Present

Committee Members 

Paul Boyle (Chair)

Mark Brewin (HM Revenue & Customs)

Michael Chapman (NHS Digital)

Chris Dibben (Independent member)

Tricia Dodd (Independent member)

Emma Gordon (UK Research and Innovation)

Roger Halliday (Scottish Government)

Sarah Henry (ONS)

Alex Singleton (Independent Member)

Misa Tanaka (Independent member)

Advisors

Michael Cottam (Legal Services, ONS/UK Statistics Authority)

Andy Wall (Chief Security Officer, ONS)

Simon Whitworth (Data Ethics, UK Statistics Authority)

Ross Young (Data Protection Officer, UK Statistics Authority)

UK Statistics Authority

Sophie Gwillym

Lily O’Flynn

Grazia Ragone

In Attendance    

Elizabeth Fearon (London School of Hygiene and Tropical Medicine) for item 2

Deirdre Hollingsworth (University of Oxford) for item 2

Joe Edwards (Security and Information Management, ONS) for item 6

Apologies

Ann Berrington (Independent)

Stephanie Howarth (Welsh Government)

Paul Lodge (DWP)

Felix Ritchie (UWE Bristol), for item 6

      1.  Introductions  

1.1 The Chair welcomed the members to the twenty-third meeting of the Research

Accreditation Panel (RAP).

1.2 Members approved the minutes of the meeting held on 9 September 2021.

1.3 Sophie Gwillym updated the meeting with progress on actions from previous All actions were complete or otherwise in progress.

1.4 The Chair welcomed and introduced new RAP member Michael Chapman to the Michael is the Director of Research and Clinical Trials at NHS Digital.

     2.   Investigating epidemiological insights for the COVID-19 infection across the UK: Update from research team

2.1 Dr Elizabeth Fearon (London School of Hygiene and Tropical Medicine) and Professor Deirdre Hollingsworth (University of Oxford) are currently working on the DEA accredited research project: Investigating epidemiological insights for the COVID-19 infection across the

2.2 The aim of the project is to use the knowledge and expertise of the UK academic community to strengthen knowledge and understanding of the COVID-19 epidemic and provide insight to national and local decision-makers, local health protection teams and This project looks to enable academic partners to access data required to fulfil urgent research needs.

2.3 Dr Fearon and Professor Hollingsworth told the Panel how this research was used to support policy and decision-making throughout the pandemic period, including reporting to the Scientific Advisory Group for Emergencies (SAGE).

2.4 While the data accessed in this project has been made available using a number of legal gateways, the DEA has enabled administrative data such as the Annual Population Survey and the Business Impact of Covid-19 Survey, to be used alongside health data accessed via alternative legal means for this project.

2.5 The Panel welcomed this as a fantastic example of research in the public good which has been enabled under the DEA research powers. The RAP was interested in understanding the work the accredited project has enabled, and the public good benefits this has In particular, the project has provided an evidence base for public policy decision making and public service delivery in the UK. The work can also be used to draw lessons on how and when Test, Trace and Isolate interventions can be most effectively used.

2.6 The Panel requested the researchers produce a blog post or paper on their experience accessing data and carrying out the research project, to showcase how data accessed under the DEA research powers can be used most effectively used.

ACTION: Researchers to work with ONS SRS to publicise the impacts of the DEA research powers to enable access to data for public good research, using their experience around their research project.

2.7 Dr Fearon and Professor Hollingsworth shared their experience of difficulties within the data access process, in particular the waiting time to get access to data remotely.

2.8 The RAP thanked the researchers for their feedback and confirmed to the researchers that they are committed to improving the data access process where they The RAP requested written feedback from the researchers on their experience accessing this data under significant COVID-related time pressures so that the RAP can understand where blockers in the data access process may lie, and work in partnership with accredited processors to resolve these blockers where possible. The Panel also wanted further information on the Assured Organisational Connectivity (AOC) process, which this research team used to allow for remote access to the Secure Research Service (SRS) for this project.

ACTION: The Secretariat to reach out to the researchers to discuss any difficulties the researchers may have encountered getting hold of data and report back to the RAP on how this may be improved at a future Research Accreditation Panel meeting.

ACTION: UKSA to talk with the SRS around researchers’ experiences of the AOC process currently in place and present to a future meeting.

