|| Checking for direct PDF access through Ovid
(1) To evaluate the probability of antibiotic administration associated with ICD-9 diagnosis of otitis media with effusion (OME) in the absence of acute otitis media, (2) to determine whether usage varies according to visit setting, and (3) to ascertain if practice gaps are such that future practice changes might be measured.Cross-sectional analysis of an administrative database.Ambulatory visits in the United States.National Ambulatory and Hospital Ambulatory Medical Care Surveys, 2005-2010; univariate, multivariate, and stratified analyses of antibiotic usage were performed. The study population was restricted to children without acute or unspecified otitis media. The primary outcome was the probability of oral antibiotic administration when OME was diagnosed. The impact of the location of service and subspecialty care was also analyzed.Data from 1,390,404,196 pediatric visits demonstrated that oral antibiotics were administered for 32% of visits with an OME diagnosis, even in the absence of acute otitis media (odds ratio, 4.31; 95% confidence interval: 2.88-6.44; P < .001). The highest antibiotic administration was seen in the emergency department (risk difference, 37.1%; number needed to harm, 3). No significant increased risk of antibiotic usage was seen during otolaryngology visits. Diagnoses of infections at nonotologic sites were associated with a 1.98 to 26.60 increase in odds of oral antibiotic administration.Oral antibiotics continue to be administered in children with OME in the absence of acute infection, with risk varying by location of service. There is a potential opportunity for quality improvement through reducing antibiotic administration for pediatric OME.