British electronic patient record (EPR) provider PatientSource has introduced a cutting-edge AI tool at the Royal Hospital for Neuro-disability (RHN) to enhance early detection of patient deterioration.
A pilot program commenced at RHN in January 2023, utilizing five years’ worth of data stored in the PatientSource EPR. Data scientists and researchers at RHN employed machine learning and AI to develop and enhance the ‘MEMORI’ AI tool, comparing its digital results with real-world historic outcomes.
The MEMORI tool leverages the National Early Warning Score (NEWS) and analyzes current observations alongside a deep dive into a patient’s historical data within the PatientSource EPR, covering various biomarkers, test results, and clinical notes. This provides a personalized patient profile, enabling early detection of deterioration up to 48 hours earlier than the traditional NEWS system.
After successful ‘shadow analysis’ at the point of care, the live data confirmed the tool’s efficacy, leading to its implementation in the Draper’s Ward for brain injury patients in June 2024. The pilot is set to conclude in Autumn 2024, followed by a comprehensive evaluation before potential rollout across RHN’s other wards.
Dr. Mike Brooks, co-founder and chief medical officer at PatientSource, highlighted the goal of identifying patient deterioration before any vital signs show abnormalities. Active discussions are underway with NHS trusts in England regarding the tool’s implementation.
PatientSource takes a modular approach to EPR deployment, differing from larger firms like Epic and Oracle Health. By gradually rolling out modules and integrating seamlessly with existing systems, organizations can enhance their EPR functionality and future-proof their operations.
