Secondary Logo

Journal Logo

Original Article

Conventional stepwise vs. vital capacity rapid inhalation induction at two concentrations of sevoflurane

Martín-Larrauri, R.*; Gilsanz, F.; Rodrigo, J.; Vila, P.; Ledesma, M.§; Casimiro, C.§

Author Information
European Journal of Anaesthesiology: April 2004 - Volume 21 - Issue 4 - p 265-271


Inhalation induction is a commonly used technique for children, but in recent years has not been widely used in adults until sevoflurane became available. Sevoflurane, because of its lack of pungency and lack of airway irritation, compared to other vaporized agents has made inhalation induction a valid alternative to intravenous (i.v.) induction in many clinical settings. Moreover, it enables maintenance or rapid recovery of spontaneous ventilation of the lungs, an issue that can be important in some patients, especially those with a difficult airway. Sevoflurane is the anaesthetic of choice for inhalational induction of anaesthesia [1-3]. Several studies have compared i.v. induction with inhalational induction of anaesthesia using sevoflurane and found advantages in some settings [4,5]. In fact, a previous survey has shown widespread use of this technique in some countries, such as the UK [6]. Different inhalation induction techniques have been designed using several anaesthetics, halothane [7,8], isoflurane [9] and sevoflurane [10].

The conventional stepwise incremental induction technique, multiple vital capacity technique or vital capacity rapid inhalation induction have been described previously [11-13], but a consensus on which of these techniques is the most effective, the safest and best accepted by patients has not been reached. This multicentre study has been designed to test two non-standard methods of inhalation induction in our common clinical practice, comparing the conventional stepwise spontaneous induction vs. the vital capacity rapid inhalation induction method at two different sevoflurane inspiratory concentrations to evaluate the suitability of these methods in clinical practice.



Institutional Ethics Committee approval from each participating hospital and written informed consent from each patient were obtained. One-hundred-and-twenty-five adult patients of ASA Grades I-II aged 19-65 yr undergoing short surgical procedures (<90 min) under general anaesthesia with spontaneous ventilation via laryngeal mask airway were recruited. Patients with an allergy or sensitivity to volatile anaesthetics, known or suspected genetic susceptibility to malignant hyperthermia, abnormal hepatic or renal function tests, pregnant females, previous participation in a similar study or any other condition which may have increased the risk to the patient were excluded from the study. Patients were randomly assigned to one of three anaesthetic groups: conventional stepwise induction technique at normal tidal volumes (Group C) or vital capacity inhalation rapid inhalation induction technique at two different inspiratory concentrations, 4.5% (Group VC4.5) and 8% (Group VC8). Forty-two patients were allocated to Group C, 42 patients to Group VC4.5 and 42 patients to Group VC8.

Randomization and patient alignment

This was a randomized, open label, multicentre study with three groups in parallel. The sample size was calculated and it was comparable with similar published studies. Patients enrolled in the study were randomized in equal numbers to one of the three treatment groups: C, VC4.5 or VC8. Patients qualifying for entry into the study were sequentially assigned the lowest available patient number once informed consent was provided. The treatment assignment for each patient number was contained in sequentially ordered individually sealed envelopes that were opened on the day of the scheduled surgery - after-wards the study design was open. All data were collected in case report forms.

Anaesthetic procedures

When the patients arrived in the anaesthetic room, they were informed to which group they were assigned. Patients in the VC groups were given instructions what to do. Blood pressure, heart rate and oxygen saturation via pulse oximetry (SPO2) monitoring was initiated. All patients were hydrated with Ringer's lactate 1.5 mL kg−1 for every fasting hour and given fentanyl at a dose of 1 μg kg−1 through an i.v. catheter immediately before receiving the study anaesthetics.

The breathing system was a Bain circuit modified by the use of a 4 L rebreathing bag. The anaesthetic machine was set to deliver fresh gas flow at 8 L min−1 with a mixture of 2:1 nitrous oxide-oxygen (N2O:O2) and the vaporizer was adjusted to deliver an inspired fraction of sevoflurane corresponding to each induction technique.

