Dr Megan Prictor1, Dr Gun Soin2, A/Prof. Melissa Mccradden3,4
1Melbourne Law School, The University Of Melbourne, Carlton, Australia, 2Sonic Healthcare Australia Clinical Services, Sydney, Australia, 3Australian Institute for Machine Learning, University of Adelaide, Adelaide, Australia, 4SickKids Research Institute, Toronto, Canada
Biography:
Dr. Megan Prictor is a senior lecturer at Melbourne Law School, the University of Melbourne, and co-director of the Health, Law and Emerging Technologies research programme. She has conducted extensive research on legal issues in healthcare consultation recording and other clinical technologies.
Dr Gun Soin is a practicing GP with over 25 years of clinical experience across primary and secondary care in the UK and Australia and is the Chief Medical Officer of Sonic Clinical Services, the largest operator of primary care medical practices in Australia.
Dr. Melissa McCradden is the Women's and Children's Health Network Artificial Intelligence Director and The Hospital Research Foundation (THRF) Group Clinical Research Fellow in Ethics of AI at the Australian Institute for Machine Learning at the University of Adelaide. She is an Adjunct Scientist with the SickKids Research Institute.
Abstract:
Facilitator – Dr Megan Prictor, Melbourne Law School
Dr Gun Soin, Sonic Healthcare Australia Clinical Services
AI in Healthcare: Introduction
Recent progress in the capabilities of “Artificial Intelligence” driven by Large Language Models (LLMs) has led to rapid growth in the development and deployment of AI in healthcare settings. Implementations include the streamlining of administrative tasks, AI scribes for facilitating medical record keeping, and clinical decision support tools. While the use of AI has the potential to improve efficiency and the quality of healthcare delivery, we need to ensure that the core pillars of medical ethics are maintained, while traditional regulatory frameworks try to keep up with the pace of change.
Megan Prictor, Melbourne Law School
Mapping Legal Risks of Medical AI scribes
AI scribes may both heighten and mitigate legal risks. In their favour, these products can improve the accuracy and completeness of record keeping, and reduce physician burnout. Against this, privacy risks include a lack of effective patient consent; data leakage; and secondary use of data. Moreover, any errors in the patient record remain the responsibility of the healthcare practitioner.
Dr Melissa McCradden, University of Adelaide
The Evidence and the Ethics of Medical AI Scribes
AI scribes are of immense interest to reduce documentation-related burdens on clinicians. Despite this promise, the evidence at present shows mixed reviews. Imprecise implementation strategies and insufficient validation raise substantial risks for clinicians and organizations alike. Moreover, concerns like the environmental impact and biases are sliding under the radar. What would we need to use AI Scribes in an evidence-based, ethics-forward way?