In a patient whose airway is likely to become obstructed upon loss of consciousness, anesthesia may be induced using an inhaled vapor. If the airway occludes during such an inhalational induction, the speed of patient awakening is related to the rate at which anesthetic gas redistributes away from lung and brain to other body compartments. To determine whether redistribution occurs more rapidly with a more blood-soluble or a less blood-soluble agent, the authors used subanesthetic concentrations of halothane and sevoflurane to simulate inhalational induction and airway obstruction in eight healthy human volunteers.


Inhalational induction was simulated using stepwise increases in inspired halothane or sevoflurane concentration, sufficient to reach an end-tidal concentration of approximately 0.1 minimal alveolar concentration. Airway occlusion was then simulated by initiating a 90-s period of rebreathing from a 1-l bag. During rebreathing, end-tidal halothane or sevoflurane concentration was measured continuously by mass spectrometry, and a time constant for the decline in concentration was calculated using a monoexponential model.


At the onset of rebreathing, end-tidal concentrations of halothane and sevoflurane were 0.10 +/- 0.03 and 0.11 +/- 0.03 minimal alveolar concentration, respectively (mean +/- SD; P > 0.1, Student t test). During rebreathing, the time constants for the decline in end-tidal halothane and sevoflurane concentration were 22 +/- 9 and 62 +/- 16 s, respectively (P < 0.0001).


During simulated airway occlusion in healthy volunteers, the end-tidal concentration of halothane falls more rapidly than that of sevoflurane. Halothane may therefore lead to more rapid awakening, compared with sevoflurane, should the airway obstruct during an inhalational induction of anesthesia.

IN a patient whose airway is likely to become obstructed during anesthesia (e.g.  in the presence of an upper airway mass), anesthesia may be induced by using an inhaled vapor, rather than a more conventional intravenous agent. This technique is useful because, should the airway become completely obstructed, inhalation of the vapor would cease and the patient should wake relatively quickly to restore his or her own airway, avoiding the need for urgent surgical intervention.

For many years, the anesthetic of choice for inhalational induction in patients with such a difficult airway was halothane. More recently, there have been numerous case reports suggesting that the newer agent sevoflurane is the more appropriate choice.1–5Although it also causes less hemodynamic depression and airway irritation than halothane,6,7the major advantage of sevoflurane is commonly reported to be its 3- to 4-fold lower blood-gas solubility coefficient (0.65 and 2.4 for sevoflurane and halothane, respectively).8This relatively low solubility of sevoflurane in the blood allows the alveolar and arterial partial pressure to rise more quickly during induction, and it is reported by some authors that a more rapid loss of consciousness can therefore be achieved, compared with the use of a more soluble agent such as halothane.9,10 

In contrast to the substantial literature relating to induction characteristics of sevoflurane and halothane, there are few reports relating to the effect of the blood or tissue solubility on the speed of awakening from anesthesia when the airway obstructs during inhalational induction. In such a circumstance, any decrease in the alveolar and arterial partial pressure of the anesthetic agent will be primarily the result of redistribution from the lung to the body tissues. The greater solubility of halothane in the blood (i.e. , greater blood-gas coefficient) would provide a larger conduit to the tissues, and the greater tissue-gas coefficient of halothane, compared with sevoflurane, would also provide a greater reservoir into which the agent may be distributed. On the basis of these theoretical considerations, it has been argued that halothane would be removed from the lung more rapidly than sevoflurane during airway obstruction, resulting in a more rapid decrease of alveolar and arterial partial pressure.11However, this view is not universal1,4,12; in a recent clinical study in two groups of patients undergoing simulated inhalational anesthesia with halothane or sevoflurane before surgery, the end-tidal concentration of sevoflurane was found to be lower than that of halothane after 3 min of airway obstruction.13 

The purpose of the current study was to test the hypothesis that the end-tidal concentration of halothane would decline 3–4 times more quickly than that of sevoflurane during simulated airway occlusion, reflecting the difference in their respective blood-gas and tissue-gas partition coefficients.8The relationship between blood or tissue content and gas partial pressure is widely recognized to be linear for inhalational anesthetics; therefore, we were able to achieve this aim in a laboratory setting by using anesthetic concentrations of approximately one tenth of the partial pressure commonly encountered clinically. Healthy volunteers were exposed to a stepwise increase in the inspired sevoflurane or halothane concentration, sufficient to reach an end-tidal concentration equivalent to approximately 0.1 minimal alveolar concentration (MAC). Airway obstruction was then simulated by sudden initiation of a 90-s period of closed-circuit rebreathing, during which end-tidal anesthetic concentration was measured breath-by-breath using a mass spectrometer.

