Antibiotics for acute otitis media: a meta-analysis with individual patient data

A meta-analysis of four randomised trials that compared amoxicillin (or amoxicillin/clavulanic acid) with placebo, and two randomised trials that compared amoxicillin with delayed treatment, which included data from 1643 children, aged 6 months to 12 years, used the presence of fever and/or ear pain at 3-7 days after onset as markers of poor outcome. The meta-analysis found that antibiotic treatment significantly reduced the risk of poor outcomes for: children aged less than 2 years with bilateral otitis media (NNT =4), and children with otitis media and otorrhoea (NNT=3). In all other children with otitis media the NNT to reduce the rate of fever or ear pain at days 3-7 was >10. The authors recommended that antibiotic treatment should not be prescribed for the majority of patients.