World map with data points

Photo by Clay Banks on Unsplash

How new data institutions could be the key to radically improving our health and wellbeing

Mon May 4, 2020
$download_content = get_field('download_content');

We talk to geneticist Dr Wen Hwa Lee and lawyer Rob Bryan about how data institutions can help to enable trustworthy data sharing in healthcare

Data institutions are organisations whose purpose involves stewarding data on behalf of others, often towards public, educational or charitable aims. By stewarding data responsibly, they have the potential to bring about innovation, from scientific discoveries to the rapid commercialisation of products and services.

Here, we talk to geneticist Dr Wen Hwa Lee (CEO) and lawyer Rob Bryan (Chair) of the charity Action Against Age-related Macular Degeneration (AAAMD). They’re working with the ODI to explore how data institutions can help to enable trustworthy data sharing such as the INSIGHT Health Data Research Hub for eye health.

What problems do you think that data institutions can help to solve?

Lee: I’m always looking for metaphors to explain how I feel about data and I can’t think of a better one than the mythical dragon. In the west, dragons are often seen as a malevolent force in the same way that there’s a lot of suspicion and concern around how data about us is used and sometimes, misused. I’m of Chinese heritage and for the Chinese, the dragon represents strength and unlimited potential. In fact, with data there is so much potential that it’s sometimes hard knowing how to approach it, how to tap into it.

We also have a habit of trying to look at new concepts within old frameworks. For example, data isn’t like oil, in that data can be used again and again. I see data more like language. Individual words encapsulate a specific meaning – thus each on its own has very limited value but not much. Their power is only realised when you bring them together to make a book, a speech, an article – thus more complex ideas and insight. Similarly, the more data points you have, the more powerful they become. At the same time, you have this concern that people see data as their individual property and are concerned about the potential for it to be used to unlock one complete individual. This works against the benefits that can come from putting lots of data together, for example to find new diagnostics and cures for eye diseases. That’s where the new data institutions come in, including data trusts.

Rob: In simple terms, they can provide wider access to data in a manner that engenders greater trust – that’s it in a nutshell. This is what data institutions do, in their various forms and at the moment, we’re pleased to be working with the ODI to experiment with data trusts in practice, rather than just in theory. So it’s a very exciting time to be involved.

Lee: These new data institutions are very different to more traditional models in the way that they allow us to combine datasets to drive more value – innovation has always thrived near the boundaries of established ‘expertise’. Traditionally we have asked doctors to look after health data, geologists or cartographers to look after location data, and bankers to take care of financial data. Integrating all this data can only work within a new framework where we actually drop away all the old misconceptions and think from a ‘first principle’ approach to organisational design – with data trusts for example, what types of skills and experience will ‘data trustees’ need to have?

Why is your organisation interested in data institutions and data trusts?

Rob: At AAAMD, we are an R&D based charity, operating in healthcare and life sciences and so it’s important to provide a strong evidence base. The data institution we’re creating as part of the INSIGHT project involves the public, patients and other stakeholders in decision making about who gets to use the data held by the partners in the project. We’re confident that this, in turn, will increase trust in the process and lead to more data becoming available for research. And that means we will hopefully stand the best possible opportunity of finding an early-stage intervention to macular degeneration that will have the greatest chance of success.

Lee: The problem here, as with most diseases, is that the approach is pretty much predicated on ‘firefighting’ or trying to fix it after it has happened. For age-related macular degeneration (AMD) only a subset of late stage disease patients with ‘wet’ AMD can be helped with an expensive drug injected into the eyeball every other month. With an ageing population, and this being an age-related condition – and with the high individual price of the therapy at around £800 per injection, per eye – the cost to the NHS will rocket. In the near future, many more people will end up with sight loss as the remaining type of AMD can’t yet be treated.

What the public doesn’t appreciate is that sight loss brings a lot of other knock on effects: not just loss of independence but deteriorating mental health, loneliness and isolation, at a time when people are already fragile, as well as the increased impact on healthcare and social care.

So, if we want to stop the early stage AMD developing further into its late, sight-threatening stages, we need to have data from the general population – to study and understand the disease. But data privacy concerns can prevent people wanting data about them to be shared. So we need institutions that steward data in open, trustworthy ways to reassure people. I see data institutions, including data trusts, as organisations that are not in the system to represent specific vested interests – they are opportunities to diversify who is involved in governing data. In the context of our work, we’re interested in involving the public and patients of these eye diseases in decisions about the use of data.

Rob: This challenge is actually more acute in the health space because here, you’re dealing with sensitive personal data. And so assuring people that data about them will be safe and won’t be misused is one of the greatest challenges and one that we’re taking very seriously.

What are the opportunities?

Rob: There are colossal opportunities! We’re faced with one here and now with the COVID-19 pandemic. A specific example is where countries from around the world are producing data about the virus. If we were able to better combine and analyse data from across these different sources, our response to the pandemic may be more informed and coordinated. That would mean that researchers could understand the particular traits and nuances of particular cases and identify trends in a way that you otherwise couldn’t do. But you would only want that kind of data access and sharing through a trusted institution.

Lee: If people share their data, we hope to be able to gain a greater understanding of the different dimensions of disease. Data institutions give us and the general population a trusted place for data to be shared because they help to ensure that data is used safely and in the most unbiased way possible. The way we have been thinking about data institutions, and data trusts, is that they aren’t yet another faceless institution that claims it can do things on your behalf. The concept of a data trust, for us, is that we can truly integrate the voices of patients and the public into a governance model where they have a real say. It goes beyond an advisory role.

What do you envision for data institutions in the next twelve months?

Rob: We are living in exceptional times but momentum needs to be maintained to maximise the likelihood of data institutions’ potential being fully utilised. I think that if the general public can see that medical research can be progressed more rapidly as a result of individuals’ willingness to share their data, then that will be exceptional progress. The COVID-19 pandemic is a good example because everyone is affected by it. In a more general sense I think that data sharing enabled by data institutions can build greater understanding of medical conditions. Whether it’s sight loss, cancer, dementia or COVID-19, I think this understanding will lead to a positive correlation with individuals’ willingness to volunteer data.

Lee: With Ebola we learned that the more we can share, the more quickly we can act, and the sooner we can solve problems and fight pandemics. With COVID-19 the various restrictions that have been introduced have only been understood and implemented because the UK government has been able to see epidemiological and health datasets from other countries. Imagine if we were not sharing that kind of information, what would have been the consequences of keeping it hidden? This will be a great lesson and a chance to show people why it is so important to share data. So I do have big hopes for data institutions in the coming year because, while we have all been experiencing the positive impact of using data without knowing, COVID-19 really brought in the limelight how data sharing can lead to fast coordination and response, ultimately saving lives.

Podcasts: data institutions for healthcare data