Sepsis is a complex clinical situation that may not be well defined and, for some adult acute care units at UCSF, a rare event that makes it difficult for nursing staff to convey urgency to physicians--and for physcians to respond in a timely manner. 

Despite efforts to improve sepsis detection and management within both the adult acute and critical care settings over the past several years, including the implementation of best practice alerts (BPA), gaps in the continuum of care persist. One-third of sepsis cases are acquired during the patient's hospital stay, resulting in more than 500 deaths per year. 

An inter-disciplinary team with expertise in clinical/nursing workflow, service deisgn, data analytics and operations came together to understand the current state of sepsis in adult patients. The team used mixed methods to understand the human and system factors associated with the current sepsis signaling and management workflows across different service lines. 

The outcome from the initial discovery work frames a care path way that includes signal design using AI and a more cohesive management system.