COVIDENCE/CORONAVIT - 1141 - 12/05/2021 16:46

NOV 2021

V1, Amendment to correct charges. No other changes have been made to the application.

__________________________________________________________________________________

PURPOSE

The COVIDENCE UK Research Study is a longitudinal cohort study that was established to look into how circumstances of peoples’ lives might influence the risk of catching COVID-19, the severity of symptoms, speed of recovery and any longer-term effects on health. People aged 16 years and over from all parts of the UK are being asked to provide their informed consent and some baseline information about the circumstances of their lives and health using an online questionnaire. If consented, they are contacted once a month to check if they have developed symptoms of coronavirus infection or if they have attended a hospital for treatment. The data they provide will be linked to their health records (GP, Hospital, Covid-19 Test & Vaccine data, Deaths).

The COVIDENCE UK Research Study was developed in response to the novel coronavirus outbreak (COVID-19) and as such is not a continuation of previous work. All requested data will be used to answer the research questions stated below.

The aim of the project is to investigate how circumstances of peoples’ lives might influence the risk of catching COVID-19, the severity of symptoms, speed of recovery and any longer-term effects on health. Anticipated outcomes include:

- advance understanding of risk factors for coronavirus disease among UK adults;

- find out how quickly people recover from coronavirus disease and whether there are any long-term complications of this illness;

- evaluate the impact of coronavirus disease on the physical, mental and economic wellbeing of the UK population;

- establish a platform for future research on coronavirus disease in the UK.

Whilst interesting, determinants of self-isolation avoidance is not something QMUL and Swansea intend to investigate. By ‘economic wellbeing’ QMUL and Swansea are referring to an investigation into the impact of C19 on the UK economy.

This request will enable the addition of routinely collected COVID-19-specific data from primary and secondary care settings to the non-routinely collected, self-reported questionnaire and COVID-19 antibody test data from the COVIDENCE study (https://www.qmul.ac.uk/covidence/), in order to build a contiguous timeline of participant events during the pandemic, which will enable the analysis of study outcomes.

The purpose of this request is to link routinely collected COVID-19-specific data to non-routinely collected data from participants of COVIDENCE UK study in order to achieve the following detailed study objectives:

- Determine risk factors for incident COVID-19 and for adverse outcomes of COVID-19 in the UK population

- Characterise the natural history of COVID-19 in the UK population

- Evaluate the impact of COVID-19 on the physical and mental health of the UK population

- Provide a resource from which to identify potential participants for future clinical trials, and to use data collected in COVIDENCE UK as comparison or control data for trial participants who have been randomised to receive one or more interventions.

- Establish the prevalence and determinants of low vitamin D status in a subset of cohort participants

- Establish the prevalence and determinants of sero-positivity to SARS-CoV-2 infection in a subset of cohort participants

- Establish the prevalence and determinants of T cell sensitisation to SARS-CoV-2 infection in a subset of cohort participants

- Establish determinants of antibody responses to COVID-19 vaccination

- Establish determinants of cellular immune responses to COVID-19 vaccination

- Establish incidence of, and risk factors for, adverse reactions to COVID-19 vaccination

Queen Mary University of London (QMUL) and Swansea University (Secure Anonymised Information Linkage (SAIL)) are joint data controllers who are permitted to process data – QMUL is the hosting institute of COVIDENCE UK study; SAIL is a data safe haven centre within Swansea University. SAIL are data processor and will provide QMUL researchers secure access to the data. There are no honorary contracts, or committees/boards advising on this study. Other organisations are involved in the wider project but will not provide input on this aspect of the project or on the NHS Digital data request (see here for other organisations involved - https://www.qmul.ac.uk/covidence/).

The justification for inclusion of the requested datasets is they contain fields specific to COVID-19 and any other respiratory disease outcomes which will enable investigation of the above stated outcomes. Recruitment is taking place in Scotland, however only the IDs of participants residing in England will be sent to NHS-digital for linkage. A separate linkage application for Scottish participants will be made.

