Ten RCTs (n=2,766 patients, sample size 47 to 553) were included in the review. All RCTs were prospective. Four trials received a Jadad score of 3 or more and six received a score of 2. GRADE assessments indicated that the quality of the evidence base was moderate.
Risk of failure: Macrolide antibiotics were associated with a higher risk of failure than amoxicillin-containing antibiotics (RR 1.31, 95% CI 1.07 to 1.60; 10 studies) with no evidence of heterogeneity. When analysed separately in subgroup analyses, there was no evidence of a statistical difference in subgroups of younger or older children, high-dose amoxicillin-containing regimens, studies with Jadad score of 3 or more or studies that evaluated clarithromycin. Subgroup results for low-dose amoxicillin-containing regimens, studies with Jadad scores less than 3 and studies of azithromycin supported the overall results that indicated a benefit for amoxicillin-containing antibiotics. Calculation of the number needed to harm indicated that 32 patients would need to be treated with azithromycin rather than first-line antibiotics to cause one additional failure.
Adverse events: Macrolide antibiotics were associated with a significantly lower risk of any adverse reaction (RR 0.74, 95% CI 0.60 to 0.90; unknown number of studies) or developing diarrhoea (RR 0.41, 95% CI 0.32 to 0.52; unknown number of studies). There was no evidence of a difference between treatments in rates of vomiting, abdominal pain, nausea and generalised rash.