Why Mayo Clinic Leverages Real-World Data to Promote Precision Medicine

Highlights

• The Mayo Clinic uses healthcare-specific big language models in a generative AI conversation app to enhance patient care and clinical decision-making
• An AI chatbot helps clinicians explain how AI models created on evidence-based healthcare data may accelerate research and improve patient access

Mayo Clinic clinicians are exploring the use of healthcare-specific large language models (LLMs), accessed through a generative artificial intelligence (AI) chat application, to improve patient care and clinical decisions. Dr. Peter Noseworthy, Chair of Cardiac Electrophysiology at Mayo Clinic, suggests that utilising data from millions of patients could aid in identifying similar patients and understanding their outcomes.

Assessing Healthcare-Specific Data Reliability

Atropos Health, a California-based healthcare data network has partnered with Mayo Clinic to provide physicians plus researchers with access to the deidentified data repository and its analytical tools. This collaboration enables real-time interaction with real-world data, providing insights at the point of care for critical care patients. The platform can save time by allowing care teams to find answers to research questions in a matter of days, unlike traditional methods that could take weeks to determine treatment.

The Potential Value of Patient Information Originates with AI

Mayo Clinic is implementing a decentralised clinical trial program, piloting ChatRWD, a real-time data generation tool that creates real-world clinical data at the point of care. This approach, unlike ChatGBT or other LLMs, has provided valuable insights, although it has not been deployed at scale at Mayo Clinic.