General Study Design

The study was approved by the Oxfordshire Clinical Research Ethics Committee (OxREC, Oxford, United Kingdom), and volunteers gave written, informed consent on each day of participation. Eight healthy volunteers (six men, two women) each visited the laboratory twice. The mean (± SD) age was 33 ± 11 yr (median 29 yr). During one visit, a standard simulated induction/rebreathing protocol (see Simulated Inhalational Induction and Simulated Airway Obstruction sections) was performed twice using halothane. During the other visit, the same protocol was performed twice using sevoflurane. Within each day, the two repeats of the protocol were separated by at least 60 min of breathing room air. Four volunteers were exposed to halothane during their first visit and sevoflurane during their second visit, and four volunteers vice versa . In accordance with standard anesthetic practice, volunteers ate no food for 6 h and drank no fluids for 2 h before taking part; as a result of the small risk of hepatotoxicity, they had not been exposed to halothane within 6 months of commencing the study.

Gas Delivery System

During experiments, each volunteer sat in a comfortable chair and breathed through a mouthpiece with his or her nose occluded. A T-piece was positioned close to this mouthpiece, to which 100% oxygen was delivered via  two routes: (1) a constant bias flow of 100% oxygen was delivered to the T-piece via  a mass flow controller (MKS Instruments Ltd, Altrincham, Cheshire, United Kingdom) at a rate of at least 40 l/min and (2) a constant flow of 10 l/min of 100% oxygen was also delivered to the T-piece via  a sevoflurane or halothane vaporizer, according to protocol.

Throughout all phases of the study, inspired and expired gas were sampled continuously (at a rate of 20 ml/min) by using a capillary tube positioned close to the mouth and analyzed by using a mass spectrometer (Airspec QP900; CASE Scientific, Biggin Hill, Kent, United Kingdom). Sevoflurane and halothane were detected at relative molecular masses of 79 and 118, respectively. Before each experiment, the mass spectrometer was calibrated by using a gas cylinder containing a known concentration of halothane (0.208%) or sevoflurane (0.695%). Ventilatory volumes and timings were measured using a combination of a turbine and a pneumotachograph, connected in series. Data were acquired and recorded at 100 Hz using a desktop computer. A three-lead electrocardiogram and arterial oxygen saturation were monitored throughout all experiments.

To simulate airway obstruction, a two-way tap was included in the system, which could be turned to interrupt gas delivery and initiate rebreathing from an initially empty 1-l bag. This arrangement mimics airway occlusion by producing a closed ventilatory system in which any decrease in lung anesthetic concentration must represent redistribution to other body compartments. In addition, it has the advantage of preserving tidal gas flow, and it thereby facilitates breath-by-breath measurement of end-tidal anesthetic concentration.

Simulated Inhalational Induction

Volunteers initially breathed 100% oxygen for several minutes. Induction of anesthesia was then simulated by increasing the inspired anesthetic concentration by 0.05 MAC every third breath. Owing to its higher blood solubility compared with sevoflurane, a higher inspired concentration of halothane is required to produce a given end-tidal concentration. Therefore, induction was continued for 20 breaths in the case of halothane (maximum inspired concentration, approximately 0.35 MAC) and for 15 breaths in the case of sevoflurane (maximum inspired concentration, approximately 0.25 MAC). This difference was predicted to produce an end-tidal concentration of approximately 0.1 MAC for both agents on the basis of our own preliminary observations and on previously published data.13MAC was taken to be 2.6% for sevoflurane14and 0.7% for halothane.15 

Simulated Airway Obstruction

Airway obstruction was simulated in each protocol by repositioning the two-way tap immediately before the final expiration in the induction protocol, i.e. , following the 15th inspiration in the case of sevoflurane and the 20th inspiration in the case of halothane. Volunteers then rebreathed from the 1-l bag for 90 s, throughout which the mass spectrometer trace was carefully inspected for evidence of entrainment of air around the mouthpiece. Such entrainment would be visible as brief changes in gas composition towards that of air, namely very low carbon dioxide and vapor concentration and a reduction in the oxygen concentration, compared with alveolar gas. No such episodes were detected, demonstrating that the protocol was successful in modeling airway obstruction.

