- Design
- Target trial emulation within one academic health system
- Population
- 742 admissions of children with medical complexity and non-severe pneumonia
- Primary outcome
- Time to discharge
- Effect
- Broad vs narrow spectrum: HR 0.84 (95% CI 0.68–1.04)
No guideline addresses antibiotic choice for community-acquired pneumonia in children with medical complexity, and many are started on broad-spectrum drugs. This study emulated a target trial using records from one US academic health system from 2016 to 2024, comparing initial broad-spectrum antibiotics with aminopenicillins in children aged 3 months to 21 years with complex chronic conditions and non-severe pneumonia.
After weighting, 742 admissions were analysed. Broad-spectrum treatment did not shorten time to discharge (HR 0.84, 95% CI 0.68–1.04), reduce progression to severe pneumonia (OR 1.29, 0.7–2.4) or reduce 30-day revisits (OR 0.39, 0.1–1.2).
Children with tracheostomies, chronic lung disease, sickle cell disease, immunosuppression or recent pneumonia admissions were excluded, so this does not apply to them. For the rest, it supports starting with amoxicillin or ampicillin, as for other children, and reserving broader cover for severe illness or specific risks. It is observational and from a single system.
- For non-severe pneumonia in a medically complex child, consider an aminopenicillin first.
- Do not use complexity alone as a reason for broad-spectrum antibiotics.
- Reserve broader cover for severe illness, tracheostomy, chronic lung disease or immunosuppression.
- Review the response at 48 hours and escalate if the child is not improving.
Why it matters
Complexity is often treated as a reason for broad-spectrum antibiotics without evidence that it helps.
Don't overread it
This is a single-centre observational study, not a randomised trial.
The statistics, in plain English
All three intervals cross 1, meaning no difference was shown. The revisit result (OR 0.39) looks favourable for broad spectrum but its interval is very wide, from 0.1 to 1.2, so it is uncertain. Weighting balances measured differences between groups, not unmeasured ones.
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