Patient-centric solution to secure medical data Collecting and securely sharing medical data remains a key challenge in ensuring high-quality healthcare. New data integration and privacy-preserving tools could provide the answer. Security © Gorodenkoff/Shutterstock.com In Europe it is not uncommon for citizens to travel and live in different countries for periods of time. When receiving healthcare in a different country, it is essential therefore that a patient’s previous medical history is made available for clinical purposes. However, this is not always straightforward. “One reason for this is that medical data is collected and stored in different ways, using different formats and systems,” explains SERUMS project coordinator Juliana Bowles from the University of St Andrews in the United Kingdom. “Many people for example now wear personal medical devices. How can we gather this data, while ensuring that information is secure?” A further complicating factor is the fact that medical information is highly confidential, and quite rightly cannot be shared unless patient approval is explicitly given. Any system for aggregating and sharing medical data must therefore prioritise patient privacy and preferences. Collecting medical information The SERUMS project sought to develop new high-tech ways of integrating this information, and making it accessible to all relevant health practitioners. This was achieved by bringing healthcare providers together with software specialists. To begin, the project identified data collection needs in three test cases. The first was a group of hip replacement patients, who were being fitted with sensors to monitor their movements at home. The second focused on elderly patients with heart conditions and diabetes, who were also being fitted with sensors and often in and out of hospital. Finally, medical data was collected from a group of cancer patients, who might come into hospital for chemotherapy but recuperate at home between treatments. Bowles notes that in all three test cases, patients could have appointments with a number of different specialists, who might not have a global view of their healthcare. The goal was therefore to ensure that all valuable data was gathered, and made accessible to all relevant medical personnel. Synthetic data technology To test this, new data aggregation and machine learning analytic techniques were applied, not to this real-life data but rather to synthetic, anonymised data. This was generated to ensure that the privacy of real patients could not be compromised. Bowles describes the end result as a ‘data lake’ of medical information, containing everything from X-rays to data gleaned from personal health devices. Used in the real world – such as in the three use cases studied – this synthetic data could be replaced with real data, protected by privacy-preserving technology. Patients in control Indeed, another key element of the project was ensuring that patients remain fully in control. A web-based, user-friendly interface, with picture and text-based passwords, was developed, providing patients with access to their aggregated medical records. The system was successfully trialled with the three real-life use cases, with feedback helping to fine-tune the tool. Such a system could one day enable a patient to decide whether, say, doctors in a foreign hospital can access their data if they need to be treated abroad. Before that happens, a number of challenges still need to be addressed, not least ensuring that such a system is in full compliance with various laws. Nonetheless, the project has demonstrated that such a data system is conceptually possible. What’s more, the success of the data aggregation tools – and effectiveness of using synthetic data to test these tools – could benefit other sectors where data is equally sensitive, such as banking. Keywords SERUMS, medical, healthcare, data, patients, cancer, X-rays
