A new project led by Dr Anisur Rahman and Professor Damminda Alahakoon, in partnership with St Vincent’s Hospital Melbourne, will use patient records to build AI models with the potential to support decision making in emergency departments.
“A prolonged length of stay in an emergency room can result in negative outcomes, from increased mortality and ambulance ramping to higher healthcare costs and staff burnout,” Dr Rahman explains.
“Our project looks specifically at patient disposition – how a clinician determines if the patient is ready to go home, or needs to be admitted to an inpatient ward or short stay unit. This is an important part of improving patient outcomes in emergency rooms.”
The project, the first and largest multi-site study of its kind, will use AI to predict patient disposition, supporting more accurate and efficient decision-making.
Associate Professor Hamed Akhlaghi, Head of Emergency Medicine Research and Emergency Consultant at St Vincent's Hospital, said the project aim is to “leverage AI to enhance patient care in emergency departments through collaboration and partnership with La Trobe University.”
“We collected a unique set of data from various interstate and intrastate health services, including WA and Tasmania, to truly capture the heterogeneity in Emergency Department triage notes.”
“The project's outcome investigates AI decision-making bias and provide clinicians with more accurate predictions, benefiting both patients and healthcare providers," he adds.
Dr Rahman says he hopes the study will assess whether AI can assist in improving patient assessment in emergency departments, with a focus on exploring its potential role in supporting timely and appropriate care for high-risk individuals.
“We also hope to explore whether guidelines could support the integration of AI into clinical settings, assessing its impact on cost savings and emergency department outcomes.”
