Characteristics of Antibiotic Use at Ambulatory Visits for Children in Emergency Departments in the United States, 2009-2011.

Background: Few emergency departments (EDs) incorporate antibiotic stewardship program (ASP) activities, and overall antibiotic prescribing for children in U.S. EDs has not been described. We compared systemic antibiotics used for children by pediatric versus non-pediatric EDs nationwide to identify targets for ASP efforts.

Methods: We analyzed 2009-2011National Hospital Ambulatory Medical Care Survey (NHAMCS) ED data and estimated visit numbers and characteristics using NHAMCS sampling weights. Patients 0-17 years old discharged from the ED were included. We defined EDs as pediatric if >75% of visits were for children, and as non-pediatric otherwise. ED use was estimated by antibiotic mentions: those provided, prescribed or continued at the visit. Visit diagnoses were assigned based on the most likely antibiotic indication using a previously-developed system. Multivariable logistic regression was used to identify characteristics associated with guideline-concordant antibiotic use (GCAU).

Results: An estimated 30 million (95% CI 27-33 million) U.S. ED visits for children occurred annually, of which only 14% (9-21%) per year occurred at pediatric EDs. An estimated 7 million (95% CI 6-8 million) visits per year, or 24% (22-25%) of annual visits, resulted in antibiotics. Antibiotic use by diagnosis group and antibiotic class (Figure 1) demonstrated increased use for skin and soft tissue infections at pediatric EDs (12% vs 9%, p<0.001), and increased macrolide use at non-pediatric EDs (16% vs 7%, p=0.05). For suppurative otitis media, sinusitis, and pharyngitis, a guideline-recommended antibiotic was used 80%, 69%, and 71% of the time, respectively. Increased GCAU was observed in pediatric EDs, albeit this was not statistically significant (adjusted odds ratio=1.47, 95% CI 0.96-2.24). All EDs had high rates of antibiotic use during visits for which antibiotics are rarely indicated (non-pediatric EDs 32%, 95% CI 30-34; pediatric EDs 26%, 95% CI 20-32, p=0.11 for comparison).

Conclusion: Most ED visits by children occur in non-pediatric hospitals. Opportunities exist to improve GCAU in both pediatric and non-pediatric EDs. Many antibiotics used for children in all EDs are likely unnecessary, highlighting the need to improve antibiotic use.