Health Data Research Service investment success for Scotland

1DQ: Primary care data quality toolset to improve observability

Led by: DataLoch. Partner organisations: RDS, Grampian Data Safe Haven (DaSH), Health Informatics Centre (HIC), West of Scotland Safe Haven

Focusing on primary care data quality, the 1DQ project aims to create a toolkit to address challenges in understanding longitudinal GP data and make GP data as useful as possible for researchers. The project will build on existing solutions developed by DataLoch, expanding them to address challenges such as lack of standardisation, data loss during patient movement and practice changes and migrations.

The project will complement the SPARK project, which also seeks to improve access to GP data.

ANCHOR: Agreed National framework for Consistent Handling of Opportunities and Risks

Led by: RDS. Partner organisations: DaSH, DataLoch, eDRIS, HIC, West of Scotland Safe Haven

The ANCHOR project will develop a framework to help assess the risks and mitigations of research projects, enabling a consistent approach across different approvals pathways for securely accessing data for research.

BEACON: Building Enhanced metadata Access, Consistency and Observability across the Network

Led by: RDS. Partner organisations: DaSH, DataLoch, eDRIS, HIC, West of Scotland Safe Haven, Metadataworks

This project will focus on standardisation and interoperability of metadata across the Scottish Safe Haven Network. BEACON aims to improve the consistency of information published on the RDS metadata catalogue, where it will then be pushed to UK-wide platforms such as the HDR UK Gateway.

This means that metadata across Scotland will meet the FAIR principles: Findable, Accessible, Interoperable and Reusable, making it more valuable for researchers looking to discover appropriate datasets for their research projects.

HARMONY: Harmonisation and standardisation of dermatology data and its governance for federated machine learning

Led by: HIC. Partner organisations: RDS, Thames Valley and Surrey Secure Data Environment, South-West SDE, Wessex Secure Data Environment, NHS Tayside, Oxford University Hospitals NHS Trust, Royal Cornwall Hospitals NHS Trust, British Association of Dermatologists

Working in collaboration with partners across the UK, the HARMONY project will use dermatology data to deliver data harmonisation and standardisation solutions, for use in federated machine learning (FML). The project aims to establish the data and governance foundations necessary for an ongoing cross-UK FML partnership network.

SCALA: Scottish Consolidated Asset for Laboratory Analytics

Led by: DataLoch. Partner organisations: RDS, DaSH, eDRIS, HIC, West of Scotland Safe Haven

SCALA will build on proof-of-concept work around harmonising laboratory data across the Scottish Regional Safe Havens. It aims to produce an on-demand laboratory dataset, created and maintained from secure, detailed regional data.

SCAMPI: Scalable architecture for medical image processing and management

Led by: HIC. Partner organisations: RDS, DaSH, DataLoch, eDRIS, West of Scotland Safe Haven, SAIL Databank, Eastern SDE, EPCC

Medical imaging data is a valuable resource for research, but it requires significant storage capacity and fast processing.

SCAMPI builds on the Medical Research Council-funded projects PICTURES and Scottish Medical Imaging (SMI), which developed processes for managing and archiving medical imaging for research. The project will deliver a standalone, open-source solution for medical imaging processing and management, with enhanced metadata querying, faster processing and lower storage requirements.

SPARK: Scotland's Primary Care Data Architecture for Research and Knowledge

Led by: RDS. Partner organisations: DaSH, DataLoch, eDRIS, HIC, West of Scotland Safe Haven, Public Services Delivery Scotland

GP data is a crucial resource for researchers to understand some of the country’s most urgent health concerns. The SPARK project aims to support the technical and governance aspects of making GP data accessible for research through the Scottish Safe Havens and ensure that Scotland-wide solutions consider regional needs and expertise.

SPARK will achieve this by funding two Scottish Safe Havens to deploy the existing technical architecture solution developed by DataLoch, engaging with GP practices and defining what’s needed to move from a regional approach to a national-regional system for making GP data available for research.

The project will complement the 1DQ project, which also seeks to improve access to GP data.

STAR-Lab: Safe and Trustworthy Access to Pathology and Microbiology Free Text

Led by: DataLoch. Partner organisations: RDS, DaSH, HIC, West of Scotland Safe Haven

Free-text information (such as notes typed into systems by clinicians and technicians) can give vital contextual information when used alongside other health data.

The STAR-Lab project will use laboratory pathology and microbiology data to develop a standardised approach for de-identifying clinical free-text information, making it available for secure research.

It will evaluate Large Language Model (LLM)-based de-identification solutions across Scottish Regional Safe Havens, aiming to develop a solution that is accurate, scalable and capable of preserving research utility while maintaining the safety and security of data.

TETHER: Trusted Environment Tooling for Health-data Exchange and Record-linkage

Led by: DataLoch. Partner organisations: RDS, DaSH, eDRIS, HIC, West of Scotland Safe Haven

The TETHER project will develop improvements in three key areas of making data available for research. Firstly, it aims to develop a reusable process for data ingest and monitoring data quality.

Secondly, it will pilot a process through which data can flow from Public Health Scotland to one Regional Safe Haven, with the aim to assess whether this could be replicated in other Safe Havens.

Finally, it will develop a collaborative process to improve data indexing and data linkage across the Scottish Safe Haven Network that is adaptable when individuals move between different NHS boards and TRE boundaries.