Launching the Barts Health Data Platform

The healthcare landscape is undergoing a profound transformation with the integration of advanced data science and AI technologies. Leading this change is Barts Health NHS Trust, one of the largest in the UK, which has launched the Barts Health Data Platform (BHDP) in April 2025. This innovative Secure Data Environment (SDE) service provides access to over 2.5 million anonymised patient records, offering unprecedented opportunities for research and clinical insights.  

Sven Bunn, Life Sciences Programme Director at Barts Life Sciences, discusses the critical role of SDEs, the opportunities and challenges they bring, and how Barts Health is setting the standard in secure, responsible data use. 

Note: The views expressed herein are solely those of the author. 

Can you share what the launch of the secure data platform means for Barts Life Sciences and its partners, and what excites you most about this initiative?  

It’s been a longstanding ambition of ours to enable a much wider range of partners to access our NHS health data in a secure, responsible and accessible way. 

There are three key dimensions to this. The first is security – ensuring the data is used strictly for its authorised purposes and that people can trust it’s being handled appropriately. The second is technical usability – making the data accessible and ready for researchers without them needing to spend time cleaning it up to make it usable. The third is the governance framework – having clear guidelines on what the data is being used for, the purposes it serves, and robust safeguards and approvals to manage it responsibly. 

This platform brings all those elements together into a single, organised system. After years of work, we’ve reached the point where partners can access data through a smooth and cohesive process. Now, we can start using this wealth of health data for broader research purposes. 

What excites me most is the potential to address some of the toughest questions in healthcare: Why do people develop certain diseases? How can we detect them earlier? What’s the best way to treat and follow up with patients? And, ultimately, how can we deliver better outcomes? Secure access to these large-scale datasets will be key to answering these questions and transforming healthcare on a much broader scale. 

This platform will bring together all of the Trust’s data in one secure environment. From your perspective, what are the most significant benefits this will bring to both internal and external researchers? 

Fundamentally, it’s about the breadth, depth, and quality of the data. When we say 2.5 million records, that reflects the rough catchment population of Barts Health, but the reality is far broader. Our services operate on a national and international scale, which means our data encompasses a wide range of populations. It’s also longitudinal, covering at least 12 years of history. This allows researchers to track how different patient populations respond to various interventions and understand their long-term outcomes. 

Another key strength is diversity. Our data reflects a wide range of social, economic and ethnic characteristics, making it truly representative. It provides insights into a broad and varied population, rather than being limited to a single demographic. 

Then there’s the level of detail. We’ve integrated imaging data, pathology results, genomic information, cancer treatments, and much more. This granularity sets our dataset apart from others, such as national or regional datasets, which often cover larger populations but lack this depth and richness. 

Finally, one of the most practical benefits is how the data is organised. Historically, researchers would spend significant time ‘data wrangling’, cleaning and formatting datasets to make them usable. We’ve done that work for them, making the data much more accessible and usable for research. That, combined with the richness and diversity of what we offer, is a game changer for both internal and external researchers. 

So, what are the primary benefits of using big data in medical research, and what challenges do you anticipate? How does the secure data platform tackle these challenges? 

The benefits revolve around the ability to analyse large-scale, longitudinal datasets, both retrospectively and prospectively. This allows us to see the impact of specific treatments or interventions over time and explore how factors like age, ethnicity, or socioeconomic background influence disease susceptibility or treatment response. With the scale, breadth, and depth of the data, we can uncover patterns that smaller datasets simply cannot reveal. 

One of the most exciting aspects is the potential for precision medicine. The platform enables us to identify highly specific groups of patients based on characteristics such as age, gender, ethnicity, and disease type. This level of granularity allows us to study how these groups respond to certain treatments and to fine-tune future interventions based on those insights. 

However, this doesn’t come without challenges. Data quality is a significant one – ensuring clinical data is recorded accurately and consistently is a work in progress, requiring ongoing engagement with clinicians. Another major challenge lies in public trust. There’s often concern around how health data is used, as we’ve seen with some previous national initiatives. 

So, to tackle these challenges, we’ve developed a robust patient and public engagement programme. It’s about being transparent – explaining what we’re doing, why we’re doing it, and demonstrating the tangible benefits that emerge from this work. Building trust is essential, and by focusing on communication and outcomes, we’re aiming to be proactive in our approach. 

In your view, what are the emerging trends in data science that will significantly impact the future of precision medicine, and how is Barts positioning itself to be at the forefront of these advances? 

AI is clearly the biggest thing right now. I think there are three broad categories of what’s broadly called AI. First, AI-driven data analytics, which can help predict and anticipate patient needs based on a set of characteristics, such as the likelihood of GP visits or hospital admissions, enabling better care planning. The second category is using AI to identify previously unrealised connections. For example, we discovered through our research that a drug commonly used as a standard treatment for heart disease is significantly less effective in South Asian populations compared to white populations. Identifying these types of differences is crucial because it allows us to tailor treatments more precisely. 

Finally, large language models, like ChatGPT, are already being adapted for clinical use. While they’re not yet suited for medical applications, there’s ongoing work to fine-tune them using medical data to provide more accurate insights. 

