- Design
- Systematic review and meta-analysis of randomised trials, GRADE assessed
- Population
- 8 trials, 4,106 children with Kawasaki disease
- Primary outcome
- Coronary artery abnormalities
- Effect
- Within 1 month RR 0.49 (0.17 to 1.42); after 1 month RR 0.81 (0.43 to 1.55); very low certainty
A meta-analysis pooled eight randomised trials (4,106 children) adding corticosteroids to IVIG and aspirin for Kawasaki disease, with GRADE certainty assessment.
Coronary artery abnormalities did not differ significantly within 1 month (RR 0.49, 0.17 to 1.42) or after 1 month (RR 0.81, 0.43 to 1.55). Fever duration and adverse events also did not differ. Hospital stay was about 1 day shorter. Subgroups hinted at benefit with prednisolone-based or longer regimens, but the authors caution against these exploratory findings. Overall certainty was very low.
The routine addition of steroids for every child is not supported. The open question remains the high-risk child, for whom trials such as RAISE in Japan used risk scores that may not transfer to other populations.
- Give IVIG within 10 days of fever onset as standard
- Do not add corticosteroids routinely to first-line treatment
- Consider steroids for IVIG-resistant or high-risk disease with specialist input
- Arrange echocardiography at diagnosis, 2 weeks and 6 weeks
Why it matters
It removes support for a common add-on in all children while leaving the high-risk question open.
Don't overread it
Very low certainty and wide intervals mean a real benefit in selected children cannot be excluded.
The statistics, in plain English
A risk ratio of 0.49 looks like a halving, but the interval runs from 0.17 to 1.42, so the data are also compatible with 42% more coronary abnormalities. That width, plus inconsistent trials, is why certainty was graded very low. Subgroup 'promise' across many comparisons often disappears in new trials.
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