As the NHS approaches its 70th birthday several reports and commentaries are bringing into sharp focus its achievements and the challenges that lie ahead.
Innovative technologies have featured extensively this week, including a joint briefing by the Health Foundation, Institute for Fiscal Studies, The King’s Fund and the Nuffield Trust that considers the question of; ‘what new technologies will mean for the NHS and its patients?’. Another is the interim report of a review being led by Dr Eric Topol of the Scripps Research Institute. This review focuses specifically on the impact of technological developments on the healthcare workforce, and what’s needed to prepare the workforce for these changes. It sets out the context and key themes for its wider discovery phase, with an open call for evidence to inform the final report.
Both reports pay specific attention to the impact of advances in genomics, data/digital medicine, and artificial intelligence. These developments have enormous potential to improve healthcare, but their full benefits cannot be realised without addressing pressing questions around regulation, policy, infrastructure, and the workforce.
Genomics
The Topol review interim report acknowledges that widespread introduction of genomics will require its mainstreaming across the wider health system and a shift toward multidisciplinary working. Ensuring all NHS clinical specialists can easily and appropriately access genomic tests for their patients is a knotty challenge and one which PHG Foundation has a well recognised track record in untangling. This includes our work on Genomics in mainstream medicine, and involvement with the Genomics in Mainstream Medicine Working Group to explore how genomics is relevant to different clinical specialities.
The Topol paper cites a report of a workshop led by the PHG Foundation last year which convened clinicians, clinical geneticists and the clinical genomics science community, with the aim of highlighting changes essential to widening access to genomic medicine. Essentially these changes are needed across the clinical pathway, from getting the right test for the right patients, to getting and reporting the correct result.
Data and digital medicine
Digital medicine, within the Topol review, covers a spectrum of technologies including telemedicine, wearables, digital diagnostic tests, bionanotechnology, digital therapeutics and virtual and augmented reality. The Health Foundation et al report considers these through the lens of technologies that could enable more remote care. They point out that digital technology is increasingly supporting patients to better manage and understand their condition. Driven by patient demand and technology supply, much of this is happening outside of the NHS, with patients increasingly collecting and storing information about their condition.
The Topol interim report specifically refers to this ‘patient-generated data’ observing that in the future:
- Patient generated data will be interpreted by algorithms enabling personalised self-management and self-care
- Unified personal records that integrate patient generated data along with electronic health records will benefit patients
- Healthcare professionals will have to be trained in how to interpret patient-generated data (e.g. from wearables)
- Clinical bioinformatics training will evolve to include patient generated data
These are all essential points for consideration. However, given the rapid pace of progress in smart devices that are enabling individuals – before they become patients – to generate data about their health, PHG Foundation is going wider – to consider the impact of ‘citizen generated data’, i.e. data not only from patients, but also from those in good health and those not yet interacting with the health services. We all become patients at some point in our lives, however around 80% of an individual’s life expectancy is spent in good health. Understanding the pathological and physiological changes that occur during the greater part of our lifetime, is key to better understanding the mechanism of disease.
Artificial intelligence (AI)
AI based technologies are not yet in routine use in the NHS. However they do show great promise for medical image analysis and in the future could transform many aspects of healthcare. According to the Health Foundation et al, how the development and then deployment of AI technologies should be governed is a fundamental question yet to be adequately addressed. The answer to this question will encompass the legal framework surrounding the use of algorithms in healthcare including their regulation, and issues of liability and intellectual property, work that is being undertaken by PHG Foundation.
In our invited oral evidence to the House of Lords AI Committee last November, we highlighted that whilst the NHS holds valuable assets i.e. data, for developing machine learning algorithms for health care, it does not have the infrastructure and AI expertise, a point reflected in the Topol Review interim report which notes that ‘the NHS needs AI specialists who understand the requirements of patients and healthcare professionals and are keen to be embedded in all care settings, so that AI systems designed for healthcare are fit for purpose’. How the NHS is to acquire or access these AI expertise remains to be resolved. In the short term at least cross-sector collaboration will be essential to bridging these skills and infrastructure gaps.
Harnessing technological advances for health – a complex gift?
The opportunities presented by advances in genomics, data/digital medicine and AI could be enormously transformative for the NHS and its patients. The greatest impact could be seen through the coalescence and combinatorial use of these groups of technologies and many others. There is no doubt however that these technologies present complex challenges, and as the Health Foundation et al point out these are ‘not a silver bullet for the pressures facing the NHS’. With the right supporting infrastructure, policy – including data policy – standards, and workforce development, these technologies have the potential to have continuous and growing positive impact, for example as we understand more about our genetic variants, or as better and evolving datasets drive improvements in the performance of algorithms. Ideally these technologies would become the gifts that keep on giving.