From January 21 to 25, medical specialists and researchers from UMC Utrecht and data analysts from the financial sector in Utrecht will work together to find a solution to better detect serious cardiovascular diseases in men and women. The specialists will make analyses on large amounts of data collected in recent years by the UMC Utrecht. They will investigate whether correlations can be found between the outcomes from simple medical tests, such as an ECG, and outcomes from follow-up examinations such as an MRI. The goal is to find an algorithm that shows (or rules out) whether someone with symptoms has problems in the small or large vessels of the heart.
Simple tests such as an ECG already work well to see damage to large vessels and are widely used. However, if a patient has damage to the small vessels, which we now assume are women, it is not easy to see with current diagnostics. In addition, diagnostic tests generally perform worse in women. The analysis incorporates the differences between men and women. The assumption is that an algorithm will help in more targeted patient referral, especially in female patients.
And that has important benefits. Making the right diagnosis in cardiovascular disease is far from easy. For example, women often have vaguer heart symptoms than men and do not always realize that their symptoms may indicate heart problems. On the other hand, it is difficult for doctors to determine that something is wrong with the heart or vessels. It now sometimes happens that patients are sent home after an ECG, even though it later turns out that there is a heart problem.
A reliable algorithm can help reduce this group and ensure that these people undergo further examination and treatment sooner. In addition, patients are now sometimes referred for an MRI when nothing turns out to be wrong afterwards. An MRI is relatively expensive and requires a lot of hospital staff time. This will not be completely avoidable, but help from a reliable algorithm could save hospitals money here.
“This way the people who need it get the care they deserve and unnecessary testing in people who do not have heart problems is avoided,” explains Hester den Ruijter, UMC Utrecht researcher. “That eases the pressure on care, the cost of care and reduces the chance that we miss something in a patient.”
Previous ABN Amro and healthcare hackathons have found such predictive algorithms for cancer treatment. Should it succeed again now, a period of validation will first follow to establish the predictive value of the algorithm with certainty.
Marina Senten, Head of Alliances, Science & Innovation at the Heart Foundation is enthusiastic about the collaboration: “In view of the growth in the number of patients, it is important that we recognize and diagnose cardiovascular diseases earlier, so that people can be treated in time. We believe collaboration is the way to accelerate better and effective solutions for doctors and patients. This hackathon, in which specialists from ABN Amro and UMC Utrecht are working as a team to find an algorithm and build a better risk model for doctors, is a great example of that.”
