Research Data Scotland: outline business case

Management Case

Introduction

This chapter details arrangements for the programme of work necessary to establish RDS, adopting a programme management approach. It sets out a delivery plan with clear milestones, documents project planning, governance structures, risk management, communications and stakeholder management, benefits realisation and assurance mechanisms. It demonstrates that robust arrangements are in place for the delivery, monitoring and evaluation of the scheme, including feedback into the organisation’s strategic planning cycle.

The FBC will include development of a more detailed delivery and operational plan and organisational strategy over the first two-to-three years.

Project Management, Governance, Roles and Responsibilities

Project management arrangements are led by the Data Sharing and Linkage Unit and oversee the activities of a core delivery team working on the implementation of RDS.

Programme governance is provided through the RDS Transition Board that meets regularly to review progress and provide advice and oversight for the overall direction of the project. The Board also oversee management of project risks and timescales.

A number of themed working groups have been established to work on the more substantive considerations for the delivery of RDS:

• Legal Working Group
• Financial Working Group
• Regional Safe Havens short life working group

Terms of Reference for each of the Groups have been agreed and documented.

The project reporting structure is shown in the following diagram.

Project roles and responsibilities are as follows:

• Senior Responsible Officer
• Project Executive Manager
• Project Lead
• Project Manager
• Lead Business Case
• Lead Funding and Finance
• Lead Service Design/Transformation Manager
• Lead Communications and Stakeholder Engagement
• Business Analysts

Project Plan

A project plan has been developed to capture tasks, deliverables and timescales across the working groups and project roles. This maps out the set of activities to reach a final draft OBC and provides for a stock-take and further gap analysis to identify decisions and further work to reach the FBC, the Minimum Viable Product (MVP) and the future Target Operating Model.

A summarised version of this is presented to the Transition Board at each meeting.

Communications and Stakeholder Engagement

There is high interest in RDS across several key stakeholder groups. Project communications and stakeholder engagement will be an important aspect of implementing RDS. The SRO, Chief Statistician Roger Halliday has already commenced engagements with key delivery partners and produced public-facing summaries highlighting the longer-term benefits of RDS and the opportunity it presents. The key stakeholders are as follows:

Table 18: Key RDS stakeholders

Stakeholders	Role in RDS	Remarks
Scottish Government	Funder of RDS. Governance role in RDS.	Sets national outcomes, and Scottish Digital Strategy. Existing SILC partner.
Public Health Scotland	Governance role in RDS.	New organisation set-up in April 2020.
National Service Scotland	Oversight role for eDRIS.	Lead agency for SILC, previous home of eDRIS.
National Records Scotland	Indexing.	Existing SILC partner.
eData Research and Innovation Service (eDRIS)	Data Linkage provider.	Existing SILC partner within PHS.
University of Edinburgh	Governance role in RDS.	Provides EPCC to SILC. Existing SILC partner.
Edinburgh Parallel Computing Centre (EPCC)	Computing infrastructure/capability.	Part of the national safe haven. Existing SILC partner.
Health Data Research UK	Funder of RDS.	Successor to FARR Inst. Which was formal SILC partner.
ADR-Scotland	Funder of RDS. Service recipient.	Existing SILC partner.
Chief Scientific Office		Funds five regional save havens as well as National Safe Haven.
Regional Safe Haven service partners	Regional indexing/linkage service and secure analytical platforms.
Academics, researchers, analysts	Users.	Existing users of SILC.
Data controllers	Providers of datasets.	Existing providers of datasets to SILC.
Data professionals: statisticians/data scientists	Users.	Existing users of SILC.

Change Management Plan

Transition from the current service model to a Minimum Viable Product for RDS will enable changes occurring in response to Covid-19 to be built into the service and launch in 2021. This will require careful planning and change management to ensure continuity of service delivery. A more detailed future service model for RDS will be developed at the FBC stage, allowing a detailed change management plan to be developed and agreed with relevant stakeholders.

Benefits Realisation Plan

A profile of benefits has been captured in the preparation of this OBC, set out in the Strategic Case under spending objectives. During development of the FBC, this will include a detailed plan of how each identified benefit will be measured so its realisation can then be monitored.

The benefits realisation plan will be aligned with the project delivery plan: it is currently profiled so that full values will not be realised until at least Year 5 of the programme.

Contract Management Plan

This will set out arrangements where contracts are required and specify the accompanying documentation, in line with the Commercial Case.

Risk Management and Risk Register

Risk Registers and an overarching Risk and Issues log have been developed for the RDS Programme. These are updated regularly and used to capture and manage risks across the work-streams and are shared with the Transition Board.

A full risk management plan will be presented in the FBC.

Assurance

Whilst there are governance structures in place to oversee the planning and delivery of RDS, it may be prudent to consider independent assurance and/or a gateway review process. SG guidance should be consulted and implemented as required.

Post project evaluation

In accordance with SG guidance, a post implementation review will be included in the overall project and undertaken approximately nine months after RDS is operational.

A more detailed specification for the type and nature of evaluation (economic, impact or process) to be conducted and the information requirements for this will be set out in the FBC, building on the benefits realisation plan and broadening out.

Contingency plans

In the event that this project fails, the current operating model can be continued through the existing operational arrangements. It is planned to initially offer RDS on a minimum viable product (MVP) basis. This means that, if required, there would be the opportunity to operate on this basis for a longer period than anticipated if that were required.

Successful delivery

As an organisation, RDS will service two separate groups of stakeholders through its work and operations; those who contribute data to the repository and those who wish to access data held in the repository (or available via the repository).

