Connected Conversation: What would it take to secure full public support for NHS data re-use?

On 30th January 2024 we hosted a conversation to share findings from, and have a wider discussion about, the implications of “Our Health Data Stories”; a report written by Just Treatment and Connected by Data. The report aims to bring out the personal stories of a range of patients as they encounter, think about, and respond to health data collection and use, to ground understanding and action in lived experience and help deliver person-centred care. It exposes how patients think about data on a day-to-day basis, and how this affects their approach to their health data and the way they use health services. And it recommends further work around data opt-outs, data service procurement, community engagement, and digital health apps. In our conversation we sought reactions to report, and also discussed action to be taken in response to the findings.

This conversation had invited contributions from:

• Diarmaid McDonald and Hope Worsdale from Just Treatment
• Jeni Tennison from Connected by Data
• Nicola Hamilton from Understanding Patient Data

We were joined by participants who, between them, brought research, legal, and campaigning expertise; lived experience of health conditions and health data breaches; and experience working on behalf of patient groups.

Conversation themes

We opened the conversation with a round of opening thoughts on the current UK landscape of health data. We heard that:

• The power dynamics around health data collection and use need to be considered and addressed by anyone seeking to engage the public or patients.
• Patient voice is needed, but it can be hard to engage people well in discussions around health data: data literacy is a particular challenge in this area.
• The route to impact from public engagement needs to be better established and demonstrated.
• Wider publics need to be included in health data decision making, including young people.
• There is a desire for data (and the technology it enables) to be used, but this must be done with strict and trusted regulation;.
• Trustworthiness is important: and it needs to be built by trustworthy institutions demonstrating that they are following good practices.
• A collective approach is needed: the NHS is a collective endeavour (and operates within a much wider ecosystem of health data users including academia, charities, and the private sector), therefore approaching data governance via individual data protection is not enough.
• Our conversations have been predominantly UK/England-focused, but we also need to consider the wider international context, as well as sharing learning internationally.
• Health data is an important public resource, but the public need clarity on its use so that they support efforts to realise its potential.
• Current opt-out approaches are not adequate, and complex opt-outs risk not being meaningfully used.
• There is no “one size fits all approach”, instead we need collaborative working across different organisations with communities to develop trust.

Reflections on the “Our Health Data Stories” report

The findings of the Our Health Data Stories report, and participants’ reflections on it in our conversation, highlight a context where we can see the intersection of corporate interests and patient interests at play. This data-focussed discussion has strong parallels with the embedded private-public ways in which the pharmaceutical industry operates, and highlights how questions of business model substantially impact on patient outcomes and the wider working of the NHS. However the current growth of data driven technologies in health presents a timely opportunity to shape what they will become, and the governance behind them. Good governance will be essential in making sure that the health inequalities we have seen in other areas are not replicated or deepened as a result of monopolistic industry structures and practices.

Concerns about health data exist within the zeitgeist of increased awareness of all sorts of data handling. If perceptions of health data are subsumed within this wider framing of risk, it may leave few levers to shift how patients feel about data use.

Health data literacy is a key concern. Our conversation discussed the lack of understanding, even among experts, when it comes to patient data. Where people are worried about how their health data can be collected and used, concerns range from those based on solid understanding, though to those based on misperceptions.

Prevalent (yet false) narratives paint anyone who raises concerns about health data as “anti tech”, and by extension anti-modern. In reality, many patients are positive about health data and the potential of tech in the health space, and want their data to be used for the benefit of all, so long as they understand how and why.

This said, many concerns around health data are significant and well founded. Health data collection is ubiquitous, and the choices around data collection can be problematic. Our discussions explored how patients often feel they must consent in order to receive treatment. The use of apps can also complicate consent, and it is not always clear where or to whom consent is being given.

Involving communities in decision making around their health data may be most effective at the local level: and can interface with existing NHS governance. However, conversation participants reported seeing local authorities having less input and power in NHS governance, and NHS Trusts reducing their public engagement. The balance between local and national decision making will be an important factor in shaping how community voice is, or is not, heard in health data decision making.

Health data is some of the most valuable, private, and rich information about our lives. Trust is the fundamental core issue when it comes to sharing this. Issues around health data will impact how people feel about the NHS, and how they will act in accessing health services; any breakdown in trust in this domain is critical.

Lived experience

Our conversation drew on input from participants with lived experience of complex health conditions, and of health data breaches. Through these inputs we heard about the need for an individual, personal approach to health data sharing: recognising not just majority views, but also the needs of particular communities.

While many express a strong desire to help other people by sharing health data, this cannot come at the expense of patient relationships and interactions. Patients with challenging health conditions need to speak freely about extremely personal, sensitive information, and they make themselves vulnerable in having these conversations. The choice of who to share this information with is key, and having private conversations is essential.

Our conversation surfaced high levels of concern about schemes like the Palantir-run Federated Data Platform. There are risks that patient worries about who might access information will create barriers to seeking medical help, and may prevent patients sharing sensitive personal information. This could lead to many suffering in silence.

What is needed?

Our Health Data Stories sets out a number of recommendations for addressing the risks of low public trust in health data use. Our discussions explored and built on these:

• Opt out: we need a co-designed replacement for the National Data Opt-Out.

In the conversation, participants discussed that granular opt-outs are unlikely to be well engaged with, and there is a tension between simplicity and the need for flexibility on which data is shared and when.

• Procurement: we need a strong requirement for revised ethical procurement system that includes public voice.

We discussed the importance of diversity in public voice, and the challenge in including the perspectives of those who may not be highly engaged or interested in data issues to begin with.

• Federated Data Platform: local deployment brings local community control opportunity, earning social licence to operate.

During the conversation, there was a tension raised between the opportunity for action at the local level, with the trend of disempowerment of local anchor institutions. We discussed the imbalance of power between local decision makers and tech providers, which requires broader national policy and action to be redressed.

• Health data in the private sector: we need to think beyond the NHS (e.g. health apps) and to provide better controls on data use in the private sector.

We heard how, from a patient perspective, understanding which apps are owned by which providers, and when and to whom consent is given, can be highly challenging.

• Civil society orgs have a role in assisting those who handle health data in becoming more trustworthy in what they’re doing and supporting the ability of public voice to reach decision-makers.

We focused the final part of our conversation on action arising from the report and the discussion around it, and asked participants to discuss: what should our demands be, to make health data collection and use meet patients’ concerns and needs? What action is needed, and by who?

Key points raised included:

• Public awareness and education campaigns to build health data literacy.
• Transparency across all power holders.
• Sharing real health data stories: for example the experience of a health data breach.
• Collaboration with researchers who use health data.
• Working with the Labour party and Lib Dems to shape their approach to health data, in particular to include public voice.
• Build space and resourcing for civil society input into what acceptable use of health data should look like.
• Public engagement needs to be sufficiently well financed as well as politically influential, with ongoing continuous engagement built in.
• Better investment into the public sector, to allow for meaningful competition with private sector in attracting and retaining talent, which would enable more ‘in-house’, not-for-profit solutions.

Open questions

Our conversation also left a number of open questions:

• How do we understand public benefit? Perceptions of public benefit change over time, and depend on context, meaning people will make different cost/benefit decisions around their health data at different times (for example during Covid).
• How do we ensure a full range of public opinion is represented in public engagement activities? We know that the people who respond are more likely to be interested and engaged already.
• What is the right level of clarity and simplicity, versus flexibility and personalisation, in opt out processes?

We will continue to explore many of these themes in our work over the year to come.