Data standards in Scotland's public sector - framework for action: consultation analysis

3 Proposals from the consultation activities

Section 3 collates ideas that were proposed during the workshops and matches them to the challenges identified in Section 2. The workshops were the primary source of proposed actions but are augmented with others made by survey respondents, and examples of action identified in the case studies. Not all proposals for action are included in Section 3, please see Annex Report A for further detail on the workshops, and Annex Reports B and C for the case studies and findings from the survey.

The challenges described in Section 2 are:

• a lack of strategic leadership on public sector data;
• a fragmented landscape;
• limited learning or sharing of best practice;
• skills gaps;
• data quality (and data maturity);
• legacy systems;
• barriers to data sharing;
• disinterest; and
• investment required.

The proposals collated below were not provided in direct response to a specific challenge, they have been allocated to a challenge subsequently during the reporting process. Some proposals will be capable of responding to more than one challenge.

Providing strategic leadership on public sector data

The work being undertaken by the Data Standards Team and the development of a new Digital Strategy for Scotland by the Scottish Government are significant steps toward addressing a lack of strategic leadership. Proposals for action put forward by consultees include:

• Put someone, or an organisation, in place to take overall control of data standards. For example:
  • Create a Data Conduct Authority to take responsibility for data regulation.
  • Appoint a data Tzar.
  • Pass the 'Data Scotland' Act.
  • Create a group that can review data standards through Data Dictionaries, seek improvement where appropriate and discuss the way forward.
• Develop a national Scottish vision for public sector data, including a clear ambition around interoperability.
  • Create a shared intent for collaboration and define the differences between what is required to make the intent happen locally versus making it happen nationally.
• Scottish Government participates as a full voting member of global standards organisations.
• Strengthen cross-Government working.
• Establish effective governance of data gathering and sharing.

Connecting the landscape

Proposals that respond to issues associated with reducing fragmentation in the public sector data landscape include:

• We need to deepen our understanding of the existing landscape. Audit/map existing data standards.
• Develop data custodian posts in public sector organisations to be a catalyst to create and support good governance, and to be a gateway between individual organisations and collaborative efforts.
• Create central infrastructure, for example:
  • Create national data registers.
  • Establish a properly resourced data warehouse for the public sector from where metrics can be calculated centrally and not left to local councils to complete and return.
  • Develop unique identifiers for all organisations in Scotland.
  • For specific topics, develop centralised datasets that Councils can access. This would support consistency and good governance.
• Ensure Scotland has comprehensive and good quality metadata within a dedicated metadata catalogue to facilitate data discovery reuse. For example:
  • Agree a metadata standard within Scotland, reusing what already exists, and adopt the standard into existing frameworks, for example Digital First.
  • Openly license metadata.
• Increase alignment. For example:
  • Identify must use data standards and define minimum standards.
  • Create a controlled register of vocabularies.
  • Agree consistent definitions and recording to allow disaggregation of datasets by protected characteristics, including impairment. Begin by auditing the different definitions already in use with the aim of integration.
  • Go FAIR.
  • Join – fully align with – the EU programme on digital standardisation.
  • Invest in technology systems so that different organisations can speak to each other (co design shared/compatible systems).
  • Seek opportunities to rationalise data standards.
• Be a pathfinder for AI based services.

Increasing learning and sharing of best practice

Proposals that focus on expanding and enhancing learning and the sharing of best practice include:

• Provide guidance and create a common approach to implementation. For example:
  • Customise the process so that whatever anyone needs is provided for them. Build in feedback loops and make this a learning system.
  • Develop guidance to support better quality specification of systems and datasets before implementation.
  • Develop guidelines to support understanding and use of data standards.
• Conduct a knowledge mapping exercise to identify use cases and help the audience to select appropriate standards.
• Create a learning culture. For example:
  • Establish a culture where different partners work together to help each other out.
  • Learn from existing examples of good practice. Some of what is already happening in Scotland (health and care) is ground-breaking.
  • Create a directory of who can be contacted to provide support for data sharing.
• Create a data standards community that can be tapped into to help organisations without enough capacity to extend the reach of their datasets and access other data sources.
• Create linkages between health and social care and other public service providers and third sector providers.

Developing skills

Proposals from the consultees that seek to support skills development include:

• Action needs to be taken to address the following skills gaps:
  • How to use and interpret data wisely.
  • Data capture, data entry and the application of data standards.
  • Ethical data management.
• Develop systems that are simple and fit for purpose to improve engagement and support for their use.
• Support the development of data analytical skills across partners to ensure evidence-based policy making is built on a shared capability to analyse and report on data sets.
• Educate citizens to expect data portability.
• Establish rules and norms for best practice use of data sets from an information governance perspective. This may require policing.
• Fix the skills gap in social care that means data is still being issued in Excel.
• Bring two disparate organisations together work to develop common data standards (for example, the Judiciary and Tourism Scotland). Pull people who do not yet have the skills into this process and ensure necessary skills are developed as part of it.

