Secondary Logo

Journal Logo


Inhaled anaesthetics and nitrous oxide

Complexities overlooked

things may not be what they seem

Hendrickx, Jan; Peyton, Philip; Carette, Rik; De Wolf, Andre

Author Information
European Journal of Anaesthesiology: September 2016 - Volume 33 - Issue 9 - p 611-619
doi: 10.1097/EJA.0000000000000467
  • Free



The title of this review was inspired by that of an editorial written by Ted Eger 20 years ago,1 encouraging us to continue to challenge existing concepts: pharmacokinetics and pharmacodynamics of inhaled anaesthetic drugs have reached the status of ‘all has been said, done, and written’. In this review, we elaborate on some of these existing concepts, and point out where we may have overlooked some complexities, simplified some concepts to such an extent that we hamper progress and have been slow to incorporate some new knowledge into our textbooks. Although not all comments may be clinically relevant, all ought to be thought provoking. The sequence of topics is somewhat arbitrary, and consecutive items do not necessarily relate to one another.

The abbreviation conundrum

The consistent use of abbreviations is essential to allow meaningful communication, especially when discussing pharmacokinetics of gases where items such as water vapour pressure and temperature corrections in and by themselves cause confusion. Despite the existence of international conventions, the abbreviation of ‘inspired oxygen concentration’ continues to be misspelled as ‘FiO2’ or ‘FiO2’ or ‘FIO2’, both in manuscripts and on anaesthesia machine displays (it should be FIO2).2,3 Although concentrations of gases in both liquid phase (dissolved) and gas phase can be expressed as partial pressures in mmHg or kPa, gas concentrations in the gas phase are also often described as the proportion they make up of the total pressure (e.g. 21% oxygen = 160 mmHg of 760 mmHg or 21.3 kPa of 101.3 kPa) or the proportion of a certain volume (volume/volume, e.g. 210 ml out of 1000 ml of air is oxygen, or 21%). This proportion is expressed in either percentage (0 to 100%) or a fraction (0 to 1), for which the letter ‘F’ is used. If ‘F’ is used to denote the partial pressure of a gas dissolved in a liquid, it refers to the fraction of gas present in the imaginary gas phase that would be in equilibrium with that liquid phase (Table 1).

Table 1
Table 1:
List of abbreviations

Anaesthetic uptake and the FA/FI curve

When a patient inhales an anaesthetic gas, the partial pressure in the lungs rises toward the inspired partial pressure. To compare agents with a different minimum alveolar concentration (MAC) or partial pressure, we can observe the rise of the ratio of the alveolar (FA) over the inspired (FI) partial pressure over time, the FA/FI curve. The novice is likely to be shown this graph for the different agents during the first few weeks of residency. The FA/FI curve is usually shown when anaesthetic uptake is discussed, but exactly how does it describe uptake? Uptake is concealed in the area above the curve: uptake is approximated as the product of 1−FA/FI, minute ventilation and FI (Fig. 1). Even though this area is smaller for an agent with a lower blood/gas partition coefficient (λB/G), total body uptake at equipotent partial pressure of a less soluble agent is higher than that of a more soluble agent because its MAC and thus FI is higher: with an FA of 0.75% isoflurane, 1.6% sevoflurane, 4.5% desflurane or 65% nitrous oxide (N2O; all about 0.65 MAC), cumulative uptake after 1 h is approximately 586, 904, 1254 and 14 000 ml of vapour or gas, respectively in an average adult.5

Fig. 1
Fig. 1:
Quantitative interpretation of the F A/F I curve. The F A/F I versus time curve (thick black line) describes how the end-expired partial pressure F A rises toward the inspired partial pressure F I, but it does not directly represent uptake. Uptake is proportional to 1 − F A/F I, which is the area above the F A/F I curve (see formula in the figure). Wash-in of the FRC and uptake by the VRG, MG and FG can then be presented schematically as the separate-coloured areas above the curve. Uptake across the alveolo-capillary membrane is represented by the combined uptake of the VRG, MG and FG (it does not include FRC wash-in). Times to 95% saturation of each modern agent are displayed at the bottom of the figure. Note that the time axis is not linear to allow saturation of the different groups to be schematically displayed. 95% saturation of the CNS (part of the VRG) is displayed separately between brackets. The derivation assumes that the TVIN is equal to the TVEX and that there is no dead space ventilation. Uptake also depends on the absolute value of F I and MV. FG, fat group; FRC, functional residual capacity; K, proportioning constant; MG, muscle group; MV, minute ventilation; RR, respiratory rate; τ, time constant; TVIN, inspired tidal volume; TVEX, expired tidal volume; VRG, vessel-rich group. Reproduced with permission from 4.

