Using Artificial Intelligence to Improve Emergency Care of People with Chest Pain
RAPIDx AI aims to test whether the use of computer algorithms in hospital emergency departments can help doctors provide better care for patients who have symptoms that may relate to their heart.
Nearly 1 million people visit Australian hospital Emergency Departments (ED) because of chest pain or similar possible heart attack symptoms. Although most people do not ultimately end up getting diagnosed with a heart attack, current ED assessment of these patients is resource intensive and often inefficient which contributes to hospital congestion and may in fact cause harm.
This project will test whether using computer algorithms in hospital EDs can help doctors provide better care for patients with symptoms that may due to their heart.
We have designed computer algorithms that compare a specific patients health information with a very large database of patients who have been treated in Emergency Departments before and can identify those that have similarities to the current patient, such as similar age, symptoms and health status. By looking at the care they were provided and how they recovered, the computer algorithm may be able to support doctors by assisting with the diagnosis and providing recommendations for personalised care
Whilst we believe that artificial intelligence can never replace clinicians, we will test its usefulness as a decision support tool to enable more consistent delivery of high-quality evidence-based care that improves patients lives. There are many factors that could affect whether these computer algorithms are able to help doctors improve patient care and outcomes which is why we need to conduct this study before we start using in routinely in all emergency departments across the country.
For more information on the project please read the Executive Summary.
- Prof Derek Chew
Chief Investigator, Professor of Cardiology, Flinders University
- Prof Tom Briffa
Head, Cardiovascular Research Group
University of Western Australia
- Prof Louise Cullen
Pre-Eminent Emergency Medicine Specialist, Royal Brisbane Hospital
Clinical Trialist/Outcomes Researcher, University of Queensland
- A/Prof Stephen Quinn
Chair, Department of Statistics, Data Science and Epidemiology
Swinburne University of Technology
- Prof Jon Karnon
Deputy Director, Healthy Communities Theme, Flinders Health and Medical Research Institute
- A/Prof Cynthia Papendick
Senior Emergency Physician, Royal Adelaide Hospital
Central Adelaide Local Health Network
- Dr Phil Tideman
Clinical Director, Integrated Cardiovascular Clinical Network SA
Clinical Lead, State-wide Cardiac Clinical Network
- Prof Anton Van den Hengel
Director, Australian Institute of Machine Learning
- Dr Johan Verjans
Deputy Director, Medical Machine Learning
Australian Institute of Machine Learning
- Dr Maria Alejandra Pinero de Plaza
Post-doctoral Research Fellow in Consumer Engagement and Knowledge Translation
We will be conducting a randomised controlled trial in 12 hospitals throughout South Australia. Six hospitals will be randomly assigned to provide normal care with the additional support of computer algorithms and the other six will provide normal care only. By comparing individuals and their outcomes between these two groups, we will know if doctors can improve the care and outcomes of patients with the help of these algorithms.