policy-makers, healthcare organisations and the scientific community) through preprints and other rapid forms of communication, as well as through traditional publication routes.

The results from the analyses with the SARS-CoV-2 testing data from Public Health England and information on pre-existing conditions and new diagnoses based on Hospital Episode Statistics are already contributing to the international COVID-19 Host Genetics Initiative (https://www.covid19hg.org/), and therefore increasing global knowledge of the biology of SARS-CoV-2 infection and disease.

Research findings will also be disseminated to the public through various routes including the INTERVAL and COMPARE study websites, newsletters and through the active public engagement programmes.

---

The initial focus for INTERVAL and COMPARE has been working to improve NHSBT’s core services (e.g. safety and efficiency of blood donation). Since our initial application, University of Cambridge have published results from a further two years follow up of INTERVAL participants (Kaptoge, Lancet Haematol 2019). This showed that more intensive reminders increased whole blood donation rates and allowed evaluation of the longer-term risks and benefits of varying inter-donation intervals. Outputs are shared with participants through regular updates in the form of web publications, email communications and newsletters. Results are also disseminated to the public and donors through study website, newsletters, national and international press releases and media interviews. The realised and expected benefits of these studies to blood donors and the wider patient population are described below.

The renewal of this agreement will allow for continued follow up of study participants to identify new health events. Findings will be published and disseminated through publications in high impact journals and through dissemination to academic, health service, and general public audiences. Publications arising from these studies will be available on the studies website at http://www.intervalstudy.org.uk/publications/ for INTERVAL and http://www.comparestudy.org.uk/publications/ for COMPARE.

Subsequent stages of both the INTERVAL and COMPARE studies are related to the creation of a comprehensive resource that will enable detailed studies of health-related questions by linking health outcomes data to genetic, biological and lifestyle information. For example, University of Cambridge have published a novel analysis of the human plasma proteome (Sun, Nature 2018) combining genomic and proteomic measurements from the INTERVAL study to identify potential therapeutic targets, opportunities for matching existing drugs with new disease indications, and potential safety concerns for drugs under development.

Use of the linked data from NHS Digital is at an earlier stage but, as examples of the types of output to be expected, manuscripts describing the shared underpinnings of five distinct cardiometabolic conditions (coronary disease, atrial fibrillation, stroke, type 2 diabetes, chronic kidney disease) are at advanced stages. These use novel methods that combine polygenic risk scores, molecular ‘omics, and disease databases and expect this to result in multiple high impact scientific publications. As such, it is expected that findings from this resource will extensively advance biomedical research and inform public health policy. Given that health outcomes will accrue over time and University of Cambridge intend to track participants’ health over many years University of Cambridge anticipates that the outputs of this research will be realised for a considerable time into the future. The renewal of this agreement will allow the university to continue and extend this work.

One of the data sets requested in this application (GPES Data for Pandemic Planning and Research - GDPPR) has been collected under a specific COVID-19 related direction, and as such can only be used for COVID-19 research related purposes. In the absence of a continuing legal basis for NHS Digital to provide this data - the long term follow up of the cohort will not apply to the GDPPR data.

As described above, researchers will communicate results through both traditional academic routes (papers, seminars, etc.) but also through the study websites, newsletters, press releases and media interviews. University of Cambridge will continue to work closely with the Patient and Public Involvement and Engagement panel managed by the NIHR Blood and Transplant Research Unit in Donor Health and Genomics, who provide both advice on plans and on the dissemination of results and the unit has an active public engagement programme (see http://www.donorhealth-btru.nihr.ac.uk/btru_events/).

When sharing research findings, results will be displayed as aggregate data only (with small numbers suppressed, in line with the HES analysis guide), therefore individual data cannot be recognised. For the purposes of this application, only researchers from the University of Cambridge/visiting academics with the appropriate honorary contracts will access the data products requested in this application.