New technology is bound to make us rethink the old concepts. During target-controlled delivery, the anaesthesia provider enters the desired FA of the inhaled agent, and the anaesthetic machine will manage fresh gas flows (FGF) and agent administration rates to achieve this target rapidly with very low FGF. This technology forces us to rethink the theoretical effects of cardiac output, ventilation and solubility on the FA/FI curve.6 A higher cardiac output (with a proportional increase in cerebral perfusion) has classically been claimed to slow induction by inhalational anaesthetics because it increases uptake and thus lowers FA, causing the partial pressure in the central nervous system (CNS; FCNS) to rise faster but toward a lower plateau. But when FA is target-controlled and, by definition, is held constant, FCNS will rise faster toward the target FA and hasten induction, just as with i.v. agents. Also, liquid agent injection directly into the circle breathing system [used by the Zeus (Dräger, Lübeck, Germany) and the FLOW-i (Maquet, Solna, Sweden)] can be so fast that the rate of rise of FA becomes independent of λB/G because of the overpressure that can be provided (Fig. 2).5 Finally, a change in ventilation would not alter FA (or only momentarily): target control means that if FA tends to increase, the FGF is briefly increased to wash out agent, and if FA tends to decrease, more agent will be injected; all this results in a stable FA.

Fig. 2
Fig. 2:
Solubility of modern inhaled anaesthetics has no effect on the rate of rise of F A if they are administered by liquid injection. When a liquid injection technique is used, the degree of overpressure that can be provided is such that the rate of rise of the end-expired isoflurane, desflurane and sevoflurane partial pressures (here expressed as MAC equivalents) are identical, and thus become independent of λB/G. Commonly used pharmacokinetic principles of uptake of potent inhaled anaesthetics do not apply under these conditions or warrant a different interpretation. MAC, minimum alveolar concentration. Reproduced with permission from 5.

Even during conventional delivery, the effect of ventilation on FA has to be qualified, because it depends on the FGF: it ranges from no effect during closed-circuit anaesthesia to ‘full’ effect according to the universal gas equation [FA = FI − (anaesthetic uptake/alveolar ventilation)], with a vast and complex ‘no man's land’ in between.7 Increased ventilation will always shorten the wash-in (or wash-out) time constant of the lungs, but this is only relevant after a step change in the delivered concentration (FD).

Blood/gas partition coefficients are overrated

A lower λB/G causes FA to rise faster toward FI, but this does not necessarily translate into anticipated clinical advantages such as faster induction. Desflurane has a low λB/G, but is rarely used to induce anaesthesia because of airway irritation, laryngospasm, hypertension and tachycardia. In adults, anaesthesia is most often induced intravenously, and there may be no need for a rapid rate of rise in the period after securing the airway and before surgical incision when adequate propofol and opioid concentrations are still present.

As previously mentioned, liquid injection of inhaled agent can make the rate of rise of FA independent of λB/G. In addition, the clinical relevance of λB/G of an agent cannot be considered separate from that of the other tissue solubilities. Even though the λB/G of N2O and desflurane are similar, immediate recovery after maintaining anaesthesia with desflurane is still slower than when part of the desflurane is replaced by N2O, because the solubility of N2O in other tissues is lower.8 Also, the wash-in and wash-out time of the CNS is determined by its time constant (τCNS): capacity of the CNS for the agent divided by agent transfer to the CNS, or (volumeCNS × λCNS/G)/(perfusionCNS × λB/G). Therefore, τCNS is proportional to λCNS/GB/G. This ratio is smallest for N2O, making it the fastest agent in our routine clinical armamentarium (isoflurane 1.56; sevoflurane 1.69; desflurane 1.22; N2O 1.06). Although wake-up times are influenced by tissue solubilities, they can be shortened by isocapnic hyperventilation and the use of N2O during the last part of the anaesthetic. With isocapnic hyperventilation, lung wash-out increases whereas isocapnia and therefore cerebral perfusion are maintained through the addition of carbon dioxide (CO2) to the inspired gas mixture.

