To determine the relationship between the delivered gas flows via nasal cannulas and face masks and the set gas flow and the breathing circuit pressure when connecting to the Y-adapter of the anesthesia breathing circuit and using the oxygen blender on the anesthesia machine, relevant to surgery when there is concern for causing a fire. The flow rates that are delivered at various flow rates and circuit pressures have not been previously studied.

Laboratory investigation.

Academic medical center.

None.

The gas flows from each of 3 anesthesia machines from the same manufacturer were systematically increased from 1 to 15 L/min with changes to the adjustable pressure limiting valve to maintain 0–40 cm water pressure in the breathing circuit for nasal cannula testing and at 20–30 cm water circuit pressure for face masks.

The delivered gas flows to the cannula were determined using a float-ball flowmeter for combinations of set gas flows and circuit pressures after connecting the cannula tubing to the Y-piece of the anesthesia circuit via a tracheal tube adapter. Decreasing the supply tubing length on the delivered flow rates was evaluated.

There was a highly linear relationship between the anesthesia circuit pressure and the delivered nasal cannula flow rates, with 0 flow observed when the APL valve was fully open (i.e., 0 cm water). However, even under maximum conditions (40 cm water and 15 L/min), the delivered nasal cannula flow rate was 3.5 L/min. Shortening the 6.5-ft cannula tubing increased the flow at 20 and 30 cm water by approximately 0.12 L/min/ft. The estimated FiO₂ assuming a minute ventilation of 5 L/min and 30% FiO₂ ranged from 21.7% to 27.0% at nasal cannula flow rates of 0.5 to 4.0 L/min. When using a face mask and the APL fully closed, delivered flow rates were 0.25 L/min less than the set flow rate between 1 and 3 L/min and equal to the set flow rate between 4 and 8 L/min.

When using a nasal cannula adapted to the Y-piece of the anesthesia circuit, the delivery system is linearly dependent on the pressure in the circuit and uninfluenced by the flow rate set on the anesthesia machine. However, only modest flow rates (≤ 3.5 L/min) and a limited increase in the inspired FiO₂ are possible when using this delivery method. When using a face mask and the anesthesia circuit, flow rates close to the set flow rate are possible with the APL valve fully closed. Patients scheduled for sedation for head and neck procedures with increased fire risk who require more than a marginal increase in the FiO₂ to maintain an acceptable pulse oximetry saturation may need general anesthesia with tracheal intubation.