Treatment of Acute Otitis Media: Five Days May be Sufficient


Synopsis: Treatment for five days is sufficient in uncomplicated acute otitis media (AOM).

Source: Kozyrskyj A, et al. Treatment of acute otitis media with a shortened course of antibiotics. A meta-analysis. JAMA 1998;279:1736-1742.

Drug resistance in community-acquired infections has increased of late and has heightened concern about the indiscriminate prescribing of antibiotics. There has been recent movement by pediatric infectious disease specialists toward recommending decreasing the duration of treatment for AOM to five days.

Kozyrskyj et al identified 236 clinical trials of AOM treatment in the literature, using strict criteria for analysis, and pared the studies to 32 trials divided three ways: 1) 17 trials of oral short-acting antibiotics (amoxicillin, amoxicillin-clavulanate, cefaclor, and various cephalosporins), 2) 11 trials of oral azithromycin, and 3) four trials of intramuscular ceftriaxone. Key methodological differences in the 32 trials had to do with the inclusion of children with chronic or recurrent OM, ages of the children, and the definitions of satisfactory outcome or treatment failure. Standard meta-analysis methodology was used to derive odds ratios and summary risk differences for treatment failures.

Among 3138 children treated for AOM, the odds ratio of failure for five-day treatment was 1.38 compared to children treated for 10 days, but only when the outcome evaluation was performed 8-19 days after the initiation of therapy. Children who were evaluated later, closer to one month after initiation of therapy, showed no significant difference in outcome between treatment groups. When only those studies with large sample sizes and strict methodological quality were included, the marginal risk of treatment failure in the five-day group was not observed.

When Kozyrskyj et al looked specifically at odds ratios for antibiotic failure at 30 days in children younger than 2 years, there were no differences between groups. Insufficient data were available for children with perforated tympanic membranes and recurrent or chronic otitis media. Kozyrskyj et al also looked at the side effects in the short and long course. When the trials using amoxicillin-clavulanate were excluded, there was no difference in the likelihood of gastrointestinal side effects.

Kozyrskyj et al point out that even if the small increased risk of failure at 8-19 days is true, 44 children would require treatment with a longer course of antibiotics to prevent one failure following five-day treatment. This is not likely to be important, because children with persistent symptoms at 10-14 days are likely to be taken for a repeat evaluation. There is also some bias in group comparisons at 8-19 days, in that those children who were treated for 10 days had fewer days to experience a relapse prior to evaluation. In summary, this meta-analysis supports the use of five days of antibiotic therapy in uncomplicated AOM.


The limitations of meta-analyses are widely recognized and I will not discuss them further here. However, only the snootiest of academics would assert that a carefully performed meta-analysis using strict inclusion and exclusion criteria is devoid of value. Prospective clinical studies of AOM in children are extremely difficult to perform. The definition of the disease is not uniform among clinicians, diagnostic techniques such as tympanograms may seem burdensome to parents (my doctor can tell just by looking at his ears!), and only antibiotic manufacturers are willing to sponsor large and expensive prospective trials of the disease. Therefore, this meta-analysis, even with its inherent limitations, is a valuable addition to the literature of AOM.

The most common chief complaint in community EDs is "fever," and it is no surprise that AOM is the leading pediatric diagnosis in EDs. Because of emerging resistance to antibiotics, issues of poor compliance, and a never-ending stream of recommendations promulgated by manufacturers, science must be applied to the diagnosis and treatment of this clinical entity. We need to move past that old paradigm of pediatric fever management: "amoxicillin or admission." The practical application of AOM treatment in EDs boils down to a few basic questions.

Does the child have AOM?

AOM is defined as the presence of middle ear fluid and associated signs or symptoms of acute local or systemic illness, such as otalgia and fever. Presence of fluid in the middle ear or "otitis media with effusion" (OME), without signs or symptoms of acute infection, does not require treatment with antibiotics.

Is the AOM complicated or uncomplicated?

Infants younger than 60 days, children with recurrent or chronic otitis media, children with perforated tympanic membranes, and those with impaired immunity require special attention. In the absence of risk factors, children can be successfully treated with a five-day course of short-acting antibiotics.

How does one deal with persistent otitis media in the ED?

Access to primary care issues aside, antibiotic prophylaxis probably should not be undertaken from the ED. Many experts believe that prior to suppression therapy, tympanocentesis should be performed to identify the organism, and the child should receive close monitoring for resolution of effusion. Approximately 70% of children will have a persistent middle ear effusion at two weeks after initiation of therapy for AOM, 50% at one month, 20% at two months, and 10% at three months. Therefore, the mere detection of fluid behind the tympanic membrane is insufficient for the initiation of antibiotic therapy, particularly in children without symptoms.

One concern with this meta-analysis is the choice of short-acting antibiotic. There were no trials included that used amoxicillin (without clavulanate) in the 5-7-day treatment group. The majority of studies used a cephalosporin (cefaclor, cefpodoxime, cefuroxime, or cefprozil). There were two studies from other countries with penicillin VK and two with amoxicillin-clavulanate. Therefore, the decision to use amoxicillin for a shortened course should be guided by the prevailing pathogen in AOM in the community, if known. Several studies documented similar outcomes between a five-day course of azithromycin and a 10-day course of other antibiotics. Primary use of azithromycin should be limited by concerns about cost. In addition to concerns about broad-spectrum drugs and cost, the use of ceftriaxone may be further limited by the necessity for intramuscular (IM) administration.

This meta-analysis provides further evidence that a five-day course of short-acting oral antibiotics is sufficient for the treatment of uncomplicated AOM. It is consistent with recommendations published from other reviews in the recent literature. Children should get relief from their acute symptoms over the same time period as with longer courses of antibiotics, and compliance problems should diminish. One-month follow-up to monitor the child for middle ear effusion is preferable in the absence of further acute symptoms.


    1. Dowell SF, et al. Treatment of acute otitis media with a shortened course of antibiotics. A meta-analysis. Pediatrics 1998;101:165-171.