One disadvantage of a high λB/G that has been underappreciated during emergence is that the lower lung clearance of an agent with a higher λB/G not only has an effect on the speed of emergence (Fig. 3), but also puts the patient at higher risk of rehypnotisation with a lesser degree of hypoventilation compared with agents with a lower solubility.9

Fig. 3
Fig. 3:
Effect of blood solubility on lung clearance. Partial pressure in the blood and gas phase before (left) and after (right) equilibration for three hypothetical agents with different blood/gas partition coefficients (a, b, c). Reproduced with permission from 4.

Rebreathing and the FDFIFA relationship

The classical FA/FI over time graph sufficed to describe the kinetics of inhaled anaesthetics when high FGFs were used, because under these conditions the vapouriser setting (FD) matches FI, which is perceived by the clinician as ‘being in control’. When a circle breathing system is used with reduced FGF, FI will no longer match FD because of rebreathing of exhaled gas. The difference FDFI is determined by the degree of rebreathing, and the difference FIFA by the amount of uptake. Lowe and Ernst7 mathematically described the FD, FI, FA, FGF and time relationship by ‘the general anaesthetic equation’. For reasons unclear to us, this concept has been ignored in textbooks. Figure 4 depicts these relationships for sevoflurane.5,7,10

Fig. 4
Fig. 4:
Empirically derived general anaesthetic equation for sevoflurane in oxygen/air with an ADU (Anaesthesia Delivery Unit: GE, Madison, WI, USA) anaesthesia machine to maintain F Asevo = 1.8%. F A = yellow surface; F I = green surface; F D = orange surface. The difference between F I and F A is caused by uptake; the difference between F D and F I is caused by rebreathing. The first 5 min are not presented. Most clinicians intuitively use a FGF of 1.5 to 2 l min-1, because then the difference between F D and F I is not very large; however this difference is more pronounced with FGF < 1.5 l min-1. Data such as these can be used to develop administration schedules by averaging the F D with a certain FGF over a specific time period. Two regimens to attain F Asevo = 1.8% are presented as examples: FGF of 0.75 l min-1 (light grey line) and F D = 4.5% between 5 and 15 min, followed by F D = 3.5%; or FGF of 2 l min-1 (dark grey line) and F D = 3.2% between 5 and 15 min, followed by F D = 2.7%. The search for these sequences has been dubbed ‘the search for the ideal FGF-F D sequence’, that reduces consumption yet most promptly achieves the target F Asevo. ADU, anaesthesia delivery unit (GE, Madison, WI, USA); FGF, fresh gas flow. Reproduced with permission from 4,5,9.

During the maintenance phase of anaesthesia, the difference between FI and FD remains fairly small as long as FGF is at least 1.5 l min−1, which explains why many anaesthesiologists intuitively use at least a 1 to 2 l min−1 of FGF. Unfortunately, this 1 to 2 l min−1 FGF range also happens to be the FGF range in which hypoxic guards provide the least protection against the formation of an hypoxic FIO2 because of rebreathing, a fact which remains surprisingly poorly appreciated by the anaesthesia community.11,12

Why inhaled agents are so easy to use: mystery and physics

Because the dose–response curves of inhaled anaesthetics for the clinical end points, ‘unconsciousness’, ‘immobility’ and ‘suppression of sympathetic response’, are steep with a narrow spread, the FA of the inhaled anaesthetics can be used as a measure of anaesthetic depth, that is, 1 or 1.3 MACawake, MAC and MACBAR (at steady state) define the 50 or 95% likelihood of hypnosis (loss of response to verbal command), immobility and blockage of adrenergic response (BAR) (or haemodynamic control), respectively.13,14 Why inhaled anaesthetics have this property is a mystery: ‘Someday, when we understand the mechanism of inhaled anaesthetic action, we will look back on this low variability in MAC and think that it was so obvious that the mechanism had to be ‘X’, because only that could have accounted for the low variability’.13

In addition, physics (Henry's law) explains why their administration is so easy. Because transport of inhaled anaesthetics is partial pressure-driven, the partial pressure in all tissues will automatically (governed by the physics laws) rise toward the FA – and we can measure the latter continuously! Some patients will take up more than others (because of differences in size or body composition), but have you wondered why we do not have to be bothered by this (in contrast to i.v. drug delivery)? Elementary physics, my dear Watson. Steering the measured FA toward the desired FA by adjusting the vapouriser suffices.

