We thank Dr. Kempen for his interest in our study1and questions and we have attempted to address them in order. The baseline minute ventilation was as given in table 1 for both groups. The ventilator was a Datex-Ohmeda Aestiva ventilator set (GE Healthcare, Madison, WI) during surgery and emergence to volume control with an inspired:expired ratio of 1:2. It has compliance compensation functionality, but despite this there were significantly higher expired tidal volumes in the N2O group during the first 5 min of emergence phase (118 ml higher at 1 min and 110 ml at 5 min, P < 0.05 at each time point with Bonferroni correction), explained by the rapid volume washout of N2O.
In the control group there were six abdominal, three head and neck, and one orthopedic surgeries, with one patient in the prone position. In the N2O group there were five abdominal, two spinal, and three orthopedic surgeries, with two patients in the prone position.
The means (SD) of the blood/gas partition coefficients were 0.73 (0.14) for the control group and 0.69 (0.18) for the N2O group (P = 0.37 on the two-tailed Student t test). The trend toward a higher partition coefficient in the control group might suggest that the difference in blood partial pressures between the two groups we measured was a slight underestimate.
Ventilatory and hemodynamic disturbances due to coughing, straining, etc. , would have disrupted gas elimination and affected our results, as Dr. Kempen points out. We deliberately restricted data sampling during emergence to the first 5 min after cessation of anesthetic gas administration, and at 30 min in the postanesthesia care unit so that this was successfully avoided. We agree that an unblinded study of time to emergence is prone to observer bias, but we found time to eye opening to command a robust endpoint, and time to extubation correlated closely with this. We reiterate that these were secondary endpoints in the study, but were prospectively studied in our protocol, and so were reported.
We believe our intention to fashion the study around a typical general anesthetic protocol with similar depth of anesthesia across the two groups was a sensible one. The different sevoflurane concentrations in the two groups were an inevitable consequence of this, as in standard anesthetic practice. Given that the relative change from baseline in sevoflurane partial pressures was the primary outcome variable, we believe this approach was appropriate. Dr. Kempen's suggestion for a study using identical sevoflurane concentrations and balancing depth of anesthesia with propofol in the control arm is a valid one, and we look forward to the results of such a study. However, we would encourage researchers in the field to take the trouble to measure blood partial pressures rather than just expired concentrations in these types of pharmacokinetic investigations, because of the significant effect of ventilation-perfusion scatter on alveolar-arterial partial pressure gradients. More meaningful quantitative data are obtained and the implications of the findings of a study are clearer.
Finally, we agree entirely that it is possible to accelerate elimination of volatile agent and emergence by deliberately increasing expired alveolar ventilation, which emulates the effect of N2O washout. As Dr. Kempen proposes, turning on N2O near the end of surgery is a useful maneuver that the primary author frequently uses to gain the benefit of the minimum alveolar concentration-sparing and washout effects without the potential side effects of prolonged N2O administration. The intent of our study was to demonstrate the pharmacokinetic principles underlying the administration of inhalational anesthesia with N2O, and its potential implications for speed of emergence. Clearly, the detailed conduct of anesthesia so as to achieve rapid and smooth emergence is a larger issue than simply whether or not to include a single agent such as N2O, and this complex formulation should always be left to the judgment of the skilled anesthesiologist.