Harnessing health data to address health inequalities in North East London

Professor Carol Dezateux is a leading expert in Clinical Epidemiology and Health Data Science at Queen Mary University of London. With expertise in paediatrics and population health, Professor Dezateux has been at the forefront of research addressing the factors contributing to children’s health, particularly through the analysis of routinely collected data in electronic health records. Her involvement with Barts Life Sciences from its inception has made her an influential advisor in data science and strategic direction.

The conversation delves into the significant impact of the Barts Health Data Platform on her research, particularly in tackling childhood obesity, and explores its broader implications for Barts Health, Queen Mary, and the local community. Through her work, Professor Dezateux highlights the urgent need to address health inequalities and improve outcomes for vulnerable populations in North-East London.

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

Your research has significantly focused on childhood obesity. What are the global and national trends in this area?

The latest data from the 2023-24 school year paints a concerning picture of childhood obesity in England. At the start of primary school, 1 in 5 children aged 4-5 are already overweight or living with obesity, with 1 in 10 living with obesity. The situation is worse for children leaving primary school at ages 1-11.  1 in 3 are overweight or living with obesity, and 1 in 5 are living with obesity.

There are pronounced inequalities in these figures. Children in the most deprived areas are more than twice as likely to be living with obesity compared to those in the least deprived areas, highlighting a stark disparity.

What trends have been observed in North-east London?

In North-east London, which includes some of England’s most deprived areas, the situation is even more alarming. For instance, in Barking and Dagenham—home to a large childhood population—1 in 3 children are leaving primary school with obesity. The region has also seen a rise in severe obesity, with nearly 1 in 10 children affected – significantly higher than the national average. Similar trends can be seen in Newham and Tower Hamlets.

Over the past decade, childhood obesity rates have been steadily increasing both in the UK and globally. A recent review in The Lancet highlights that obesity has risen in most countries, while underweight and thinness remain prevalent in South Asia and parts of Africa. It is crucial to recognise that both obesity and undernutrition are forms of childhood malnutrition. Many people are unaware that obese children are malnourished.

These trends underscore the inequalities arising from childhood poverty, which had stagnated and is now on the rise, alongside a lack of access to affordable, healthy food for children. Addressing these issues is essential to halt or reverse the growing trend of childhood obesity.

What are the priorities from your perspective in reducing the rate of childhood obesity in your region?

One key focus should be controlling sugar and salt in foods, both locally and nationally. The 2021 National Food Plan, authored by Henry Dimbleby, outlines effective strategies that are largely overlooked by governments. There’s strong evidence that incentivising the food industry to lower sugar content in soft drinks has had a positive impact. Since its introduction in 2018, this initiative has been linked to a relative reduction in obesity rates among girls, particularly in disadvantaged areas.

Another crucial measure is limiting the advertising of unhealthy food and drinks. Advertising restrictions on the Transport for London network have led to a 7% decrease in weekly household purchases of unhealthy products, especially those high in sugar and unhealthy fats. Extending these policies to other platforms, like TV and social media, is essential as children are exposed to marketing in various places. Local authorities also need to review their advertising strategies within their communities.

Expanding free school meals is another impactful strategy. Providing healthy, appetising and nutritious meals during school and holiday periods is vital. London has extended free school meals to all primary school-aged children, a programme that has been successfully offered in Newham and Tower Hamlets.

I, along with colleagues, have been evaluating this initiative in Tower Hamlets and collaborating on research to improve school food quality. While it’s important to ensure children aren’t hungry, proving a direct link between free meals and weight management or learning readiness is challenging. However, I believe the absence of concrete proof shouldn’t hinder efforts to provide free school meals. Although the government promotes breakfast clubs, there is ongoing advocacy for free school lunches and holiday food programmes, which were essential during the pandemic.

We also need to deepen our understanding of the consequences of childhood obesity. While we know its long-term effects on adult health, we must investigate how it impacts children’s daily lives. By linking school and GP data across four local authorities, we have gained insights into health outcomes, including how excess weight affects children’s muscles and joints.

We’ve also explored at Barts Health NHS Trust whether childhood obesity is a common factor in severe tooth decay, where the rates of tooth extractions under anaesthesia in young children are alarmingly high. Tooth decay is entirely preventable. Removal of teeth in early childhood can affect speech and language development as well as dental health in adult life. In response, Barts is now investing in free community dentistry for children, emphasising the importance of regular dental visits alongside good oral hygiene and reduced sugar intake.

Can you describe the research conducted by your group and the impact it has had?

I am one of the clinical leads of the Clinical Effectiveness Group (CEG) at Queen Mary. Our team, comprising clinicians, health data analysts, facilitators, and academics, has trusted access to primary care data from 266 general practices, covering 2.4 million people. We use this data, alongside data-driven tools, to enhance population health and reduce health inequalities, primarily in North-East London, though our impact extends beyond this region.

