Implementing a screening algorithm for early recognition of sepsis in hospitalized children: a quality improvement project

• Published in: International journal for quality in health care 2023-02, Vol.35 (1)
• Main Authors: Feinstein, Yael, Kogan, Slava, Dreiher, Jacob, Noham, Ayelet, Harosh, Shimrat, Lecht, Jenia, Sror, Tzipi, Cohen, Nurit, Bar-Yosef, Eileen, Hershkowitz, Eli, Lazar, Isaac, Schonmann, Yochai, Greenberg, David, Danino, Dana
• Format: Article
• Language: English
• Subjects: Algorithms; Child; Child, Hospitalized; Humans; Intensive Care Units, Pediatric; Quality Improvement; Sepsis - diagnosis
• ISSN: 1353-4505; 1464-3677
• DOI: 10.1093/intqhc/mzad006

Abstract Sepsis is a leading cause of mortality in children. Utilizing a screening tool for early recognition of sepsis is recommended. Our centre had no screening tool for sepsis nor a standardized protocol for sepsis management. In December 2020, a screening algorithm for sepsis was implemented. The algorithm consisted of vital signs measurements in children with an abnormal body temperature, a pop-up alert, nurse’s and physician’s evaluation, and activation of a workup protocol. The project’s primary aim was to increase vital signs measurement rates in hospitalized children with abnormal body temperature from 40% to >90% within 6 months, by 1 June 2021, and sustain until 31 December 2021. Adherence to the algorithm and performance were monitored during 2021, and the outcomes were compared to the preceding 5 years and a control ward. The alert identified 324 children and 596 febrile episodes. Vital signs measurement adherence increased from 42.7% to >90% in 2 months. A nurse evaluated 86.4% of episodes, and a physician evaluated 83.0% of these. Paediatric intensive care unit (PICU) admission rates were lower in the intervention period vs. the pre-intervention period vs. the control ward (4.6% vs. 5.6% vs. 6.0%, respectively); the median PICU length of stay was shorter in the intervention vs. the control ward [2.0 (IQR 1, 4) vs. 5.5 (IQR 2, 7), respectively]. These differences were not statistically significant. During the intervention period, the adherence to vital signs measurements reached the goal of >90%. The alert system prompted an evaluation by caregivers and management according to the protocol. Further monitoring is needed to improve outcomes.