Vital capacity rapid inhalation induction groups

In Groups VC4.5 and VC8, the breathing system was primed with sevoflurane 4.5% or 8% in 2:1 N2O:O2 for 30-45 s. Patients were told that they would have to breathe in as deeply as they could. Following a maximum expiration, we instructed the patient to take a vital capacity breath from the face mask connected to the anaesthetic system and hold their breath as long as possible. Extra vital capacity breaths were taken if necessary. The sevoflurane vaporizer was maintained at 4.5% or 8% for each corresponding group until loss of eyelash reflex.

Conventional stepwise inhalation induction group

Patients in the tidal breathing group were told that for induction of anaesthesia they would just have to continue breathing normally from the face mask. The vaporizer was set at sevoflurane 1% in 2:1 N2O:O2 and then increased by 1% every three tidal breaths until loss of the eyelash reflex occurred.

Laryngeal mask airway placement and assessment variables

After the eyelash reflex had been lost, the inspired vaporizer concentration was decreased to 4.5% in all groups and respiration was manually assisted until sufficient jaw relaxation (defined as spontaneous drop of the jaw and maximal mouth opening without resistance) was achieved to tolerate insertion of a laryngeal mask. Inspired oxygen concentration and inspiratory, end-tidal concentrations of CO2 and anaesthetic gases were monitored. The haemodynamic variables were again registered when the laryngeal mask had been inserted; we registered any changes in haemodynamic variables greater than 25% of baseline values. After the insertion of the laryngeal mask, the sevoflurane concentration was reduced to a 1.3-1.8% depending on clinical characteristics of the patients.

The primary end-point was loss of eyelash reflex time, defined as the interval from application of the face mask (Group C) or the end of the first inspiration (Groups VC4.5 and VC8) until loss of eyelash reflex. Eyelash reflex was ascertained every 3 s in each patient until it was lost. Secondary time end-points were: time to cessation of finger tapping on the trolley, time to the end of laryngeal mask insertion (i.e. the time until the mask was completely placed in situ) and time to spontaneous breathing after mask insertion. Furthermore, time from mask insertion to spontaneous breathing and time from loss of eyelash reflex to mask insertion were also determined.

Ease of laryngeal mask placement was assessed and any side-effects occurring during the induction process were recorded. In the recovery room, the patient's experience during the anaesthetic induction (scored as pleasant, indifferent, unpleasant or no memory) and the patient's opinion about receiving the same type of anaesthesia in the future, should it be necessary, were recorded. Postoperative nausea and vomiting (PONV) were recorded only if severe.

Statistical analysis

Quantitative data followed a normal distribution and complied with homoscedasticity (i.e. the assumption that the variance around the regression line is the same for all values of the predictor), allowing for the use of analysis of variance (ANOVA) tests to compare the study groups. The Bonferroni-Dunn correction for multiple comparisons was applied when the general ANOVA test yielded a statistically significant result. Quantitative data were analysed using Pearson's χ2-test. Kaplan-Meier survival curves were used to plot time variables. A value <0.05 was required for statistical significance in the general tests and <0.017 in the multiple comparisons tests. Values are reported as mean ± SD, unless otherwise stated.


The three groups were comparable in terms of gender, age, weight, height and ASA Grade (Table 1). The distribution of surgical operations was similar in the three groups and consisted primarily of gynaecological and general surgery procedures (uterine fibroids and inguinal hernia). Length of surgical procedures was also similar in the Groups VC4.5, VC8 and C (50 ± 32 min, 50 ± 26 min and 49 ± 32 min, respectively).

Table 1
Table 1:
Patient characteristics in the three groups.

The time to loss of eyelash reflex was significantly shorter in the Groups VC8 and VC4.5 than in the Group C (Table 2). Pairwise comparisons were also significant (Groups VC8 vs. VC4.5, P = 0.006; Groups VC8 vs. C, P < 0.0001; Groups VC4.5 vs. C, P = 0.011) (Fig. 1). The Kaplan-Meier plot shows that the three groups were clearly differentiated in the times to loss of eyelash reflex (Fig. 2). Group VC4.5 required more additional vital capacity breaths to reach unconsciousness than Group VC8 (4.3 ± 0.5 vs. 3.3 ± 0.5 breaths, P < 0.001). Other timing data are shown in Table 2. Significant differences were observed in all three pairwise comparisons for time to cessation of tapping (Groups VC8 vs. VC4.5, P = 0.011; Groups VC8 vs. C, P < 0.0001; Groups VC4.5 vs. C, P = 0.0005); in Groups VC8 vs. VC4.5 comparison for time to laryngeal mask insertion (Groups VC8 vs. VC4.5, P = 0.015; Groups VC8 vs. C, P > 0.05; Groups VC4.5 vs. C, P > 0.05). The other end-point times assessed did not show significant differences in the general test.

