Am Fam Physician. 2011;83(1):37
Author disclosure: Dr. Ebell is Editor in Chief of Essential Evidence, published by John Wiley & Sons, Inc., who also publish The Cochrane Library.
Are short courses of antibiotics as effective as standard courses for children with acute otitis media?
Antibiotic courses between two and seven days in duration are associated with a slightly higher risk of treatment failure than longer courses (number needed to treat = 33), but also a somewhat lower risk of adverse gastrointestinal effects (number needed to harm = 29). Ceftriaxone (Rocephin) and azithromycin (Zithromax) appear to be similarly effective when given for shorter or longer durations. (Strength of Recommendation = B, based on inconsistent or limited-quality patient-oriented evidence)
The usual duration of a course of antibiotics ranges from 10 days in the United States to six or seven days in the Netherlands, and even less elsewhere. Many clinicians advocate severely restricting antibiotics because acute otitis media is typically a self-limited condition and the benefit of antibiotic treatment is modest.1,2
This Cochrane systematic review identified randomized controlled trials that compared regimens of less than seven days in duration with regimens of seven days or longer in children with acute otitis media. In some of the studies, participants in both arms received the same antibiotic, but other studies compared two different antibiotics. The authors identified a total of 49 studies with 12,045 participants, including 22 studies that were not included in the original publication of this review in 2000. Most of the studies enrolled children younger than one year (n = 39), whereas eight studies only included children who were two years or older. Study quality was mixed—most were not blinded, most did not clearly describe allocation concealment, and the authors considered only one-third of them to be at low risk of other sources of bias.
Results based on two small studies (n = 118) showed that regimens of two days or less were not as effective as longer regimens (odds ratio = 2.99; 95% confidence interval, 1.04 to 8.5). Short regimens that were at least 48 hours in duration were also associated with an increased risk of treatment failure at up to one month of follow-up (odds ratio = 1.34; 95% confidence interval, 1.15 to 1.55). The absolute risk of treatment failure was 21 percent with short-course treatment versus 18 percent with long-course (number needed to treat = 33). This effect was greatest within three weeks of treatment; with longer follow-up, there was less evidence of benefit from longer regimens.
Regarding specific antibiotics, short courses of ceftriaxone and azithromycin were as effective as longer courses. Gastrointestinal adverse effects were reported by 13 studies with 4,918 children, and shorter regimens were associated with fewer upset stomachs (number needed to harm = 29). This benefit of shorter regimens was particularly true in studies of treatment with amoxicillin/clavulanate (Augmentin). The authors found no evidence of publication bias, which may occur when small studies showing no difference in effect are not published.