Background:
Despite widespread awareness of best practices in the early management of sepsis, sepsis continues to burden our healthcare systems with high mortality, prolonged length of stay and excessive cost. Early recognition of sepsis and adherence to evidence-based initial resuscitation protocols is known to reduce both mortality and costs associated with care. In 2013, a multidisciplinary team was formed at Rush University Medical Center which included leadership from critical care medicine, emergency medicine, internal medicine, nursing, pharmacy, information systems, quality improvement and finance to discuss potential strategies to improve and standardize care for patients with sepsis. In addition, the Centers for Medicare & Medicaid Services (CMS) announced a new measure, the Severe Sepsis/Septic Shock Early Management Bundle (SEP-1), to monitor the quality of sepsis care within hospitals in 2015.
Purpose:
Our goal was to design and integrate clinical decision support tools within current inpatient workflows to promote early recognition and prompt initiation of evidence-based treatment for sepsis.
Description:
To facilitate early recognition of sepsis, we developed a real-time alert within our electronic health record (EHR) to notify both the nursing staff and physicians involved in the care of specific patients with a concern for sepsis. Our alert attempts to mimic the clinical definition of sepsis. The logic which triggers our alert looks at discrete data elements, such as vital signs and laboratory data, to determine if a patient has evidence of a systemic inflammatory response but also considers physician ordering practices to determine if an infectious process is suspected. By using these ordering practices as a surrogate for clinical suspicion of an infection, we were able to minimize false positive events and subsequent alert fatigue. An order set is then suggested to physicians who receive this alert to facilitate the timely use of evidence-based management for sepsis and to ensure compliance with the CMS core measure. This sepsis management order set includes orders for fluid resuscitation, empiric antibiotic regimen, blood culture collection, and lactate monitoring. Since the CMS core measure for sepsis management mandates detailed, time-sensitive reassessment documentation for patients in septic shock and continued lactate monitoring when the initial lactate value is elevated, we also developed a subsequent prompt within our EHR to remind our clinicians to complete the full bundle of care for our patients with sepsis. These alerts and order set are intended to augment our current clinical workflows and help to coordinate standardized care for our patients with sepsis.
Conclusions:
We describe the development of integrated clinical decision support tools to facilitate early recognition of sepsis and prompt initiation of standardized, evidence-based care. Our tools were designed by a multidisciplinary team to align with current clinical workflows and not oppressively burden our clinicians involved in patient care. Future efforts are aimed at demonstrating improved compliance with the SEP-1 core measure due to these clinical decision support tools and improved abstraction based on these automatically identified patient cases.