Acute cardiac care & resuscitationCohort study
Rescue-breathing versus chest compression-only bystander cardiopulmonary resuscitation in non-cardiac out-of-hospital cardiac arrest: a nationwide retrospective cohort study
- retrospective nationwide cohort study
Authors' conclusion
Thirty-day mortality and unfavorable neurological outcomes were lower after RB-CPR than after CO-CPR for non-cardiac OHCA, particularly in pediatric patients. Although CO-CPR remains important in bystander CPR, rescue breathing may still have benefit, particularly in pediatric patients and those with cardiac arrest related to drowning or asphyxia.
Study context
- Population
- Japanese patients with non-cardiac out-of-hospital cardiac arrest (OHCA) in the All-Japan Utstein Registry, 2012–2023, who received bystander CPR with documented chest compressions
- Participants
- 154,137 patients
- Exposure
- bystander-performed RB-CPR vs compression-only CPR (CO-CPR)
- RB-CPRn = 20,124
- CO-CPRn = 134,013reference
Main finding
30-day mortality · 30 days
Adjusted relative risk
0.98
95% CI 0.98–0.99
RB-CPR vs CO-CPR
adjusted RR 0.98 (95% CI 0.98–0.99)
As reported in the abstract
Observational data: associations, not effects.
Further findings
| Outcome | adjusted RR (95% CI) | |
|---|---|---|
| 30-day unfavorable neurological outcomes | 0.99 (0.99–0.99) | |
| 0.512Adjusted relative risk (log scale) |
Both outcomes by age group (subgroup)
- RB-CPR was associated with lower risks of both outcomes in pediatric patients but no clear differences were observed among adult or elderly patients
Associations by subgroup
- Associations were strongest for pediatric patients and drowning-related arrests and modest in patients with asphyxia
Original publication
Rescue-breathing versus chest compression-only bystander cardiopulmonary resuscitation in non-cardiac out-of-hospital cardiac arrest: a nationwide retrospective cohort study. Resuscitation. Published 25 September 2026. doi:10.1016/j.resuscitation.2026.111332
Summarised from the published abstract; every number is checked against it