A robust learning software platform capable of combining both patient physiologic monitor alarms and data in EMR (e.g., laboratory tests) to more precisely monitor patients.
Patient physiologic monitors are an essential tool for medical staff. However, the excessive number of monitor alarms leads to desensitized or failed responses to true clinically significant events and compromises the quality of patient care. The presented learning algorithm meets this unmet need for reliable patient monitoring systems by providing more accurate prediction of patent deterioration and the opportunity for early medical intervention. For example, it shows 90% sensitivity for predicting code blue two hours ahead of time and 85% reduction in false alarms.
Additional advantages of this invention include:
Researchers at University of California, San Francisco have developed an advanced analytic software platform for more precise patient monitoring and prevention of alarm fatigue in medical staff. It monitors the streaming data from monitors and EMR (e.g., alarms and lab test results) and identifies patterns among these data that signal patient deterioration.
To develop & commercialize the technology as an analytic software module that can extend the functionality of patient monitors and EMR systems
Proof of Principal
|United States Of America||Published Application||20170046499||02/16/2017||2014-081|
Digital Health, Health Informatics, Health Analytics, EMR, Alarm Fatigue, Patient Monitoring, Patient Deterioration, Rapid Response, Predictive Model, Machine Learning