22 - Implementation of decision support for the evaluation of new fever in the pediatric intensive care unit
Saturday, April 23, 2022
3:30 PM – 6:00 PM US MT
Poster Number: 22 Publication Number: 22.201
Anna Sick-Samuels, Johns Hopkins University School of Medicine, Baltimore, MD, United States; Lauren D. Booth, Johns Hopkins Hospital, Baltimore, MD, United States; Aaron Milstone, Johns Hopkins University School of Medicine, Baltimore, MD, United States; Christina Schumacher, Johns Hopkins University School of Medicine, Baltimore, MD, United States; Jules P. Bergmann, Johns Hopkins University School of Medicine, Balitmore, MD, United States; Lindsey Gnazzo, Johns Hopkins University School of Medicine, Baltimore, MD, United States; David C. Stockwell, Johns Hopkins Children's Center, Baltimore, MD, United States
Assistant Professor Johns Hopkins Children's Center baltimore, Maryland, United States
Background: There is variability in diagnostic testing practices among pediatric intensive care unit (PICU) patients and clinical decision support tools can standardize testing practices. Previously, our PICU implemented blood culture and endotracheal culture guidelines. We developed a comprehensive infectious disease testing algorithm for critically ill children with new onset of fever incorporating guidance for blood, endotracheal, and urine cultures and other considerations.
Objective: The objective of this study is to describe the algorithm development, implementation process and impact on testing practices.
Design/Methods: This was a mixed-methods study to evaluate the impact of a quality improvement project that implemented a novel “PICU Fever algorithm” at a single center quaternary children’s hospital in July 2020. Impact on monthly blood culture, endotracheal culture, and urine culture rates per 1,000 ICU patient-days was monitored with statistical process control charts, and incident rate ratios (IRR) comparing one year before and after algorithm implementation. Monthly electronic safety surveys of attending and fellow physicians were conducted for one year post-implementation.
Results: Introduction of a PICU fever algorithm was associated with reductions in blood cultures by 13% (incident rate ratio (IRR) 0.87, 95%CI 0.80-0.95), endotracheal cultures by 31% (IRR 0.69, 95%CI 0.55-0.85), and a urine cultures by 23% (IRR 0.77, 95%CI 0.66-0.91). Forty-six of 108 (43%) invited physicians replied to the safety surveys with 39 (85%) reporting having used the algorithm during the prior service week, none reported a patient safety concern, 2 (4%) provided constructive feedback, and 28 (61%) felt the algorithm improved patient care.Conclusion(s): A comprehensive PICU fever algorithm maintained prior reductions in avoidable blood and endotracheal cultures and reduced avoidable urine cultures without detection of patient harm by physician report. This study suggests that algorithms guiding evaluation of patients with fever promote judicious testing - both avoiding testing over-use and promoting diagnostic differentials to improve patient management in critically ill children.