A Balancing Act: Safer Decisions for Stroke Prevention in Atrial Fibrillation

Doctor Hannah Stevens, The University of Melbourne

Postdoctoral Fellowship

Years funded: 2026 - 2028

Atrial fibrillation (AF) is a common heart condition affecting over half a million Australians. People with AF have a high risk of stroke, which causes long-term disability and significant healthcare costs. Blood-thinning medications, called anticoagulants, are effective at preventing stroke but come with a risk of bleeding complications.

Currently, doctors use scoring systems such as the CHA2DS2-VASc which considers factors like age, blood pressure, and diabetes, to assess stroke risk and guide anticoagulant use. However, these scores leave many people - around 75,000 Australians - at "intermediate risk", where there is uncertainty about whether the benefits of anticoagulants outweigh the bleeding risks.

The solution:

This project aims to address this uncertainty and provide personalised guidance by: 1) Discovering new risk markers: Using advanced methods to identify new markers of stroke and bleeding risk by evaluating blood proteins, metabolites, heart imaging, and demographic factors such as rural or remote living. 2) Create an enhanced prediction tool: Combining these new markers with existing clinical information to clearly identify who would most benefit from anticoagulants. 3) Implement an app for smartphones to assist with clinical decision-making, and patient education: Using the prediction tool and turning this into a user-friendly application to help doctors make clear, informed decisions and educate patients about treatment options.

Impact of this research: 1) Improved risk prediction for stroke and bleeding: By identifying new biomarkers and combining them with clinical and demographic data, this project will develop a more accurate risk prediction score. Based on previous studies, this new tool could improve accuracy by up to 40%, potentially reclassifying 30,000 Australians who currently receive uncertain treatment advice. 2) Safer treatment decisions: With better risk prediction, doctors can tailor anticoagulation to ensure it is only used for people who will truly benefit. This will reduce the number of avoidable strokes as well as unnecessary bleeding complications through improved risk assessment. 3) Empower doctors and patients through education and technology: The prediction tool will be built into a smartphone app to provide clear information to doctors, and improve patient confidence in treatment choices. 4) Reduce healthcare costs by preventing strokes and bleeding complications that require hospital inpatient treatment.

Overall, this research will transform new types of data into clear and practical recommendations. By using this refined prediction tool incorporated into a smartphone app, this research will clarify treatment recommendations for thousands of patients with AF, resulting in improved outcomes and quality of care.