Outpatient management of most afebrile infants with acute otitis media who haven’t been tested for invasive bacterial infection may be reasonable given the low occurrence of adverse events, said Son H. McLaren, MD, MS, of Columbia University, New York, and colleagues.
Dr. McLaren and associates conducted an international cross-sectional study at 33 emergency departments participating in the Pediatric Emergency Medicine Collaborative Research Committee of the American Academy of Pediatrics (AAP): 29 in the United States, 2 in Canada and 2 in Spain.
The researchers sought first to assess prevalence of invasive bacterial infections and adverse events tied to acute otitis media (AOM) in infants 90 days and younger. Those who were clinically diagnosed with AOM and presented without fever between January 2007 and December 2017 were included in the study. The presence of fever, they explained, “is a primary driver for more expanded testing and/or empirical treatment of invasive bacterial infection (IBI). Secondarily, they sought to characterize patterns of diagnostic testing and the factors associated with it specifically in this patient population.
Of 5,270 patients screened, 1,637 met study criteria. Included patients were a median age of 68 days. A total of 1,459 (89.1%) met AAP diagnostic criteria for AOM. The remaining 178 patients were examined and found to have more than one of these criteria: 113 had opacification of tympanic membrane, 57 had dull tympanic membrane, 25 had decreased visualization of middle ear structures, 9 had middle ear effusion, 8 had visible tympanic membrane perforation and 5 had decreased tympanic membrane mobility with insufflation. None of the 278 infants with blood cultures had bacteremia, nor were they diagnosed with bacterial meningitis. Two of 645 (0.3%) infants experienced adverse events, as evidenced with 30-day follow-up or history of hospitalization.
Dr. McLaren and colleagues observed that despite a low prevalence of IBI and AOM-associated adverse events, more than one-fifth of patients were prescribed diagnostic testing for IBI and subsequently hospitalized, a practice that appeared more common with younger patients.
Significant testing and hospitalizations persisted despite low prevalence of IBIs
Although diagnostic testing and hospitalizations differed by site, they were, in fact, “substantial in contrast to the low prevalence of IBIs and adverse events,” the researchers noted. “Our data may be used to help guide clinical management of afebrile infants with clinician-diagnosed AOM, who are not included in the current AAP AOM practice guideline,” the authors said. They speculated that this practice may be due, in part, to young-age risk of IBI and the concern for IBI in this population based on febrile infant population data and a general hesitance to begin antibiotics without first evaluating for IBI. They also cited a low prevalence ranging from 0.8% to 2.5% as evidence for low risk of IBI in afebrile infants with AOM.
Also of note, given that roughly three-fourths of infants included in the study were reported to have symptoms of upper respiratory infection that can lead to viral AOM, including these infants who could have a lower likelihood of IBI than those with known bacterial AOM, may have led the researchers to underestimate IBI prevalence. Because existing data do not allow for clear distinction of viral from bacterial AOM without tympanocentesis, and because more than 85% of older patients with clinically diagnosed AOM also have observed bacterial otopathogens, the authors clarify that “it is understandable why clinicians would manage infants with AOM conservatively, regardless of the presence of concurrent viral illnesses.” They also acknowledged that one major challenge in working with infants believed to have AOM is ensuring that it is actually present since it is so hard to diagnose.
Dr. McLaren and colleagues cited several study limitations: 1) completeness and accuracy of data couldn’t be ensured because of the retrospective study design; 2) because not all infants were tested for IBI, its prevalence may have been underestimated; 3) infants whose discharge codes did not include AOM may have been missed, although all infants with positive blood or cerebrospinal fluid cultures were screened for missed AOM diagnosis; and 4) it is important to consider that any issues associated with testing and hospitalization that were identified may have been the result of management decisions driven by factors that cannot be captured retrospectively or by a diagnosis of AOM.