      3.   Integrated Data Service: Research Governance Update

3.1 Lily O’Flynn presented a revised Digital Economy Act (DEA) project application process, which will be trialled during the testing phases of the Integrated Data Service (IDS). The IDS has the potential to maximise the use of the DEA Research powers, as the DEA will be the default legal gateway under which the IDS will provide research access to linked data controlled by multiple public authorities.

3.2 The application process, and accompanying form, has been designed for the IDS on the assumption that the IDS will become a DEA accredited processing environment in 2022. Once tested in the initial development stages of the IDS, the UK Statistics Authority plan to roll out this revised project accreditation application process to the network of DEA accredited processing environments to support more streamlined access to data under the DEA.

3.3 To streamline the data request process for the research community, the new application form has been designed to both uphold the legal requirements for access to data under the DEA and measure a research project’s compliance with the UK Statistics Authority’s (UKSA) ethical principles. An overview of the changes are:

i. The DEA research project accreditation application and the UKSA’s ethics self-assessment tool have been merged into one form, meaning the data ethics of a project is considered during the project accreditation process without researchers needing to complete an additional ethics assessment.

ii. All information collected that was duplicated within other areas of the data access process have been removed to improve

iii. Questions have been designed so that responses can be provided by researchers through predominantly multiple choice and tick box answers to ensure the UKSA and the RAP are getting the quality information they require to make an efficient accreditation decision on the research

3.4 This proposal is based on the further automation of the DEA project accreditation To mitigate against the potential risk of further extending the risk-based accreditation model that the RAP currently operates, whereby only those most risky projects come to the RAP for independent scrutiny, this process will be supported by a suite of proposed safeguards, including:

i. Continued output checks by the accredited processor to ensure project outputs meet accredited aims, meaning the researchers must provide clear and adequate information on their research aims when requesting access to data to ensure their outputs can be removed from the environment.

ii. An ongoing randomised audit function established by the UK Statistics Authority to ensure live research projects are operating within the remit of the accredited research project, as set out in the application, while live in the processing

3.5 UKSA suggested a revised way of assessing a project’s research methodology, whereby researchers that have been badged to a government analytical profession would not be required to provide an in-depth explanation of their research methodology, and instead provide a high-level overview of the methods to be used.

3.6 The RAP understood the rationale behind this revision, but made clear that, where possible, any streamlining of the research application process under the DEA should be opened out to the entire research community, to ensure the DEA process is improving research accreditation for all.

3.7 The RAP suggested the Secretariat reconsider the way in which research methodology information is collected under this revised application process to ensure that researchers can benefit from the expertise that the RAP can provide in terms of methodological advice and support, and to enable those projects that make use of potentially experimental methods to be identified and reviewed by the RAP where required.

ACTION: UKSA to share the proposed audit function for assessment of live DEA project’s compliance with their accreditation research project application at a future meeting of the Research Accreditation Panel.

3.8 The Panel noted they approved the process; however, the RAP noted that it would be useful to understand the extent to which data providers within other government departments are supportive of this process, and whether they feel it provides the assurance that they require to make their data available in safe and secure ways for public good research. Lily O’Flynn confirmed this engagement is part of the forward plan for the Secretariat.

ACTION: The Secretariat to engage with other government departments to understand the extent to which there is buy in for the revised DEA research project accreditation process and report back to the RAP at a future meeting on the findings of this work, and whether this process should be tweaked to further assure data providers of its robustness.

      4.   Research Accreditation: Progress from the last year

4.1 Sophie Gwillym (UKSA) presented the Panel with an update of the progress made in the research accreditation space within the last year, areas for improvement, and a strategy for continuing improvements and impact in the next year.

4.2 The introduction of the online platform system, which allows Panel Members to accredit projects online outside of sitting RAP meetings, and the Project Accreditation Tool (PAT), which allows projects to be accredited at official level, have resulted in a more efficient and faster accreditation process.

4.3 The RAP confirmed its support for the implementation of these developments which has created significant efficiencies over the last year, and thanked the Secretariat for the work that has supported these improvements.