Data Analysis and Statistics

Ventilatory timings were used to identify accurate inspired and end-tidal concentrations of halothane and sevoflurane. Values represent the average of 50 data points (i.e. , 500 ms). To estimate the time course of the decline in halothane or sevoflurane levels during rebreathing, end-tidal measurements were modeled using a simple monoexponential function in the form y = ae−t/τ , and a time constant (τ) was calculated for each experiment. Unless otherwise stated, data are presented as mean ± SD and statistical comparisons were performed using a paired Student t  test. P < 0.05 were considered statistically significant.

All volunteers remained awake and alert at all times. Most reported a degree of light-headedness for a few moments after the end of the protocol and a degree of dyspnoea during rebreathing, but these symptoms resolved within minutes. No adverse events occurred during the study, and no volunteer withdrew at any stage.

Simulated Inhalational Induction

Mean inspired and end-tidal anesthetic concentrations during the induction phase of the study are shown in figure 1. Data for individual participants are shown in table 1. At the onset of rebreathing, there was no difference between the end-tidal concentration of halothane (0.10 ± 0.03 MAC) and sevoflurane (0.11 ± 0.03 MAC, P > 0.1).

The duration of induction was significantly longer in the halothane protocol (96.1 ± 25.3 s) compared with the sevoflurane protocol (77.8 ± 22.3 s, P < 0.0001). This difference reflects the greater number of breaths during induction in the halothane protocol.

Simulated Airway Occlusion

The major finding of this study is that the end-tidal sevoflurane concentration during the rebreathing phase of the experiment fell with a significantly slower time course than that of halothane. The monoexponential time constants (τ) for the decline of end-tidal halothane and sevoflurane concentration were 22.1 ± 8.6 and 62.1 ± 15.7 s, respectively (P < 0.0001). Example data and model fits for one participant are shown in figure 2, which demonstrates the very close matching of the monoexponential model to our experimental data. Model fits for all experiments are shown in figure 3, and parameters for individual participants are given in table 1.

The main finding of this study is that the end-tidal concentration of halothane declines around three times more rapidly than the end-tidal concentration of sevoflurane after simulated inhalational induction and acute airway obstruction in healthy volunteers. If the rate at which the partial pressure of anesthetic decreases in the lung is assumed to reflect the rate at which it declines at the site of anesthetic action, this result suggests that a patient with an obstructed airway may wake more rapidly after induction with a more soluble agent such as halothane, compared with a less soluble agent such as sevoflurane.

Comparison with Existing Literature

The findings of this study are in the line with a number of reports in the literature. Fenlon and Pearce, for example, predicted on the basis of blood and tissue solubility characteristics that the partial pressure of halothane in the blood would decrease more rapidly than that of sevoflurane during airway obstruction, when the decrease would primarily result from redistribution of the agent from blood to tissues.11By using standard blood and tissue solubility data from the literature and a four-compartment mathematical model of the adult human, it can be calculated for a range of anesthetic depths that the partial pressure of halothane would not only decline faster than that of sevoflurane, but that it would also remain lower for at least 5 min after inhalational induction and airway obstruction (data from mathematical modeling provided by Andrew D. Farmery, D.M., F.R.C.A., Oxford, United Kingdom, 2009).

In contrast to these reports, some authors appear to suggest that the lower solubility of sevoflurane may in fact lead to a more rapid decline in arterial partial pressures during airway occlusion, compared with halothane.1,4However, although there is considerable experimental and clinical evidence for faster elimination of sevoflurane in the context of a patent airway, when low solubility may enhance elimination via  the lung,16,17the rationale for suggesting a more rapid decline of alveolar sevoflurane concentration during airway obstruction is unclear.

One possibility supported by Girgis et al.  is that redistribution of anesthetic during airway obstruction is related to the duration of induction.13In an experimental study on 40 patients before surgery, in whom anesthesia was initially induced using an intravenous agent, these authors mimicked inhalational induction using either halothane or sevoflurane. The inspired anesthetic concentration was increased by 0.5 MAC every three breaths until an end-tidal concentration of 2 MAC was achieved. This required a mean of 36 breaths for halothane, compared with 27 breaths for sevoflurane. When the airway was subsequently occluded for 3 min and the end-tidal gas sampled immediately thereafter, it was found that the end-tidal sevoflurane concentration was significantly lower than the end-tidal halothane concentration. This result was interpreted as evidence of a more rapid redistribution of sevoflurane than halothane, and the authors suggested that the reason for this difference was the longer duration of induction when using halothane. Given the similar blood-tissue solubility coefficients for halothane and sevoflurane,8the longer induction may have allowed the concentration of halothane in body compartments other than brain and blood to rise higher than the corresponding concentrations of sevoflurane. When the airway was then acutely obstructed and the anesthetic redistributed away from the blood to other compartments, the concentration gradient for sevoflurane was suggested to be greater and redistribution more rapid.

