Evidence assessment of management of acute otitis media: I. The role of antibiotics in treatment of uncomplicated acute otitis media

Context. In 1995, >5 million episodes of acute otitis media (AOM) accounted for $3 billion in health care expenditures. Objectives. To synthesize the literature on the natural history of AOM, the effectiveness of antibiotic treatment in uncomplicated AOM, and the relati ve effectiveness of specific antibiotic regimens. Data Sources. Seven electronic databases for articles published between 196 6 and March 1999 and reference lists in proceedings, published articles, re ports, and guidelines. Study Selection. Two physicians independently assessed each article. Studie s addressing AOM in children 4 weeks to 18 years old were included; those a ddressing children with immunodeficiencies or craniofacial abnormalities we re excluded. Randomized, controlled trials (RCTs) were used to assess antib iotic effectiveness, and RCTs and cohort studies were used to assess the na tural history of AOM. Among the 3491 citations identified, 80 (2.3%) met ou r inclusion criteria. Data Extraction. Two physicians independently abstracted data and assessed the quality of studies using a validated scale for RCTs and 8 quality compo nents for cohort studies. Data Synthesis. Random-effects estimates of pooled absolute rate difference s of outcomes were derived, and heterogeneity of both the rates and rate di fferences was assessed. Children with AOM not treated with antibiotics expe rienced a 1- to 7-day clinical failure rate of 19% (95% confidence interval : 0.10-0.28) and few suppurative complications. When patients were treated with amoxicillin, the 2- to 7-day clinical failure rate was reduced to 7%, a 12% (95% confidence interval: 0.04-0.20) reduction. Adverse effects, prim arily gastrointestinal, were more common among children on cefixime than am ong those on ampicillin or amoxicillin. They were also more common among ch ildren on amoxicillin-clavulanate than among those on azithromycin. Conclusions. The majority of uncomplicated cases of AOM resolve spontaneous ly without apparent suppurative complications. Ampicillin or amoxicillin co nfers a limited therapeutic benefit. There is no evidence to support any pa rticular antibiotic regimens as more effective at relieving symptoms. Certa in antibiotics are more likely than others to cause diarrhea and other adve rse events.