Six randomised trials and 576 children, mean age three to five years, compared honey with pharmacological treatment for acute cough over three days of follow-up. Just over half the participants received honey.
The primary analysis showed high heterogeneity and no significant difference across most endpoints — cough frequency, severity, bothersomeness, caregiver sleep. The one significant result favoured honey: the cough's effect on the child's sleep, a mean difference of -0.73 with an interval from -1.45 to -0.01. After a post-hoc leave-one-out sensitivity analysis, honey came out ahead on every outcome, and subgroup analysis found the result held whichever pharmacological comparator was used.
Read the sensitivity analysis cautiously — post-hoc analyses that turn a null into a positive are the least reliable kind, and the authors say so themselves. But the decision this informs does not need honey to win. Paediatric antitussives have a documented adverse-effect profile and a long history of inappropriate prescribing in exactly this age group. An intervention that performs comparably, costs almost nothing and carries no such risk is the better choice on equivalence alone. The one absolute contraindication is age: no honey under twelve months, because of infant botulism.
- Never in infants under twelve months — botulism risk, and this is absolute
- Follow-up was three days, so this is about symptom relief, not illness course
- The primary analysis was null; the positive result came from a post-hoc analysis
- Equivalence is a sufficient argument given antitussive adverse effects
- Mean age three to five years — the group most often prescribed antitussives
The statistics, in plain English
The sleep result, a mean difference of -0.73 with an interval from -1.45 to -0.01, only just excludes zero — a p value of exactly 0.05 on one of several endpoints. Post-hoc sensitivity analyses that convert null results into significant ones are the weakest form of evidence in a meta-analysis, because removing studies until the result changes is a decision made after seeing the data. The primary analysis is the honest one, and it showed equivalence.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for paediatrics, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free