AI to help GPs identify those at high risk of heart-related death

The OPTIMISE tool is able to identify those people at high risk of developing conditions that can subsequently lead to heart related death, suggests new research.

By spotting conditions such as kidney failure and diabetes, it can enable preventative or revised treatment at an earlier stage to safeguard against subsequent mortality risks. The research at the University of Leeds was recently presented to the European Society of Cardiology Congress, hosted in London.

The research was led by consultant cardiologist at Leeds Teaching Hospitals NHS Trust and University of Leeds professor Chris P Gale, together with Dr Ramesh Nadarajah, health data research UK fellow at Leeds.

Training OPTIMISE by analysing the health records of more than 2 million patients resulted in the identification of more than 0.4 million individuals at high risk of dying from a heart-related cause. Records established that within a 10 year period, this group subsequently comprised nearly three quarters (74 per cent) of participants who died of a heart-related condition. 

Piloting the AI tool on a group of 82 high risk-patients, the researchers diagnosed kidney disease in that would not have been picked up otherwise in 20% of the individuals. Additionally, more than half of those with high blood pressure were given different medication to better manage their risk.   

Professor Bryan Williams, chief scientific and medical officer at the British Heart Foundation, which funded the research, commented on the value of the work:  

“A quarter of all deaths in the UK are caused by heart and circulatory diseases and this new and exciting study harnesses the power of ever-evolving AI technology to detect the multitude of conditions that contribute to it,” he said.  

"Early diagnosis is key to reducing hospital admissions and heart-related deaths, allowing people to live longer lives in good health. We look forward to seeing how this will help accelerate and inform clinical decision-making, ensuring patients receive timely and effective treatment and support." 

Nearly eight in 10 people with heart and circulatory diseases have at least one other comorbidity, often exacerbated by the impact of a poorly functioning heart and circulatory on the supply of oxygen and nutrients, so affecting other organs

The researchers say that implementing OPTIMISE into GP systems would enable identification of high-risk patients by the push of a button. A larger clinical trial is now planned, with hopes that the tool can be rolled out for GP use within two years. 

Research co-leader Dr Nadarajah said of the team’s plans to scale up its work for adoption within the health system:

“Heart-related deaths are often caused by a constellation of factors. This AI uses readily available data to gather new insights that could help healthcare professionals ensure that they are providing timely care for their patients. 

“We hope our research will ultimately benefit patients living with heart and circulatory diseases, as well as helping relieve pressure off our NHS systems, as prevention is often a cheaper solution than treatment. Next, we plan to perform a clinical trial where we are providing doctor led care to patients.”