Influencing outcomes with automated time zero for sepsis through statistical validation and process improvement

Area of Science:

• Medical Informatics
• Clinical Decision Support Systems
• Healthcare Quality Improvement

Background:

• Sepsis affects over 1.5 million Americans annually, with a mortality rate of at least 250,000 deaths each year.
• Early sepsis identification and intervention are crucial for reducing mortality and intensive care unit (ICU) resource utilization.
• Current electronic medical record (EMR) based sepsis algorithms require enhancement for improved performance and timeliness.

Purpose of the Study:

• To enhance an existing EMR-based sepsis algorithm through the integration of a real-time, automated time zero calculation.
• To improve the accuracy and timeliness of sepsis alerts for early identification.
• To increase adherence to evidence-based sepsis treatment bundles for early intervention.

Main Methods:

• Utilized enterprise-wide health information system data within a data lake platform for statistical validation of sepsis algorithms.
• Introduced a novel algorithm for automated time zero calculation.
• Conducted a post-hoc analysis comparing test performance, alert timing, bundle compliance, ICU length of stay, and mortality before and after the algorithm enhancement.

Main Results:

• The automated time zero algorithm improved the positive predictive value of sepsis alerts from 37% to 52%.
• Alerts became more timely, with 47% occurring within one hour of the calculated time zero, and 77% within the critical 180-minute SEP-1 window.
• Observed a 14% increase in bundle compliance, a 10% reduction in ICU length of stay, and a decrease in sepsis-related mortality.

Conclusions:

• The integration of a real-time, automated sepsis time zero calculation significantly improved the performance and timeliness of predictive sepsis alerts.
• Enhanced alerts delivered via a mobile application support clinicians and administrators in timely sepsis management.
• This technological advancement contributes to improved patient outcomes and more efficient healthcare delivery in sepsis management.