Development, Implementation, and Impact of an Automated Early Warning and Response System for Sepsis

In the Journal of Hospital Medicine, Craig Umscheid, Benjamin French, and other Penn colleagues investigate electronic health record (EHR)-based interventions aimed at reducing sepsis-related mortality. Severe sepsis affects as many as three million patients in the U.S. annually and kills 750,000. Earlier intervention could help to lower the mortality rate, but identifying at-risk patients is a challenge. Umscheid and colleagues propose that a better screening mechanism would help providers recognize and treat sepsis right away. To test this, they conduct a study where the EHRs of adult non-ICU patients in acute inpatient units are programmed to alert the provider, nurse, and rapid response coordinators whenever a patient’s vital signs and laboratory metrics show certain predetermined abnormalities. The use of this automatic prediction tool was found to promote a statistically significant increase in early sepsis care, ICU transfer, and sepsis documentation. The authors also find a related decrease in sepsis mortality, but this data was not statistically significant.