Background: Acute otitis media is the most common childhood infection for which antibiotics are prescribed in the United States. In children, treating otitis media, which includes both acute otitis media and otitis media with effusion, costs an estimated $2 billion annually. A 2001 evidence report from the Agency for Healthcare Research and Quality focused on the management of uncomplicated acute otitis media in children, and concluded that treatment with ampicillin or amoxicillin reduced clinical failure rates when compared with observation. However, the microbiology of acute otitis media has shifted following widespread use of the 7-valent pneumococcal conjugate vaccine (PCV7) in children. Shekelle and colleagues systematically reviewed evidence on the accuracy of diagnosis, the effect of PCV7, and the effectiveness of different treatments for acute otitis media.

The Study: The authors conducted a meta-analysis of six key questions that addressed the following topics: accuracy in the diagnosis of uncomplicated acute otitis media; effect of PCV7 immunization on acute otitis media microbial epidemiology; effectiveness of treatment options for uncomplicated acute otitis media; effectiveness of treatment options for recurrent otitis media; treatment outcomes for specific subpopulations; and the adverse effects of available treatment options. Articles included in the evidence report were found by a literature search of PubMed, Cochrane Central Register of Controlled Trials, Cochrane Databases of Systematic Reviews, Web of Science, and the Science Citation Index using specific search terms and strategies. Systematic reviews, randomized controlled trials, controlled clinical trials, and observational studies (only when more rigorous studies were insufficient to answer the study question) were included. Two pediatricians who had been trained to critically analyze scientific literature independently reviewed each study for inclusion in the evidence report based on whether it reported original data (or was a systematic review) and if it answered one of the report questions.

Results: Each study defined clinical success by one of the following findings: absence of symptoms, improvement in acute symptoms, absence of otorrhea, resolution of otoscopic findings, or cumulative clinical resolution. The random effects pooled rate difference for clinical success was calculated for each comparison by looking at clinical outcomes by day 14 or day 16. All but one study were of moderate quality; therefore, further high-quality research will likely change the confidence in the estimated difference and the calculated estimated difference itself.

Ampicillin or amoxicillin versus placebo. Seven studies reported a pooled difference for clinical success by day 14 of 12 percent (number needed to treat [NNT] = 9).

Ampicillin or amoxicillin versus ceftriaxone (Rocephin; single dose). Four studies estimated the pooled difference in clinical success by day 14 to be 0 percent; no advantage or equivalence of antibiotics could be established.

Amoxicillin/clavulanate (Augmentin; seven to 10 days) versus ceftriaxone (single dose). Five studies determined that the pooled difference for clinical success by day 16 was 3 percent in the amoxicillin/clavulanate group; neither antibiotic has a notable advantage.

Amoxicillin/clavulanate (seven to 10 days) versus azithromycin (Zithromax; five days or less). Nine studies reported a pooled difference for clinical success by day 14 of –0.3 percent. Neither antibiotic has been established to be superior, and equivalence could not be determined.

Azithromycin (less than five days) versus cefaclor (seven to 10 days). Three studies estimated the pooled difference for clinical success by day 14 to be –0.7 percent. The two antibiotic treatments were equally effective and, because the quality of evidence was high, further high-quality studies are unlikely to change the results.

Antibiotics versus wait-and-see and prescription- to-hold. Four studies compared various antibiotics with delayed treatment approaches. Two studies of amoxicillin compared with wait-and-see and prescription-to-hold found a clinical success rate difference of 15 and 16 percent, respectively, demonstrating that immediate treatment has a higher rate of clinical success than the delayed treatment approach. However, the other two studies were inconclusive because otalgia and fever were improved by the trials' end points, regardless of antibiotic use. In all four studies, compliance in the prescribed antibiotic groups was high; additionally, up to 38 percent of patients in the delayed treatment groups ultimately received antibiotics.

Short- versus long-duration treatment. One study comparing a five-day and a 10-day course of antibiotics concluded that the treatment lengths were equally effective. However, longer antibiotic use was associated with less risk of signs and symptoms, relapse, or reinfection during days 8 to 19 (NNT = 17).

Conclusion: The authors conclude that treatment of otitis media with ampicillin or amoxicillin demonstrates greater clinical success rates than placebo. The results are mixed about whether immediate treatment improves clinical success rates compared with delayed treatment strategies. In this review, short- and long-duration courses of antibiotics were essentially equally effective. No difference in clinical success was reported for ampicillin or amoxicillin versus single-dose ceftriaxone, amoxicillin/clavulanate versus single-dose ceftriaxone, or amoxicillin/clavulanate versus short-duration azithromycin. Additionally, treatment with azithromycin was found to be equivalent to treatment with cefaclor. Future high-quality research will likely change the estimated difference or confidence in the estimated difference for the studied comparisons.