TOP 61 - The role of corticosteroids in pediatric orbital cellulitis

Publication Number: 1820.TOP 61

Amanda Cooper, Kaiser Permanente - Oakland Medical Center, Pleasanton, CA, United States; Miranda Weintraub, Kaiser Permanente - Oakland Medical Center, Oakland, CA, United States; Matthew Weymer, Kaiser Permanente Division of Research, Oakland, CA, United States; Tara L. Greenhow, Kaiser Permanente San Francisco, San Francisco, CA, United States; Zahra Samiezade-Yazd, Kaiser Permanente Division of Research, Pleasanton, CA, United States; Lea Bornstein, Kaiser Permanente - Oakland Medical Center, Oakland, CA, United States; Tran Nguyen, Kaiser Foundation Hospital - Roseville Women's and Children's Services, Roseville, CA, United States; Emily Kragel, Seattle Children's, Seattle, WA, United States

Background: Orbital cellulitis is an infection of the post-septal tissues of the eye that can be associated with serious complications including intracranial extension and vision loss. While the cornerstone of treatment is antimicrobial therapy, there is no overarching consensus on the role of corticosteroids in the management of orbital cellulitis. Existing literature examining the use of corticosteroids to improve outcomes in orbital cellulitis is limited by the sample size and generalizability of results.

Objective: Among pediatric patients with orbital cellulitis, we aimed to 1) describe demographic and clinical characteristics; 2) describe variations in corticosteroid use; and 3) assess association of corticosteroid use with differences in patient outcomes including length of stay in the hospital, intravenous antibiotic course, duration of elevated pain scores, and documented fevers.

Design/Methods: This is an IRB-approved retrospective cohort study of 190 pediatric patients with imaging-confirmed orbital cellulitis treated in an integrated health system between 1/1/2010 and 12/31/2024. All charts were reviewed manually for eligibility criteria and severity classification, presenting symptoms, medication data, pain scores, and fever data were manually extracted. Demographics and lab values were electronically extracted. Descriptive statistics were used to summarize study population clinical and demographic characteristics, variations in treatment, and outcomes. Chi square analyses were used to assess differences between the antibiotic only and antibiotic plus steroid group with regard to length of hospitalization, duration of elevated pain score (>3), duration of fever, antibiotic duration, surgery, and readmission within 30 days. Models will be adjusted for severity of infection (based on imaging and presenting symptoms) at presentation using both linear and logistic regression. All analyses were performed using RStudio Version 2025.05.1+513 and Statistical Analysis Systems (SAS) 9.4.
Data collection and Table 1: preliminary results attached. Initial results do not show significant differences in outcomes between patients who received steroids and those who did not. Steroid dosing will be described.
Regression analyses (adjusting for severity at presentation): By Dec 1. Preliminary results indicate when adjusted for category of severity, steroid use was not significantly associated with reduced or prolonged length of stay or pain duration. After adjustment it was also not associated with readmission.

Conclusions: Jan 1

Cohort Demographic Characteristics, Presenting Signs/Symptoms, and Unadjusted Outcomes
Table1.pdf