The justification for requesting record level, pseudonymised data is to be able to link respiratory health events to the COVIDENCE UK cohort at the participant level. Patient identifiable data (PID) will be provided, along with a unique study identifier, only for the purpose of linkage; medical records data associated with our unique study identifier and stripped of PID will then be returned back to us, in order to maintain participant anonymity.

The justification for a 24-month period of data capture is to coincide with the duration of sampling for COVIDENCE UK study, as this represents the most complete window of events on which to base our analysis.

The justification for data capture over the whole of England is due to the wide geographical recruitment of participants to COVIDENCE UK study.

The following efforts were made to ensure data minimisation, where possible: the selection of datasets are restricted to only those which capture respiratory health outcomes including COVID-19; variables within these datasets are restricted to only those we need to assess our study outcome measures; the period of data capture is restricted to the 24-month period of sampling for our study; we are able to filter records by those aged 16 years and older, as this was an inclusion criterion for participation in our study. We have not requested any elective or maternity episodes, and do not have a timeframe around an index event that would require us to capture data outside of our 24-month window.

Health records for COVIDENCE participants residing in England will be provisioned to SAIL for the project, including ingress of health records for COVIDENCE participants residing in Scotland, Wales and Northern Ireland, which will be obtained by separate applications to their respective Public Health Bodies. The project will from time to time request pseudonymised extracts of data to be removed from SAIL for various justified purposes, under the informed consent of the cohort which they have obtained. The purpose of pseudonymised extracts is to be able to follow-up with participants if necessary, and to extend an invitation for future studies, if eligible – participant informed consent has been provided for this. Funding for this extraction has been provided by BREATHE (the HDR UK respiratory health hub), which uses SAIL as its Trusted Research Environment.

The COVIDENCE UK PPI team is led by a prolific medical journalist and communications consultant who is a regular contributor to the UK National Press. Their wealth of experience as a communicator will maximise ability to engage with all sectors of the UK population to maximise recruitment of a representative population sample. She will be joined on the PPI team by representatives of the Asthma UK PPI group, who have also been instrumental in commenting on study design and development of participant-facing materials.

With regards to GDPR legal basis for processing, Article 6(1)(e) has been selected on the basis that the task is in the public’s interest and aims to answer specific high-priority research questions at a time of national crisis. Given that the requested health records include flows of sensitive information, Article 9(2)(j) has also been selected in accordance with Article 89(1), where the purpose is to conduct statistical scientific research. The case for this research being in the public interest is established through balancing the strong policy driver of improved understanding of the pandemic with consideration of potential harms to the patients/participants whose records are involved.

To provide clarity on the role of the academic collaborators in this project: they will not have access to any of the NHS Digital data from this linkage exercise. Institutes other than QMUL are collaborating institutes which due to their involvement in the project will be afforded the ability to disseminate findings via their own press offices, these organisations will not see any NHS Digital data.

NOV 2021

V1, Amendment to correct charges. No other changes have been made to the application.

__________________________________________________________________________________

The following data products are requested from NHS Digital:

COVID-19 Hospitalization in England Surveillance System

COVID-19 Second Generation Surveillance System

GPES Data for Pandemic Planning and Research (COVID-19)

COVID-19 Vaccination Status

Hospital Episode Statistics Admitted Patient Care

Medicines dispensed in Primary Care (NHSBSA data)

Electronic Prescribing and Medicines Administration (EPMA) data in Secondary Care for COVID-19

Civil Registration (Deaths) - Secondary Care Cut

HES:Civil Registration (Deaths) bridge

DATA FLOW

Data flow from SAIL Databank, Swansea University into NHS digital will contain the following data items to identify the cohort:

- Study ID

- Gender

- Date of Birth

- Postcode

Data flow out of NHS digital will be pseudonymised extracts of data products linked to the Study ID. Linked pseudonymised NHS Digital data will be securely transferred and stored in SAIL databank, Swansea. All data will be stored in the SAIL databank (Swansea) and secure remote access will be provided to the data controller at QMUL in order to work on the data within SAIL databank.