Minimum Alveolar Concentration semantics

MACawake, MAC and MACBAR have been defined in the previous paragraph. For some reason, MAC never acquired the tag ‘immobility’, ‘minimum’ should have been denoted ‘median’ and ‘concentration’ should be read as ‘partial pressure’. And why is there no MACpain? There can be no pain when one is unconscious. The stimuli used to define anaesthetic depth ‘would’ be painful ‘if the patient was awake’. The use of the word ‘pain’ is an eternal source of confusion when discussing anaesthetic depth and should be abandoned: when a person is unconscious, there is no pain because pain is a conscious perception. The terms ‘nociception’ and ‘noxious stimuli’ are to be preferred in this context, and the terms ‘opioids’ or ‘antinociceptive drugs’ should be used instead of ‘analgesics’ when discussing their intraoperative use.

How to get the maximum out of the minimum alveolar concentrations: hysteresis, differential opioid effects, context-sensitive half-times

The different MACs as defined above will only be found to be clinically useful if they are age adjusted, incorporate hysteresis and take into account the differential effect of opioids on each of the different MACs as well as the drugs’ context-sensitive half-times.

The MAC concept is based on the measurement of FA. However, when FA changes, the partial pressure in the CNS changes with a specific delay according to τCNS. For sevoflurane and desflurane, the values are approximately 3.3 and 2.4 min, respectively; 95% equilibration is obtained after 3 τ. Therefore, patients may still move even though MAC may have been above 1.3 MAC if the incision has been within 3 τ (Fig. 5). During emergence, patients may not wake up even though MAC has been less than 0.3 for a while. But during the maintenance phase and during slow alveolar wash-out time (‘coasting’), the displayed MAC will reflect the probability of response suppression more accurately because there is less hysteresis under these conditions.15 Combining slow alveolar wash-out time with the continued use of N2O ensures an exceedingly rapid, consistent emergence after turning off N2O and increasing FGF, with MAC and MACawake excellent guides to anaesthetic depth. In the near future, effect site partial pressure/concentration and drug interaction displays such as the SmartPilot (Dräger) and the Navigator (GE, Madison, Wisconsin, USA) will help the clinician fine-tune emergence from anaesthesia.16

Fig. 5
Fig. 5:
F A as an estimate of response suppression. The definition of ‘equilibrium’ and ‘steady state’ for inhaled anaesthetics is confusing because of the existence of an F A-F a gradient as well as a delay until F CNS reaches F a. Steady state and equilibrium are said to exist if F A, F a and F CNS all have levelled off (plateau phase) and are more or less parallel once F CNS = F a; they do not require that F A = F a. Because F CNS by definition correlates directly with clinical effect, it is also labelled the ‘effect site concentration’ or C e (the better term would be ‘effect site partial pressure’). F CNS can be derived from F A with an empirically derived t 1/2effect site, which describes the F AF CNS delay. This delay is shorter for the ‘faster’ inhaled anaesthetics (those with lower solubility). The use of C e (e.g. by the SmartPilot) helps the clinician recognise that even though F A suggests anaesthetic depth to be, for example, 1.3 MAC, FCNS may not yet have equilibrated with F a: if the stimulus is applied before F a = F CNS, more patients will move than anticipated and the MAC concept will appear not to be valid. F, fraction or partial pressure; F A, alveolar fraction (end-expired fraction); F I, inspired fraction; F D, delivered fraction (vapouriser setting); F a, arterial fraction; F CNS, fraction in central nervous system; MAC, minimum alveolar concentration. IAD, inhaled anesthetic drug. Reproduced with permission from 4.

Drug interactions between opioids and inhaled agents can have a pronounced effect. A remifentanil equivalent effect site concentration of 1.5 to 2 ng ml−1 (a concentration typically attained with a bolus of 10 μg sufentanil or 100 μg fentanyl) reduces the required FA to suppress proper response to a verbal command, movement in response to a noxious stimulus or sympathetic response to tracheal intubation in 50% of study participants by 10 to 15, 50 and 75%, respectively (Fig. 6).17–20 Maximally exploiting this synergy allows for a significant FA reduction (minimising agent usage) and may result in faster emergence. A proper understanding of the underlying pharmacokinetics and pharmacodynamics turn MAC and MACawake into extremely useful monitoring and guidance tools. Systems such as the SmartPilot and the Navigator incorporate differential opioid effects, hysteresis and context-sensitive half-times, and will be used routinely in the future to help improve drug titration.