As described above, the results from the analyses with the SARS-CoV-2 testing data from Public Health England and information on pre-existing conditions and new diagnoses based on Hospital Episode Statistics are already contributing to the international COVID-19 Host Genetics Initiative (https://www.covid19hg.org/), and therefore increasing global knowledge of the biology of SARS-CoV-2 infection and disease.

---

Results of the INTERVAL and COMPARE studies are already shaping policy, nationally and internationally, and leading to improvements in the health of about 1 million blood donors in England.

Results from the INTERVAL trial published in The Lancet (Di Angelantonio et al, 2017), are already shaping policy nationally and internationally, and leading to the improvements in the health of donors. In particular, results from INTERVAL have shown that more frequent blood donations from donors can be done without causing harm to donor health. It has provided policy-makers with evidence that more frequent collection from donors than is now standard can be done over two years without causing harm to donor health, allowing better management of the supply to the NHS of units of blood with in-demand blood groups. Furthermore, INTERVAL has led to NHSBT’s adoption of comprehensive multi-modal reminders (eg, SMS messages) to help donors make and keep appointments.

Since the initial application the University of Cambridge have published results from a further two years follow up of INTERVAL participants (Kaptoge, Lancet Haematol 2019). The findings have provided policy-makers with two key evidence-based options to meet blood supply needs; the use of frequent reminders to help donors keep appointments and shorter inter-donation intervals than are now standard. This study has also quantified the extent of iron depletion within four years of repeated donation, thus informing safety guidelines. Results have been disseminated to the public and donors through study website, newsletters, national and international press release and media interviews.

The COMPARE study was designed to evaluate the optimum method to measure haemoglobin levels in ~30,000 potential whole blood donors in advance of each donation. Results from COMPARE have led to an evidence-based decision by NHSBT to replace the copper sulphate-based haemoglobin screening with finger-prick haemoglobin testing, thereby preventing ~350 female donors being inappropriately bled each day in England (ie, female donors with haemoglobin levels <12.5g/dL, the minimum threshold mandated by regulators). The COMPARE manuscript is due to be submitted for publication by the end of 2019.

This v5 amendment is an urgent request to support COVID-19 research which will increase data dissemination frequency from bi-annual to monthly. There is an urgent need for a better understanding of the risk factors for COVID-19, clinical trajectories after infection with SARS-CoV-2 and the associated health outcomes.

University of Cambridge have linked the INTERVAL and COMPARE studies to SARS-CoV-2 testing data from Public Health England. Together with the existing genomic, molecular and other data held about participants, this will allow research to:

1) Clarify clinical risk factors associated with COVID-19 status and prognosis in order to inform targeted preventative measures (e.g. prioritisation of vaccinations when available) by linkage of e-health records and COVID 19 status/outcome.

2) Establish molecular factors associated with susceptibility to, and clinical trajectory of, COVID-19.

3) Understand resilience to (and recovery from) severe clinical consequences of SARS-CoV-2 infections.

To maximise the value of this research, University of Cambridge need timely information about study participants’ health status, both prior to diagnosis with COVID-19 and, as the outbreak progresses, during and after treatment. Moving from biannual to monthly HES updates will enable this.

---

INTERVAL and COMPARE studies are multi-purpose, multi-stage research projects involving blood donors. These efforts are national flagship research projects supported by leading public funders and research charities, including the UK Medical Research Council (MRC), National Institute for Health Research (NIHR), NHS Blood and Transplant (NHSBT), British Heart Foundation (BHF), and Wellcome Trust. Collectively, these funders have invested more than £15 million into these studies since 2012. It is worth noting that none of these organisations have access to any record level data and will only access aggregate outputs with small numbers suppressed in line with HES analysis guide. Furthermore, none of these organisations have any decision-making powers regarding the use of the data.