Both groups of stakeholders will interact with RDS through a defined service model. This model must meet the needs of both these groups and be fit for the future in terms of anticipating future demand (of both data controllers and end users) and building a service to support this. Innovation and performance improvement will be at the heart of how the service model matures.

RDS will need to keep pace with changes to standards and cyber security practice to continue to gain the support of data controllers. Similarly data curation methods and systems must strive to innovate to keep pace with the requirements of increasingly sophisticated analytical techniques applied by end-users. RDS will need to devote resource to this horizon-scanning and relationship-building work.

RDS will lead an ongoing programme of public engagement and public-facing communications that explains what RDS is about, what it aims to achieve and how and takes views from the public on these issues. It will be transparent with users and the public concerning how and why data is processed and regularly provide examples of impactful work where projects have informed and evidenced Scottish policy and public services.

Data Controllers, Information Governance – Privacy at the Centre

It is recognised that different data have varying levels of sensitivity - from those which can be made freely available (i.e. open data) through to those that must be carefully controlled to protect citizens’ privacy.

It is also recognised that the public hold subtly different views regarding the acceptability of people from the public/private/voluntary sectors accessing data for various uses – from development and improvement of public services through to the generation of private sector profit.

It is proposed that a mixed model of IG arrangements will be adopted under RDS across a broad spectrum of data. This approach will build in the flexibility to accommodate statutory requirements relating to some data holdings, while supporting a different approach for other types of less sensitive data.

IG under RDS will work toward ensuring robust, holistic and proportionate assessment of requests to access data and work with data controllers to explore how best this can be delivered within an environment where many different data sources are being combined for single analyses.

There will be the flexibility for data controllers to specify which data can be used, in which ways and by which types of users. In some cases this will involve data controllers depositing whole datasets into the secure computing environment, with the potential for permission for these to be used in specified ways. In other cases data controllers may wish to retain an approach wherein they retain source datasets within their local computing environments and only provision data after assessing each project. In still other situations, data controllers may agree to the creation of programme level datasets (involving linkage of more than one dataset) and sanction use for specified purposes.

RDS will offer all of these options, using standardised data sharing and data processing agreements. It will remain the responsibility of the data controllers to make the assessment concerning the risks and benefits of providing access to the data for specific requests.

Where data is made available for research, RDS will develop a data prospectus that outlines the datasets and what they contain (their metadata). This will allow researchers to search the prospectus to decide whether using this data would be worthwhile and if so to start to plan their research. Clear permissions around the use of such data will be agreed between RDS and data controllers and made available to researchers.

RDS Service Users, Data Catalogues

RDS will seek to engage with professionals from the:

• Public sector
• Academia
• Private sector
• Voluntary sector.

This will ensure that the potential of Scottish public sector data are released for public benefit. In order to develop viable and worthwhile projects users will need to understand what data are available and the likely logistics involved in provisioning that data.

RDS will develop a data prospectus that outlines the datasets that are available for research and linkage and what they contain (their metadata). This will allow researchers to search the prospectus to decide whether using this data would be worthwhile and if so to start to plan their research. RDS will commission work to set-up and maintain a web-based resource.

RDS Service Providers – Research Co-ordinators, Customer Support

RDS will commission research co-ordinator services to:

• Liaise with experts in the required datasets.
• Build knowledge of various datasets that are available (including meta data, coding structures etc.).
• Support study design and assessment of the logistics involved in creating the required dataset.
• Implement agreed information governance and data access processes – ensuring that agreements with data controllers are followed.
• Advise on the creation of study cohorts from single and multiple datasets
• Maintain appropriate project documentation (e.g. IG and researcher approvals).
• Regularly communicate across the technical parts of the data linkage service to drive progress (customer support, indexing and secure computing environment).
• Data provisioning and ad hoc/bespoke linkages.

Within the current data linkage system this function is currently provided by the eData Research and Innovation Service (eDRIS) within Public Health Scotland.

Current experience has demonstrated that not all users of RDS will be technically able to undertake data analysis and may wish to commission support from another organisation – this would be available (or sourced) through the RDS user service. RDS will therefore commission, and audit, this service currently provided by the eData Research and Innovation Service (eDRIS) within Public Health Scotland.

RDS Service Providers – Indexing Service

Identifiable personal information is not always required for the vast majority of research and innovation purposes. The current experience is that worthwhile projects often require large volumes of data and/or the joining of multiple datasets. However, this does not commonly require a concomitant need to provide personal identifiable information to researchers.

The new linkage model will continue to be based upon a technical approach which is known as the ‘separation of function’ model. Within this approach the linkage of individuals takes place separately, in a different organisation, from the joining of the information required for a research projects.

RDS will therefore commission, and audit, the services of the well-established de-identification and linkage service already in existence at National Records of Scotland.

This service replaces individual personal information with de-identified index numbers and allows datasets to be joined and provisioned without data custodians needing to exchange personal identifiable information. This approach will be central to the development of RDS and form the key underpinning for the de-identified data that is provisioned for data linkage projects.

RDS Service Providers – Secure Storage

In order to ensure public benefit is realised a secure, high-performance computing environment is required, which will have two key functions:

a) Secure storage of data for research and innovation

b) Provision analytic environments.

For this reason RDS will enforce contractual controls upon those accessing data in combination with state of the art computing security. To ensure public trust the secure computing environment will be subject to external, independent, scrutiny and be expected to achieve relevant kite marks relating to industry best practice.

The design of the IT architecture will be developed collaboratively between the IT service provider and RDS. This will ensure that the design takes account of multiple data controllers’ requirements.

RDS will ensure the design and its operating processes are subject to external scrutiny and regular compliance testing.