P2D188: The Sepsis Response: Improving Early Identification

Purpose/Objectives: Sepsis is a leading cause of hospitalization and among the top 10 causes of death in children < 10 years of age. There are >75,000 pediatric sepsis hospitalizations per year and approximately 7,000-10,000 pediatric deaths as a result. As many as 8% of pediatric sepsis cases go unidentified. Our project aim is to improve early identification and time to treatment of patients meeting sepsis criteria to prevent morbidity and mortality.

Design/Methods: Our EMR notification system triggers a “sepsis alert (SA)” when vitals and labs are suggestive of sepsis. In August 2022, in addition to the electronic alert, a “sepsis huddle (SH)” was implemented throughout the hospital, where physician and nursing providers assess the patient at the bedside and determine as a team whether the patient meets sepsis criteria and requires intervention. During cycle 1, resident floor teams were educated regarding the newly implemented SH and associated documentation. Cycle 1 data analysis included a review of 127 charts pre-intervention (April 16th, 2022-August 15th, 2022) compared to 202 charts post-intervention (August 16th, 2022-January 31, 2023) that triggered a sepsis alert. Charts were chosen for review if a SA was triggered and the following inclusion criteria were met: age 29 days-20 years, temperature >38.5℃ or < 36℃, and the patient had a suspected or proven infection. Cycle 2 focuses on education regarding use of a sepsis order set (SOS) and timely intervention with IV fluid resuscitation and antibiotics ( < 1 hour). Cycle 2 is currently ongoing and will run from February 1, 2023-May 30, 2023.

Results: Cycle 1 data showed a 100% improvement in huddle compliance and 16% improvement in documentation completion. Average time to documentation decreased by 39%. These measures showed a >10% improvement following intervention, exceeding our goals. While cycle 2 is currently ongoing, promising results are seen with an improvement in the percentage of patients receiving intervention in < 60 min compared to baseline data. There has been an increase of 49% for blood culture collection in < 60 min, 52% for antibiotic administration in < 60 min, and 29% for fluid bolus administration in < 60 min.

Conclusion/Discussion: Resident education and active participation in the implementation of a sepsis early identification and management pathway led to an increase in SH compliance in response to SAs and decreased time to associated documentation. Ideally, SAs along with implementation of SHs will increase identification of sepsis, ultimately leading to decreased morbidity and mortality. Additionally, through preliminary analysis of cycle 2 data, we have identified trends suggestive of an increased percentage of patients receiving intervention in < 60 min compared to baseline data.