INTERVAL and COMPARE study has received support from the organisations mentioned above in different areas. The INTERVAL and COMPARE trials were funded by NHSBT, the NIHR Blood and Transplant Research Unit in Donor Health and Genomics with the linkage of data from NHS Digital funded by the NIHR Blood and Transplant Research Unit in Donor Health and Genomics (NIHR BTRU-2014-10024). The trials’ coordinating centre at the Department of Public Health and Primary Care at the University of Cambridge, Cambridge, UK, has received core support from the UK Medical Research Council (G0800270), British Heart Foundation (SP/09/002), and the NIHR Cambridge Biomedical Research Centre. DNA sequencing of trial participants was carried out by the Wellcome Sanger Institute using core funding from the Wellcome Trust.

The INTERVAL and COMPARE studies were set up and are managed by the University of Cambridge, who are the data controller for this agreement. The University of Oxford and NHS Blood and Transplant (NHSBT) provide advice on analyses performed by University of Cambridge across the studies, but do not have access to the data disseminated by NHS Digital and will only receive aggregate outputs with small numbers suppressed. Originally, the University of Cambridge designed the methodology for processing data within the studies and will decide on any future methods in carrying out study practises, including the processing of NHS Digital data.

These studies have been specifically designed from their inception to have a multi-purpose strategy to be delivered in multiple stages. This strategy has led to generation and uptake of new evidence-based policies by NHSBT (see below) as well as rapid completion of major-multiple purpose studies.

The initial stage of these studies had been related to blood donation research aiming to improve NHSBT’s core services (e.g. safety and efficiency of blood donation):

- The INTERVAL study recruited ~50,000 blood donors in a randomised controlled trial aiming to assess the impact of varying the frequency of blood donation on donor health and the blood supply.

- The COMPARE study was designed to evaluate the optimum method to measure haemoglobin levels in ~30,000 potential whole blood donors in advance of each donation.

Cohort:

Across the two projects, the University originally recruited a total of ~80,000 participants. Under previous versions of this agreement, sub-cohorts of the total population have been confirmed with NHS Digital for the purposes of receiving hospital data via HES and to compile NHS Numbers of participants using list clean exercises via MRIS reports.

Under version 4 of the agreement, the University of Cambridge has requested the full cohort be used for tracing the data outline below. The previous total cohort was confirmed as 77,500, which the University confirms to now be 77,477 to account for participants who have withdrawn from the study. The total cohort is split across the two projects as below:

INTERVAL: 48448

COMPARE: 29029

Participants for both the INTERVAL and COMPARE studies were recruited across England from blood donor centres and mobile teams. Prior to donating blood, donors were invited to join the study where informed consent was gained.

Subsequent stages of both INTERVAL and COMPARE studies are related to the creation of a resource of healthy volunteers to enable study of associations of genetic, biological, lifestyle, and other exposures with health outcomes. For the purposes of this agreement, only researchers from the University of Cambridge will access the data products requested in this application.

As outlined in the section above, INTERVAL and COMPARE studies have been specifically designed to have a multi-purpose strategy to be delivered in multiple stages .

The overall objective is to create a multi-dimensional, multi-purpose resource by linking detailed lifestyle and biological information collected on INTERVAL and COMPARE participants with health-related records. The establishment of such a comprehensive resource of healthy volunteers will enable detailed study of the health of blood donors and, more generally, allow studies of cardiovascular disease and other health-related outcome at the University of Cambridge (ie, only researchers from the University of Cambridge will access the data requested in this application ). The datasets requested will be used to update the records held about participants and to identify health events to further characterise the study participants. Mortality data will be used to update records held about participants. HES and MRIS data will be used to identify conditions and health events using both existing phenotyping algorithms (such as those developed by the CALIBER initiative or UK Biobank) and study specific code lists. This information will be combined with genetic and biological characteristics of the participants to address a range of research questions.

For example, this resource will allow researchers at the University of Cambridge to:

- Identify genetic and lifestyle determinants of iron homeostasis and its potential health-related consequences in blood donors, which may in the future be used to predict donation outcomes and personalise the national blood service. The study will initially focus on iron-related biomarkers that have been measured in INTERVAL and COMPARE. Researchers will characterise associations of iron-related biomarkers measured in INTERVAL and COMPARE and quantifying relationships with several health outcomes and will define the shapes of dose-response relationships with health outcomes, and explore whether associations with outcomes vary in key subgroups (eg, by age and sex).

