Presentation
Engineering system safety: Liability informed healthcare system safety design
DescriptionGlobally and within Canada, each year patients die from preventable patient safety incidents, insomuch that harm arising from patient safety incidents continues to be one of the leading causes of preventable death. Of these preventable patient safety incidents, factors/characteristics that influence risk outcomes in Canadian healthcare are largely unknown.
There is a paucity of aggregated patient safety incident data in Canada and there is not a standardized mechanism to monitor or learn from past patient safety incidents nor prevent or predict future harm. Analysis of patient safety incidents largely occurs at the institutional level, which does not afford a pan-Canadian systemic lens or system improvement based on aggregated learning across incident response data. At present, there are no prescriptions on how to engineer system safety to reduce top preventable patient safety risks in Canada.
Medico-legal insurance data may provide valuable knowledge for frontline healthcare decision-makers including human factors practitioners, but at present there is an insufficient systems approach in healthcare safety design. An increase in accessibility and new opportunities in applying machine learning to generate insights and risk prediction models exist.
This oral presentation will explore applying active machine learning to Canadian medico-legal claims data to seek understanding on systemic factors that may improve patient safety and reduce risk. This session will explore current approaches to the application of systems theory in patient safety, share lived experience on the design of machine learning risk prediction models using medico-legal data and the application of risk prediction models in identifying preventable systemic factors for human factors practitioners, healthcare leaders, administrators, and researchers interested in improving system safety.
Takeaway Points:
• Liability informed system safety design adds a data driven approach, based on true risk outcomes, to identify and prioritize safety and risk reduction.
• Share a set of features and research questions that address patient safety at a systems level
• Explore a ranking of Canadian risk outcomes to support human factors practitioners, healthcare leaders, administrators and researchers in system safety design.
There is a paucity of aggregated patient safety incident data in Canada and there is not a standardized mechanism to monitor or learn from past patient safety incidents nor prevent or predict future harm. Analysis of patient safety incidents largely occurs at the institutional level, which does not afford a pan-Canadian systemic lens or system improvement based on aggregated learning across incident response data. At present, there are no prescriptions on how to engineer system safety to reduce top preventable patient safety risks in Canada.
Medico-legal insurance data may provide valuable knowledge for frontline healthcare decision-makers including human factors practitioners, but at present there is an insufficient systems approach in healthcare safety design. An increase in accessibility and new opportunities in applying machine learning to generate insights and risk prediction models exist.
This oral presentation will explore applying active machine learning to Canadian medico-legal claims data to seek understanding on systemic factors that may improve patient safety and reduce risk. This session will explore current approaches to the application of systems theory in patient safety, share lived experience on the design of machine learning risk prediction models using medico-legal data and the application of risk prediction models in identifying preventable systemic factors for human factors practitioners, healthcare leaders, administrators, and researchers interested in improving system safety.
Takeaway Points:
• Liability informed system safety design adds a data driven approach, based on true risk outcomes, to identify and prioritize safety and risk reduction.
• Share a set of features and research questions that address patient safety at a systems level
• Explore a ranking of Canadian risk outcomes to support human factors practitioners, healthcare leaders, administrators and researchers in system safety design.
Event Type
Oral Presentations
TimeTuesday, March 243:30pm - 3:50pm EDT
LocationMurray Hill East
Patient Safety Research and Initiatives