Object Oxygen (O₂) therapy is first line treatment in paediatric respiratory failure. Further deterioration may require additional respiratory support. There is growing clinical evidence that helium–oxygen gas mixtures (‘heliox’) may reduce the work of breathing, improve oxygenation, ventilatory efficiency and reduce need for additional respiratory support. However, heliox delivery techniques are poorly researched. This study aimed to identify the best facemask for heliox delivery.

Methods Heliox21 gas (21% O₂/79% helium) was used. Lung simulation experiments (Ingmar Medical ASL5000) were undertaken replicating apnoea, normal respiration and increasing levels of respiratory distress, at flow rates of 0–15 L/min. Four types of facemasks were compared: a 3-valve facemask with reservoir bag (3V+R), a 2-valve facemask with reservoir bag (2V+R), a 1-valve facemask with reservoir bag (1V+R), and a facemask with no valves or reservoir bag (0V−R). The fractional inspired concentration of helium deliverable (FiHe) and O₂ (FiO₂) was measured with each facemask.

Results 1. In descending order, the face masks that performed best were: 3V+R >2V+R >1V+R >0V−R. The effects were more pronounced for heliox.

2. Addition of a reservoir bag greatly increased FiHe.

3. Increasing gas flow rate up to 10 L/min consistently resulted in increasing FiHe.

4. Increasing respiratory distress resulted in lower FiHe, regardless of facemask type.

Conclusions 1. 3V+R is the ideal facemask to use for optimised heliox delivery.

2. There is little advantage in using flow rates greater than 10 L/min.

3. In severe respiratory distress there is considerable air entrainment and dilution of the inspired helium.