- Study genetic and lifestyle determinants of several metabolic traits and their potential relevance to cardiovascular diseases: Participants in INTERVAL and COMPARE have already provided information about their lifestyle and several metabolic traits have been measured, including >1100 metabolites and >1200 lipids. Using developed methods, researchers will study in detail associations of several metabolic traits with incident cardiovascular diseases and other health outcomes. These analyses will: (i) quantify relationships by correcting associations for potential confounders, and (ii) define shapes of dose-response relationships, explore subgroup effects and heterogeneity, and suggest lifestyle and biological determinants of novel metabolic factors.

- Study of genetic determinants of several soluble plasma proteins and their potential relevance to chronic diseases, such as cardiovascular diseases: >4000 plasma proteins have been measured in INTERVAL and COMPARE. These analyses will: (i) quantify relationships with several health outcomes, and (ii) define shapes of dose-response relationships, explore subgroup effects and heterogeneity, and suggest lifestyle and biological determinants of circulating proteins.

Data Summary

The University of Cambridge requests additional Hospital Episode Statistics & Medical Research report data on a bi-annual basis for the term of the agreement. The University has requested further data within the following datasets:

• HES – Admitted Patient Care

• HES – Outpatients

• MRIS – Cause of Death report

• MRIS – Cohort Event Notification report

• MRIS – Flagging Current Status report

The University has requested record level data with identifiable mortality & embarkation data as well as record level HES pseudonymised at NHS Number. Historic HES data is included within the data already held by the University going as far back as 2002 (10 years before the recruitment of participants). For HES data, the University of Cambridge has received historical data (going back up to 10 years from the recruitment of the first INTERVAL participant, 2002-2018/19) and is requesting future disseminations every 6 months. Linkage to HES records will be used to enhance information already recorded at baseline in the INTERVAL and COMPARE studies about the donors’ prior medical history. Given the age of participants enrolled in these cohorts it is believed that 10 years will provide a reliable timeframe to capture pre-existing medical conditions.

The data controller for the INTERVAL and COMPARE studies is the University of Cambridge, which processes the data as part of its public task under GDPR Article 6(1)(e). Processing special category personal data is necessary for scientific research purposes under GDPR Article 9(2)(j). Appropriate safeguards to the processing of data are in place, as are appropriate technical and organisational measures, described elsewhere in the application. The public interest nature of the processing is assessed as part of the research ethics committee review of the studies and through peer review from the public bodies funding the research.

Additional outputs for urgent COVID-19 work: University of Cambridge will rapidly disseminate ensuing evidence on COVID-19 risk factors to key stakeholders (e.g., policy-makers, healthcare organisations and the scientific community) through preprints and other rapid forms of communication, as well as through traditional publication routes.

---

The initial focus for INTERVAL and COMPARE has been working to improve NHSBT’s core services (e.g. safety and efficiency of blood donation). Since our initial application, University of Cambridge have published results from a further two years follow up of INTERVAL participants (Kaptoge, Lancet Haematol 2019). This showed that more intensive reminders increased whole blood donation rates and allowed evaluation of the longer-term risks and benefits of varying inter-donation intervals. Outputs are shared with participants through regular updates in the form of web publications, email communications and newsletters. Results are also disseminated to the public and donors through study website, newsletters, national and international press releases and media interviews. The realised and expected benefits of these studies to blood donors and the wider patient population are described below.

The renewal of this agreement will allow for continued follow up of study participants to identify new health events. Findings will be published and disseminated through publications in high impact journals and through dissemination to academic, health service, and general public audiences. Publications arising from these studies will be available on the studies website at http://www.intervalstudy.org.uk/publications/ for INTERVAL and http://www.comparestudy.org.uk/publications/ for COMPARE.

