University of Edinburgh

Wellcome Leap

DL-2022-028

Many people attend hospital with symptoms of a possible heart attack, such as chest pain or breathlessness. Getting the right diagnosis is not easy – there is a lot of information including blood results and heart traces for clinicians to consider. Advances in computing and science can now find patterns in data that are not always obvious to humans. This is sometimes called Artificial Intelligence, or AI. In this project, we aim to develop an AI risk calculator that generates a “score” based on clinical information that is available from the patient during his/her stay at the emergency department. The score that is being calculated from the AI risk calculator will provide information about the patient’s risk of having a heart attack, and also on the future risks of that patient such as needing to come back to hospital. We expect that this approach will provide better and more individualised predictions of risk and that this application could improve the quality and safety of care delivered to patients in the emergency department.

Acute cardiac conditions are the leading cause of deaths across the world. Up to 1 in 4 of all Emergency Department (ED) attendances are for symptoms such as chest pain or breathlessness, where the accurate and rapid diagnosis of heart attacks and related conditions is critical to safe patient care. However, this is complex due to symptoms overlapping with other conditions and the need to integrate multiple pieces of relevant information (e.g. patient history, electrocardiograms, blood tests, X-rays). This project will develop artificial intelligence prediction models to improve insights from these data. We aim to provide clear estimates of the risks of a heart attack, early hospital reattendance if discharged and future cardiovascular risk. This project is supported by Wellcome Leap and NHS Lothian with a clear roadmap to service implementation if successful. Clinicians and patients would be involved in the future co-design of the NHS-deployment tool. This has the potential to improve the accuracy and effectiveness of care delivered to many people in Lothian – approximately 1,500 patients are seen for suspected acute coronary syndrome by our ED staff every month. By providing smarter information to clinicians, our aim is to reduce the need for these patients to reattend the hospital by 20%.

Public Health Research

17/01/2023

De-Personalised

TRE