Background Guidelines recommend avoidance of excessive oxygen administration during neonatal resuscitation. Blenders are used in some but not all hospitals. It has been suggested that self-inflating bags without a reservoir deliver around 40% oxygen and could be used to provide an inexpensive and effective technique of avoiding oxygen toxicity.

Objective To explore how much oxygen is delivered when using two different brands of neonatal self-inflating resuscitation bags without a reservoir.

Methods In a benchtop setting, the smallest non-disposable self-inflating bags from the Laerdal and Ambu ranges were tested. Oxygen concentration delivered by these devices under a variety of conditions was measured. 108 combinations of oxygen flow rates (10; 5 to 1 litre/min), ventilation rates (30, 60, 100 inflations/min) and peak inspiratory pressure ranges (20 to 25 cm H₂O, 35 to 40 cm H₂O or pop-off valve range, 55 to 60 cm H₂O) were tested.

Results Delivered oxygen concentration varied depending on three parameters: gas flow rate, ventilatory rate and pressure. At a pressure of 20 to 25 cm H₂O, mean oxygen concentration delivered by both bags exceeded 70% at any gas flow rate except for 1 litre/min (where delivered oxygen concentration was 60% to 70%). When the pop-off valve was opened at 35 to 40 cm H₂0, oxygen concentrations fell to 30% to 45% at gas flow rates ≤2 litres/min. The Ambu bag delivered a lower oxygen concentration than the Laerdal bag but this difference was not clinically important.

Conclusion When using the Laerdal and Ambu infant resuscitation self-inflating bags without a reservoir, delivered oxygen concentration is >70% for currently recommended flow and pressure settings.