Eleven RCTs (n=2,320) were included in the review. All trials scored at least 2 points on the Jadad scale and 10 studies scored three or higher. All trials that used a placebo comparator were double-blinded; two of those used watchful weighting blinded investigators and two were open label.
Antibiotic treatment significantly improved the likelihood of clinical success compared to both placebo (RR 1.11, 95% CI 1.05 to 1.18; seven RCTs, n=1,405) and watchful waiting (RR 1.18, 95% CI 1.07 10 1.32; four RCTs, n=915). A similar result was found when all trials were combined (RR 1.13, 95% CI 1.08 to 1.19; 11 RCTs, n=2,320). Sensitivity analysis revealed no differences between trials that used rigorous diagnostic criteria and those that did not.
Short-term persistence of symptoms was less likely in groups treated with antibiotics compared to placebo (RR 0.75, 95% CI 0.64 to 0.88; four RCTs, n=1,014), to watchful weighting (RR 0.47, 95% CI 0.29 to 0.76; one RCT, n=285) and in all trials (RR 0.68, 95% CI 0.54 to 0.85; five RCTs, n=1,299). There was a higher incidence of diarrhoea in the antibiotic treatment groups for watchful waiting comparisons (RR 1.99, 95% CI 1.32 to 3.01; three RCTs) and for all trials (RR 1.50, 95% CI 1.16 to 1.95; seven RCTs, n=1,807).
No statistically significant differences were found for any other outcome. Statistical heterogeneity was low to moderate for the reported analyses (I2 range: 0% to 42%).