Subsequent stages of both the INTERVAL and COMPARE studies are related to the creation of a comprehensive resource that will enable detailed studies of health-related questions by linking health outcomes data to genetic, biological and lifestyle information. For example, University of Cambridge have published a novel analysis of the human plasma proteome (Sun, Nature 2018) combining genomic and proteomic measurements from the INTERVAL study to identify potential therapeutic targets, opportunities for matching existing drugs with new disease indications, and potential safety concerns for drugs under development.

Use of the linked data from NHS Digital is at an earlier stage but, as examples of the types of output to be expected, manuscripts describing the shared underpinnings of five distinct cardiometabolic conditions (coronary disease, atrial fibrillation, stroke, type 2 diabetes, chronic kidney disease) are at advanced stages. These use novel methods that combine polygenic risk scores, molecular ‘omics, and disease databases and expect this to result in multiple high impact scientific publications. As such, it is expected that findings from this resource will extensively advance biomedical research and inform public health policy. Given that health outcomes will accrue over time and University of Cambridge intend to track participants’ health over many years University of Cambridge anticipates that the outputs of this research will be realised for a considerable time into the future. The renewal of this agreement will allow the university to continue and extend this work.

As described above, researchers will communicate results through both traditional academic routes (papers, seminars, etc.) but also through the study websites, newsletters, press releases and media interviews. University of Cambridge will continue to work closely with the Patient and Public Involvement and Engagement panel managed by the NIHR Blood and Transplant Research Unit in Donor Health and Genomics, who provide both advice on plans and on the dissemination of results and the unit has an active public engagement programme (see http://www.donorhealth-btru.nihr.ac.uk/btru_events/).

When sharing research findings, results will be displayed as aggregate data only (with small numbers suppressed, in line with the HES analysis guide), therefore individual data cannot be recognised. For the purposes of this application, only researchers from the University of Cambridge will access the data products requested in this application.

Results of the INTERVAL and COMPARE studies are already shaping policy, nationally and internationally, and leading to improvements in the health of about 1 million blood donors in England.

Results from the INTERVAL trial published in The Lancet (Di Angelantonio et al, 2017), are already shaping policy nationally and internationally, and leading to the improvements in the health of donors. In particular, results from INTERVAL have shown that more frequent blood donations from donors can be done without causing harm to donor health. It has provided policy-makers with evidence that more frequent collection from donors than is now standard can be done over two years without causing harm to donor health, allowing better management of the supply to the NHS of units of blood with in-demand blood groups. Furthermore, INTERVAL has led to NHSBT’s adoption of comprehensive multi-modal reminders (eg, SMS messages) to help donors make and keep appointments.

Since our initial application we have published results from a further two years follow up of INTERVAL participants (Kaptoge, Lancet Haematol 2019). The findings have provided policy-makers with two key evidence-based options to meet blood supply needs; the use of frequent reminders to help donors keep appointments and shorter inter-donation intervals than are now standard. This study has also quantified the extent of iron depletion within four years of repeated donation, thus informing safety guidelines. Results have been disseminated to the public and donors through study website, newsletters, national and international press release and media interviews.

The COMPARE study was designed to evaluate the optimum method to measure haemoglobin levels in ~30,000 potential whole blood donors in advance of each donation. Results from COMPARE have led to an evidence-based decision by NHSBT to replace the copper sulphate-based haemoglobin screening with finger-prick haemoglobin testing, thereby preventing ~350 female donors being inappropriately bled each day in England (ie, female donors with haemoglobin levels <12.5g/dL, the minimum threshold mandated by regulators). The COMPARE manuscript is due to be submitted for publication by the end of 2019.

