NHS England Podcast Machine learning and LFPSE – revolutionising how we learn from patient safety events
Feb 15, 2023
In this insightful discussion, Maia Cassis, a patient safety data scientist, and Marcus Mann-Harris, head of LFPSC and NRLS, explore the groundbreaking impact of machine learning on patient safety. They explain how machine learning can unveil hidden patterns and enhance data processing to identify safety risks faster than manual reviews. Delving into technical models being tested and user engagement strategies, they highlight the promise of real-time feedback and improved outcomes for patients, revealing an exciting future for healthcare safety.
AI Snips
Chapters
Transcript
Episode notes
ML Unlocks Hidden Patient Safety Patterns
- Machine learning will reveal hidden patterns and themes across huge patient safety datasets.
- This enables earlier detection of risks that humans alone could take decades to find.
Models Will Be Built Into LFPSE Workflow
- LFPSE will embed ML models into production so users interact through live applications.
- Topic and novelty models are already being tested with clinicians and linked to taxonomy fields.
Iterate With Clinician Feedback
- Continue clinician evaluation and fine-tuning of models before wide rollout.
- Expand user engagement so feedback improves and refreshes model performance.