590 - Treating Pediatric Deep Neck Infections: A Cohort Study Comparing Early, Late, and Non-Surgical Approaches

Publication Number: 2575.590

Lauren G. Tapp, Medical University of South Carolina College of Medicine, Charleston, SC, United States; Daniel L.. Brinton, MUSC Children's Hospital, Charleston, SC, United States; Elizabeth Oddo, Medical University of South Carolina College of Medicine, Charleston, SC, United States; Phayvanh P. Pecha, Medical University of South Carolina College of Medicine, Charleston, SC, United States; Kelly Roelf, Medical University of South Carolina College of Medicine, Charleston, SC, United States; Daniel Williams, Medical University of South Carolina College of Medicine, Charleston, SC, United States

Background: Deep neck infections (DNI) in children cause significant morbidity and often require hospitalization, with estimated annual charges of $75 million. Management includes both medical and surgical approaches however hospital courses vary due to limited evidence to suggest if and when surgery is indicated. Understanding how treatment affects length of stay (LOS) may guide clinical practice and improve system efficiency.

Objective: To evaluate the impact of treatment groups (medical management only, early drainage, late drainage) on LOS among children hospitalized with DNI.

Design/Methods: This retrospective cohort study used 2015-2024 Pediatric Health Information System data. Patients were identified based on diagnosis codes of retropharyngeal/parapharyngeal abscess, peritonsillar abscess, and other abscess of pharynx. Patients receiving antibiotics on day 0 or 1 of hospitalization were included. Transfers and those with complex chronic conditions were excluded. Procedure codes were extracted to identify presence of surgery. Patients were separated into treatment groups of early drainage ( 48 hours from admission), late drainage (> 48 hours), and medical management (no drainage). Bivariate analyses compared demographic and clinical variables across groups. Variables associated with treatment group were included in multivariable Poisson regression models to estimate associations with LOS. Secondary analyses among surgically managed patients evaluated post-drainage LOS.

Results: Our cohort included 22,140 children. 4,912 underwent surgical drainage. Independent predictors of longer LOS included higher disease severity (p < 0.001), ICU admission (p < 0.001), and mental health comorbidity (p < 0.05). Younger age and non-Hispanic Black race were associated with longer LOS (p < 0.001). Late drainage was associated with a 79% increase in LOS when compared to early drainage (3.4 vs 1.9 days; p < 0.001). LOS was similar between medical management and early drainage groups (p=0.055). In those managed surgically, late drainage was associated with marginally shorter post-operative LOS (1.3 vs. 1.4 days; p=0.026).

Conclusion(s): In children hospitalized with DNI, treatment groups of exclusive medical management and early surgical drainage were associated with shorter total LOS after controlling for factors including illness severity. Recovery time did not differ meaningfully between early and late surgical groups. These findings suggest that effective medical management and early coordination between pediatric hospitalist and surgical teams may help improve patient outcomes and costs associated with pediatric DNI.

Table 1
Demographics, characteristics, and select outcomes

Figure 1
Length of Stay