November 8, 2024

expert reaction to The Sudlow Review – Uniting the UK’s Health data

Scientists comment on The Sudlow Review, commissioned by the Chief Medical Officer for England, the UK National Statistician and NHS England’s National Director for Transformation.

Professor James Naismith FMedSci, Vice-President (Non-Clinical) at the Academy of Medical Sciences, said:

“The publication of the Sudlow Review marks a significant step towards making health data work better for the UK. Strong, trustworthy data systems, designed and delivered with input from patients, researchers and healthcare providers, will lead to better care and support innovations in treatment and prevention. Embedding an effective health data infrastructure is crucial to building an NHS that is fit for the future.  

“Success depends on active and equitable partnerships. Agreements between the NHS and its partners must ensure that everyone, especially patients and the NHS, reap the benefits of harnessing data. 

“The Academy of Medical Sciences and its Fellows will support these changes, bringing expertise and experience to help develop data systems that work for everyone. The Sudlow recommendations can make the UK’s health data a truly national asset, improving both patient care and driving economic development.” 

Michelle Mitchell, Chief Executive of Cancer Research UK, said:

“The UK has huge amounts of data which could be used to improve prevention, detection, diagnosis and treatment of cancer. But there are major barriers in how this data is accessed and linked, which means that we aren’t making the most of this vast resource to beat cancer sooner.

“The Sudlow Review has correctly identified the need to improve linkage and access to datasets which support cancer diagnosis and treatment, such as primary care data and screening data. This will be warmly welcomed by the researchers we fund, who frequently cite these as barriers to their efforts to improve the lives of people with cancer.  

“A National Data Service would allow us to make far better use of the high quality data we have in the UK. Engagement with people affected by cancer must be a top priority and we welcome the detailed proposals in the Review to ensure the National Data Service will be accountable and transparent about its role and purpose. Governments across the UK must now make it a priority to implement the review recommendations.”

Nicola Hamilton, Head of Understanding Patient Data, said:

“Understanding Patient Data welcomes the publication of the Sudlow Review and its assertion that health data should be recognised as critical national infrastructure. The UK is uniquely positioned to use insights from patient data to provide better care, improve NHS planning, and drive research and innovation.

We are pleased to see UPD’s work referenced in the chapter on patient, public and healthcare professional views on the uses of health-relevant data, and particularly support the recommendation about the need for an ongoing dialogue with these groups to co-create a trustworthy system. This should build on public deliberations that have been happening recently across all UK nations, and we are keen to continue to be involved in this space.

Whilst we strongly support reducing complexity and the central message of making the simple easy and the difficult possible, this needs to happen carefully and without removing people’s rights and protections. Sustained recent public attitudes research has demonstrated that, generally, we have the social licence to move towards a more permissive system, but only with more involvement, security, transparency and visible repercussions for those who break the rules and undermine public expectations. Otherwise, this trust that takes a long time to build can be lost in an instance.”

Prof Sir Martin Landray, Professor of Medicine and Epidemiology, University of Oxford, said: “Professor Sudlow provides a thorough and thoughtful examination of issues that have blighted efforts to improve healthcare for decades.

“I have witnessed many discussions and heard many promises to improve the safe and constructive use of health information to improve the quality of healthcare and the treatment of patients. Yet, so often despite many good intentions there has been little substantive action. For example, I have lost track of the number of times I have been told that GP data will be made available for consented clinical trials and cohort studies ‘shortly’, ‘within the next 6 months’ or ‘by Christmas’. 

“As Professor Sudlow highlights, much of this is not about wishing for more data to be collected, but about fixing the protracted, opaque and uncoordinated processes that prevent and delay access to data that is already sitting there.

“It is hard to think of another sector or service industry that makes so little opportunity to analyse the information it collects, identify challenges and deficiencies, and drive improvements through research and development. 

