Treatment of Acute Otitis Media in Children Under 2 Years of Age

Hoberman A, Paradise JL, Rockette HE, et al
N Engl J Med. 2011;364:105-115

Study Summary

A current clinical practice guideline in the United States suggests watchful waiting as a treatment option for AOM in children 6-23 months with non severe infection.
However, the investigators of the current study comment on the relative lack of high-quality clinical trial data that have included children younger than 2 years. This trial, conducted at 2 centers in the United States, enrolled children aged 6-23 months during 2006-2009.
All children had received at least 2 pneumococcal vaccine doses and had AOM diagnosed on the basis of 3 criteria:

* Onset within previous 48 hours of a symptom score above an established threshold;
* Middle ear effusion; and
* Bulging of the tympanic membrane.

The children were otherwise well, without chronic illnesses that might affect outcomes.
Before randomization, the children were stratified by past AOM history and by exposure to other children for at least 10 hours per week.
The children were then randomly assigned within each stratum.
Treatment-arm children received a 10-day course of amoxicillin-clavulanate at a daily dose of amoxicillin 90 mg/kg.
The placebo preparation was similar to the study drug in taste and color.
Parents and researchers were blinded to group assignment.
Use of acetaminophen was allowed for symptom relief.
Daily symptoms were assessed by phone until day 4 or 5 (first follow-up visit) and then at the end of therapy (days 10-12, second follow-up visit) and at the end of the study (days 21-25, third visit).
The children were examined at each of the follow-up visits at which point tympanic membrane appearance was scored and symptoms were assessed
Clinical failure on or before the visit at day 4-5 was considered to occur if the child's symptoms had worsened, symptoms had not substantially improved, the tympanic membrane had not substantially improved, or tympanic membrane findings had worsened.
Failure at the visit on day 10-12 visit was considered to have occurred if the child had not experienced almost complete resolution of symptoms and otoscopic findings, without regard to the presence or absence of middle ear fluid.
Children who experienced failure were started on a rescue regimen.
Finally, reappearance of AOM at or by the day 21-25 visit was considered a relapse.

The primary outcome of interest was time to resolution of symptoms, measured by both time to reach a minimal symptom score and total symptom burden, determined by calculating the daily score averaged over the first 7 days after initiation of treatment.
Acetaminophen use, adverse events, and use of healthcare resources were also assessed.
More than 1300 children were screened, 498 were eligible, and 291 were enrolled (58% of eligible children).
Of interest, the tympanic membrane findings as assessed by the clinical team did not correlate with parental reports of symptoms at enrollment.
More than half of the children were 6-11 months old; 53% were boys, 45% were white, 41% were black, and 70% were on Medicaid.
Approximately half were exposed to other children, and 52% had bilateral AOM at enrollment. The intervention and placebo groups did not differ in any of the measured or demographic variables at enrollment.
Symptom improvement was not different between the 2 groups.
For example, at day 2, 35% of treated children and 28% of children taking placebo had experienced initial resolution of symptoms. By day 4, the frequencies of symptom resolution were 61% for treated children and 54% for the children on placebo. By day 7, these frequencies were 80% and 74%, respectively.

However, when looking at total symptom burden, the mean symptom scores were lower in the treated children at each timepoint.
The reduction in clinical failure at day 10-12 was as large among children with least severe infection as it was in children whose illnesses were considered severe.
Treatment also produced quicker resolution of middle ear fluid and lower frequencies of persistent infection.
Clinical failure was markedly reduced with treatment.
At day 4-5, treatment failure had occurred in 4% of treated children vs 23% of children who received placebo.
At day 10-12, the failure frequencies were 16% for treated children and 51% for children on placebo.
Being exposed to 3 or more other children, more severe scores at entry, and bilateral AOM were all predictors of worse outcomes.
However, even after controlling for those factors, treatment produced significant improvement over placebo.
Finally, the relapse rates were similar (16% for treated children and 19% for children receiving placebo, a difference that was not statistically significant). One child in the placebo group developed mastoiditis.
The investigators concluded that treatment of AOM with appropriate antibiotics produces short-term symptomatic benefits and improved clinical findings irrespective of the apparent severity of the illness.