A white paper entitled ‘NHS data: Maximising its impact on the health and wealth of the United Kingdom’ has been published by Imperial College London’s Institute of Global Health Innovation.
The document is intended to “create a first step towards establishing a vision, strategic framework and underlying principles to underpin how health data should be used to improve patient care.” Authors Saira Ghafur, Gianluca Fontana, Jack Halligan, James O’Shaughnessy and Ara Darzi point out the need for agreement on “the areas of action needed to maximise the value of NHS data; the current situation and existing barriers for each of these areas; recommendations to explore further; and outstanding questions that should be resolved using evidence-based research.”
Looking at how we can maximise the impact of the UK’s health data, the authors highlight the opportunities available, including huge amounts of data that is collected in healthcare, the position of the NHS as a single-payer system under a common legal framework to create a single longitudinal dataset, and the numerous efforts already in place to deliver benefits for the population through use of health data. These efforts include the NHS Artificial Intelligence Lab, the NHS App, Health Data Research UK, the Digital Innovation Hubs, Genomics England and UK Biobank.
Whilst acknowledging the challenges also facing the system – such as difficulties around estimating the value and benefits of data, the need for a clear strategy and inconsistent public involvement – the authors go on to share six key areas in which improvements can be made.
Public opinion and engagement
The paper highlights existing work to understand the public’s perception of health data, but notes that there is limited understanding of opinions across demographics and socio-economic groups. It suggests a need for more proactive engagement with clinicians and other frontline staff on this topic, and points out that citizens have not been involved in setting the rules and principles by which decisions about data use are made.
Recommendations, therefore, include better engaging with citizens and NHS staff on this topic; the papers emphasises a need to understand which data licensing and value-sharing models are the most appropriate and ethical, and to ensure that the system understands the attitudes and concerns of all segments of the UK population, including NHS staff.
The authors that that the system should “use what we already know combined with what we can learn through better public engagement to develop and implement a communications strategy led by the NHS on the use of health data, prioritising communications that foster trust, not just information transfer.” This could include describing tangible benefits using real-world examples; developing a communications strategy for mass and social media, with focus on tackling misinformation; and communicating more clearly around how health data is used. In addition, NHS staff should be trained to involve patients in decisions about how their health data can be used.
Data governance and legal frameworks
The white paper notes how data governance standards in the NHS have been significantly improved in the past ten years; however, the legal framework governing the use of personal data in healthcare “remains complex and creates a number of legal and societal challenges”.
The authors recommend that legal frameworks relevant to health data usage should be clarified, particularly around “discrepancies and misunderstandings as envisaged by the European Data Protection Supervisor (EDPS) on a pan-EU level, by increasing dialogue with the research community.”
In addition, there should be clarification around data usage and data-sharing in healthcare for both primary and secondary uses, which could be done in conjunction with the National Data Guardian and NHS England, and should involve parters from industry, academia and research.
The paper also notes a need to clarify the scope of the “public interest in the area of public health” exemption in Data Protection Act 2018, along with “the extent to which a new legal basis for processing is required where the purpose of subsequent processing is compatible with the purpose of the original processing”.
Data quality and infrastructure
“Outside of primary care, there is a marked difference in data quality, standard adherence and interpretation, and infrastructure,” the papers state. “Data often requires significant effort to be post-processed, as accurate data are very often not captured real-time.”
The authors point to legacy technology and infrastructure and their role in delaying the ability to move to the cloud, further impeding data quality.
Recommendations here include enabling codified, real-time data to be captured at source; this will involve enforcing common standards for data capture across the NHS, increasing digital maturity and moving away from paper-based processed, and ensuring that NHS staff are properly trained in using relevant systems.
The system should invest in standards-based infrastructure and cloud-based services across the NHS, the authors add. They suggest that the scope of cloud services should be outlined in local service agreements, including who is responsible for what and who holds insurance; in addition, agreements should avoid vendor lock-in by ensuring that providers can lift and share data from one cloud provider to another as contracts end.
Finally, they say, the mechanisms for funding IT in the NHS should be reviewed, and suggest encouraging a move towards capital funding of Software-as-a-Service (SaaS) solutions.
A more effective approach is required to deliver the recommendations of the Wachter and Topol reviews, the white paper states. It notes that there are “few senior digital leaders in the NHS with the knowledge and skills to deliver large and complex IT programmes” and adds “there needs to be significant investment to attract data science and engineering talent to the NHS, particularly from the private sector.”
The recommendations related to IT and analytical capability in the NHS from the Wachter and Topol reviews should be implemented, the authors say. This includes developing a workforce of trained clinician-informaticists at NHS trusts, investing 1 percent of the £4.2 billion to be spent on digitising the NHS to support workforce development and deployment, and focusing on recruitment of science, mathematics and computing graduates to fill leadership roles in bioinformatics. There is also a need to attract graduates in data analytics to a career in health, they add, in order to create and implement technological solutions capable of improving care and productivity. On a similar note, another recommendation is to ensure that genomics and data analytics are prominent in undergraduate curricula for healthcare professionals, along with offering opportunities for healthcare students in areas such as engineering or computer science.
In addition, the authors call for a review into how IT and analytical talent in the NHS is paid in order to increase competitiveness, and also to determine which approach to building data capability fits the needs of the NHS.
The paper states that “a major increase in government investment is needed to achieve the required improvements in data quality, infrastructure and capabilities”, and that for the UK to “maintain its position as a top destination for research and development, clear rules need to be in place along with appropriate incentives for innovators.”
Along with increased government investment, the paper recommends development of a “clear UK-wide strategy to attract investment specifically in data-driven health innovation” and notes that the system should focus on “significant new opportunities such as the Accelerating Detection of Disease and other cohort studies that pioneer new models of consent and active patient participation.”
Here the authors acknowledge that the NHS “should receive fair value for the use of its data, whether financial or in kind; failing to ensure that the NHS is properly compensated may influence public trust.” They note that it is currently unclear who, within the NHS, should receive the benefits generated from licensing data, and add that it is “urgent to have a public debate on this topic as many NHS organisations are currently engaging with organisations without central guidance.”
As such, recommendations include providing NHS organisations with a clear framework and guidance for entering into value-sharing arrangements when licensing data. In addition, they point to a need to strengthen the support available to NHS organisations for the negotiation of data and value-sharing agreements, stating that expertise in structuring complex data-sharing arrangements should be built within the NHS at central and local levels.
The white paper provides a framework that describes the value-sharing arrangements that could be used by NHS organisations, available to view from page 30 here.