Shortening antibiotic therapy duration for hospital-acquired bloodstream infections in critically ill patients: a causal inference model from the international EUROBACT-2 database
An erratum to this article was published (Intensive Care Med, Oct 2025 ↗)
Main finding · 28-day treatment failureResult — go to the resultOdds ratio 0.64 95% CI 0.44–0.93Study context
- Population
- ICU patients with hospital-acquired bloodstream infection eligible for shortened therapy (excluding no antibiotics within 3 days or requiring prolonged therapy)550 eligible
- Participants
- 2,600 patients
- Exposure
- Shortened vs long antibiotic therapy
- Shortn = 213
- Longn = 337reference
- Outcome
- Treatment failure (death, persistent infection, or subsequent infectious complications)
- Follow-up
- Day 28
Methodological detailssecondary analysis of the EUROBACT-2 cohort, international, multicentre, inverse-probability of treatment weighted (IPTW), causal inference model
- Secondary analysis of the EUROBACT-2 cohort
- International
- Multicentre
- Inverse-probability of treatment weighted (IPTW)
- Causal inference model
Main finding
28-day treatment failure
Odds ratio
0.64
95% CI 0.44–0.93
P = 0.019
Short vs Long
OR 0.64 (95% CI 0.44–0.93); P = 0.019
As reported in the abstract
Observational data: associations, not effects.
Further findings
| Outcome | OR (95% CI) | P value | |
|---|---|---|---|
| Subsequent infectious complications · Day 28IPTW logistic regression · main driver of reduced treatment failure | 0.58 (0.37–0.91) | P = 0.018 | |
| 0.20.5125Odds ratio (log scale) |
Mortality
- Similar between short and long treatment: OR, 95% CI, P
- OR 0.92 [95% CI 0.59, 1.43], p = 0.7
Persistent infection
- Similar between short and long treatment: OR, 95% CI, P
- OR 0.47 [95% CI 0.17, 1.14], p = 0.12
Original publication
Shortening antibiotic therapy duration for hospital-acquired bloodstream infections in critically ill patients: a causal inference model from the international EUROBACT-2 database. Intensive Care Med. Published 7 April 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →