Introduction
Macrolides, including azithromycin, are rarely considered first-line therapy for pediatric acute respiratory infections (ARIs) due to increasing resistance among common respiratory pathogens.Reference Jenkins and Farrell1 Between June and December 2024, there was a nationwide surge in Mycoplasma pneumoniae cases in children 0–17 years old.2,Reference Diaz, Hersh, Olson, Shah, Hall and Edens3 Although macrolides are generally active against M. pneumoniae which can present with upper or lower respiratory symptoms, national guidance recommends against treating most cases, as the infection is generally mild and self-limited,4 and there is little evidence supporting antimicrobial therapy for non-hospitalized children with M. pneumoniae-associated respiratory tract disease.4–Reference Bradley, Arguedas, Blumer, Saez-Llorens, Melkote and Noel6 Since 2022, the Sharing Antimicrobial Reports for Pediatric Stewardship Outpatient (SHARPS-OP) Collaborative has been providing quarterly outpatient antibiotic prescribing benchmarking reports to 22 pediatric institutions recruited through our larger collaborative.Reference El Feghaly, Herigon and Kronman7 In these reports, azithromycin prescribing rates for ARIs have consistently remained below 5% in emergency departments (ED)s, urgent care clinics (UC)s, and primary care clinics (PCC)s.
This study aimed to evaluate outpatient azithromycin prescribing during the nationwide M. pneumoniae surge, using a nationally representative sample of pediatric institutions.
Methods
Each institution uploaded de-identified, quarterly aggregate electronic medical record (EMR) reports on antibiotic prescribing metrics from EDs, UCs, and PCCs to a Research Electronic Data Capture (REDCap)® site managed by the coordinating center, which compiled and analyzed the data.Reference El Feghaly, Herigon and Kronman7,Reference El Feghaly, Herigon and Kronman8 We collected the proportions of acute ARI encounters, defined as acute encounters with International Classification of Diseases 10th version (ICD-10) ARI discharge diagnosis codes, that resulted in any oral antibiotic prescription. Among those, we collected the percentage of encounters prescribed amoxicillin (amoxicillin index) and azithromycin (azithromycin index).Reference El Feghaly, Herigon and Kronman7 The study period was dichotomized into presurge (January 2021–June 2024) and M. pneumoniae surge (July 2024–December 2024) periods. We compared prescribing metrics between these time periods and stratified aggregated data by practice settings (i.e., ED, UC, and PCC). We used unadjusted multilevel logistic models to compare prescribing patterns, with the institution as the random-intercept. We also completed difference-in-difference [DID] models. All analyses were performed using R software.
Individual institutions completed Institutional Review Board (IRB) approvals and/or data use agreements per their institutional requirements.
Results
Twenty SHARPS-OP institutions contributed data for the entire study period, totaling 5.8 million ARI encounters. We observed an increase in overall antibiotic prescribing for ARI encounters over the last 8 quarters (Q1 2023–Q4 2024) across all practice settings (Odds ratio [OR] 1.66; 95% confidence interval [CI] [1.65, 1.67]) (Figure 1A). During the M. pneumoniae surge period, the amoxicillin index declined by 5.8% (from 67.1% to 61.3%) (OR: 0.78; 95% CI [0.77, 0.79]), while the azithromycin index tripled (from 4.1% to 20.3%) (OR: 5.88; 95% CI [5.82, 5.95]). Compared to the azithromycin index increase in the ED (11.6%), the increase in PCC (19.0%) was significantly higher (DID: 1.16; 95% CI [1.05, 1.13]), whereas the increase for UCC (10.3%) was significantly lower (DID: 0.76; 95% CI [0.72, 0.79]) (Figure 1B). Effect coding of institution in the DID model demonstrated significant variability in azithromycin index increases (deviance χ2: P < .001), with some institutions showing no increase, while others reached an azithromycin index of over 30% in certain settings (Figure 1C).
A. trends in the rate of antibiotic prescribing for ARI encounters in the different practice settings. B. Trends in azithromycin index in the different practice settings C. Changes in azithromycin index among the different institutions between presurge period (Q1 2021–Q2 2024), and M pneumoniae surge period (Q3–Q4 2024). A. Yellow dots represent markers outside of the confidence intervals while blue dots represent markers within the confidence intervals. B. Yellow dots represent markers outside of the confidence intervals while blue dots represent markers within the confidence intervals. The M. pneumoniae surge period (last 2 dots) were grayed out because they were excluded when creating the centerline of the control chart (as to not skew the centerline). C. Blue dots represent presurge study period while red dots represent the M pneumoniae surge period. *Not all sites submitted data for all care settings.
Discussion
This multicenter study of nearly 6 million pediatric ambulatory ARI encounters demonstrates a striking rise in azithromycin prescribing accompanied by a notable decrease in amoxicillin prescribing during the M. pneumoniae surge.
Several factors may explain why prescribing increased despite guidance discouraging routine antibiotic use for M. pneumoniae.4,Reference Kimberlin, Banerjee, Barnett, Lynfield and Sawyer5 Clinicians may have perceived heightened parental expectations, observed failures with traditional therapy, or feared complications and extra-pulmonary manifestations. Additionally, media coverage of the surge and historical reliance on macrolides for atypical pneumonia may have reinforced prescribing habits. The higher azithromycin index in PPCs, where illness severity is expected to be lower than EDs and UCCs, may reflect diagnostic uncertainty since laboratory testing is often limited.
Despite growing resistance among common respiratory pathogens,Reference Jenkins and Farrell1 azithromycin remains the second most commonly prescribed antibiotic for all infections,9 and the most commonly prescribed antibiotic for community acquired pneumonia in children in the US.Reference Poole, Shapiro, Kronman and Hersh10 Increased macrolide use during surges increases risk of adverse events, and could accelerate the growing resistance among M pneumoniae Reference Kim, Jung, Kim, Park, Yang and Lee11 and other pathogens. These findings underscore the need for proactive stewardship messaging during respiratory pathogen surges.
Variation across sites was notable, with some institutions exceeding a 30% azithromycin index despite presurge baselines below 5%. Differences may reflect local testing availability, institutional response strategies, or prescriber perceptions of illness severity. Understanding these drivers is critical for tailoring interventions.
This study highlights the value of nationally representative benchmarking collaboratives, which enable the identification and tracking of prescribing trends-data not otherwise available through national platforms. Our results inform preparedness for future surges. Targeted education for primary care clinicians, rapid dissemination of evidence-based guidance, and integration of stewardship alerts into EMRs could mitigate inappropriate prescribing. Broader public health messaging to manage expectations may also help curb unnecessary antibiotic use.
This study has limitations. Data collection was limited to aggregate counts of total antibiotic, amoxicillin, and azithromycin prescriptions; specific diagnoses, diagnostic use, and other antibiotics were not captured. We were also unable to determine how many encounters resulted in prescriptions for more than one antibiotic (e.g., both amoxicillin and azithromycin). Finally, participating institutions were large pediatric centers involved in a collaborative focused on improving outpatient antibiotic use, which may limit generalizability to community pediatric practices.
In conclusion, the surge in azithromycin prescribing during the M. pneumoniae outbreak highlights vulnerabilities in outpatient antibiotic stewardship. Benchmarking collaboratives can play a pivotal role in identifying trends and guiding timely interventions to preserve antibiotic effectiveness.
Acknowledgments
Sharing Antimicrobial Reports for Pediatric Stewardship Outpatient (SHARPS-OP) Collaborative Authors: Holly Maples PharmD; Lauren Kushner MD; Hayden Schwenk MD; Meghan Birkholz MSPH; Saleh Ezzeldin MD, Michael Buhnerkempe PhD; Ana Vaughan MD, MPH; Mari Nakamura MD, MPH; Elizabeth Lloyd MD; Nicholas Dillman PharmD; Nipunie Rajapakse MD, MPH; Laura Dinnes PharmD; Anthony Diez; Kelly Flett MD, MMSc; Christine Tricarico, PharmD, MBA; Shankar Upadhyayula MD; David Haslam MD; Josh Courter PharmD; Diana Yu PharmD; Sophie Katz MD, MPH; Diana Nguyen DO; Jordan Burdine PharmD, MBA; Brendan Bettinger, PhD; Katie Ray PharmD; Michelle Mitchell MD; Louis Vernacchio MD, MSc; Jonathan Hatoun MD, MPH, MS; Jason Newland MD, MEd.
Additionally, we would like to acknowledge the Children’s Mercy Medical Writing Center for providing editing of the manuscript.
Financial support
This work was partially supported by the Heroes in Implementation Research Scholar Award from the Association for Professionals in Infection Control and Epidemiology (Funds receiving to institution to support REE and BL).
Competing interests
REE – institution receives grant support from Merck for unrelated antimicrobial stewardship projects. MPK – institution receives grant support from AHRQ for an unrelated antibiotic stewardship project. All other authors report no conflicts of interest relevant to this article.
