Programming the Future of Sepsis: Developing a Pragmatic, Real-Time Sepsis Prediction Model for Non-ICU Patients

The Opportunity

Sepsis is the dysregulated host response to an infection, which can result in life-threatening organ dysfunction. Nosocomial sepsis, though rare, is associated with worse outcomes. Early, effective therapy reduces mortality and cost. The goal of this project is to augment current sepsis prediction model performance to produce high sensitivity without sacrificing positive predictive value (PPV).

Our Approach

Dr. Michelson and his team augmented the sepsis prediction algorithm performance by optimizing for prospective real-time performance. They also developed an infrastructure to enable a sustainable and continuous, real-time electronic health record (EHR)-algorithm interface.

The Solution

Additional outcomes are not complete for this project, so please check back at a later date for more information.