WIP 29 - Outcomes of Antibiotic Prescription for Respiratory Infections Among Hospitalized Children with Tracheostomies with a Positive Viral Test

Publication Number: WIP 29.7542

Katarina Manzi, Seattle Children's, Seattle, WA, United States; Kawsar A. Hirsi, Seattle Children's, Seattle, WA, United States; Naoko Kono, Keck School of Medicine of the University of Southern California, Los Angeles, CA, United States; Teoman Toprak, Seattle Children's, Bellevue, WA, United States; Christopher Russell, Stanford University School of Medicine, Palo Alto, CA, United States; Sarah C. Hofman DeYoung, University of Washington School of Medicine, Seattle, WA, United States; Jonathan Cogen, University of Washington School of Medicine, Seattle, WA, United States

Background: Children with tracheostomies often are hospitalized with respiratory symptoms and receive antibiotics to treat tracheitis and/or pneumonia. When these children test positive for a viral infection, antibiotics are still frequently prescribed, even though limited data exists to inform this practice.

Objective: Among children with tracheostomies who have a positive viral panel and require hospitalization for respiratory symptoms, this study aimed 1) To describe the proportion who received antibiotics, 2) To determine factors associated with antibiotic prescription on admission, and 3) To determine if children who received antibiotics had a lower readmission rate compared to children who did not receive any antibiotics.

Design/Methods: This study was approved by the Seattle Children’s IRB and is a retrospective cohort study including children with tracheostomies ≤ 21 years of age admitted to the hospital with a respiratory PCR confirmed infection. Other inclusion criteria include having a tracheal aspirate obtained within 24 hours of admission. We divided children into two groups: those who received ≥1 dose of an antibiotic (Group 1) and those who did not receive any antibiotics (Group 2). Clinical and demographic characteristics will be summarized with descriptive statistics (e.g., counts, proportions, means (SD) and medians (IQR) when appropriate). To look at factors associated with antibiotic prescription, a logistic regression model will be performed that includes relevant covariates of interest (e.g., ventilator use (yes/no), white blood cell count elevation (yes, no), enteral steroids prescribed (yes/no), etc.) In addition, a logistic regression model with generalized estimating equations (GEE) will be performed to look at the odds of readmission (yes/no) between Groups 1 and 2 (with the same covariates of interest included). Analyses to be completed by early 2025.