INTERVAL and COMPARE studies are multi-purpose, multi-stage research projects involving blood donors. These efforts are national flagship research projects supported by leading public funders and research charities, including the UK Medical Research Council (MRC), National Institute for Health Research (NIHR), NHS Blood and Transplant (NHSBT), British Heart Foundation (BHF), and Wellcome Trust. Collectively, these funders have invested more than £15 million into these studies since 2012. It is worth noting that none of these organisations have access to any record level data and will only access aggregate outputs with small numbers suppressed in line with HES analysis guide. Furthermore, none of these organisations have any decision-making powers regarding the use of the data.

INTERVAL and COMPARE study has received support from the organisations mentioned above in different areas. The INTERVAL and COMPARE trials were funded by NHSBT, the NIHR Blood and Transplant Research Unit in Donor Health and Genomics with the linkage of data from NHS Digital funded by the NIHR Blood and Transplant Research Unit in Donor Health and Genomics (NIHR BTRU-2014-10024). The trials’ coordinating centre at the Department of Public Health and Primary Care at the University of Cambridge, Cambridge, UK, has received core support from the UK Medical Research Council (G0800270), British Heart Foundation (SP/09/002), and the NIHR Cambridge Biomedical Research Centre. DNA sequencing of trial participants was carried out by the Wellcome Sanger Institute using core funding from the Wellcome Trust.

The INTERVAL and COMPARE studies were set up and are managed by the University of Cambridge, who are the data controller for this agreement. The University of Oxford and NHS Blood and Transplant (NHSBT) provide advice on analyses performed by University of Cambridge across the studies, but do not have access to the data disseminated by NHS Digital and will only receive aggregate outputs with small numbers suppressed. Originally, the University of Cambridge designed the methodology for processing data within the studies and will decide on any future methods in carrying out study practises, including the processing of NHS Digital data.

These studies have been specifically designed from their inception to have a multi-purpose strategy to be delivered in multiple stages. This strategy has led to generation and uptake of new evidence-based policies by NHSBT (see below) as well as rapid completion of major-multiple purpose studies.

The initial stage of these studies had been related to blood donation research aiming to improve NHSBT’s core services (e.g. safety and efficiency of blood donation):

- The INTERVAL study recruited ~50,000 blood donors in a randomised controlled trial aiming to assess the impact of varying the frequency of blood donation on donor health and the blood supply.

- The COMPARE study was designed to evaluate the optimum method to measure haemoglobin levels in ~30,000 potential whole blood donors in advance of each donation.

Cohort:

Across the two projects, the University originally recruited a total of ~80,000 participants. Under previous versions of this agreement, sub-cohorts of the total population have been confirmed with NHS Digital for the purposes of receiving hospital data via HES and to compile NHS Numbers of participants using list clean exercises via MRIS reports.

Under version 4 of the agreement, the University of Cambridge has requested the full cohort be used for tracing the data outline below. The previous total cohort was confirmed as 77,500, which the University confirms to now be 77,477 to account for participants who have withdrawn from the study. The total cohort is split across the two projects as below:

INTERVAL: 48448

COMPARE: 29029

Participants for both the INTERVAL and COMPARE studies were recruited across England from blood donor centres and mobile teams. Prior to donating blood, donors were invited to join the study where informed consent was gained.

Subsequent stages of both INTERVAL and COMPARE studies are related to the creation of a resource of healthy volunteers to enable study of associations of genetic, biological, lifestyle, and other exposures with health outcomes. For the purposes of this agreement, only researchers from the University of Cambridge will access the data products requested in this application.

As outlined in the section above, INTERVAL and COMPARE studies have been specifically designed to have a multi-purpose strategy to be delivered in multiple stages .

The overall objective is to create a multi-dimensional, multi-purpose resource by linking detailed lifestyle and biological information collected on INTERVAL and COMPARE participants with health-related records. The establishment of such a comprehensive resource of healthy volunteers will enable detailed study of the health of blood donors and, more generally, allow studies of cardiovascular disease and other health-related outcome at the University of Cambridge (ie, only researchers from the University of Cambridge will access the data requested in this application ). The datasets requested will be used to update the records held about participants and to identify health events to further characterise the study participants. Mortality data will be used to update records held about participants. HES and MRIS data will be used to identify conditions and health events using both existing phenotyping algorithms (such as those developed by the CALIBER initiative or UK Biobank) and study specific code lists. This information will be combined with genetic and biological characteristics of the participants to address a range of research questions.

