HUS – the Hospital District of Helsinki and Uusimaa – is developing preventive healthcare and medical research using the HUS-DataLake.
Head of Data & AI, Tietoevry Care
Developed in collaboration with Tietoevry, the solution enables the analysis of huge datasets to create healthcare-related predictions and cutting-edge innovations. According to Mikko Rotonen, ICT Development Director at HUS, and Visa Honkanen, Director of Strategic Development at HUS, the newly available HUS-DataLake enables increasingly higher-quality healthcare while ensuring even more cost-effective operations.
HUS is a trailblazer in digital healthcare data systems, having already collected data digitally for decades. Electronic health records have been stored since 1980, and laboratory tests, X-rays and other health records of the 2.8 million patients treated at HUS over the past 10 years are currently accessible in digital format. In order to exploit this data to its fullest – in combination with the huge amount of data from other sources to benefit healthcare, research and businesses – a data lake is needed.
'The big data solution created by Tietoevry and the HUS Data Administration department integrates the data scattered among different HUS data systems; for example, patient data repositories and administrative systems. The data is sorted and stored so that data sources can be linked in real-time for different purposes and in various desired ways. This enables the highlighting of all relevant information needed for each situation,' explained Mikko Rotonen, ICT Development Director at HUS.
HUS wants to be the leading provider of world-class healthcare services and preventive care. The ability to refine and improve existing and externally gathered data plays an integral part in achieving this ambition.
'By utilising the data generated by health and social care systems, we are able to create algorithms that can identify at-risk patients with different types of potential illnesses and predict epidemics as well. Applications of this solution will make treatments increasingly more successful and affordable to patients. We can offer patients increasingly more high-quality, customised care,' describes Visa Honkanen, Director of Strategic Development department at HUS.
The HUS-DataLake solution speeds up the data-gathering process for the end users – doctors, researchers and the administration of HUS – as they are now able to gather patients’ medical data more easily than ever. It is now possible to focus on the analytics itself, as there is no longer a need to individually harvest data from each of the different data systems. In the future, doctors will be utilising the predictions generated by the HUS-DataLake algorithms to deliver better care.
The benefits of the HUS-DataLake have already been seen in the treatment of premature babies. Since 2000, data gathered from each newborn treated at the Neonatal Intensive Care Unit K7 (NICU) at HUS Women's Hospital has been integrated into an algorithm that predicts the risk of sepsis in newborns. With this algorithm, the risk of sepsis can be predicted up to 48 hours before onset of the infection, and thus doctors and the nursing staff now have two days to devise an appropriate treatment plan for this serious condition.
'The future vision of HUS entails that data will benefit both the individual and society as a whole. On the individual level, we can help people prevent the onset of medical conditions by identifying potential risks. For the elderly, we can provide longer and safer residencies at their own private homes. On a societal level, we can make predictions of the risks affecting whole populations – for example risks of epidemics – enabling forecasts of healthcare service needs,' Honkanen continued.
In addition to contributing to the well-being of Finns, the integrated HUS-DataLake also creates business opportunities. Through collaborations with IT professionals and organisations, healthcare data gathered by HUS can be used directly in research and development for cutting-edge healthcare services and products. However, this development process does not require changes in ownership of individual patient health records. The data uses pseudonyms, so personal medical information cannot be used to identify individual people.
At present, we are continuing to develop the DataLake with HUS to create even more high-quality, innovative and cost-effective healthcare for Finns – and medical innovations for all of society.
Data-driven solutions for health and care.
ChatGPT and beyond: What is the impact of generative AI on information work?