Ventilation/perfusion relationships and lung gas uptake: moving beyond Riley's model

The concept of MAC gave us an excellent clinical tool for real-time monitoring of the contribution of inhaled agents to anaesthetic depth through FA measurements. But anaesthetic depth results from the effect of FCNS, and the latter will be more closely reflected by arterial partial pressures (Fa) than by FA of anaesthetic gases. How well does FA reflect Fa? In fact, they differ more often than is realised (Fig. 5), for three reasons. First, for any gas, the FA-Fa difference is a product of ventilation/perfusion (V/Q) scatter in the lung, as is routinely seen for CO2 and oxygen whenever an arterial blood gas sample is taken. Second, inhalational anaesthesia itself ‘increases’ V/Q scatter and the FAFa differences for all gases. Finally, volatile agents are heavy molecules with finite diffusibility and develop longitudinal partial pressure gradients, which add to the difference between partial pressures in the alveoli and pulmonary blood.

Under the familiar three compartment ‘Riley’ model of V/Q scatter (shunt, dead space and ‘ideal’ compartments), alveolar dead space is calculated from the Bohr–Enghoff equation, using inspired, end-tidal and arterial partial pressures of CO2. However, because of their different solubilities and diffusion characteristics, the FAFa differences for anaesthetic gases are not the same as those for CO2. The Bohr equation gives us a vastly larger alveolar dead space for isoflurane compared with that for CO2 in the same patient.21 This discrepancy presents a real issue when determining the pharmacokinetics of inhalational anaesthetics between vapouriser and body tissues solely based on FI and FA measurements.

So how much do these pharmacokinetic niceties matter in clinical practice? In fact, the MAC concepts work well, precisely because their values have been derived from FA measurements and thus ignore the FAFa difference. However, these issues cause great confusion when we examine other, more complex phenomena. An example is the second gas effect, where rapid uptake of N2O is said to exert a concentrating effect on the other alveolar gases, increasing FA/FI for an accompanying volatile agent (Fig. 7a). In the Riley model, alveolar dead space measured using the Bohr equation for CO2 is relatively small in healthy lungs and all gas uptake occurs in the large ‘ideal’ lung compartment (Fig. 7b). Reliance on this model convinced many that these concentrating effects must be trivial, and that the second gas effect must therefore be clinically unimportant.23 Indeed, most studies consistently show only a small acceleration of rise in FA/FI for volatile agent after induction of anaesthesia when N2O/oxygen is the vehicle gas mixture.24,25

Fig. 7
Fig. 7:
Pitfalls of V/Q scatter: moving beyond Riley's model to explain the second gas effect. (a) When 70% N2O (the first gas) is the carrier gas along with oxygen, the alveolar partial pressure of sevoflurane (the second gas) rises faster toward the inspired partial pressure. However, the effect on arterial partial pressure of sevoflurane (sevo) is 2 to 3 times greater. (b) The traditional diagram depicts the mechanism for the second gas effect in a simplified model of the lung with F IN2O = 80%, F IO2 = 19% and F I ‘second gas’ = 1%. The increase in F AO2 and second gas concentration is shown to the right, where half of the inspired volume of N2O is taken up by blood (heavy lines at bottom). The concentrations of the gases in blood leaving the lung are shown at the bottom. This model, however, cannot explain the much greater second gas effect observed on arterial partial pressures. (c) Modified version to explain that ‘the second gas effect’ is caused by the uptake of N2O in a subset of alveoli with a particular V/Q mismatch that is typically present during anaesthesia. The hypothetical lung is divided into two compartments, so that the left-hand compartment now receives three fourths of the blood flow. Uptake of N2O in the compartments is kept proportional to their blood flow. The resulting concentrations of the alveolar gases in each compartment are shown, as well as the flow-weighted concentration in blood leaving the lung. Inhomogeneity of blood flow has increased the second gas effect. The effect is more pronounced in the arterial blood (and thus end-organs) than in alveoli. V/Q, ventilation/perfusion. P/PI, partial pressure relative to the inspired pressure. Reproduced with permission from 22.