For example, this resource will allow researchers at the University of Cambridge to:

- Identify genetic and lifestyle determinants of iron homeostasis and its potential health-related consequences in blood donors, which may in the future be used to predict donation outcomes and personalise the national blood service. The study will initially focus on iron-related biomarkers that have been measured in INTERVAL and COMPARE. Researchers will characterise associations of iron-related biomarkers measured in INTERVAL and COMPARE and quantifying relationships with several health outcomes and will define the shapes of dose-response relationships with health outcomes, and explore whether associations with outcomes vary in key subgroups (eg, by age and sex).

- Study genetic and lifestyle determinants of several metabolic traits and their potential relevance to cardiovascular diseases: Participants in INTERVAL and COMPARE have already provided information about their lifestyle and several metabolic traits have been measured, including >1100 metabolites and >1200 lipids. Using developed methods, researchers will study in detail associations of several metabolic traits with incident cardiovascular diseases and other health outcomes. These analyses will: (i) quantify relationships by correcting associations for potential confounders, and (ii) define shapes of dose-response relationships, explore subgroup effects and heterogeneity, and suggest lifestyle and biological determinants of novel metabolic factors.

- Study of genetic determinants of several soluble plasma proteins and their potential relevance to chronic diseases, such as cardiovascular diseases: >4000 plasma proteins have been measured in INTERVAL and COMPARE. These analyses will: (i) quantify relationships with several health outcomes, and (ii) define shapes of dose-response relationships, explore subgroup effects and heterogeneity, and suggest lifestyle and biological determinants of circulating proteins.

Data Summary

The University of Cambridge requests additional Hospital Episode Statistics & Medical Research report data on a bi-annual basis for the term of the agreement. The University has requested further data within the following datasets:

• HES – Admitted Patient Care

• HES – Outpatients

• MRIS – Cause of Death report

• MRIS – Cohort Event Notification report

• MRIS – Flagging Current Status report

The University has requested record level data with identifiable mortality & embarkation data as well as record level HES pseudonymised at NHS Number. Historic HES data is included within the data already held by the University going as far back as 2002 (10 years before the recruitment of participants). For HES data, the University of Cambridge has received historical data (going back up to 10 years from the recruitment of the first INTERVAL participant, 2002-2018/19) and is requesting future disseminations every 6 months. Linkage to HES records will be used to enhance information already recorded at baseline in the INTERVAL and COMPARE studies about the donors’ prior medical history. Given the age of participants enrolled in these cohorts it is believed that 10 years will provide a reliable timeframe to capture pre-existing medical conditions.

The data controller for the INTERVAL and COMPARE studies is the University of Cambridge, which processes the data as part of its public task under GDPR Article 6(1)(e). Processing special category personal data is necessary for scientific research purposes under GDPR Article 9(2)(j). Appropriate safeguards to the processing of data are in place, as are appropriate technical and organisational measures, described elsewhere in the application. The public interest nature of the processing is assessed as part of the research ethics committee review of the studies and through peer review from the public bodies funding the research.

The initial focus for INTERVAL and COMPARE has been working to improve NHSBT’s core services (e.g. safety and efficiency of blood donation). Since our initial application, University of Cambridge have published results from a further two years follow up of INTERVAL participants (Kaptoge, Lancet Haematol 2019). This showed that more intensive reminders increased whole blood donation rates and allowed evaluation of the longer-term risks and benefits of varying inter-donation intervals. Outputs are shared with participants through regular updates in the form of web publications, email communications and newsletters. Results are also disseminated to the public and donors through study website, newsletters, national and international press releases and media interviews. The realised and expected benefits of these studies to blood donors and the wider patient population are described below.