Data will be loaded to the SAIL Databank, at Swansea University, for analysis as per the above objectives. Data is hosted on a DB2 database system and provisioned to a dedicated project area and view for use only by the COVIDENCE/CORONAVIT team, who are researchers and substantive employees at QMUL. Data ingressed from NHS Digital will be matched to the COVIDENCE study using a common Study ID, and provided to the analysts in a pseudonymised format (encrypted identifiers, data suppression and masking to ensure minimal risk of identifiability).

With regard to the CORONAVIT trial (a clinical trial of vitamin D that is nested within the COVIDENCE UK study): this is supported by donations from charities, individual philanthropists and companies manufacturing and selling vitamin D supplements. None of these entities has any influence over the design and conduct of the trial, or on the decision to submit results for publication. Raw NHS Digital data will not be made available to any of these funders; neither will any party involved in the application receive any form of commercial benefit from the use of the data.

Querying of the data will take place primarily within the SAIL Databank virtual desktop infrastructure, managed by an ISO:27001, NHS DSP Toolkit, and UKSA Digital Economy Act accredited Information Security Management System and associated operational policies. Members of the COVIDENCE team at QMUL will be permitted to retrieve pseudonymised extracts of data from the SAIL environment due to their informed consent, which enables them to use relevant linked data extracts at their own institution (Queen Mary University of London), such uses will be for analysing data for re-contact as per the COVIDENCE and CORONAVit ethical approval and consents.

NOV 2021

V1, Amendment to correct charges. No other changes have been made to the application.

__________________________________________________________________________________

Findings of the research will be disseminated via publication in peer-reviewed journals, presentation at academic conferences and by communication with the general public in collaboration with press offices at Queen Mary University of London, King's College London, University of Edinburgh, Swansea University, Queen’s University Belfast and London School of Hygiene and Tropical Medicine. Institutes other than QMUL are collaborating institutes which due to their involvement in the project will be afforded the ability to disseminate findings via their own press offices, these organisations will not see any NHS Digital data.

All outputs from data processing will be in the form of tables and figures showing aggregate data with small numbers suppressed in line with the HES analysis guide i.e. no individual participant will be identified or identifiable from the material presented.

Intended modes of disseminating findings:

a. Reports : reports to be sent to study funders

b. Submission to peer reviewed journals: Manuscripts presenting results to be submitted to pre-print servers and to peer-reviewed journals e.g. The New England Journal of Medicine

c. Presentations: Oral and poster presentations of study findings to be made to conferences/learned societies e.g. the British Thoracic Society

d. Conferences: Oral and poster presentations of study findings to be made to national and international conferences

All outputs from data processing will be in the form of tables and figures showing aggregate data, with small number suppression in line with the HES analysis guide.

Statistical analysis plans have been developed to ensure that all specified outputs are analysed and disseminated, and no un-specified outputs are analysed and disseminated: Cox regression will be applied to identify factors independently associated with risk of developing incident COVID-19 disease and adverse outcomes; estimates of effect will be presented as adjusted hazard ratios with 95% confidence intervals. To explore participants’ characteristics at recruitment that are associated with COVID-19 outcomes (ventilator use, death, etc), generalised linear regression with a log link and binomial distribution will be used to estimate adjusted risk ratios with 95% confidence intervals. Multiple linear regression will be used to test factors associated with severity (peak symptom score during episode) and symptom duration (date of symptom resolution minus date of onset) possibly with log transformation as these outcomes are expected to have a skewed distribution. Ordinal regression will be utilised for analysis of individual items of symptom severity.

Study findings will be disseminated by submission of manuscript to journals. Webinars featuring plain language presentations will be prepared for study participants, who will be sent to these via email and via the study twitter account (@CovidenceUK).

Results will be shared with the rapid C19 oversight group, the science advisee on nutrition (SACN), public health England, the Chief Medical Officer and the Chief Scientific Officer.

The rapid C19 oversight group is a collaboration between: NHS England and NHS Improvement, the Medicines and Healthcare products Regulatory Agency (MHRA), the National Institute for Health Research (NIHR), the National Institute for Health and Care Excellence (NICE), Scottish Medicines Consortium (Healthcare Improvement Scotland), All Wales Therapeutics and Toxicology Centre, All Wales Medicines Strategy Group, and Northern Ireland Health and Social Care Board.