Table 2
Table 2:
Induction times in the three groups.
Figure 1
Figure 1:
Box and whiskers plot of times to loss of eyelash reflex in the three groups Groups C, VC4.5 and VC8. The top and bottom edges of the box show the 25th and 75th‰ of data, while the horizontal line in the middle of the box corresponds to the median value. The ends of the whiskers correspond to the range of values that fall between the 75th and 90th‰ values, and the 25th and 10th‰ values.
Figure 2
Figure 2:
Kaplan-Meier plot of study groups, Groups C, VC4.5 and VC8 showing that times to loss of eyelash reflex were clearly different in the three groups. ▿: Group C; ○: Group VC4.5; ▴: Group VC8.

Insertion of the laryngeal masks was performed with no difficulty in 78% of cases and, without differences between groups studied (4% of patients had difficulties during insertion). There were no significant differences between groups in the incidence of side-effects (Table 3). Induction of anaesthesia was uneventful in 32 of 41 patients in Group VC4.5, in 34 of 42 patients in Group VC8 and in 35 of 42 patients in Group C. Sixty-one percent of all patients recalled the beginning of induction and there were no significant differences between the groups. This recollection was unpleasant in 7% of patients (Table 4), but 94% of patients would undergo the same anaesthetic technique in future (no differences between groups). No severe PONV was recorded in any patient.

Table 3
Table 3:
Side-effects during sevoflurane induction in the three groups.
Table 4
Table 4:
Patient acceptance of inhalation induction with sevoflurane. Number of patients in the four categories of a semiquantitative satisfaction scale in the three study groups.


Study rationale

Inhalation induction with sevoflurane does not cause anaphylaxis, is associated with a lower incidence of apnoea, hypotension and hangover than i.v. induction. Moreover, inhalation induction, is less expensive, has a similar time to loss of consciousness compared with i.v. induction with propofol and can safely be used in smokers [4,5,14]. Although some concern about contamination of the operation theatre has been associated with inhalation induction, sevoflurane has only shown dangerous exposure levels (>20 ppm) under very extreme conditions. Mean exposure levels measured during inhalation induction in the operation theatre and the anaesthesiologist breathing area have been found low [15].


The type of breathing and premedication modifies inhalation induction with sevoflurane. Nasal breathing slows induction, while oral breathing accelerates it by up to 25% and is associated with higher end-expiratory values of sevoflurane [16,17]. Premedication with fentanyl (2.4 μg kg−1), midazolam or a combination of both shortens induction time [18]. The lower doses of fentanyl, 1 μg kg−1 used in this study, have no effect on induction times or laryngeal mask insertion, but are associated with a reduction of adverse airway events [19]. The addition of i.v. midazolam 0.1 mg kg−1 provides more stable haemodynamics and increases patient satisfaction [20] - nevertheless we decided not to use it in this study because it might interfere with the induction time.

Induction times

The conventional stepwise technique (used in Group C patients) is commonly used for inhalation induction with halothane. However, due to low degree of pungency of sevoflurane, patients can quickly tolerate high inspiratory concentrations (8%). In our study, the mean time to loss of eyelash reflex with the conventional stepwise technique was longer than the mean time reported by other authors in non-premedicated volunteers [12], and shorter than the mean time in another study in patients premedicated with midazolam and fentanyl [21]; however, it was similar to other series of patients premedicated with higher doses of fentanyl [22]. However, mean times to loss of eyelash reflex in this study were longer compared to other techniques using constant high inspiratory concentrations (7-8%) of sevoflurane in both healthy volunteers [11] and patients [4]. Our results are in agreement with the population and procedures studied.

Ruffle and colleagues [7] and Wilton and colleagues [8] first described the single breath forced vital capacity technique in the mid-1980s as a technique for accelerating induction with halothane in adults. This technique resulted in fewer respiratory side-effects and excitatory movements and provides more rapid loss of consciousness but requires some patient training, priming of the anaesthesia breathing system and a need for a different oxygen source for patient preoxygenation [23]. Also, breath holding at vital capacity volume produces a Valsalva manoeuvre that can lead to untoward cardiovascular events [10].

