A nationwide study of diseases that occur in the years following a heart attack

The BHF-Turing Cardiovascular Data Science Awards (Second Call): Post myocardial infarction disease trajectories: process mining of 145 million hospitalised events (joint funding with The Alan Turing Institute)

A heart attack occurs when the blood supply to part of the heart muscle is blocked. In the UK more than 100,000 hospital admissions each year are due to heart attacks and they can be fatal. However, survival rates have improved dramatically over recent decades. People who survive a heart attack are at greater risk of having a subsequent heart attack, stroke, or developing heart failure. They are therefore offered treatments which aim to prevent these conditions from happening. We are not yet aware of all the complications and diseases that could occur after a heart attack, and this could be leading to many unnecessary premature deaths. A team of cardiovascular experts and data scientists at the University of Leeds, led by Dr Marlous Hall and Dr Jianhua Wu, will analyse 145 million anonymous electronic hospital records for people who have suffered a heart attack to reveal the full spectrum of adverse effects that can occur in the following years. They will use an artificial intelligence method called process mining. Process mining has the potential to extract a wealth of useful and clinically-relevant information from this complex and vast dataset, where there will be huge variation in the problems that occur after a heart attack, and the order and timeframe in which they occurred. The research could help save lives and reduce healthcare costs by leading to the discovery of previously unknown consequences of a heart attack.