407 - Impact of a Multi-Center Collaborative on Antibiotic Duration for Acute Respiratory Infections in Pediatric Urgent Care

Publication Number: 2395.407

Amanda Nedved, Children's Mercy, Lenexa, KS, United States; Brian R. Lee, Children's Mercy Hospitals and Clinics, Kansas City, MO, United States; Luis E.. Sainz y Diaz, Children's Mercy, Kansas City, MO, United States; Matthew Kronman, University of Washington School of Medicine, Seattle, WA, United States; Adam Hersh, University of Utah School of Medicine, Salt Lake City, UT, United States; Victoria Parente, Duke University School of Medicine, Durham, NC, United States; Rana F. Hamdy, Children's National Health System, Washington, DC, United States; Rana E. El Feghaly, Children's Mercy Kansas City, UMKC, Kansas City, MO, United States

Background: Updated national guidelines for many common infections encourage shorter antibiotic durations.

Objective: To evaluate if our national quality improvement (QI) collaborative designed to reduce disparities in first-line antibiotic prescribing for common acute respiratory tract infections (ARIs) in pediatric urgent care centers (PUCs) impacted the use of antibiotic durations ≤ 7 days.

Design/Methods: The QI collaborative included 92 PUC sites across 9 healthcare organizations, spanning 22 states and Washington, D.C. Each site formed a dedicated QI team and employed Plan-Do-Study-Act (PDSA) cycles to address disparities in guideline-concordant first-line antibiotic selection. We performed a secondary analysis of encounters of patients aged 6 months to 18 years with ICD-10 diagnoses of ARIs resulting in an antibiotic prescription from April 2022 to September 2024. We consider the period when PDSA cycles occurred (October 2023 to April 2024) as a washout. We excluded patients who received azithromycin (typically prescribed for 5 days). We performed a multivariable, multilevel logistic model to compare the odds of having a duration ≤ 7 days during the 3 time periods. We evaluated differences in antibiotic duration across sociodemographic groups using a multinomial model.

Results: We included 108,874 encounters. The most common diagnosis for which antibiotics were prescribed was acute otitis media (AOM) (65,422; 60.1%). Table 1 describes the demographics of the encounters. After initiation of the collaborative, we saw an increase in antibiotic duration ≤ 7 days for all diagnoses, particularly pneumonia (13.7% to 70.3%) (Table 2). When compared with the pre-period, significantly higher odds were observed for washout (odds ratio [OR]: 2.97 [2.85, 3.08]; p< 0.001) and post-period (OR: 3.79 [3.63, 3.95]; p< 0.001). We found differences in duration associated with race, language, and insurance (Table 3).

Conclusion(s): Antibiotic duration was not a targeted metric in our QI collaborative. However, we observed a significant and sustained increase in prescriptions durations ≤ 7 days for common ARIs following project initiation. We also identified differences in duration across sociodemographic groups. These findings reinforce the benefits of collaborative efforts to improve antibiotic use to extend beyond the intended outcomes as participants embrace a culture of stewardship. However, more work is needed to understand underlying cause of differences in prescribing behaviors and develop interventions that promote equitable antibiotic use.

Table 1: Demographics of Included Encounters


Table 2: Frequency of Prescriptions ≤ 7 Days and > 7 Days by Race/Ethnicity, Payor Type, Language, and Diagnosis


Multinomial Model Comparing Antibiotic Duration Lengths Before and After the Intervention Period
Coefficients based on multinomial model with '5-7 days' as the reference outcome. Largest group used as reference for race/ethnicity, language, payor type and diagnosis.