In our study, mean time to loss of eyelash reflex in the Group VC4.5 (94 ± 6.5 s) was longer than published in non-premedicated volunteers, ranging from 54 [12] to 81 s [24], and longer than the mean time reported by Sloan and colleagues [9] in patients premedicated with midazolam (75 ± 3 s). Mean time to loss of eyelash reflex in Group VC8 (68 ± 7 s) was longer than the mean induction times reported in non-premedicated patients, ranging from 41 to 47 s [11,25,26]. On the other hand, mean times reported by other authors [5,27-29] are similar to the mean time to loss of eyelash reflex observed in our study.

Results of both VC groups in our study are consistent with those found by Yurino and colleagues [26] in healthy volunteers, where they identified an inverse relation between the mean time of induction and the logarithm of the inspiratory concentration of sevoflurane - up to a maximum of 6%. Time to cessation of finger tapping on the trolley is another indicator routinely used to evaluate inhalation induction. In our study, cessation of finger tapping occurred before loss of the eyelash reflex. This result is similar to that found in other studies [5,27] using the VC8 technique.

Insertion of laryngeal mask

Insertion of the laryngeal mask is less invasive and requires a lower depth of anaesthesia [10] than endotracheal intubation and permits a quicker return to spontaneous ventilation [5,27,30,31]. Laryngeal mask insertion times using sevoflurane depend on the inhalation induction technique used and the population under study. In the present study, the mask insertion time was longer than in previous studies since, after the loss of eyelash reflex, sevoflurane concentration was fixed at 4.5% in all patients. The insertion time was significantly shorter in the Group VC8 compared with the other two groups. This may reflect a higher sevoflurane brain concentration at the end of induction after loss of the eyelash reflex.


The incidence of side-effects was similar among the three groups in our study. The total number of side-effects for the Group C was different to those previously published. Yurino and colleagues [11] found a lower rate of side-effects in non-premedicated healthy volunteers anaesthetized with the VC8 technique compared to those exposed to stepwise inhalation induction similar to Group C in the present study. In our study the incidence of side-effects was similar for the two groups and matching those obtained by Yurino in Group VC8. The highest incidence of side-effects was observed in Group VC4.5 (22%). Excessive salivation was the most frequent side-effect recorded with no difference among groups. Excitement is generally associated with a longer time of induction [12]; however, in our study, although Group C had the longest time to loss of eyelash reflex the incidence of excitement was only 5%, while Group VC4.5 had the highest incidence of excitement (17%) and Group VC8 the lowest (2%).

Premedication with fentanyl should be associated with a higher rate of respiratory side-effects according to Muzi and colleagues [18]. The low incidence of side-effects in our study could partially be explained by the lower dose of fentanyl used in this study (1 μg kg−1 vs. 2.4 μg kg−1). Another possible explanation for the lack of significant differences between groups in the rate of side-effects is that inspiratory concentration was fixed at 4.5% in all groups after the eyelash reflex had been lost.

Patient acceptance

The acceptance rate was excellent in all three groups. Ninety-four percent of the patients would accept to repeat the same induction technique. These results are similar to those published by Hall and colleagues [5] with the VC8 technique (94%) but better than those obtained by Thwaites and colleagues [4] for normal tidal volume induction with sevoflurane 8% (43%). The rate of induction recall was 61%. Odour was unpleasant in 10% of those patients recalling induction in Group VC8 and in 8% of patients from Group C. These results are better than those published by Thwaites and colleagues [4] using sevoflurane 8% and those published by Yurino and colleagues [12] with the stepwise and VC4.5 techniques.

General remarks

Conventional stepwise induction with sevoflurane was the slowest technique for those patients evaluated. However, the number of side-effects and acceptance rate was not significantly different from either of the VC techniques. Although the VC techniques have shorter times to loss of the eyelash reflex, additional time is needed for priming the breathing system and patient training and they require an auxiliary source of oxygen, all of which may increase the total time of the anaesthetic procedure [32]. The conventional stepwise induction technique with sevoflurane starting from a 1% concentration that we used in our study may not be the best option in view of the results we have obtained. An alternative might be a tidal volume technique using a constant high inspiratory concentration (7-8%) from the start - that also avoids the drawbacks associated with VC techniques (need for patient collaboration, and requirement for an additional oxygen source), causes fewer haemodynamic changes than the stepwise technique and lacks the inhibitory effects on sympathetic activity seen with both [33,34] VC techniques.