Our results appear to contradict this hypothesis directly. In the current study, the time course of the decline in end-tidal halothane concentration was significantly more rapid than that of sevoflurane, despite a substantially longer period of induction. The reasons for this apparent discrepancy are not clear. One possibility relates to the lack of sampling of end-tidal gas by Girgis et al.  during airway occlusion. It is possible, for example, that a relatively higher washout rate of sevoflurane in the first few breaths after airway restoration could have favored lower end-tidal values for sevoflurane, compared with halothane, in their study. Alternatively, an early, rapid decline in halothane concentration could have occurred undetected. A second theoretical possibility relates to the hepatic metabolism of halothane. The longer duration of airway occlusion in the study of Girgis et al.  (180 s), compared with the current study (90 s), would favor greater metabolism of halothane in the former study. In contrast, animal studies suggest that halothane metabolism is concentration-dependent, such that fractional removal of halothane is much greater at subanesthetic concentrations, such as those used in the current study.18In fact, we believe that metabolism of halothane is unlikely to contribute significantly to the findings in either study. Using published kinetics for halothane metabolism by human liver microsomes in vitro  19and data relating total microsomal protein content to liver and body size in humans,20,21the rate of halothane metabolism in vivo  can be estimated to be in the range 6–12 μmol/min. At an end-tidal concentration of 0.1 MAC (the value at the onset of rebreathing in the halothane protocol), the blood halothane concentration would be expected to be approximately 70 μm and total blood halothane content 350–420 μmol. Hepatic metabolism would therefore have reduced blood halothane content by significantly less than 5% over the 90 or 180 s of airway occlusion. Sevoflurane is metabolized to a lesser extent than halothane, so metabolism is unlikely to be significant for either agent.22,23 

Limitations of the Experimental Approach

As discussed above, one important interpretation of the findings of the current study is that the higher solubility of halothane, compared with sevoflurane, favors redistribution from the lung to other body compartments during simulated airway occlusion. In turn, this result may be interpreted as evidence in favor of more rapid awakening from a halothane anesthetic, should the airway obstruct during induction. There are, however, a number of assumptions inherent in these interpretations.

First, as discussed above, it is assumed that differences in the metabolic elimination of anesthetic do not contribute to our findings. Second, we assume that the levels of halothane and sevoflurane in the brain closely follow those in the arterial blood and lung. This assumption is commonly made for the steady-state, but it may not be reasonable when anesthetic concentrations are changing. Furthermore, we assume that the rate of equilibration between the blood and brain is similar for the two agents. In support of this assumption, the reported brain-blood partition coefficients for halothane and sevoflurane are very similar8,24; in both cases, equilibration would be expected to proceed very rapidly, given the high blood flow to tissue volume ratio in the brain.8Finally, our approach was to model the measured decline in end-tidal anesthetic concentration as a monoexponential function of time. Clearly, this is a simplification of the known multicompartment nature of anesthetic distribution. However, although a multiexponential model would be required over a longer time period, we feel that our monoexponential approach provides a valid approximation during the brief 90-s period of rebreathing in this study.

In conclusion, although sevoflurane is reported to have a number of distinct advantages over halothane for inhalational induction of anesthesia, our results suggest that the alveolar partial pressure of halothane may fall more rapidly than that of sevoflurane after acute airway obstruction. Assuming that speed of awakening is proportional to alveolar partial pressure in this setting, the greater blood and tissue solubility of halothane, compared with sevoflurane, may lead to more rapid awakening in the event of airway obstruction during inhalational induction of anesthesia.

The authors thank Simon Goddard, B.Sc., B.M., B.Ch., Frenchay Hospital, Bristol, United Kingdom, for assistance with data analysis; Mr. David O’Connor, Chief Technician, Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, United Kingdom, for technical assistance; and all volunteers for their participation.