“During the pandemic, there was a temporary change of attitude. Sharing and analysing data suddenly became easier. This was not due to substantial technical changes but to a change in attitude and a few straightforward changes to the information governance approach. During the height of Covid, people recognised the harm that would be caused by failing to share data with those seeking to answer relevant research questions. 

“In the RECOVERY trial, we were able to access around 25 different data sets. Doing so, reduced the need to ask sick patients for information they had already given, lessened the workload for busy clinicians, and improved the quality of the research by ensuring that we had comprehensive information about what happened to people even if they moved hospital, were discharged home or were subsequently re-admitted somewhere else. All this was done with the usual safeguards in place including consent, ethical approval and cybersecurity. If we can do this for COVID, then the case is at least as strong for the many other conditions that impact on patients, their families, the NHS and wider society.

“Professor Sudlow makes a series of five recommendations. I agree completely with the themes that she has chosen. Where I would be more forceful is in the sense of urgency and the need for action. What is needed is not simply to ‘agree a coordinated joint strategy to make England’s health data a “critical national infrastructure’, ‘ establish a national health data service for England with accountable senior leadership’ or to ‘oversee and commission a strategy for ongoing coordinated engagement with patients, public, health professionals…’ but to deliver these and to do so promptly. Now is a time action. (In the depths of her report makes much the same point and helpfully provides practical and timetabled plans.)

“Strategy without delivery would be a fundamental disservice to patients. The ability use data with an appropriate set of proportionate, coordinated and timely safeguards is fundamental to improving health and delivering a sustainable and effective healthcare system.”

Prof Ben Goldacre, Bennett Professor of Evidence-Based Medicine and Director of Bennett Institute for Applied Data Science, University of Oxford, said:

“Better use of health data is vital for healthcare, research, and the economy. It’s great to see that championed in the review. It’s also great to have strong support for our work making patients’ GP data safely accessible to analysts through the OpenSAFELY platform. This service – which we built with NHS England – shows that it is possible to manage the privacy risks of data, through engagement and technical innovation.

“NHS GP data has extraordinary breadth and depth, with huge amounts of detailed information about every one of us.  That data can be used to drive research, health service improvement and innovation.  But it also presents substantial risks: because it contains everyone’s most confidential medical information.  Because of these privacy risks, and the scale of the data, it was never previously possible to make that data accessible for researchers at national scale.

“During the pandemic our team in Oxford developed new methods for “hands-off data analysis”: those new methods have now earned the unprecedented trust of professional bodies, patients, citizens juries, and the privacy campaigners who have loudly objected to all previous national GP data programmes.  With their enthusiastic support, we have been able to implement these new privacy tools in a working platform that allows researchers to run analyses – remotely and securely – against the whole population’s GP records, for the first time in history.

“This new service has already produced a huge volume of high impact research on COVID-19 in journals such as Nature, The Lancet, the BMJ, and more, showcasing the power of UK data. OpenSAFELY is now supporting 181 projects from users at 31 organisations: not just universities, but also UKHSA, NICE, and NHS analysts. With the support of this review, we hope to deliver more.

“This productivity, with an unprecedented volume of data and trust – at comparatively low cost – shows that there really is no need for a “trade-off” between privacy and access. Where government and the NHS make the right decisions about implementation and teams, platforms can be delivered that protect patients, satisfy privacy advocates, and deliver huge volumes of outputs.”

Dr Janet Valentine, ABPI Executive Director, Innovation and Research Policy, said:

“The Sudlow review has set out a compelling rationale for why research access to health data is fundamental to patient and population health and should be considered as an essential part of our national infrastructure.

“Access to health data is essential for the research and development of new medicines and vaccines, and so we fully endorse the recommendation to establish a national health data service. In taking this recommendation forward, it will be essential that government understands user needs, engaging across all research sectors from the outset to design and implement of a shared vision.

“We also fully support the Review’s recommendation to engage with healthcare professionals, patients and the public so that the system that is created has widespread support.

