JOURNAL CLUB · LANDMARK TRIAL
Resuscitation of asphyxiated newborns with room air or oxygen (Resair 2)
Saugstad OD, Rootwelt T, Aalen O. Pediatrics. 1998;102(1):e1. An international, quasi-randomised controlled trial.
CHANGED
NEWBORN RESUS
No difference in the primary outcome
Death within 7 days or moderate/severe HIE: 21.2% (air) vs 23.7% (oxygen); OR 0.94 (95% CI 0.63-1.40).
Early recovery favoured air
Fewer 5-minute Apgar <7 (p=0.03), and shorter time to first breath and first cry (p=0.004 and 0.006).
No mortality difference at 28 days
No difference in 28-day mortality, or in the individual components of the primary outcome.
Later evidence favours air
Subsequent meta-analyses (ILCOR: 5 RCTs + 5 quasi-RCTs) show lower short-term mortality with air (RR 0.73, 95% CI 0.57-0.94).
WHY THIS MATTERS
- 100% oxygen was standard of care - with little scientific basis and no human trials
- Conventional wisdom strongly favoured oxygen ("one breath of oxygen is worth five of air") - a marked lack of equipoise
- The delivery room was a near-untested setting for trials
- Hyperoxia during resuscitation can drive oxidative injury
- It opened the door to evidence-based delivery-room care
STRENGTHS
- First large, international, multicentre trial of oxygen concentration in newborn resuscitation
- Showed it was feasible to randomise a resuscitation strategy in the seconds after birth
- Provided a template for later, higher-quality delivery-room trials
- 18-24 month follow-up showed no difference in growth or neurological impairment
- Its findings, with later studies, drove a global, life-saving practice change
LIMITATIONS
- Quasi-randomised (alternate-day by date of birth) - high risk of allocation and selection bias
- Unmasked - caregivers knew the intervention (performance / detection bias)
- Under-recruited: 703 enrolled against a planned 920
- Delivery-room consent and conditions make rigorous trials very hard
- Not definitive alone - the practice change rests on the body of evidence
Practice implications
Before Resair 2, newborns were resuscitated with 100% oxygen almost by default. This trial - and the meta-analyses that followed - showed that starting with room air is at least as safe, with some early markers (Apgar, time to breathe) favouring air, and later evidence showing lower mortality in term infants. It underpins current guidance: ANZCOR and ILCOR now suggest starting resuscitation of infants ≥35 weeks with 21% oxygen, titrating to pre-ductal saturation targets. Preterm infants start on a low (not necessarily 21%) oxygen, guided by SpO₂. A landmark shift from dogma to evidence in the delivery room.
Why was 100% oxygen the standard for so long despite little evidence?
How does a quasi-randomised, unmasked trial change how much you trust the result?
What starting oxygen do you use for a term vs a preterm resuscitation, and why?
How do you weigh a single flawed landmark trial against the meta-analysis that followed?
Pediatrics 1998 ANZCOR
Newborn resuscitation HIE
The Hub condition page LANDMARK TRIALS
More trials to know
Take-home message: Resair 2 was the first large international trial to challenge the dogma that asphyxiated newborns must be resuscitated with 100% oxygen. Despite its quasi-randomised, unmasked design, it showed room air was at least as effective - with better early recovery (Apgar, time to breathe) - and, with later meta-analyses showing lower mortality with air, it changed practice worldwide. Guidelines now recommend starting term resuscitation with 21% oxygen, titrated to saturation targets. A landmark move from dogma to evidence in the delivery room.
For educational purposes only. Journal club appraisal - figures paraphrased from the published trial and later reviews; read the full article for complete data. Always align practice to current ANZCOR/NRP guidelines.