Wang CY, Chiu CL, Delilkan AE: Sevoflurane for difficult intubation in children (Letter). Br J Anaesth 1998; 80:408
Barker I: Is sevoflurane replacing halothane? (letter). Br J Anaesth 1998; 80:123
Kandasamy R, Sivalingham P: Use of sevoflurane in difficult airways. Acta Anaesthesiol Scand 2000; 44:627–9
Mostafa SM, Atherton AM: Sevoflurane for difficult tracheal intubation. Br J Anaesth 1997; 79:392–3
Ip-Yam PC: Sevoflurane in acute airway obstruction (Letter). Br J Anaesth 1998; 80:273
Wodey E, Pladys P, Copin C, Lucas MM, Chaumont A, Carre P, Lelong B, Azzis O, Ecoffey C: Comparative hemodynamic depression of sevoflurane versus  halothane in infants: An echocardiographic study. Anesthesiology 1997; 87:795–800
Doi M, Ikeda K: Airway irritation produced by volatile anaesthetics during brief inhalation: Comparison of halothane, enflurane, isoflurane and sevoflurane. Can J Anaesth 1993; 40:122–6
Eger EI 2nd, Saidman LJ: Illustrations of inhaled anesthetic uptake, including intertissue diffusion to and from fat. Anesth Analg 2005; 100:1020–33
Black A, Sury MR, Hemington L, Howard R, Mackersie A, Hatch DJ: A comparison of the induction characteristics of sevoflurane and halothane in children. Anaesthesia 1996; 51:539–42
Lerman J, Davis PJ, Welborn LG, Orr RJ, Rabb M, Carpenter R, Motoyama E, Hannallah R, Haberkern CM: Induction, recovery, and safety characteristics of sevoflurane in children undergoing ambulatory surgery. A comparison with halothane. Anesthesiology 1996; 84:1332–40
Fenlon S, Pearce A: Sevoflurane induction and difficult airway management (Letter). Anaesthesia 1997; 52:285–6
Ip-Yam PC: Sevoflurane for difficult tracheal intubation (Letter). Br J Anaesth 1998; 81:104
Girgis Y, Frerk CM, Pigott D: Redistribution of halothane and sevoflurane under simulated conditions of acute airway obstruction. Anaesthesia 2001; 56:613–5
Fragen RJ, Dunn KL: The minimum alveolar concentration (MAC) of sevoflurane with and without nitrous oxide in elderly versus  young adults. J Clin Anesth 1996; 8:352–6
Miller RD, Wahrenbrock EA, Schroeder CF, Knipstein TW, Eger EI 2nd, Buechel DR: Ethylene–halothane anesthesia: Addition or synergism? Anesthesiology 1969; 31:301–4
Sury MR, Black A, Hemington L, Howard R, Hatch DJ, Mackersie A: A comparison of the recovery characteristics of sevoflurane and halothane in children. Anaesthesia 1996; 51:543–6
Stern RC, Towler SC, White PF, Evers AS: Elimination kinetics of sevoflurane and halothane from blood, brain, and adipose tissue in the rat. Anesth Analg 1990; 71:658–64
Sawyer DC, Eger EI 2nd, Bahlman SH, Cullen BF, Impelman D: Concentration dependence of hepatic halothane metabolism. Anesthesiology 1971; 34:230–5
Spracklin DK, Hankins DC, Fisher JM, Thummel KE, Kharasch ED: Cytochrome P450 2E1 is the principal catalyst of human oxidative halothane metabolism in vitro . J Pharmacol Exp Ther 1997; 281:400–11
Barter ZE, Bayliss MK, Beaune PH, Boobis AR, Carlile DJ, Edwards RJ, Houston JB, Lake BG, Lipscomb JC, Pelkonen OR, Tucker GT, Rostami-Hodjegan A: Scaling factors for the extrapolation of in vivo  metabolic drug clearance from in vitro  data: Reaching a consensus on values of human microsomal protein and hepatocellularity per gram of liver. Curr Drug Metab 2007; 8:33–45
Chouker A, Martignoni A, Dugas M, Eisenmenger W, Schauer R, Kaufmann I, Schelling G, Lohe F, Jauch KW, Peter K, Thiel M: Estimation of liver size for liver transplantation: The impact of age and gender. Liver Transpl 2004; 10:678–85
Kharasch ED, Karol MD, Lanni C, Sawchuk R: Clinical sevoflurane metabolism and disposition. I. Sevoflurane and metabolite pharmacokinetics. Anesthesiology 1995; 82:1369–78
Carpenter RL, Eger 2nd EI Johnson BH, Unadkat JD, Sheiner LB: The extent of metabolism of inhaled anesthetics in humans. Anesthesiology 1986; 65:201–5
Yasuda N, Targ AG, Eger EI 2nd: Solubility of I-653, sevoflurane, isoflurane, and halothane in human tissues. Anesth Analg 1989; 69:370–3