“The UK is not starting from a blank sheet, as there are already internationally recognised health data research services in this country. The government should build on these successes and consolidate the strengths of the UK to create a truly world-leading health data research infrastructure. The ABPI looks forward to being part of these discussions as plans evolve.”

Prof Krishnan Bhaskaran, Professor of Statistical Epidemiology at the London School of Hygiene & Tropical Medicine (LSHTM) said:

“This is a thorough and impressive report. Professor Sudlow and her team clearly set out the huge untapped potential of health data in the UK.

“The organisation of health care through the NHS means that the UK is in a unique position to bring together health records on a large scale to better understand disease and benefit the nation’s health. Analysing individual and linked datasets from GPs, hospitals, and other health settings has already generated important insights and evidence to help improve care, but the report shows that we could be doing so much more.

“The team have clearly consulted widely, getting to the heart of the barriers that researchers and public health professionals face in accessing and linking our rich datasets in a timely way to benefit public health. The report proposes positive solutions and ways forward with timelines that are ambitious, but achievable with political will.”

Prof Sheila M. Bird, Formerly Programme Leader, MRC Biostatistics Unit, Cambridge & Honorary Professor, College of Medicine and Veterinary Medicine, University of Edinburgh, said:

“In her Executive Summary, Professor Cathie Sudlow reminds readers about the internationally accepted “Five Safes Framework” (safe data, safe research – make data available only for approved uses for public benefit, safe people – appropriately trained, certified and authenticated analysts, safe settings, safe outputs).

“Because transparency matters, may I suggest a sixth: safe protocol – to ensure that the approved study-protocol and analysis-plan are put into the public domain at the time of, and as a condition of, data access.

“Scotland’s record-linkage has two advantages: Community Health Index number as a basis for no-names linkage of health and other data; and prompt registration of fact-of-death.

“The Sudlow Review’s first recommendation is: agree a coordinated joint strategy to make England’s health data a critical national infrastructure. Prompt registration of fact-of-death for all deaths in England is de rigueur for England’s health data to qualify as a critical national infrastructure. Let me explain.

“In Chapter 3 of her excellent review, Professor Sudlow maps our health (and other) data landscape and acknowledges that for England, Wales and Northern Ireland: “Delays for inquest verdicts on suicides and drug-related deaths cause inaccuracies in the estimation of calendar-year trends”. And she regrets that those delays have increased across the 21^st century.

“Importantly, as the Royal Statistical Society pointed out over a decade ago, all record-linkage studies need to know whether a follow-up sequence of events – whether hospitalizations, incarcerations, truanting, benefit claims – ceases due to rehabilitation (say) or because the subject has died.

“Prompt registration of fact of death is thus a cornerstone of our critical national infrastructure for record-linkage but fails us in England, Wales and Northern Ireland.

“Legislation is overdue to remedy this registration ailment, overdue perhaps because it affects three Ministries: Department of Health and Social Care because the Chief Medical Officer cares, Home Office because the Registrar General for England is located in the Home Office rather than death-registration being a responsibility of the National Statistician who cares very much; and Ministry of Justice because responsibility for inquests which are ongoing at death-date plus one year resides with the Chief Coroner.

“Professor Sudlow remarks that 93% of deaths in England were registered in the same year as they occurred (not within 5 or 8 days of their occurrence). Since when did UK consider that a death-registration system that fails 7% of the deceased was adequate?

“Prompt registration of fact-of-death for all deaths in England is required for its health data to merit the accolade of critical national infrastructure.

“Until then, analysts need to request both death-date and death-registration-date so that they can assess to what extent and how late registration of fact-of-death affects their cohort or randomized controlled trial.

“By way of illustration: in October 2024, Andrew Rosindell MP posed this question¹:

“To ask the Minister for the Cabinet Office, how long the 20 longest delays between (a) occurrence date and (b) registration date were for covid-19 related deaths in England that (i) occurred in 2020 and (ii) were subject to an inquest.