So, I do see in the next few years, the combination of all three of those areas driving significant change in the way that clinical healthcare practice is delivered. We’ll reach a point where we have a much clearer understanding of the risks people face, and once they develop an illness, we’ll be better equipped to predict its progression. We’ll also have more precise information on how individuals will respond to treatments and the best ways to approach their care. 

Right now, AI exists as a support tool, providing additional information that clinicians can use to make decisions, while they remain legally and ethically responsible for those choices. However, AI algorithms are advancing quickly, and they may soon be able to make decisions that are more accurate than decisions made by clinicians. This raises the question: should we allow AI to act autonomously? If AI consistently delivers better results, should it be used more widely? It’s similar to the way chess algorithms now outperform human players –something once considered impossible. If we reach that level in clinical decision-making, managing the shift will be a major challenge. 

There’s also the financial pressure, especially with the NHS. If AI can take over tasks currently done by clinicians, it’s likely there will be financial incentives to make that shift, raising both ethical and practical concerns. Balancing these factors, possibly within the next 5 to 10 years, will be crucial and complex. 

Data can drive research, but public trust is key. How can we ensure that people feel confident that their data is being used and protected responsibly? 

Transparency is essential. We need to be clear about what we’re doing, why we’re doing it, and how people can engage with and understand the process. This means not only explaining the safeguards in place to protect data but also communicating the tangible benefits of using this information for research and patient care. 

There are two main risks we need to address. First, people may be concerned about how their information is being used or worry that they lack control over it. To build trust, we need to explain exactly how data is anonymised so that individuals cannot be identified and outline the strict controls we have in place. 

Second, it’s about emphasising the benefits. Most people expect health data to be used in an aggregated form to improve healthcare services – whether that’s discovering better treatments, enhancing diagnosis, or providing more personalised care. By clearly showing the positive outcomes and the role this data plays in advancing healthcare, we can build greater confidence and understanding. 

Can you provide an example of a successful pilot project that has used the secure data platform? What were the key outcomes and impacts of this project? 

A good example comes from when we were still in the prototype phase of the data platform. We developed a tool that identifies undiagnosed diabetes and, more importantly, its complications – particularly vascular issues that can lead to foot and lower limb amputations. 

Often, people visit their GP or hospital with symptoms that might initially be linked to an existing condition or attributed to a different one. By analysing their records holistically, the tool can identify factors that may suggest an undiagnosed complication of diabetes and prompt early intervention. 

As a result, we’ve identified hundreds of people with previously undiagnosed complications. These patients are now receiving the appropriate care, which is already leading to significant outcomes: fewer amputations, reduced disease progression, and overall better management of their health. 

That’s a really great example of the platform’s ability to offer insights from data that might otherwise remain fragmented, directly benefiting patients and paving the way for further research innovations. 

How do you anticipate the secure data platform will change the way care is delivered? What potential impacts could it have on patient outcomes? 

The secure data platform offers the opportunity for a highly data-driven and analytically guided approach to healthcare delivery. A first port of call is the use of AI algorithms to identify risks and treatment options early in the patient journey. By providing clinicians with these insights upfront, they can focus on confirming diagnoses and making swift, informed decisions, helping them identify issues and start appropriate treatments much sooner. 

This speed translates to practical benefits, such as reduced waiting times in emergency departments or for specialist treatments, and earlier intervention, which can help prevent complications or disease progression. The result is better-managed health conditions and a healthier population overall. 

Beyond immediate care improvements, the platform has vast potential for driving innovation. It supports the development of precision medicine tools, identifying which treatments work best for specific patient groups and paving the way for advanced therapies like cell and gene therapies or immunotherapies. These approaches are particularly promising for diseases like certain cancers that are currently complex and life-limiting. 

Looking ahead, what is your vision for the future of the secure data platform? How do you see it evolving? 

Our vision is to expand the platform’s reach and capabilities to create a more comprehensive and valuable resource. This includes bringing in more partners, starting with those in North East London, and potentially extending across London and beyond. The more data we can incorporate, the better equipped we are to achieve both depth and breadth of information, which are essential for driving insights. 

We also aim to enhance the platform’s analytical services significantly. While the data resources are vast, our current capacity for skilled data science and analytics inevitably has its limits. By growing these teams, we can collaborate even more effectively with researchers and industry partners, unlocking greater potential from the data we already have. 

Ultimately, our goal is to continue innovating and driving meaningful research and clinical advancements that improve patient outcomes and transform healthcare at scale. 

About the author  

Sven is the Life Sciences Programme Director for Barts Health NHS Trust and Queen Mary University of London. He is responsible for the development and delivery of the Barts Life Sciences programme, a set of research, education and commercial development activities which aims to accelerate the adoption of health care innovation. In this role he works with a wide range of government, commercial and community partners. 

Recent work includes the development of a new clinical research facility at the Royal London Hospital (opening 2025), new data asset and data science services at Barts, and the creation of life sciences skills and training programmes with local colleges in east London. 

Sven has held senior leadership positions in health care strategy, operational management and large organisational development programmes.