4.4 However, the UKSA has identified a few areas for further improvement and has created a forward plan to understand how the RAP and the UKSA can continue to support, guide, and advise on improving access to secure data for public good This forward plan includes:

i. The Secretariat commencing a programme of work with DEA accredited processing environments to agree a regular flow of standardised management information reporting to allow the RAP to gain a full picture of the end-to-end data access process. This will allow the UKSA and the RAP to gain a better understanding of researchers’ experiences when gaining access to data, and support in providing clarity to researchers on what to expect when requesting access to data; and,

ii. The Secretariat commencing a programme of engagement with data providing government departments and DEA accredited processors to gain their feedback on the DEA research powers and continuing to improve communications about what data is available under the DEA. The UKSA plans to present their results of government engagement to the RAP by mid- 2022, which will provide the RAP with an overview of results from our engagement programme.

4.5 The Panel is supportive of the changes introduced in the last year and of the Secretariat’s plan of action to enhance engagement.

ACTION: UKSA to report back to RAP with a paper on engagement with other government departments by mid-2022

    5.   RAP Self-Assessment

5.1 Sophie Gwillym presented the RAP with the findings of the RAP Self- Assessment, which Panel members filled out in November to provide the Secretariat of an overview of their feelings towards the work of the Research Accreditation Panel. The Secretariat wanted to gain an understanding of areas for improvement and areas of success in the functioning of the RAP processes.

5.2 From the Self-Assessment it emerged that Panel members were positive about the roles and responsibilities of the RAP, clearly understand the RAP’s governance arrangements, felt that supporting documentation was provided in a timely manner and that the roles of the Chair and Secretariat were effectively undertaken.

5.3 However, according to the Self-Assessment, the Panel would like to see improvements in the following areas:

i. Keeping the mix of research experience of the RAP under review to ensure the RAP always has the required expertise to support the wide range of research projects the RAP provides accreditation of under the DEA;

ii. Ensuring it is clear to RAP members the areas of a research project application that the RAP is being asked to scrutinise to provide DEA accreditation when research projects are accessing data via multiple legal gateways;

iii. Implementing a strategy to better equip RAP members to assess processors’ accreditation reports to ensure adherence to ongoing best practice for accredited processing environments; and,

iv. Refocusing the RAP’s expertise on assessing the public good of projects, rather than a detailed overview of research methodologies.

5.4 The UKSA set out action plans to address the RAP’s requests, including investigating the feasibility of convening a RAP sub-committee to advise specifically on matters relating to processor accreditation. The Panel was content with the suggested approach and requested further information on the potential for this sub-committee for processor accreditation at a future RAP meeting.

ACTION: The Secretariat to produce a paper on the feasibility of a RAP sub- committee for processor accreditation.

ACTION: UKSA to act on the plans for improvement presented to the RAP following members’ feedback in the self-assessment.

4.5 The RAP praised the Secretariat for the fantastic job they do and are very supportive of suggested improvements they want to make in the future.

     6.   SRS Accreditation Report

6.1 Lily O’Flynn presented the Panel with an evidence report of the ONS SRS’s second DEA accredited processor annual review. The Accreditation teams recommended that the RAP continues ONS SRS’s accreditation under the DEA. Levels of compliance with the security and capability controls selected as part of this review are either ‘Good’ or have progressed to ‘Mature’.

6.2 The Panel agreed to validate the continued accreditation of ONS SRS for the provision of data under Chapter 5 of Part 5 of the Digital Economy Act, based on the evidence provided in the annual review report.

ACTION: The Secretariat to write to ONS SRS to confirm the continuation of accreditation under the Digital Economy Act, following the successful completion of this annual review.

    7.   Any Other Business

7.1 The Chair noted the ‘for information’ These included:

i. A paper outlining the rationale and proposed prioritisation for projects that were previously approved by the Microdata Release Panel (MRP) to be accredited under the Digital Economy Act 2017 (DEA). The Panel supported this approach.

ii. An update regarding long-term remote access arrangements to the Secure Research Service (SRS). Currently, remote access will be continued for six months and the Secretariat will keep the RAP informed if remote access to the SRS is discontinued.

iii. The usual reports of accreditation processes undertaken by the UKSA and overseen by the RAP.

ACTION: The Secretariat to update the RAP if remote access to the Secure Research Service (SRS) is discontinued.

7.2 The next RAP meeting is on 8 March 2022.