In our study, the VC4.5 method does not appear to have any advantage over VC8 procedure due to the longer time for induction and the higher rate of excitatory events. Even though faster times to induction cannot be attained with concentrations higher than 6%, as noted by Yurino and colleagues [26], the highest possible concentration should be recommended for use with the VC technique (8% with current vaporizers), thus allowing a high inspiratory concentration in spite of not achieving complete priming of the anaesthetic circuit.

In conclusion, the three methods tested were fast, reliable and safe. No relevant side-effects were observed with any of the methods employed. Sevoflurane induction enables easy placement of the laryngeal mask. The VC8 technique is the fastest of the three methods tested in this study. Sevoflurane induction provides high degree of patient satisfaction.


The authors thank Dr. A. Biarnés (Hospital Vall D' Hebrón, Spain), Dr. JI Gómez-Herreras (Hospital Clínico Universitario de Valladolid, Spain) and Dr. G. Blanc (Hospital de la Princesa de Madrid, Spain) for their contributions to the study. This study was supported by a grant from Abbott Laboratories S.A., Madrid, Spain.


1. Doi M, Ikeda K. Airway irritation produced by volatile anaesthetics during brief inhalation: comparison of halothane, enflurane, isoflurane and sevoflurane. Can J Anesth 1993; 40: 122-126.
2. Goresky GV, Muir J. Inhalation induction of anaesthesia. Can J Anesth 1996; 43: 1085-1086.
3. Wong AY, Lawmin JC, Irwin MG. Anaesthetic management of a patient with low tracheal obstruction requiring placement of a T-Y stent. Anaesth Intens Care 2000; 28: 196-198.
4. Thwaites A, Edmends S, Smith I. Inhalation induction with sevoflurane: a double-bind comparison with propofol. Br J Anaesth 1997; 78: 356-361.
5. Hall JE, Stewart JIM, Harmer M. Single-breath inhalation induction of sevoflurane anaesthesia with and without nitrous oxide: a feasibility study in adults and comparison with an intravenous bolus of propofol. Anaesthesia 1997; 52: 410-415.
6. Moore EW, Davies MW. Inhalational versus intravenous induction. A survey of emergency anaesthetic practice in the United Kingdom. Eur J Anaesthesiol 2000; 17: 33-37.
7. Ruffle JM, Snider MT, Rosenberg JL, Latta WB. Rapid induction of halothane anaesthesia in man. Br J Anaesth 1985; 57: 607-611.
8. Wilton NCT, Thomas VL. Single breath induction of anaesthesia, using a vital capacity breath of halothane, nitrous oxide and oxygen. Anaesthesia 1986; 41: 472-476.
9. Sloan MH, Conard PF, Karsunky PK, Gross JB. Sevoflurane versus isoflurane: induction and recovery characteristics with single-breath inhaled inductions of anesthesia. Anesth Analg 1996; 82: 528-532.
10. Muzi M, Robinson BJ, Ebert TJ, O'Brien. Induction of anesthesia and tracheal intubation with sevoflurane in adults. Anesthesiology 1996; 85: 536-543.
11. Yurino M, Kimura H. A comparison of vital capacity breath and tidal breathing techniques for induction of anaesthesia with high sevoflurane concentrations in nitrous oxide and oxygen. Anaesthesia 1995; 50: 308-311.
12. Yurino M, Kimura H. Induction of anesthesia with sevoflurane, nitrous oxide, and oxygen: a comparison of spontaneous ventilation and vital capacity rapid inhalation induction (VCRII) techniques. Anesth Analg 1993; 76: 598-601.
13. Djaiani GN, Hall J, Pugh S, Peaston RT. Vital capacity inhalation induction with sevoflurane: an alternative to standard intravenous induction for patients undergoing cardiac surgery. J Cardiothorac Vasc Anesth 2001; 15: 169-174.
14. Mendonca C, Thorpe C. Effect of smoking on induction of anaesthesia with sevoflurane. Anaesthesia 2001; 56: 19-23.
