The publication of the UK Synthetic Data Community Group report, Perspectives & Recommendations on the Development of Synthetic Datasets in Trusted Research Environments, represents a significant step forward in the responsible use of synthetic data across the UK research landscape.
The UK Synthetic Data Community Group is co-chaired by SeRP Director Professor Simon Thompson, and led by Lewis Hotchkiss from the DPUK Data Portal. Funded by DARE UK, the report brings together insights and recommendations developed over the past year by experts working across synthetic data, Trusted Research Environments (TREs), and administrative data research.
Central to the report is the VSTAR Framework, a practical tool designed to support organisations in developing and deploying synthetic data in secure research settings. The framework addresses key considerations, including governance, public trust, disclosure control, and real-world implementation, helping ensure that synthetic data is used safely and effectively. Strong engagement at the report’s launch and showcase reflected a growing consensus around the importance of clear standards and shared principles for synthetic data.
These priorities closely align with ongoing work at the Secure eResearch Platform (SeRP), which is advancing synthetic data capabilities in collaboration with its partner tenants, Dementias Platform UK (DPUK) and the SAIL Databank. Together, the partners are investing in approaches that expand access to data while maintaining the high standards of security and privacy expected within TREs.
Synthetic data is artificially created data, typically generated using statistical or machine-learning models trained on real-world datasets. By learning the underlying statistical properties and relationships within the original data, these models can produce artificial data that are statistically similar without containing identifiable information about real individuals. This makes synthetic data particularly valuable in sensitive domains such as health research.
Within health and administrative data research, synthetic data has a range of practical applications. It can be used to support researcher and student training, allowing users to learn how to analyse complex datasets without navigating lengthy governance approval processes. Synthetic data also enables researchers to develop and test analytical algorithms and workflows while waiting for access to real data, improving efficiency and reducing project delays. In more advanced scenarios, it can support federated research where real data cannot be directly accessed or shared.
SeRP’s synthetic data collaborations have also delivered tangible outcomes, including the development of SynthOpt, an open-source Python package available through the DPUK Data Portal. SynthOpt allows users to generate different levels of synthetic data tailored to specific research needs, from training datasets to more analytically useful synthetic versions for method development.
The UK Synthetic Data Community Group report reinforces the importance of aligning technical innovation with robust governance and public trust.
Discover more about SeRP’s Synthetic Data advancements on our Innovation Lab.
New report delivers pathway for the responsible use of synthetic data.
