Trusted Research Environments (TREs), also referred to as “Data Safe Havens” or “Secure Data Environments,” are extremely secure computing environments that allow authorised researchers, scientists, and other experts remote access to access valuable data via a stringent approval process for research in the public interest and to better people’s lives.

Health Dara Research UK (HDRUK) are committed to enabling the trustworthy use of health data for research – find out more on the HDRUK website.

DATAMIND TRE supports a broad range of studies and research questions, including social determinants of mental health, multimodal biomarker discovery, adolescent mental health trajectories, treatment effectiveness, mechanistic studies that explore underlying biological, psychological and social processes, as well as research on early intervention, prevention, and treatment. Research can draw on routinely collected electronic health records or linked cohort data.

Approved researchers will be able to access, analyse, and link health and administrative data in a safeguarded setting, without compromising public trust or data security.

Please explore our ‘Training and Support’ section for more information on required and recommended special training to gain approval to use the DATAMIND TRE.

Your research project will also need to be approved before you gain access to the platform. All applications are reviewed to ensure they meet legal, ethical, and public-benefit requirements. For more information visit our Apply for Access page.

Yes. All users must complete the training pathway listed below, before accessing DATAMIND’s secure research environment. This ensures researchers can use the platform confidently and securely.

The training pathway includes the following steps:

1. Good Research Practice (Mandatory for all)

Complete the Medical Research Council (MRC) Good Research Practice e-Learning module and quiz:

MRC Good Research Practice module and quiz

This MRC-developed module tests your understanding of ethical and responsible research. It includes 10 randomly generated questions with a 70% pass mark. You can retake it until you pass and obtain a completion certificate.

Note: Completion is required before receiving your YubiKey – a secure, physical device needed for two-factor authentication and access to the DATAMIND TRE environment.

2. ONS Data Access (Mandatory – if applicable)

If your project involves access to the Office for National Statistics (ONS) data through DATAMIND, you must register via the ONS Research Accreditation Service:

Apply for ONS Researcher Accreditation

3. Introduction to the Gateway (Recommended)

Complete the Datacise Open Learning Introduction to the SAIL Gateway module:

Introduction to the SAIL Gateway

If you’re unfamiliar with Trusted Research Environments (TREs) or virtual desktops, this optional course walks you through what the DATAMIND Gateway looks like, how it functions, and what to expect during your first login session.

If you’re unsure where to begin or have any questions you can also email our team on DATAMINDTRE@swansea.ac.uk and we’ll be happy to guide you through the process.

DATAMIND TRE includes both de-identified datasets and where there is consent and a lawful basis, identifiable data. Data types include electronic health records (EHRs), education data, census data, genomic and molecular data, neuroimaging, social care data, wearables and survey responses.

These are stored in anonymised or pseudonymised form and always handled under strict safeguards. This ensures researchers never have direct access to identifiable information.

Identifiable data undergoes a split-file process and de-identification, often using NHS third-party services, before entering the research environment. The DATAMIND TRE follows the same anonymisation pathway as the SAIL Databank, which is explained in detail on SAIL’s Privacy by Design page.

We accept identifiable, pseudonymised, and fully anonymised data depending on governance.

This information will be provided within the Data Deposit Form.

Data must be in a machine-readable, structured format such as CSV, Excel, SQL, JSON, or DICOM. This includes digital images (e.g., MRI, CT, or microscopy files), genomic and other omics datasets, and data from wearables or sensors. Paper-based or unstructured files (e.g., PDFs or handwritten notes) will need to be converted before submission.

DATAMIND supports secure transfers via encrypted upload portals, ingestion tools, or other agreed routes depending on file type and sensitivity.

Visit our Apply for Access page, where you can follow our step-by-step process, designed to make applying straightforward.

You’ll be asked to provide key details such as your project aims, proposed datasets, ethical approval status (if applicable) and any technical or linkage requirements.

If you’re unsure where to begin or have any questions you can also email our team on DATAMINDTRE@swansea.ac.uk and we’ll be happy to guide you through the process.

Yes, changes to timelines, methodology, or data needs can be discussed with our team. In some cases, these changes may require updated governance approvals.

Data is hosted on ISO27001-certified servers at Swansea University which includes the use of encryption. Researchers use a secure virtual desktop with two-factor authentication to access the data.

Datasets are normally retained for the lifetime of the approved project. With the permission of the data owner, or if required by funders, data may be held for longer periods. In some cases, anonymised data can be stored for extended durations.

As a rule, data will be handled according to the Data Sharing Agreement signed as part of the onboarding process and in line with relevant DATAMIND TRE policies.

Users can explore and analyse datasets securely within the TRE using pre-installed tools such as R, Python, SQL, SAS, SPSS, Stata, HAIL, and Plink, as well as make use of advanced analytics, including AI, Machine Learning, and Natural Language Processing. Additional tools may be available upon request.

Depending on the specifics of your project, research teams may be able to collaborate in shared project areas regardless of geographic location – but this may not be applicable to all projects. If this is something you would like to explore in advance please email our team on DATAMINDTRE@swansea.ac.uk and we’ll be able to advise you further.

As early as possible, ideally during project planning so we can advise on datasets, governance, and linkage.

Briefly outline your aims, dataset types, timelines, funding source, and tech team. For more information, see the full outline on our Deposit your Data page.

Complete the Data Deposit Form and our team will be in touch. Alternatively, you can contact us for more information.

If governance and opt-out checks are complete, de-identified data can be transferred directly to the DATAMIND Hub. Please follow the same steps outlined on our Deposit your Data page.

We apply strict security controls and governance processes to safeguard all data. Personal identifiers are removed or replaced with unique codes before researchers access the data. Any identifiable information is stored separately under enhanced access controls. All access requests are reviewed and approved through formal governance procedures and all research outputs are checked before release to ensure they meet privacy standards.

Yes. All research using DATAMIND data must have appropriate ethical and governance approvals in place before access is granted. Researchers should provide evidence of relevant ethics committee approval or confirmation that their project is covered by existing permissions.

Yes. All projects requesting access to DATAMIND data are reviewed by the Independent Governance Review Panel (IGRP). The IGRP ensures that each project meets ethical, legal, and scientific standards and that data use aligns with participant consent and public benefit The IGRP does not grant ethical approval.

Data providers sign a DATAMIND Data Sharing Agreement confirming roles and the lawful basis for sharing with operational compliance supported by DATAMIND programme team.

Data handling is governed under the Five Safes model:

Safe Projects: The Information Governance Review Panel (IGRP) reviews/approves projects to ensure legal, ethical and public‑benefit use.
Safe People: Only trained and accredited users are granted access; researchers sign user terms and project agreements.
Safe Data: Identifiers are removed or coded and data minimisation is applied; de‑identification/split‑file methods are used where applicable.
Safe Settings: Access happens only within the secure DATAMIND TRE (no raw downloads) with logging and 2‑factor authentication.
Safe Outputs: All results are subject to disclosure control (output checking) before release to prevent re‑identification.

All data use must have a valid legal basis under UK GDPR. For identifiable data, this may include informed consent from participants or a public interest basis (for example, NHS or university research). Pseudonymised datasets may rely on other lawful bases such as Public Task or Legitimate Interests, depending on the context. You must provide documentation to demonstrate the legal basis for your project.

All the information on the TRE is available on our website.

Reach out to our team at DATAMINDTRE@swansea.ac.uk. We’re here to help.