Thanks to the PA Research Foundation, PhD student Hetiao (Slim) Xie at The University of Queensland has been able to advance his PhD research in improving emergency care performance using an AI based tele-triage system.
The Rethinking Emergency Medicine Research (REMR) project, has just been approved for more funding, allowing Slim and the research team to expand their study.
“Many of us have had the experience of waiting for hours in the emergency department (ED). Most people’s conditions aren’t serious, so they get assigned a lower triage category and end up suffering unnecessary waiting and anxiety. Recently, Queensland established Minor Injury and Illness Clinics (MIIC) to specifically treat these patients. However, according to both statistical reports and insights from clinical experts, a lot of patients still end up presenting to either the ED or MIIC inappropriately,” Slim said.
“That really got us thinking about whether we can design a way to coordinate patient flow and strengthen the collaboration between EDs and MIICs; these emergency care facilities aren’t being fully utilised according to their intended purpose. Many patients sometimes self-refer to the wrong setting that doesn’t match the severity of their condition, which affects the quality of emergency care they receive as well as adding to the overall pressure on the emergency system.
“Through our exploration and interviews with clinical experts, we narrowed our solution to designing an AI tele-triage system that can guide patients to the emergency facility best suited to their clinical needs.
Our focus on using AI is driven by two main reasons: (1) Powerful AI algorithms can uncover hidden patterns within large scale patient data that even experienced clinicians cannot identify, thereby improving the prediction of whether a patient should be assigned to the ED or MIIC. (2) The scalability of AI allows us to significantly reduce the need for additional investment of medical personnel in online triage, helping to ease the strain on healthcare resources and lower operational costs.
“What makes this approach particularly promising is that it enables patient flow and triage at the most appropriate stage. Once a patient has already presented to the ED, the current policy makes it practically infeasible to redirect them to another facility, even if they recognise the patient would be better suited elsewhere. Our approach instead provides patients with decision support before self-referral, helping them to identify the most appropriate emergency facility and ensuring the effectiveness of pre-hospital triage.”
Slim expressed his gratitude, explaining how the project is showing promising developments and now, thanks to the Foundation’s donors, can move forward and be further developed.
“The Foundation played a pivotal role in making this project possible. Before receiving the REMR Grant from the Foundation, I was a junior research student hoping to bring my background in data science and information systems into the healthcare context. After receiving the grant, I gained the funding and the opportunity to collaborate with experienced clinical experts. The Foundation offered the resources and opportunities to transform my research ideas and knowledge into a real-world healthcare innovation,” he said.
“To the healthcare system, our proposed system can help both MIICs and EDs be utilised more effectively, allowing each to fulfill its intended function. The optimised patient flow will reduce overcrowding in ED and unnecessary healthcare costs. Meanwhile, MIIC can play a more effective role in community healthcare.
“Additionally, low acuity and high acuity patients can receive appropriate emergency care at MIIC and ED respectively. This will reduce their waiting time and risk of not receiving timely treatments, as well as improving the quality of care.
“To hospital staff, the optimised patient distribution will allow them to focus more effectively on their core responsibilities, reducing unnecessary workload and alleviating the emotional stress associated with an overcrowded environment.
“My key objective for the next phase is to deliver a proof of concept on the design of an AI system that safely and efficiently predicts assignments of patients to ED vs MIIC in tele triage. Patients can type a description of their symptoms into the system to receive an online triage recommendation indicating whether they should attend the ED or MIIC.
“I would like to once again express my sincere gratitude to the PA Foundation for their generous sponsorship and strong support. As a PhD student, the research grant and industry collaboration opportunities provided by the Foundation have enabled me to take on this real and emerging healthcare challenge as the focus of my PhD research project. This support has empowered me to work with dedication to translating research knowledge into meaningful and practical outcomes.”
