Isabella

www.nesta.org.uk/feature/our-tech-our-future-dsi-2030/isabella/
Skip to content
Isabella

For years, health data existed in silos. At best, data supporting scientific research was published in tables in obscure academic journals or held in clunky online repositories. At worst, it wasn’t published at all. Individuals were generating data all the time too, with gigabytes of data from wearables kept within the walls of big tech companies. There wasn’t a shortage of data – what was missing was a way to pull it all together.

Today, there’s more data than ever and, for the first time, it’s freely available and linked. Where once the challenge was accessing the data, for Isabella, an epidemiologist and health data scientist with the Pan-European Health Network (PEHN), the challenge now is deciding what to do with it all.

The seismic shift in approaches to health data management had been arduous. Agreeing EU-wide standards for data management with universities, research institutions and drug companies was hugely complex, but was achieved in the timeframe set out by the European Commission. The real game changer had been successfully agreeing the approach to storing, distributing and governing that data. Centralised data hubs just didn’t work, and when the approaches were being designed Isabella spoke out strongly against them. As an expert advisor, she argued that the responsibility for data management was too great and too costly for just one organisation, that it didn’t solve the siloing problem, and that the ramifications of any breach would be catastrophic.

Instead, she advocated for a model of distributed data storage, which was eventually adopted. Of course, all data originated somewhere, but there would be nothing to stop other organisations having a copy. To protect individuals, a blockchain-based ledger system was introduced, allowing individuals and data gatherers to set limits on how data can be used. People can choose what data they want to share and don’t want to share, under which circumstances.

As a result, data scientists like Isabella have access to findings on a global scale, but without the ability to personally identify anyone. It’s revealed correlations between a huge number of factors and health outcomes over the short and long term, which researchers have only just begun to examine.

Isabella’s job is to manage the AI systems which run the queries. As everyone knows, correlation doesn’t always equal causation, so she focuses on delving deeper into the links identified, investigating them from different angles and using her medical knowledge to decide which queries should go forward to cohort studies, what to recommend to policymakers, and how to best serve at-risk groups.

But for Isabella, health data is not just for epidemiologists. It can make a meaningful difference to the lives of individuals too. With almost everyone now wearing general health data gathering devices, and those with long term conditions wearing more elaborate sensors accruing condition-specific data, people have access to information to understand their own health better.

Isabella understood that this potential remained unachieved - after all, most people aren’t data scientists. A year ago, Isabella proposed an experimental training programme for individual health data, for which PEHN gave her a small grant and two days a week to develop. Now Afya, a social enterprise founded by Isabella, two doctors and two nurses, runs training sessions for members of the public to help them interpret their health data from various sources.

Afya has two types of session. The first is for people who are interested in using data to maintain good health. Participants are mostly middle-aged or older, and keen to keep up their current lifestyles. They’re interested in simple tips and insights, like correlations between exercise and blood pressure, or nutrition and mental health. They’re taught how to draw out insights and can choose to compare their own data with other people’s.

The second type of session is for people with long-term health conditions. For example, rather than taking daily blood samples, as people with diabetes did in the past, tiny sensors implanted within the body allow people with diabetes to make more precise judgements about how they eat and the insulin they inject. Combined with electronic food diaries and data on exercise, precise predictions can be made about how the body will react. The sessions provide training in how to manage, interpret and respond to all this data.

Isabella lives by the old adage that prevention is better than cure. She’s pleased that these new uses of data are helping more people live longer and healthier lives. And the rich insights that health data offer gives her hope for her own future too. She knows it’s not a panacea, but she’s confident that the data she’s collecting about her own body will serve her well in the years to come.

Projects and organisations which inspired us