The renewal of this agreement will allow for continued follow up of study participants to identify new health events. Findings will be published and disseminated through publications in high impact journals and through dissemination to academic, health service, and general public audiences. Publications arising from these studies will be available on the studies website at http://www.intervalstudy.org.uk/publications/ for INTERVAL and http://www.comparestudy.org.uk/publications/ for COMPARE.

Subsequent stages of both the INTERVAL and COMPARE studies are related to the creation of a comprehensive resource that will enable detailed studies of health-related questions by linking health outcomes data to genetic, biological and lifestyle information. For example, University of Cambridge have published a novel analysis of the human plasma proteome (Sun, Nature 2018) combining genomic and proteomic measurements from the INTERVAL study to identify potential therapeutic targets, opportunities for matching existing drugs with new disease indications, and potential safety concerns for drugs under development.

Use of the linked data from NHS Digital is at an earlier stage but, as examples of the types of output to be expected, manuscripts describing the shared underpinnings of five distinct cardiometabolic conditions (coronary disease, atrial fibrillation, stroke, type 2 diabetes, chronic kidney disease) are at advanced stages. These use novel methods that combine polygenic risk scores, molecular ‘omics, and disease databases and expect this to result in multiple high impact scientific publications. As such, it is expected that findings from this resource will extensively advance biomedical research and inform public health policy. Given that health outcomes will accrue over time and University of Cambridge intend to track participants’ health over many years University of Cambridge anticipates that the outputs of this research will be realised for a considerable time into the future. The renewal of this agreement will allow the university to continue and extend this work.

As described above, researchers will communicate results through both traditional academic routes (papers, seminars, etc.) but also through the study websites, newsletters, press releases and media interviews. University of Cambridge will continue to work closely with the Patient and Public Involvement and Engagement panel managed by the NIHR Blood and Transplant Research Unit in Donor Health and Genomics, who provide both advice on plans and on the dissemination of results and the unit has an active public engagement programme (see http://www.donorhealth-btru.nihr.ac.uk/btru_events/).

When sharing research findings, results will be displayed as aggregate data only (with small numbers suppressed, in line with the HES analysis guide), therefore individual data cannot be recognised. For the purposes of this application, only researchers from the University of Cambridge will access the data products requested in this application.

Results of the INTERVAL and COMPARE studies are already shaping policy, nationally and internationally, and leading to improvements in the health of about 1 million blood donors in England.

Results from the INTERVAL trial published in The Lancet (Di Angelantonio et al, 2017), are already shaping policy nationally and internationally, and leading to the improvements in the health of donors. In particular, results from INTERVAL have shown that more frequent blood donations from donors can be done without causing harm to donor health. It has provided policy-makers with evidence that more frequent collection from donors than is now standard can be done over two years without causing harm to donor health, allowing better management of the supply to the NHS of units of blood with in-demand blood groups. Furthermore, INTERVAL has led to NHSBT’s adoption of comprehensive multi-modal reminders (eg, SMS messages) to help donors make and keep appointments.

Since our initial application we have published results from a further two years follow up of INTERVAL participants (Kaptoge, Lancet Haematol 2019). The findings have provided policy-makers with two key evidence-based options to meet blood supply needs; the use of frequent reminders to help donors keep appointments and shorter inter-donation intervals than are now standard. This study has also quantified the extent of iron depletion within four years of repeated donation, thus informing safety guidelines. Results have been disseminated to the public and donors through study website, newsletters, national and international press release and media interviews.

The COMPARE study was designed to evaluate the optimum method to measure haemoglobin levels in ~30,000 potential whole blood donors in advance of each donation. Results from COMPARE have led to an evidence-based decision by NHSBT to replace the copper sulphate-based haemoglobin screening with finger-prick haemoglobin testing, thereby preventing ~350 female donors being inappropriately bled each day in England (ie, female donors with haemoglobin levels <12.5g/dL, the minimum threshold mandated by regulators). The COMPARE manuscript is due to be submitted for publication by the end of 2019.