One notable example is our ‘Active Patient Link’ software tools, designed for primary care, which have helped us achieve the highest blood pressure control rates in England. These tools provide real-time data, enabling practices to identify next steps, such as which patients need blood pressure checks or medication reviews, rather than focusing on missed actions. This proactive approach helps prevent the harms associated with undiagnosed or untreated long-term conditions, which are prevalent in our population. My research focuses on children, and my colleagues and I have adapted this tool to improve the uptake of measles immunisations in practices.

Another significant initiative is our collaboration with hospitals to streamline healthcare delivery between hospital and primary care settings. For example, we introduced a Community Kidney Service, improved the coding of Chronic Kidney Disease in GP records, and developed a ‘falling eGFR’ software tool that simplifies the identification of declining kidney function in practice populations. This collaboration has allowed hospital kidney specialists to review records jointly with GPs. Initial evaluations revealed a dramatic reduction in the time taken to receive a specialist opinion, from 64 days to just 6, while also halving the total number of hospital appointments needed for kidney outpatients. This not only reduces the need for patients to make frequent hospital visits, saving them money on transport, but also frees up valuable clinical time.

We have also worked closely with Barts Health to prevent strokes in individuals with abnormal heart rhythms, specifically atrial fibrillation. This award-winning quality improvement project increased the prescription rates for anticoagulants from 88% to 94%, surpassing the national target of 90%. This initiative has been proven to prevent strokes and heart attacks.

Additionally, there is growing interest in using machine learning and algorithms in healthcare. For example, our team are employing machine learning to identify individuals with unrecognised atrial fibrillation, a major stroke risk factor, and treatable blood-borne virus infections, which can lead to severe liver damage and cancers. We have also created algorithms that group de-identified health records into households, link datasets, and understand the influence of environmental and wider determinants of health. This approach has been developed to protect patient privacy and confidentiality and ensure that individuals and households cannot be identified.  This open-source, highly accurate algorithm is now being used in primary care data across three of London’s five regions, covering approximately 7 million patients, and is being implemented in Wales and Scotland.

These examples demonstrate how data, innovation and close collaboration with clinicians as part of a learning health system can help reduce inequalities and deliver effective care to entire populations.

What else does a learning health system need apart from data?

While data is an essential component of a Learning Health System (LHS), it is not sufficient on its own. We need human involvement and a deep understanding of how to change the behaviour of healthcare professionals through support and incentives, which is the essence of the clinical effectiveness approach.

CEG facilitators play a crucial role in this process. These trained staff members work closely with practices to assist them in utilising data tools and implementing quality improvement initiatives. We provide a wealth of online training materials and support, and we maintain weekly communication with all our practices. It’s important to recognise that data alone will not drive change.

The Wachter Report underscored the necessity of a well-trained healthcare workforce. In response, we have established educational courses focused on health data in practice and have trained GPs in quality improvement. Queen Mary and the Integrated Care Board (ICB) are now collaborating to create 20 GP quality improvement positions with CEG over the next five years to enhance capacity.

Finally, we work closely with commissioners, and our experience has shown that local incentives, peer support and achievable targets are critical for achieving rapid improvement. This training and facilitation are key to our success.

The recent Lord Darzi report alongside Wes Streeting announcements emphasise the need for greater focus on prevention and prediction. How is this being implemented in North-East London, and what are the opportunities and challenges?

I hope the examples I’ve provided illustrate how prevention, prediction, and the reduction of health inequalities are central to our work. The findings of the Darzi report, particularly the emphasis on prediction and prevention, as well as the shift from hospital care to primary care, are highly encouraging.

In North-East London, we have remarkable opportunities to build on our existing experience and leverage the strong partnerships we’ve formed across the university, primary care, Barts Health NHS Trust, ICB, and local authorities to further transform healthcare.

Barts Life Sciences, along with funding from Barts Charity for our precision health programme, has created significant momentum. The Clinical Effectiveness Group (CEG) is fully engaged in the ambition to make North-East London a world-class example of using data to improve population health and reduce health inequalities.

We are collaborating closely with the Informatics Group at Barts Health, that is Sarah Jensen and her team, to support the development of the Barts Health Data Platform. Alongside my colleague Professor Claude Chelala, we are applying our expertise in primary care and cancer data to track patient outcomes before and after hospital attendance or admission, incorporating pathology and imaging data into our analyses.

This platform will provide trusted access to a wealth of patient data that has previously been difficult to obtain. We are on a journey where the more we utilise data for research and innovation, the more we will enhance its quality, standardisation, and the automation of coding at the point of care—such as from written information like letters and X-ray reports—while ensuring privacy is protected. This is a crucial challenge, as valuable information often remains uncoded and is not easily accessible for research.

Together, we can make North-East London a world leader in using data to improve population health and address inequalities. The Barts Health Data Platform is an exciting initiative; not only does it have ambitious goals, but it also integrates data from various sites. Historically, access to hospital data has been problematic, while GP data has been utilised for longer. The hospital data is incredibly rich—not just what patients report, but also detailed tests such as pathology reports and blood tests. This wealth of information is invaluable for our research efforts.

