Variation in the amount of leakage containing inhalational anaesthetics in the breathing circuits of anaesthesia machines

We found the article by Mulier et al. intriguing [1]. The authors cited our data [2] to estimate sevoflurane loss due to system leak in anaesthesia machines during maintenance of general anaesthesia. Their calculations suggest that leak was constant, which is inaccurate. The amount of leak is proportional to the pressure difference between the interior and exterior of the breathing circuit. The internal pressure of the breathing circuit increases and decreases during positive pressure ventilation. Thus, ‘leak level’ can be defined as the amount of leakage when the pressure in the breathing circuit is 30 cmH2O. However, airway pressure is rarely maintained above 30 cmH2O during general anaesthesia. Our experiment using a high-fidelity simulator showed that leaks seldom occur during low-flow anaesthesia if the leak level was < 300 ml.min-1 (Fig. 1) [3]. In the study, the leak level varied from 0 to 2000 ml.min-1 using artificial leak-producing devices [4] under low or high fresh gas flow conditions, and when respiratory compliance (simulated compliance of the sum of lung and thorax compliance) was arbitrarily varied between low, medium and high. The only case of a decrease in tidal volume observed at leak levels < 300 ml.min-1 was due to low compliance with a high fresh gas flow. This was the first study to show that the acceptable leak level of the breathing circuit is < 300 ml.min-1. These data were obtained using an older anaesthesia machine (Narkomed 2 B, manufactured by Dräger 20 years ago). System leaks should be negligible with modern anaesthesia machines that incorporate automatic leak testing and pressure-regulated ventilation modes.