Patients could receive a greater benefit from open heart surgery thanks to a new computer model aimed at helping surgeons to better calculate risk and decide whether it’s safe to operate.
This project, jointly funded by the British Heart Foundation (BHF) and The Alan Turing Institute, will develop a new platform using machine learning to identify patients who are most likely to have a successful operation.
Over 30,000 adult patients are considered for heart surgery every year in the UK, and risk prediction plays a major role in the decision-making process made by doctors and patients.
Assessing risk before open heart surgery is crucial due to the potential complications that can arise during and after the operation. To calculate a patient’s risk before surgery, heart surgeons currently use models such as the EuroSCORE – but this may overestimate the actual risk, partly due to improvements in patient management since the model was developed.
Consequently, people with a good chance of having a successful operation may be deemed to be inoperable and left untreated.
In cases where the risk is predicted to be too high, it can lead to risk-averse practice, meaning either the patient may refuse to have the surgery or the surgeon may decide not to undertake the operation.
Now, researchers at the University of Bristol and University of Oxford are creating a new risk calculator using a large dataset of information routinely collected from all patients undergoing surgery in the UK.
Gianni Angelini, BHF Professor of Cardiac Surgery at the Bristol Heart Institute said:
“We will apply the most advanced machine learning algorithms to identify heart patients who are likely to have a successful open-heart operation.
“By adopting current risk prediction models, surgeons in the UK are exposed to inappropriate risk-averse practice which denies surgery to ill patients who would benefit from an intervention.”
Professor Metin Avkiran, Associate Medical Director at the British Heart Foundation, said:
“It’s crucial that the likely risk versus benefit is identified as accurately as possible for each patient before proceeding with surgery. A more reliable risk assessment tool would allow a better informed choice to be made between patients receiving or being denied a potentially life-saving heart operation.
“We’re investing in innovation and developing new data-led technologies because we recognise the huge potential that data science has to transform care for the millions of people living with heart and circulatory disease in the UK.”
The EuroSCORE was developed using information collected in 1995 from a relatively small number of patients from eight European countries including the UK.
The British Society for Cardiothoracic Surgery (SCTS) recognises the urgent need to replace the EuroSCORE with a more precise risk prediction model and has endorsed the project. If successful, the SCTS will evaluate the opportunity to replace the EuroSCORE and implement the new model in all the cardiac units in the UK.
Chris Holmes, Programme Director for Health and Medical Sciences at The Alan Turing Institute and Professor of Biostatistics at the University of Oxford, said:
“This project has the potential to provide a new, precise predictive model which could play an important role in supporting healthcare professionals and patients in their decision making. It should also help healthcare providers to monitor surgical performance, conduct quality assessment and understand a patient’s risk so that patients can make an informed decision before a major procedure.”