We are delighted to share our most recent developments that improve our support for research, innovation, and service management projects with the aim to enhance health and care services.
Firstly, we have added several specialist datasets to our repository, expanding the opportunities for new projects. On a broader scale, our TransPECT and HEAL-Scot collaborations are developing new guidance and tools that will enhance research possibilities across the UK. You can also discover more about our HERON-UK collaboration alongside the Edinburgh Cancer Informatics team.
In the remaining part of this update, you can: learn about the experiences of two recently graduated PhD students whose projects relied on NHS Lothian data; see how we are supporting the Medical Physics team in NHS Lothian with their annual CT scanner audit; and find out what members of the public around the UK think about the possibility of Large Language Models being used to de-identify free-text data.
Please take a moment to enjoy our latest updates and email dataloch@ed.ac.uk if you are interested in more information or wish to start project discussions.
Latest data updates
As part of our ongoing efforts to expand the opportunities for innovative research, we have incorporated further datasets within our repository covering a variety of specialist areas. National Early Warning Score 2 (NEWS2) collates key observation data collated when patients are treated in Emergency Departments, while 4AT delirium questionnaire scores help clinical teams assess whether their patients may have delirium. Other new data include weight measurements, radiology orders, and TrakCare appointments.
Data Update – September 2026
Advancing opportunities for research: HEAL-Scot and TransPECT projects
We are collaborating on two unique projects that will enhance the possibilities for novel research conducted through Trusted Research Environments (TREs) in the UK.
TransPECT will develop methods to assess the confidentiality risks posed by new Artificial Intelligence models developed using sensitive research data, while HEAL-Scot will produce a blueprint for TRE operators to more efficiently and robustly link health with other public-sector data.
Introducing HEAL-Scot and TransPECT
HERON-UK network: supporting reproducible research
Alongside Edinburgh Cancer Informatics, we are participating in HERON-UK: the HEalth data Research OMOP Network. This collaboration enables secure, transparent, and reproducible health data research by supporting research using routinely collected healthcare data from over 50 million patients across seven data partners in the UK.
In our first project, we focused on the variability in antibiotic use across the nation and the implications for national guidance. This project also served as a proof-of-concept in analysing OMOP data across a federated network.
HERON-UK network progress and future
Insights from PhD researchers
We have interviewed two PhD researchers to learn more about their experiences of using the DataLoch service and to find out more about what they discovered through their DataLoch projects. Interestingly, they both gave the same advice to other PhD researchers: spend some time exploring and understanding your data extract to see if everything is as you expect, before diving into the analysis process!
PhD researcher perspective – Dr Rose Penfold
PhD researcher perspective – Dr Konstantin Georgiev
Collaborating with NHS Lothian Medical Physics
Every year, the Medical Physics team in NHS Lothian perform an audit to ensure their scanners are working correctly. Optimising the radiation level of their scanners is part of this process. We are currently collaborating with the Medical Physics team to support the use of real-world patient data for their optimisation process.
Optimising CT scanning in NHS Lothian
Public Perspectives: de-identification using Large Language Models
Our STAR-TRE project aims to develop a toolkit to support secure research access to de-identified free-text data. As a critical part of this project, Ipsos have delivered a comprehensive report detailing public views on the possible use of Large Language Models (LLMs) to identify privacy risks in free-text.
Among other highlights, participants emphasised how LLMs could quickly identify privacy risks, and therefore make data available for research more swiftly. However, they also raised concerns around LLM accuracy and how to handle the balance of minimising privacy risks with the value of data made available for research.
STAR-TRE deliberative workshops overview
To find out more about any of our developments, or to open discussions around possible projects, please email dataloch@ed.ac.uk.
