To the Editor:
We read with great interest the study by Peyton1 in which the effects of molar mass on the rate of diffusion of desflurane and nitrous oxide are compared. The author hypothesized the end-tidal-arterial partial pressure gradient for desflurane to be greater than nitrous oxide based on Graham’s law of diffusion.1 However, contrary to this hypothesis, the initial results showed a less than expected end-tidal-arterial partial pressure gradient for desflurane in comparison to nitrous oxide.1 This finding was attributed to the higher rate of desflurane uptake.1
After adjusting for lung uptake rate of desflurane, the results showed no evidence of end-tidal-arterial gradient difference between the two gases.1 Although this study should be prized for its sophisticated technical design, there are several reasons to be skeptical of its conclusion.
In order to achieve accurate results, we believe the study should be revised to account for the following. Our first observation relates to the patients included in the study. Table 1 in the article shows the reported oxygen uptake in this study is, note values in parentheses are SD, 166 (45) ml/min, and the reported carbon dioxide output is 166 (52) ml/min.1 This shows a calculated respiratory quotient of 1. A respiratory quotient of 1 exceeds the normal of 0.8 and makes these patients ineligible for a study of this nature and furthermore renders the results unreliable. It should be noted that the data used for calculating dead space for an anesthetic gas (VDA/VAG)1 in this study were taken from the previous study in which the reported respiratory quotient was claimed to be 0.8 and that in the subsequent study, the patient groups were described as similar.2 The higher SD for carbon dioxide as compared to oxygen shows the possibility of a respiratory quotient of more than 1 in some patients.1
The authors allowed a range of concentrations from 2 to 3% for desflurane and from 10 to 15% for nitrous oxide in this study.1 When the goal of the study is to compare diffusion of two different gases, it would be prudent to choose a single concentration for both gases to eliminate the effect of concentration on partial pressure of inspired gas, end-tidal partial pressure of gas, and arterial partial pressure of gas.1
The author reported sample collections upon “achievement of near steady-state maintenance phase anesthesia between 30 to 60 min postinduction.” While we can generally assume a clinically near steady state 30 min postinduction, if the goal is to compare the end-tidal-arterial gradient of these gases, then we should not assume a pharmacokinetically near steady state at 30 to 60 min as diffusion and redistribution of these gases follow two different rates during this period. Although during the first 10 to 15 min postinduction one can assume the same rate of diffusion and distribution due to almost identical blood/gas3 and tissue/blood (brain/blood)3 partition coefficients for these gases, the rate of muscle uptake for these gases is very different at the time of sampling. These gases have different muscle/blood partition coefficients: 2 for desflurane and 1.2 for nitrous oxide.3 This results in a 1.67 times larger time constant for desflurane compared to nitrous oxide.3 Considering 15% of cardiac output flows through muscle tissue with a mass equal to 43% of patient body weight, and muscle/blood partition coefficient of 1.2, one time constant for nitrous oxide for these patients (weight, 82.4 kg; cardiac output, 4 l/min)1 will be 72 min and for desflurane will be 120 min.3 This shows a larger rate of decline in uptake and distribution of nitrous oxide compared to desflurane at 30 to 60 min after induction.
Taking into consideration the ventilatory parameters in this study, choosing 6 l/min fresh gas flow would allow rebreathing.1 Dilution due to rebreathing of expired gases will reduce the inspired partial pressures of both gases but has a greater effect on more diffusible gas compared to less diffusible gas,3 making the achievement of steady state of inspired gas unreliable.
In conclusion, we should consider blood/gas solubility as the dominant factor for determining the gas exchange impairment4 and recognize that the absence of blood solubility results in the absence of diffusion. The tissue/blood partition coefficient and timing of sampling should also be considered when comparing diffusibility of two gases if the gases have different tissue/blood partition coefficients.1 Last, we should note the limitation of the measuring device used in the study. Datex-Ohmeda Capnomac Ultima (GE Healthcare, USA)1 has an accuracy of 2 volume percent for nitrous oxide and 0.2 volume percent for desflurane. Considering the concentrations used for nitrous oxide and desflurane in this study, we can expect greater inaccuracy for nitrous oxide (20%) in comparison to desflurane (10%) and recognize that other properties, such as molar mass (the focus of this study) of a gas, may affect the rate of diffusion. This aspect warrants further investigation.
The authors declare no competing interests.