Unable to develop formula for data collection

Sepsis is a leading cause of death worldwide, and contributes to approximately 750,000 hospitalizations per year, a third of which may die. International guidelines recommend timely interventions, including cultures, fluid resuscitation and antibiotics. Following guideline bundles is associated with a decrease in mortality. Key to timely intervention is timely diagnosis.

Using the Epic sepsis predictive analytic tool, investigators will trigger vital sign and delirium monitoring in patients determined to be at increased risk for developing future sepsis.

The primary objective of this study is to demonstrate reduced mortality in patients for whom the pre-sepsis algorithm threshold is met, and who enhanced monitoring.

A means to identify patients before they become ill may improve the effectiveness of established therapies.Epic's electronic medical record (Epic systems, Verona, WI) contains a surveillance tool that uses predictive analytics to identify patients at risk of becoming septic four hours after the alert becomes active. This affords the opportunity to intervene sooner, but it remains unclear what the best course of action should be in a population at risk of sepsis, only some of which may go on to develop the illness. We propose an automatic intervention, consisting of enhanced monitoring, that is tied to the alert. No therapeutics will be mandated. Instead, additional monitoring information will lead to faster diagnosis and therapy, and improved clinical outcomes.

Inclusion Criteria:

• any inpatient at NYU Langone Health System

Exclusion Criteria:

• not an inpatient at NYU Langone Health System