A recent BCS webinar saw a panel of experts discuss patient data in the NHS: it’s uses, applications and the forces at play which may be stopping us from making the most of data’s potential to do good.
Our experts were:
- Andrew Griffiths: CEO, FEDIP - Federation of Informatics Professionals
- Prof. Terry Young: Co-founder of founded the Cumberland Initiative
- Dr Natalie Banner: Understanding Patient Data Lead, Wellcome Trust
- Brian Runciman MBCS: Head of Content and Insight, BCS
The far-reaching discussion took in many fundamental and defining topics from the world of data-driven healthcare. These were, of course, informed by the current debate that circles the NHS’ contact tracing app.
1. Ethics questions but not novel ones
The NHS COVID-19 contract tracing app’s difficulties have put patient data, privacy and security under a microscope. There is a real need to engage the public in meaningful and solution focused debated about what ethics means.
Natalie Banner: ‘A lot of the issues that are appearing right now are not completely novel. We have a long history of exploring ethical issues on data use, privacy and consent… We have been prepared to accept quite considerable restrictions on our rights and freedoms at the moment… The way we are making trade-offs about benefits and risks has shifted. We’re accepting that we need to restrict our movements.
‘Data is playing an important part in this pandemic and it is forcing us to have those discussions in a public realm.
‘We tend to treat ethics as a slightly abstract, slightly academic topic - something for other people to think about. The pandemic has shown how embedded and how practical some of these ethical questions and challenges are. They’re for everyone to consider. You don’t need to be an ethicist or an ethics expert to weigh in and have perspectives on what the ethical use of data should be and whether we’re cross any lines.’
2. Should my data be used to make operational decisions?
Of all the recent pandemics - AIDS, MARS and SARS - COVID-19 may well be the one about which the most data has been gathered and at the greatest speed. Academics, NGOs, health bodies and even tech companies are generating data very quickly and in large volumes. We need to answer very specific questions about this data before it can be used in the theatre of clinical operations.
Is it accurate, representative and trustworthy? And, what is the context that exists around the data - are you using a set collected to answer a specific question in the hope of answering another unrelated query? All these questions and more shine a light on data.
Natalie Banner: ‘It is critical to pick up on the quality of data and how it is being collected. This matters hugely. From the IT side, knowing and understanding the limitations of that data is increasingly important. We’ve seen so many examples of wild assertions being made on the bases of data that isn’t accurate: it’s flawed and we have to very careful. As we open up more data, we have to give context. The role of the IT professional is shepherding data. If we can base operational decisions on data, the context and limitations around that data need to be articulated.’
Terry Young: ‘What has really come out of [COVID-19] is the need to act at speed and how to use data to make operational decisions. What interests me is getting the numbers right as far as making operational decisions. I don’t think the data has been clear. If we go into this situation again in, say, four years, I would like to see better data… faster mobilisation so we can work ahead of the pandemic and influence it rather than watch it evolve through an imperfect data lens. It is the operational use of data…’
Andrew Griffiths: ‘The challenge for us, now we’ve got all this data… are issues about its accuracy and its interpretation. Normally we have months of looking at data before we decide, yes, this data is okay to use. Now, we are using it within 24 hours. I’m not sure if our systems are up to being that forward-thinking, that data-driven, as opposed to looking in the rear-view mirror. But, the point is, we need it operationally to decide whether we have a lockdown.’
Andrew Griffiths: ‘[COVID-19] has made some rather simple data tech quite visible and this has challenged us. Things like: how do we use that data more proactively? I contrast it with Tesco: within two days of lockdown, they knew people stopped buying 25 different variations of baked beans and were more interested in buying flower and toilet roll. They were able to get their supply chain to react quickly. That’s a remarkable shift - and they continue to shift. The NHS doesn’t have that same flexibility and we need to develop that capability.’
3. Why doesn’t the public trust government with its data?
The NHSX COVID-19 contact tracing app’s gestation and testing problems have been well documented. A central theme and point of friction have been the use of data and concerns about privacy. At a time when, arguably, the public needs a protective shield that is founded on their data, many people don’t feel happy participating in such collective programme.
We need a solution-based debate that can enable patients and people to confidently own, curate and donate their data when they see a cause they believe in.
Andrew Griffiths: ‘Patient trust in how we use data is vital and maintaining trust is vital. Currently, all the ways that we share data - within the current legislation - shouldn’t cause anybody any concern. My experience isn’t that we’re sharing more data. The problem is the aggregate of the small bits of data that people aren’t sharing… Where individual clinicians think they can’t share data from secondary to primary care, or they can’t tell somebody a test result. The biggest problem we have is not sharing data in the course of a person’s treatment rather than sharing it everywhere.’
Terry Young: ‘We need to think through what privacy means in a data-porous world. Possibly we need to gravitate more towards a world where we have real privacy and it really is private. Or, we have a neo-village where everybody knows everything about everybody. I’m not convinced that drawing lines between those two will be successful. Thinking about extremes might help us think about what we want from privacy.’
Terry Young: ‘I think the government would do better creating apps that actively involve us in sharing data rather than forcing us to share contacts with whoever we’ve been near. I would like to see us in a place where everybody participates in being a solution rather than people being observed. I’d take a participative approach.’
Dr Natalie Banner: ‘We don’t yet know how COVID’ has [changed the public’s attitude to data sharing]. In early March, we released a report, The Foundations of Fairness. It was the result of an extensive public engagement exercise exploring what a fair partnership between the NHS and third parties should look like when it comes to sharing and using data… We found that people get that the data needs to be used by third parties… academics, charities or industry. ‘People understand that if there is a benefit coming back to the NHS, the health care system and patients, there could be really good reasons to allow this data to be shared. But, they are not involved at the moment. They don’t have a voice or a say. If there is an aim to improve care and health - all those sharing partnerships need that as their fundamental core.
‘There needs to be consistency… It is hard, from a system level to say, this is what the rules are. There needs to be a greater degree of governance, accountability and transparency. And, finally, people really want to be involved in this.
‘Once you start talking to the public, they want to be actively involved. Data can be a difficult subject to broach - it is quite technical and dry. When you get into trade-offs, benefits and risks - people want to be involved. There is huge scope for a participative model - not just observing or extractive.’
4. Is the UK public uniquely concerned about data?
COVID-19 is a global pandemic and, as such, there is a need for countries and organisations around the world to share honest, accurate and compatible data sets. The UK isn’t, however, alone in having legislation that governs data collection, storage and transit. Nor is the UK public unique in its uncertainty about entrusting governments with more information.
Dr Natalie Banner: ‘A data trust is where you have a group of people who have a fiduciary responsibility to manage that data in ways that serve the interests of the participants. It’s a legal framework. There are lots of models. The open data institute is doing a lot of work on them at the moment. There’s a lot of opportunity in this idea of data philanthropy - of people donating their data in some way. One of the big challenges is representativeness. You’re only going to get [the data from] those who are most engaged and active, who have the capacity and the willingness to explore questions around who gets the data and how it will be managed.’
Andrew Griffiths: ‘Trust is really important and you can lose it really easily. With a participatory model - when you bring people in to a project, they stop being the public. They start becoming professionals, the more they learn about the data and the way we work. They might be advocates but they are perceived as being on the inside. Trust is more important [than participation] and that’s where we need politicians involved.’
Andrew Griffiths: ‘Should we have global standards for the collection, collation and process of data? Yes... There’s a lot to be said for ensuring our data sets can be compared internationally. That happens already and there are some great benefits. Look at the World Health Organization - the UK does well at providing standardised data. The pandemic shows that we need to compare data internationally.’