15. Hall JE, Henderson KA, Oldham TA, Pugh S, Harmer M. Environmental monitoring during gaseous induction with sevoflurane. Br J Anaesth 1997; 79: 342-345.
16. Ishii M, Nishino T, Kochi T. Changes in respiratory pattern during induction of anaesthesia with sevoflurane: comparison of nasal and oral breathing. Eur J Anaesthesiol 1966: 13: 21-26.
17. Campbell IT. Breathing routes and inhalation induction of anaesthesia. Eur J Anaesthesiol 1996; 13: 1-2.
18. Muzi M, Colinco MD, Robinson BJ, Ebert TJ. The effects of premedication on inhaled induction of anesthesia with sevoflurane. Anesth Analg 1997; 85: 1143-1148.
19. Plastow SE, Hall JE, Pugh SC. Fentanyl supplementation of sevoflurane induction of anaesthesia. Anaesthesia 2000; 55: 475-478.
20. Nishiyama T, Matsukawa T, Yokoyama T, Hanaoka K. Rapid inhalation induction with 7% sevoflurane combined with intravenous midazolam. J Clin Anesth 2002; 14: 290-295.
21. Jellish WS, Lien CA, Fontenot HJ, Hall R. The comparative effects of sevoflurane versus propofol in the induction and maintenance of anaesthesia in adult patients. Anesth Analg 1996; 82: 479-485.
22. Lien CA, Hemmings HC, Belmont MR, Abalos A, Holmann Ch, Kelly RE. A comparison: the efficacy of sevoflurane-nitrous oxide or propofol-nitrous oxide for the induction and maintenance of general anesthesia. J Clin Anesth 1966; 8: 639-643.
23. Smith I, Thwaites AJ. Inhalation versus TIVA in short duration anaesthesia. Acta Anaesth Belg 1997; 48: 161-166.
24. Yurino M, Kimura H. Vital capacity rapid inhalation induction technique: comparison of sevoflurane and halothane. Can J Anesth 1993; 40: 440-443.
25. Yurino M, Kimura H. Comparison of induction time and characteristics between sevoflurane and sevoflurane/nitrous oxide. Acta Anaesthesiol Scand 1995; 39: 356-358.
26. Yurino M, Kimura H. Efficient inspired concentration of sevoflurane for vital capacity rapid inhalation induction (VCRII) technique. J Clin Anesth 1995; 7: 228-231.
27. Hall JE, Oldham TA, Stewart, Harmer M. Comparison between halothane and sevoflurane for adult vital capacity induction. Br J Anesth 1977; 79: 285-288.
28. Vanacker BF. Sevoflurane mask induction in adults: comparison of two inhalation techniques. Acta Anaesth Belg 1997; 48: 147-153.
29. Philip BK, Lombard LL, Edward RR, Drager LR, Calalang I, Philip JH. Comparison of vital capacity induction with sevoflurane to intravenous induction with propofol for adult ambulatory anesthesia. Anesth Analg 1999; 89: 623-627.
30. Smith CE, Lever JS, Sawkar S, Pinchak AC, Hagen JF. Sevoflurane-N2O versus propofol/isoflurane-N2O during elective surgery using the laryngeal mask airway in adults. J Clin Anesth 2000; 12: 392-396.
31. Hendrickx JF, Vandeput DM, De Geyndt AM, De Loof T, De Wolf AM. Coasting after overpressure induction with sevoflurane. J Clin Anesth 2000; 12: 100-103.
32. Baker CE, Smith I. Sevoflurane: a comparison between vital capacity and tidal breathing techniques for the induction of anaesthesia and laryngeal mask airway placement. Anaesthesia 1999; 54: 841-844.
33. Nishiyama T, Aibiki M, Hanaoka K. Haemodynamic and catecholamine changes during rapid sevoflurane induction with tidal volume breathing. Can J Anesth 1977; 44: 1066-1070.
34. Hall JE, Ebert TJ, Harmer M. Induction characteristics with 3% and 8% sevoflurane in adults: an evaluation of the second stage of anaesthesia and its haemodynamic consequences. Anaesthesia 2000; 55: 545-550.

ANAESTHESIA, inhalation; ANAESTHETICS, INHALATION, nitrous oxide; ANAESTHETICS, VOLATILE, sevoflurane; INTUBATION, intratracheal, laryngeal masks

© 2004 European Academy of Anaesthesiology