Design: Systematic review and meta-analysis of 10 randomised trials
Population: 8,970 immunocompetent children with non-severe community-acquired pneumonia
Primary outcome: Treatment failure, relapse and non-adherence
Effect: Failure RR 1.07 (0.93–1.24); non-adherence RR 0.35 (0.24–0.50)

A meta-analysis in the European Journal of Pediatrics (7 October) pooled ten randomised trials in 8,970 immunocompetent children with non-severe community-acquired pneumonia, comparing five days or less of antibiotics with longer courses.

Treatment failure (RR 1.07) and relapse (RR 1.04) did not differ. Non-adherence was about two-thirds lower with short courses (RR 0.35). Adverse events were similar, though trials disagreed widely on that outcome.

This fits with the trials behind it and with current stewardship advice. The authors note the analysis was not designed to prove equivalence, but the consistency across trials and the adherence gain make a short course the sensible default for non-severe disease.

Prescribe five days or less for non-severe pneumonia in otherwise healthy children
Shorter courses were much more likely to be completed
Reassess a child who is not improving by 48–72 hours rather than lengthening the course
Does not apply to severe pneumonia, empyema or immunocompromised children

Why it matters

Long courses add antibiotic exposure and missed doses without improving outcomes.

The statistics, in plain English

A failure RR of 1.07 (95% CI 0.93 to 1.24) with I² of 3% means trials agreed closely; the interval allows a small increase in failure but not a large one. The adherence RR of 0.35 (0.24 to 0.50) is a large, clear effect.