“Seven of these 20 longest delays from death-occurrence to death-registration were more than 3 years (1,322 days; 1,302 days; 1,294 days; 1,249 days; 1,162 days; 1,150 days; 1,122 days); each of the 13 others was delayed by more than 2.5 years (between 913 and 1071 days).

“In Scotland, all such deaths would have been registered within 8 days of death having been ascertained but with missing cause of death, which allows National Records of Scotland to prompt for an overly outstanding cause-of-death. In England, Office for National Statistics lacks fact-of-death registration and so cannot prompt for missing cause. A statistical nightmare.”

1 https://questions-statements.parliament.uk/written-questions/detail/2024-10-14/8764

Nicola Perrin, Chief Executive of the Association of Medical Research Charities said:

“Urgent action is needed to unlock the power of NHS data to transform research. Professor Cathie Sudlow’s review rightly concludes that the health data research ecosystem isn’t working, meaning patients and the public are missing out.

“We agree that a coordinated joint strategy is essential to help inform the development of a national health data service. Public confidence will be key to success, so we’re glad to see that the review highlights the importance of engagement with patients, the public and health professionals.

“We’re pleased to have contributed to this review and look forward to continuing to work with colleagues across the health sector to ensure that data-driven research can thrive.”

Prof Siddharthan Chandran, Director of the UK Dementia Research Institute, said:

“The UK has all the necessary elements to be a life sciences powerhouse and lead global progress towards new health breakthroughs, including in major disease areas like dementia. Critical to this are our national data assets, which include well characterised cohorts and of course our unitary healthcare system, the NHS. The Sudlow Review clearly sets out how we can harness these assets to genuinely change people’s lives for the better. We should seize the opportunity with both hands.”

Prof Sir Rory Collins, Principal Investigator and CEO of UK Biobank, said:  

“The Sudlow Review is right, we must now treat the UK’s health data like the critical infrastructure that it is. Connected health data can help the NHS deliver better care for everyone, manage the next pandemic, and rapidly get the right treatment to the right patients. We need to use the UK’s wealth of de-identified health data carefully, securely and to its full potential. Not doing so impedes scientific discovery. 

“The UK supports globally important health research projects like UK Biobank, Genomics England and the recently initiated Our Future Health. They are already having an incredible impact on our understanding of how to better prevent and treat disease, but they would be able to do so much more if we followed the steps laid out in the Sudlow Review to join up the UK’s health data.  

“UK Biobank alone has led to over 12,500 published scientific studies that describe opportunities for new diagnoses and treatments for everything from cancer to diabetes, heart disease and depression. This has been done with the help of altruistic volunteers directly providing data, which have then been linked to their hospital and cancer records. However, so much more could have been achieved if other health data were not trapped in the silos of GP practices, dentists and opticians. Diseases like arthritis, dementia, asthma and mental health conditions are almost exclusively seen and managed by GPs, whilst data from opticians and dentists would give us insights we can’t get anywhere else.  

“The Secretary of State for Health recently announced plans to make the GP data of consented participants available for studies like UK Biobank. Once this has been done, researchers will be able to make many more discoveries that lead to improvements in the way diseases are prevented and treated. The implementation of the Sudlow Report recommendations will truly be a transformative moment for health research and healthcare delivery.”  

Prof Rhian Gabe, Professor of Biostatistics and Clinical Trials, Queen Mary University of London, said:

“I really welcome these recommendations as the current system is inefficient and has long been due an overhaul. I particularly welcome this statement relating to recommendation 4: ‘The aim should be for trusted researchers and analysts conducting responsible analyses in the public good to be able to rapidly access the de-identified data they need, while ensuring that data cannot be inappropriately accessed.’

“As well as being cumbersome, the current data access process also has very high charges for researchers who are either funded by government or charities to access health data when it is for the public good and when there is no monetary gain from the information. I believe this is unreasonable and hindering some otherwise impactful research from being conducted. I do hope a new National Health Data Service will bear this in mind.”