Improving data quality and data maturity

Proposals that seek to improve data quality and address the variation in data maturity across the Scottish public sector include:

• Be draconian about implementation of data standards before we worry about data sharing. If data sharing happens first and it is not good quality data, then criticism stops good things happening.
• Ensure data standards are mutually understood and consistently applied. For example:
  • Clearly define data to enable comparison and reuse.
  • Help people understand the need for data definitions.
  • Create unambiguous standards.
• With a huge range of organisations with varying data standards and levels of data maturity; it may be better to have standards as an element of a data maturity model which gives organisations a framework with key elements (standards, governance, quality, analytics, management etc.) to assess the maturity of its data. This will help organisations see where they need to focus effort to improve data maturity. This makes it about fit for purpose, managed data.

Legacy systems

Proposals that respond to issues associated with legacy systems include:

• Develop use cases to inform solutions.

Breaking down barriers to data sharing

Proposals from participants that aim to reduce barriers to data sharing include:

• All data standards are open by default. Closed practices (licensing, non-interoperability, cost) no longer exist.
• Create a common approach. For example:
  • Establish commonly applied unique identifiers.
  • Establish unique referencing in data sets.
  • Draw up, agree, and work to guidelines that ensure consistency for shared and open data sets.
  • Start with statutory datasets and develop standards around those and then umbrella the standards across all sub-set datasets.
• Put the technical infrastructure in place to be able to share data.
• Ensure users know what data is out there.
• Ensure appropriate use of data.
  • Data security is the primary priority area of any form of data standards.
  • Scotland needs effective governance with the correct balance between data protection and data reuse/interoperability.
  • Give all professionals a level of permission about what they can and cannot access. Securing access to be a one off/on demand approach.
  • Reflect on [and consider how to ameliorate] conflict between guardians of data and demands for information
  • Create a secure approach that allays concerns about data sharing.

Increasing awareness and understanding

Proposals that focus on improving awareness and a better understanding of the value associated with reusable data and data standards include:

• Promotion of the value and need for reusable data and data standards, to Chief Executives, middle management, and operational staff. For example:
  • Share clear examples of how data standards create benefit.
  • Create a persuasive use case that demonstrates value and helps local organisations see value when contributing towards national datasets.
  • Secure cultural buy-in.
• Deploy both carrot and stick incentives:
  • Carrot: Demonstrate the benefits – and ease of change - of applying data standards; and
  • Stick: Mandatory requirements to apply data standards (e.g. INSPIRE). This requires policing or people do not do it.

Supporting investment

Proposals from consultees that seek to improve investment in data standards include:

• Reuse existing standards first before investing in new ones.
• Create a model for data standards and pilot the approach to evaluate success and build a case to encourage investment.
• Invest in definitions, infrastructure, and skills.
• Lower the cost of compliance where possible.
• Work out [and agree] funding arrangements to stay at the leading edge of the technology.
• Ring-fence investment for data standards in organisations.
• Increase awareness of the need for ongoing investment in good data and data standards.

Further proposals

During the allocation of actions to challenges in the reporting process, there were proposals made by consultees that focus on different ambitions. These additional areas are:

• Creating fit for purpose data standards; and
• Engaging the citizen.

These are both described below.

Creating fit for purpose data standards

In addition to proposals that respond to existing challenges, there were several proposals around the need to incorporate flexibility into the development of data standards so that the data standards are appropriate:

• Develop future focused data standards that do not limit opportunities for the present or for the future:
  • Constantly review tools and processes. Make the review process responsive to changing need and adapt standards in real time. Learn from the openEHR process of review.
  • Develop agility to make standards responsive (use AI).
  • Predict future standards requirements. Use AI alongside human perspectives to predict trends and changes required.
• Identify what we really want to use the data for. For example, is it to measure change, support implementation of measures, to predict need, support innovation, answer specific questions?
• Be prepared to stop collecting data that does not serve a need or cannot be objectively described.

Engaging the citizen

There were several discussions around the rights of the citizen to access or hold data that relates to them. This is a topic that goes beyond data standards but does have implications for data standards, the following comments relate to data standards.

• Develop a data app to let citizens access metadata catalogues to see what data is held on them or on wider society. Use this to drive up interest and participation in data standards.
• Ensure people can see or access the data that is held about them.
• Allow individuals to be able to identify themselves online through an attribute store.
• Make it a requirement that individuals give their consent for use of their data. Consent can be removed if data are not held or used appropriately.
• Engage widely with the population of Scotland around their health and social care data. Use CPPs to have this discussion.
• Include individuals where possible to develop systems that align with their needs.

The next section of the report brings forward key concepts from the proposals listed above and develops them into a proposed Framework for Action for data standards in Scotland's public sector.