However, when Fa is measured, the second gas effect is found to be much more powerful (Fig. 7a).22 The uptake of soluble gases such as N2O is greatest in lung units with generous blood flow (moderately low V/Q ratios). Here, the concentrating effects of N2O uptake (even of the order of magnitude of 100 ml min−1) relative to ventilation are powerful, and these units determine the gas content of arterial blood.25,26 This regional concentrating effect is powerful enough to augment arterial oxygen partial pressure even during the maintenance phase of anaesthesia.27 The effect also operates in reverse to accelerate the fall in arterial volatile agent during emergence, because of N2O washout.28 Based on FA, the true magnitude of the second gas effect is ‘hidden’ from us by the unmeasured FAFa differences of the gases we are investigating. The classical figure of the second gas effect (Fig. 7b) needs to be adjusted accordingly (Fig. 7c).

Why nitrous oxide continues to be a useful anaesthetic

A consensus panel of the European Society of Anaesthesiologists recently concluded that there is no reason not to use N2O when not specifically contraindicated.29 The ENIGMA II study effectively gave N2O a ‘clean bill of health’.30 If it were discovered now, and taking into account its obvious well known contraindications and appropriate measures such as tracheal tube cuff pressure monitoring, we could embrace it as the fastest anaesthetic agent we have at our disposal. The second gas affect is more pronounced than hitherto assumed, hastening induction and improving arterial oxygenation at a given FIO2. It lacks significant haemodynamic effects (except in the rare patient with pulmonary hypertension) and allows the doses of other agents to be reduced. A reverse second gas effect, a high MACawake and antagonistic interaction with sevoflurane for hypnosis further hasten emergence. It has no effect on the incidence of nausea and vomiting in procedures shorter than 1 h.31 Automated low-flow anaesthesia makes environmental concerns moot. Different experts have different views on the usefulness of N2O, highlighting the need for a large, comprehensive evidence-based review.


Progress can only be made by re-searching and researching: this review, the summary of a lecture at the 2016 European Society of Anaesthesiologists Annual Meeting in London, hopes to throw old (but ignored) and new (but potentially overlooked) information into the educational and clinical arenas to stimulate discussion among clinicians and researchers alike. It also hopes to serve as a warning not to let technology pass by our all too engrained older concepts.

Acknowledgements relating to this article

Assistance with the review: we would like to thank Dr Michiel D’Hondt, Dr Simon De Ridder and Dr Alexander Dehouwer for critically reviewing the manuscript.

Financial support and sponsorship: none.

Conflicts of interest: JH has received lecture support or travel reimbursements or equipment loans or consulting fees or meeting organisational support ( from AbbVie, Acertys, Air Liquide, Allied healthcare, Armstrong Medical, Baxter, Draeger, GE, Hospithera, Heinen und Lowenstein, Intersurgical, Maquet, MDMS, MEDEC, Micropore, Molecular, NWS, Philips, Quantium Medical.

Presentation: none.

Comment from the Editor: the text is a summary of a refresher course lecture presented at the European Society of Anaesthesiologists Annual Meeting, London, UK 2016.


