Why do we need a National Health Data Service?
One of the eye-catching recommendations of the Sudlow Review was to set up a National Health Data Service. But why do we need one, what would it do and how would it make a difference?
Andrew Morris, Director of Health Data Research UK, explores the potential of a national health data infrastructure and how it could transform healthcare research and patient outcomes.
A good friend of mine was unfortunately diagnosed with cancer recently and received a combination of radiotherapy and new type of chemotherapy regime. Over dinner one evening they asked me how we link cancer information to patient outcomes across the UK to give a comprehensive assessment of this and other new treatments. That way providing insights into their effectiveness, safety and enhancing personalised care strategies for patients living with cancer.
I had to tell them that despite the data being available, this information and insight was not routine across the UK.
Cathie Sudlow’s independent review was published a few weeks ago, titled Uniting the UK’s Health Data: A Huge Opportunity for Society, and goes a long way to addressing this issue. It found that complexities and inefficiencies were impeding the use of the UK’s health data, with scientists having to wait many months – or even years – to securely access data for vital research and actionable insights into diseases like dementia, cancer and heart disease.
Accelerating research through a National Health Data Service
Having identified the barriers preventing timely access to health data, the Sudlow Review made five recommendations to transform the safe use of data to improve care for diseases affecting millions of people.
One recommendation called for a ‘national health data service’ for England to be established. A body that would have the necessary clout and accountability to tackle the complexities in the system. By working in partnership with NHS, academic, regulatory and other partners, there is the opportunity to deliver a service that would enable secure access to the most important health data for England, in the various places it is held nationwide, and so speed up health research and analysis that benefits patients and their families in all parts of the country.
For example, after a new drug is approved the medicines regulator (MHRA) should then be able to monitor its use and detect any adverse effects that might be associated with the medicine. But data access issues make it difficult for the MHRA to combine data on the drugs prescribed across all health care settings with outcomes data from all GP practices and hospitals across the country. A national health data service would oversee and enable such access, benefitting us all by improving drug safety.
Improving healthcare outcomes through data linkage and curation
The power of these approaches comes from connecting or linking different datasets together to answer important questions. Cathie, in launching her review, used the example of cancer screening. The only real way to understand and improve the effectiveness of the UK’s screening programmes, she explained, is to link data collected as part of the screening programme with data from GP and hospital records on cancer cases and the outcomes from treatment. Yet this isn’t possible to do routinely, despite screening being such an important pillar of cancer early detection and diagnosis. The linking of different data types and sources would be another task for the data service.
A future data service should also prioritise providing NHS data securely to a set of our flagship, health data assets in the UK that are already enabling world-leading research. UK BioBank, Our Future Health, Genomics England, Clinical Practice Research Datalink (CPRD) and the Office for National Statistics’ Integrated Data Service are large-scale clinical studies and data services that represent significant investment by taxpayers and medical research charities, and hold great value for the UK’s health and wellbeing.
A further activity of a national health data service would be to help in ‘sorting’ the data, making it much more readily usable for researchers. Because health care systems data is complex due to the variety of tech platforms in use, different approaches to data capture, variability in which codes and data standards are used, and more – it can be very difficult for data scientists to pick their way through it all, especially with the amount of data involved.
It is frequently estimated that around 80% of time working with data can be spent on data preparation. Consolidating this activity as far as possible so it is collected and prepared once and used widely will improve the efficiency and quality of research. Currently, once one set of researchers has wrestled and wrangled data into a usable form, it is rarely shared for the benefit of others. Data curation is an under-rated part of what a data service could offer.
Lessons from the COVID-19 response: A model for other diseases
If this sounds daunting and detailed, we should not be put off. We know this can be done. And done safely by stripping away all personal identifying information to protect privacy while accessing the data via secure data environments.
Great advances were made during the pandemic to enable the rapid research needed to respond to a new virus. OpenSAFELY, run on behalf of NHS England by Ben Goldacre and colleagues at the Bennett Institute for Applied Data Science at the University of Oxford, has enabled research across the two largest electronic health record providers in the NHS (over 98% of the NHS patient population in England). Approaching 100 research papers have been published as a result.
The BHF Data Science Centre here at HDR UK, through the CVD-COVID-UK/COVID-IMPACT consortium, has enabled projects using a variety of linked health datasets for the whole population of England, Scotland and Wales to investigate the coronavirus and its connection to heart disease and stroke. Through partnership working across the UK, studies on 68 million people are now possible. This is enabled by data curation and management expertise that makes it easier for future researchers to build on.
We’ve seen the difference these data services can make in combating COVID-19, improving outcomes for patients. But what about the other pre-existing pandemics of cancer, heart disease, dementia, depression and more? These matter just as much to those affected and we should be doing the same for all those patients and their families.
Investing in health data infrastructure to improve lives
We need a national health data service to deliver this. One that operates in partnership with the NHS in both England and the other nations of the UK. One that involves patients and can maintain the confidence of the public. One that offers a single access and approval point to England’s rich health data resources for the benefit of us all.
Its success will depend on drawing in talent and expertise from across the sector. The ability to attract funding and income from a range of sources will be vital so it can grow and develop its data and service offer. This would enable it to be independent of an annual NHS budgeting process that typically causes delays and risks being ‘top sliced’ to support operational priorities.
That’s also where we agree with Cathie in her review, as she asks for health data to be seen as critical national infrastructure. Such critical infrastructure can’t exist on drip feeds of short term funding.
Setting up a national health data service will be no easy task. But we have already seen it is possible to provide all the elements needed. Now is the time to make progress on a data service to benefit us all and HDR UK is on call to contribute by mustering its expertise and resources from across >150 organisations involved in our UK-wide programmes and the UK Health Data Research Alliance. It is time to do so for the benefit of patients and citizens, and for the safe, effective and efficient design of the NHS.
