- Design
- multicentre retrospective cohort analysed as a target trial emulation with propensity-score weighting
- Population
- 196 infants aged 12 months or younger hospitalised with pertussis at 11 Italian centres, median age 86 days
- Primary outcome
- severe disease course, defined as Pertussis Severity Score above 5
- Effect
- 60% (30 of 50) with azithromycin vs 33% (48 of 146) with clarithromycin, odds ratio 1.80 (95% CI 1.01 to 3.21)
Eleven Italian centres contributed 196 infants aged 12 months or under hospitalised with pertussis between November 2023 and December 2024, analysed as a target trial emulation with propensity-score weighting. Median age was 86 days. Clarithromycin was given to 146 and azithromycin to 50.
A severe course, defined as a Pertussis Severity Score above 5, occurred in 60% of the azithromycin group and 33% of the clarithromycin group, odds ratio 1.80 (95% CI 1.01 to 3.21). Azithromycin-treated infants more often needed oxygen, had more complications and more intensive care admissions. Both deaths were in the azithromycin arm.
Guidelines treat the two macrolides as interchangeable first-line options, and azithromycin is often preferred for its shorter course and simpler dosing. This is not a randomised trial and cannot overturn that, but it is the best comparative data that exists in the group that matters most - young infants, in whom pertussis kills. Where both agents are available and the infant is unwell enough to be admitted, clarithromycin is the more defensible choice until something better exists. With pertussis resurgent, this is a decision Indian paediatricians are making regularly.
- In an infant admitted with pertussis, prefer clarithromycin where both macrolides are available
- Azithromycin remains appropriate for outpatient prophylaxis of contacts, which this study does not address
- Score severity on admission so that deterioration is measured rather than remembered
- Check pertussis vaccination status of the mother and of household contacts at every such admission
The statistics, in plain English
Target trial emulation is a method for making an observational comparison behave more like a randomised one: eligibility, treatment assignment and follow-up are specified in advance as they would be in a protocol, and propensity weighting balances measured confounders. It cannot balance unmeasured ones, and here the obvious worry runs in the direction of the finding - a sicker-looking infant might be given whichever drug the unit reserves for sicker infants. The odds ratio interval, 1.01 to 3.21, only just excludes 1.0, so this is a fragile result: a handful of events either way would erase it. Fifty infants in the azithromycin arm is a small denominator for a comparison of complications, and the two deaths cannot support any inference at all.
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