1. Eger EI 2nd. Complexities overlooked: things may not be what they seem. Anesth Analg 1997; 84:239–240.
2. Pappenheimer JR, Comroe JH, Cournand A, et al. Standardization of definitions and symbols in respiratory physiology. Fed Proc 1950; 9:602–605.
3. Quanjer PH, Tammeling GJ, Cotes JE, et al. Symbols, abbreviations and units. Eur Respir J 1993; 6 (suppl 16):85–100.
4. Hendrickx JFA, Van Zundert AAJ, De Wolf AM. Hardman J, Hopkins P, Struys M. Inhaled anaesthetics. The Oxford textbook of anaesthesia. Oxford: Oxford University Press; 2017; (in press).
    5. Hendrickx JFA. The pharmacokinetics of inhaled anesthetics and carrier gases. University of Gent, Belgium, June 17, 2004. Thesis (PhD) Freely downloadable at [Accessed April 1, 2016].
    6. Van Zundert T, Hendrickx J, Brebels A, et al. Effect of the mode of administration of inhaled anaesthetics on the interpretation of the FA/FI curve: a GasMan simulation. Anaesth Intensive Care 2010; 38:76–81.
    7. Lowe HJ, Ernst EA. The quantitative practice of anesthesia: use of closed circuit. Baltimore/London: Williams & Wilkins; 1981.
    8. Rampil IJ, Lockhart SH, Zwass MS, et al. Clinical characteristics of desflurane in surgical patients: minimum alveolar concentration. Anesthesiology 1991; 74:429–433.
    9. De Wolf AM, Van Zundert TC, De Cooman S, et al. Theoretical effect of hyperventilation on speed of recovery and risk of rehypnotization following recovery: a GasMan simulation. BMC Anesthesiol 2012; 12:22.
    10. Hendrickx JFA, De Cooman S, De Wolf AM. Do decreased sevoflurane costs offset increased canister costs when fresh gas flows are reduced? ASA Annual Meeting 2011; Abstract 1463.
    11. De Cooman S, Schollaert C, Hendrickx JF, et al. Hypoxic guard systems do not prevent rapid hypoxic inspired mixture formation. J Clin Monit Comput 2015; 29:491–497.
    12. Ghijselings IE, De Cooman S, Carette R, et al. Performance of an active inspired hypoxic guard. J Clin Monit Comput 2016; 30:63–68.
    13. Dilger JP. From individual to population: the minimum alveolar concentration curve. Curr Opin Anaesthesiol 2006; 19:390–396.
    14. Sani O, Shafer SL. MAC Attack? Anesthesiology 2003; 99:1249–1250.
    15. Katoh T, Suguro Y, Kimura T, et al. Cerebral awakening concentration of sevoflurane and isoflurane predicted during slow and fast alveolar washout. Anesth Analg 1993; 77:1012–1017.
    16. Johnson KB, Syroid ND, Gupta DK, et al. An evaluation of remifentanil-sevoflurane response surface models in patients emerging from anesthesia: model improvement using effect-site sevoflurane concentrations. Anesth Analg 2010; 111:387–394.
    17. Katoh T, Ikeda K. The effects of fentanyl on sevoflurane requirements for loss of consciousness and skin incision. Anesthesiology 1998; 88:18–24.
    18. Katoh T, Kobayashi S, Suzuki A, et al. The effect of fentanyl on sevoflurane requirements for somatic and sympathetic responses to surgical incision. Anesthesiology 1999; 90:398–405.
    19. Katoh T, Nakajima Y, Moriwaki G, et al. Sevoflurane requirements for tracheal intubation with and without fentanyl. Br J Anaesth 1999; 82:561–565.
    20. Katoh T, Uchiyama T, Ikeda K. Effect of fentanyl on awakening concentration of sevoflurane. Br J Anaesth 1994; 73:322–325.
    21. Landon MJ, Matson AM, Royston BD, et al. Components of the inspiratory-arterial isoflurane partial pressure difference. Br J Anaesth 1993; 70:605–611.
    22. Peyton P, Horriat M, Robinson G, et al. Magnitude of the second gas effect on arterial sevoflurane partial pressure. Anesthesiology 2008; 108:381–387.
    23. Lin CY. Uptake of anaesthetic gases and vapours. Anaesth Intensive Care 1994; 22:363–373.
    24. Taheri S, Eger EI. A demonstration of the concentration and second gas effects in humans anesthetized with nitrous oxide and desflurane. Anaesth Analg 1999; 89:774–780.
    25. Hendrickx JF, Carette R, Lemmens HJ, et al. Large volume N2O uptake alone does not explain the second gas effect of N2O on sevoflurane during constant inspired ventilation. Br J Anaesth 2006; 96:391–395.
    26. Peyton P, Robinson G, Thompson B. The effect of ventilation-perfusion inhomogeneity and nitrous oxide on oxygenation in anesthesia: physiological modeling of gas exchange. J Appl Physiol 2001; 91:17–25.
    27. Peyton P, Stuart-Andrews C, Deo K, et al. Persisting concentrating and second gas effects on oxygenation during N2O anaesthesia. Anaesthesia 2006; 61:322–329.
    28. Peyton P, Chao I, Weinberg L, et al. Nitrous oxide diffusion and the second gas effect on emergence from anesthesia. Anesthesiology 2011; 114:596–602.
    29. European Society of Anaesthesiology task force on use of nitrous oxide in clinical anaesthetic practice. The current place of nitrous oxide in clinical practice: an expert opinion-based task force consensus statement of the European Society of Anaesthesiology. Eur J Anaesthesiol 2015; 32:517–520.
    30. Myles PS, Leslie K, Chan MT, et al. The safety of addition of nitrous oxide to general anaesthesia in at-risk patients having major noncardiac surgery (ENIGMA-II): a randomised, single-blind trial. Lancet 2014; 384:1446–1454.
    31. Peyton P, Wu C. Nitrous oxide induced PONV is related to duration of exposure. Anesthesiology 2014; 120:1137–1145.
    © 2016 European Society of Anaesthesiology