There are numerous opportunities to use data for patient and public benefit, making this an exciting time for health data science. The launch of the data platform marks a significant milestone in our journey.

You have been heavily involved in Health Data Science, can you describe what data sets are crucial in your work and where are there gaps in data sets that would help turn the dial?

The pandemic has underscored the importance and public benefits of using NHS data to protect and enhance the health of our communities. It has also highlighted the significant role health inequalities play in determining life expectancy. We’ve seen how valuable data can be in rapidly identifying effective and ineffective treatments through the information collected during routine NHS care.

However, the pandemic also exposed several gaps. Not all hospitals have robust electronic records, and there is a lack of comprehensive data systems that capture prescribing, imaging, and pathology data. Additionally, there is insufficient linkage to information regarding the wider determinants of health, such as housing, education, and environmental factors, including air pollution.

Developing a system in North-East London that safely and securely integrates these datasets while protecting privacy and confidentiality will enable us to better understand how to prevent diseases, deliver effective healthcare, reduce inequalities, and develop new treatments.

Moreover, we know from a large deliberative public consultation involving seldom-heard residents in London that this is what the public expects and wants us to achieve. Building public trust and ensuring trustworthiness is essential, and the Barts Health Data Platform has recognised this by involving patients and the public from the outset.

How has your collaboration with the Barts Health Data Platform influenced its development?

The Barts Health Data Platform is unique in that it will encompass all Barts Health hospitals, enabling us to utilise all the information collected about patients during their routine hospital care. As partners in this collaboration, we bring 25 years of experience in trusted access to primary care data for the entire North-East London population, along with a proven track record in using this data for research, innovation and quality improvement.

Additionally, my colleague Professor Chelala at Barts Cancer Institute contributes her extensive expertise in bioinformatics and tissue biobanks, which will help us improve cancer outcomes and gain a better understanding of how to reduce recurrence.

One key aspect of establishing a data platform is demonstrating its impact, and we are well-positioned to contribute to that. For example, we are currently examining birth registration data to investigate the outcomes for children of mothers who develop diabetes during pregnancy, a condition that is relatively common in our population.

We have different expertise, and I think we work very well together.

What benefits are you hoping to achieve from the use of this Platform for patients, healthcare staff, industry, and academic groups? How important is the launch of the Barts Health Data Platform in addressing health priorities in your region?

The key benefit of this Platform is its potential to utilise data as part of a learning health system, enabling people to live healthier lives for longer and receive effective care when they do become unwell. This is a major priority for our region, which has a disadvantaged and ethnically diverse population experiencing high levels of illness early in life. Diseases that typically affect other communities later are impacting ours much sooner, often while individuals are still working, which is frequently overlooked in research.

The Barts Health Data Platform will allow us to unlock valuable information from tissue samples, imaging and genomics, helping us discover new ways to prevent, diagnose and manage chronic diseases in populations with similar characteristics. Additionally, having a robust system in place means that when we identify effective interventions, we can deliver them to the community much more swiftly. The Platform will underpin research to provide evidence of what works within our community, significantly reducing the time between discovery and the implementation of these advancements, with research and innovation at the heart of everything we do.

Nationally, do you think the current pace of innovation in health data science and the underlying infrastructure is sufficient, or are we at risk of falling behind in addressing health challenges?

The NHS, as we saw during the pandemic, has enormous potential to be at the forefront globally because we have a publicly funded national service that is free at the point of use. However, many national initiatives have stumbled due to a lack of public trust. In my view, the main challenge is to demonstrate meaningful trustworthiness to our patients, the public, and healthcare professionals who have a duty of care to patients. We have considerable experience in achieving this in North-East London. The London Data Service, as one of the largest regional NHS data hubs, is building on that foundation.

Our goal is to use this resource for research to develop evidence on what works to improve health and health equity, and then to put those findings into practice for all our population.

Scaling this at a national level to 60 million people, as opposed to 2 million in North-East London, is a challenging task. Previous attempts to do this by NHS England and the government have faltered in large part due to lack of public trust. The key message here is that the pace of innovation is limited by a lack of trust. If we focus on building that trust, as well as improving the quality of data and training our workforce, we can become a world leader in this field. Achieving this will require strong partnerships with our patients and healthcare professionals.

About the author

Professor Carol Dezateux’s research focuses on health information related to children, examining the household, social, and environmental factors contributing to childhood obesity and its health implications.

To address this, Carol is integrating data on children’s weight and health from schools, general practice, hospital, and housing records. She collaborates with clinicians, families, local authorities, and service planners in East London to identify and evaluate effective interventions. Working in the Clinical Effectiveness Group, which has over twenty years of experience of and trusted partnerships with General Practices, has shown that using patient data to enhance health services by creating a ‘learning health system’ ensures that scientific advances can be implemented at pace and supports equitable improvement in health care.