To the Editor: 

This recent excellent study on the second gas effect was extremely interesting.1I would like to ask several questions to further appreciation of the findings presented. Which ventilator type, mode, inspiratory/expiratory time settings, and fresh gas flows intraoperatively (vs.  9 l/m during emergence) were used? Some ventilator models self-correct to end expiratory volumes, whereas others (i.e. , volume-controlled Narcomed II [Draeger Medical Inc., Telford, PA]) deliver fixed inspiratory volumes, which further affect expired volumes by changes in fresh gas flow as well as the additional significant N2O egress volumes. Increasing fresh gas flow from 3 l/m to 9 l/m to fixed ventilator inspiratory volumes using 1:2 inspiratory/expiratory ratios can affect up to 200 versus  2,000 ml tidal versus  minute volume changes, respectively. Although reported minute volumes are described as nonsignificantly different in N2O versus  air/oxygen control group via  ml × min × kg (ml/Kg/min???) in your table 1, the calculated respective 6,605 versus  5,749 absolute ml/min is 15% difference (was this significant statistically?), which would be even greater if ventilator delivered versus  exhaled volumes were actually reported.

Severinghaus showed that N2O volumes of 20–29 l accumulate in the body using 80% N2O within 1–2 h and this volume is recovered and added to expiratory volumes during emergence, contributing to second gas removal. This occurs predominantly in the first 5–10 min and as much as 1 l/m N2O alveolar-arterial mass flow were described with maximum changes in the fraction of inspired N2O.2Diffusion hypoxia results, when during emergence, the combination of low tidal (ca 200 ml and just above dead space volumes via  endotracheal tube)/min volumes, airway occlusion and mass flow of nitrous oxide from the body in the face of respiration of air (79% nitrogen and 21% oxygen) does not provide adequate oxygen transfer to meet demands. Hypoxemia is not obligate with adequately maintained minute volumes using high fresh gas flows and air, however.

Surgeries and positions also were not defined (although one patient in the lateral position was excluded due to mucus plugs), whereas especially endoscopic/thoracic surgeries and nonsupine positions would expectantly significantly affect respiration parameters. The use of bispectral index delineation of anesthetic depth was an unusual distraction, resulting in two different fraction of inspired sevoflurane concentrations for emergence. This resulted in the development of a theoretic “normalized sevoflurane in N2O” curve. The reported sevoflurane blood/gas partition coefficients (Pa0-S/PA0-S) for Control versus  N2O groups are 0.76 and 0.83 (using data from table 1 and before changing from steady-state gas anesthesia at T0), quite different from the known 0.65.3To what do the authors attribute these differences? During emergence, were there significant pulmonary or methodologic problems/differences in the two groups that might compound Pa/Pa0-S relationships over time emerging, or is this more likely due to increasing methodologic inaccuracies, as the inhaled and then measured sevoflurane concentrations decline? Emergence phenomena including reversal agent hemodynamic perturbations, coughing, straining, breath-holding, switching to and anesthetic depression of spontaneous ventilation, further complicate volatile elimination, occurring typically well before eye opening, extubation, and emergence. Was this evident? The authors speculated as to the source of differences in the arterial carbon dioxide concentrations at 2 and 5, versus  30 min during emergence, but with so many possible and undescribed factors, it is interesting yet difficult to clarify as a reader. The use of (unknown) intraoperative and postoperative narcotics in the postanesthesia care unit certainly may be another factor in these surgeries requiring arterial line insertion. Why were fentanyl and morphine at induction chosen and in which group and at what rates/intervals, with such different known durations of opioid effect?

I understand the authors wished to emulate clinical emergence, but emergence itself is a significant confounding issue in this study of gas pharmacokinetics. The many confounding issues are important and remain of interest at this point. A self-subject crossover controlled study before surgery, using fixed high fresh gas flow greater than minute volumes (i.e. , 10–15 l/m), fixed inspiratory volumes using maintenance sevoflurane 1.4% and 66% nitrous oxide versus  1.4% sevoflurane in oxygen with propofol total intravenous anesthesia to ensure anesthetic depth in both groups (a true comparison test), switched to air (or oxygen) only during N2O elimination (with sustained manual ventilation, anesthetic depth, and paralysis), would have effectively removed the multiple variables questioned previously, provided clearer information and obviated any need for extrapolation of a “normalized sevoflurane in N2O curve.” There is no need to emerge from anesthesia, simply to document effects on elimination of the second gas sevoflurane at one concentration during the switch from 66% N2O versus  control to oxygen. Using bispectral index as a questionable index of anesthetic depth versus  known minimum alveolar concentration constants, only complicated the issues here (also recognized by the authors).

Finally, it is not required to use N2O to speed emergence via  second gas effect. One can maintain carbon dioxide at permissively higher concentrations toward the end of surgery (and thus ensuring patient respiration) and once the sevoflurane is terminated and high fresh gas flow is instituted, a simple increase in minute volume by 1–2 l/min toward normocapnea will more than compensate any mass flow or second gas effect of N2O egress on speeding emergence, as this effect is limited to the maximum 10–15 l N2O flowing from the body in the first 10 min. We may not be able to provide an inspired concentration of less than zero on emergence, but we can increase manual ventilation, hyperventilating from permissive hypercapnia to normocapnic levels to remove gases. Similarly, switching from pure sevoflurane to N2O monoanesthesia near surgical end (near or after peritoneum closure) can further speed emergence and avoid surgical disdain from bowel distention, as bowel is now hidden from view, inadequate time remains to accumulate significant intraabdominal volumes of nitrous oxide and emergence from N2O alone is the  most rapid emergence.

Peyton PJ, Chao I, Weinberg L, Robinson GJ, Thompson BR: Nitrous oxide diffusion and the second gas effect on emergence from anesthesia. ANESTHESIOLOGY 2011; 114:596–602
Severinghaus JW: The rate of uptake of nitrous oxide in man. J Clin Invest 1954; 33:1183–9
Eger EI: Uptake and distribution, Miller's Anesthesia, 6th edition. Philadelphia, Elsevier Churchill Livingstone, 2005, pp 131–53