Prof Andrew Morris, Director of Health Data Research UK, said:

“Safe and secure use of health data saves lives. But the UK risks falling behind in the global race to use health data for public good.

“Cathie Sudlow’s review is a landmark contribution in unlocking the potential of the UK’s health data for better care, a more effective NHS and a healthier nation. Her recommendations are given great weight by the breadth of the consultation and the depth of her review of the myriad datasets and systems held across the four nations of the UK. 

“Data is the engine underpinning progress in understanding disease, determining the best health and care, and it will power the responsible use of AI in health. Great progress was made during the pandemic in linking up the UK’s health data to tackle the new coronavirus through collaboration between the NHS, academia, policymakers and industry. But we are in danger of going backwards. The report’s recommendations show us the way to move forward again, involving the public in the best use of their data to improve all our health. This is extremely timely as the Department of Health considers its 10 year plan to restore the NHS.”

The Sudlow Review will be available after the embargo lifts at 00:01 UK Time Friday 8 November 2024 publication at: https://www.hdruk.ac.uk/helping-with-healthdata/the-sudlow-review/

Declared interests:

Nicola Hamilton: “No declarations of interest”

Prof Martin Landray “Chief Executive Officer at Protas, Professor of Medicine & Epidemiology at University of Oxford. Co-lead RECOVERY trial. Founder and previous lead of NHS DigiTrials 

Prof Ben Goldacre “I am a clinical academic, professor and director of the Bennett Institute for Applied Data Science at the University of Oxford. I led a ministerial review into better, broader, safer use of data in 2021. Our teams built the OpenSAFELY.org platform for GP data. Our work is funded by Wellcome, UKRI, NHS England and other similar funders. I receive personal income from speaking and writing books (and, infrequently in recent years, journalism) for lay audiences on the use and misuse of science and evidence.”

Dr Janet Valentine “None”

Prof Krishnan Bhaskaran “Professor Bhaskaran is a statistical epidemiologist and is active in health research using routinely collected healthcare data, including through Clinical Practice Research Datalink (CPRD) and OpenSAFELY.”

Prof Sheila M. Bird “since 2010, in the aftermath of swine-flu, SMB has led for the Royal Statistical Society on the need for legislation to end the late registration of fact-of-death in England, Wales and Northern Ireland.”

Nicola Perrin “None”

Prof Rhian Gabe “I am working on studies funded by the NIHR, Prostate Cancer UK, Cancer Research UK and Yorkshire Cancer UK to investigate cancer screening strategies for prostate, colorectal and lung cancer.”

Prof Andrew Morris “Andrew Morris is Director of Health Data Research UK, the national institute for health data science; is Professor of Medicine and Vice Principal at the University of Edinburgh; is President of the Academy of Medical Sciences, has minority (<1.5%) shareholding in Aridhia Informatics and a small number of shares in GSK (<£5,000). Cathie Sudlow is a colleague, having been Deputy Director and Chief Scientist of Health Data Research UK during the period she wrote her review. Health Data Research UK provided the secretariat to support Cathie Sudlow’s review.”

Prof Siddharthan Chandran “Siddharthan is the academic lead of Neurii, a £5M partnership to deliver patient focused digital health solutions for dementia, part funded by Eisai. The UK Dementia Research Institute holds partnerships with charities (BHF, Alzheimer’s Research UK, Alzheimer’s Society and LifeArc), and industry (Lilly, Eisai, Astex, SPARC and Ono).”

Prof Rory Collins “I am CEO and PI of UK Biobank, which is a Charitable Company established as a Joint Venture by the MRC and Wellcome. I have been in that role since September 2005, seconded 60%FTE from the University of Oxford where I am Head of the Nuffield Department of Population Health (which, along with other research organisations globally, benefits from using the UK Biobank – without any preferential access – for health-related research that is in the public interest).” 

For all other experts, no reply to our request for DOIs was received. 

This Roundup was accompanied by an SMC Briefing.