Study findings provided to SACN/Public Health England will be with a view to implement changes around nutritional guidelines

There will be no plans for commercial or other exploitation of results or other outputs.

Target data for major manuscript presenting findings on COVID risk factors is late 2021, this manuscript is written, provision of data via medical linkage is only rate limiter to submission for publication.

Webinars featuring plain language presentations are to be prepared for study participants, who will be sent to these via email and via the study twitter account. The study will liaise with the Queen Mary University of London press office to coordinate plain language dissemination of study findings to the general public, via print and broadcast media. All efforts will be made to ensure no individual (including an individual healthcare professional) can be identified (i.e. any published/shared results are statistically non-disclosive).

NOV 2021

V1, Amendment to correct charges. No other changes have been made to the application.

__________________________________________________________________________________

The first results for this study are expected within a month of the download of the NHS digital linked data. The COVIDENCE and CORONAVIT studies will hopefully lead to benefits for the following groups:

1. Benefits to the study participants

Individual study participants may benefit as a result of direct feedback of study results via the COVIDENCE UK monthly webinars (e.g. see https://www.youtube.com/watch?v=kg24a_etDIg for the latest). Linking health outcome information from NHS digital with risk factor information from their study questionnaires provides a powerful tool to maximise the value of the data they have contributed to the study, which will be used for statistical analyses to identify risk factors for COVID-19, factors influencing long-term outcomes of COVID-19, impacts of COVID-19 on physical and mental health, and factors influencing protectiveness of vaccination against SARS-CoV-2.

2. Benefits to the public

If the CORONAVIT trial shows a protective effect of vitamin D on COVID-19 morbidity and mortality, and/or a favourable effect on responses to SARS-CoV-2 vaccination, then the results are hoped to be rapidly fed back to CMOs and public health authorities for national and international dissemination. Vitamin D is safe and inexpensive, and has a high degree of acceptability, thus even modest benefits are likely to be of public health importance. If results of the trial are positive, data from the CORONAVIT trial may inform development of clear guidelines to boost vitamin D status of the UK population. Individual adults may therefore benefit from receiving the most appropriate care. This is likely to be generalisable to similar settings in other countries where vitamin D deficiency is prevalent.

Aside from the CORONAVIT trial (conducted in a subset of COVIDENCE UK participants only), the COVIDENCE UK study may generate multiple insights into risk factors for COVID-19, prognostic factors influencing any long-term outcomes of COVID-19, impact of COVID-19 on physical and mental health, and efficacy of vaccination against SARS-CoV-2. Linkage of COVID-19 data to vaccination status and post-vaccination antibody level may provide new insights into risk factors for vaccine non-response, which may aid in identifying sub-groups of the population who may be more in need of booster vaccines. All of these elements have potential to influence clinical care, public health planning and even economic planning so that it is as effective and targeted as possible, bringing multiple benefits not just to UK residents, but potentially to international populations affected by COVID-19 as well.

3. Benefits to the healthcare system

Analyses of linked datasets from COVIDENCE UK and CORONAVIT may provide public health authorities and clinicians with high quality data relating to risk factors for COVID-19, prognostic factors influencing any long-term outcomes of COVID-19, impact of COVID-19 on physical and mental health, and efficacy of vaccination against SARS-CoV-2. Two specific examples of how these data may benefit the healthcare system are as follows:

a) results from the CORONAVIT trial may inform healthcare providers about whether there are any benefits of vitamin D supplementation in terms of preventing COVID-19, reducing disease severity and/or augmenting immune responses to SARS-CoV-2 vaccination. Results may inform public health policy irrespective of whether they are positive or negative – in the former case they may prompt efforts to improve population vitamin D status, if null they may allow the CMO and other medical professionals to make a confident statement that vitamin D supplements do not play a role in host defence agaimst SARS-CoV-2

b) results from COVIDENCE post-vaccination studies may provide public health authorities with high quality data relating to relative protective efficacy of different SARS-CoV-2 vaccines, along with risk factors for non-response, which may allow booster vaccines to be targeted at groups that need them the most.

Not stated in the register.