Secondary Logo

Journal Logo

Original Article

Pressure support ventilation during inhalational induction with sevoflurane and remifentanil in adults

Banchereau, F.1; Herve, Y.2; Quinart, A.3; Cros, A.-M.1

Author Information
European Journal of Anaesthesiology: November 2005 - Volume 22 - Issue 11 - p 826-830
doi: 10.1017/S0265021505001389



The inhalational induction technique with sevoflurane is routinely used in operating rooms. It allows the use of a single hypnotic agent during both induction and maintenance of anaesthesia, thus preventing the occurrence of an anaesthetic gap between them [1]. It also allows the use of a target-controlled administration with the end-tidal sevoflurane concentration. However, during inhalational induction, depression of respiratory centres leads to a decrease in the tidal volume and a compensatory increase in respiratory rate [2]. This respiratory depression occurs early and is dose-dependent [2]. It is worsened when narcotics are used to improve intubating conditions and to limit the haemodynamic response to intubation [3]. Moreover, the amount of anaesthetic gases actually delivered to the patient is linked to alveolar ventilation [4]. Therefore, the deepening of anaesthesia might be compromised by respiratory depression, necessitating manual or mechanical support to provide good conditions for airway management. Unfortunately, manual ventilation is often difficult owing to irregular respiration. It also prevents the use of both hands to hold the mask and may result in high airway pressures [5]. Mechanical support is therefore an appealing alternative. Pressure support ventilation (PSV) is now widely used in intensive care units (ICUs) because it allows spontaneous ventilation to be maintained and provides an efficient support against increased work of breathing and respiratory depression [6]. The new anaesthesia ventilators equipped with PSV may be used during inhalational induction in order to compensate for respiratory depression. The purpose of this prospective randomized study was therefore to compare PSV and spontaneous ventilation during induction of anaesthesia with sevoflurane and remifentanil in adult patients. Our primary endpoint was the expected increase of the tidal volume. Secondary endpoints were bispectral index and ease of intubation.


After obtaining Ethics Committee approval (Comité Consultatif pour la Protection des Personnes dans la Recherche Biomédicale of Angers, France) and written informed consent from the patients, we studied 35 adult patients scheduled for elective ear nose throat (ENT) surgery. Exclusion criteria were contraindications for inhalational induction, patients with a full stomach, personal or familial history of malignant hyperthermia, oesophageal reflux, intracranial hypertension, those with predicted difficult mask ventilation [7] and patients taking respiratory depressive drugs (narcotics or benzodiazepines). Patients with predictive difficult intubation criteria were not excluded.

PSV was provided by the anaesthetic ventilator Felix (Taema, Anthony, France). Patients were randomized either to PSV (Group 1) or spontaneous ventilation (Group 2) using a block random allocation and sealed envelopes. The anaesthetists performing the study were not binded to the group, mainly because the Felix screen monitor clearly identifies the respiratory mode and also because activation of PSV is perceptible. All patients received 100 mg hydroxyzine orally 90 min before induction. Continuous monitoring included electrocardiogram (ECG), pulse oximetry, bispectral index (Aspect 2000; Aspect Medical Systems Inc., Newton, MA, USA), tidal volume and inspired and expired concentrations of oxygen, carbon dioxide and sevoflurane. Respiratory rate and non-invasive arterial pressure were recorded every minute. All patients were pre-oxygenated by mask until expired oxygen concentration reached 80%. The ventilator's anaesthetic circuit was prefilled with 8% sevoflurane in 8 L min−1 oxygen for 3 min. Induction was then performed with the vital capacity rapid inhalation induction technique using 8% sevoflurane. After a deep exhalation to the residual volume, the patients were told to take a vital capacity breath, to hold it as long as possible and then to breathe spontaneously. When the bispectral index value was lower than 50, an oropharyngeal airway was inserted. Then in Group 1, PSV was started with pressure support set at 15 cmH2O and an inspiratory trigger set at the lowest level. The minimum respiratory rate was set at 15 min−1 and the maximum inspiratory time at 1.3 s. Patients in Group 2 breathed spontaneously. Two minutes after the beginning of induction, fresh gas flow was reduced to 6 L min−1 and sevoflurane was set at 3%. Remifentanil 1 μg kg−1 was then administered over 2 min followed by an infusion of 0.1 μg kg−1 min−1. In case of apnoea in the PSV group, pressure controlled ventilation was automatically activated by the ventilator after 1 min. The ventilatory parameters were the pre-set PSV parameters, with a respiratory rate set at 15 min−1. Apnoeic patients in the spontaneous ventilation group were ventilated with pressure controlled ventilation after 30 s, settings being similar to those in the PSV group. Orotracheal intubation was performed by an anaesthetist 2 min after the end of the remifentanil bolus. If an intubation attempt failed, a bolus of 0.25 μg kg−1 remifentanil was administered before every other attempt. After intubation, fresh gas flow was reduced to 1 L min−1, the vapourizer was turned off until the expired sevoflurane concentration reached 1.5%, and patients in both groups were then ventilated with pressure controlled ventilation using the same settings. Our induction protocol was modelled with Gasman® software (Med Man Stimulation Inc., Chestnut Hill, MA, USA). The steady-state sevoflurane cerebral concentration was achieved after 5 min and remained stable thereafter.

Heart rate (HR), mean arterial pressure (MAP), respiratory rate, expired tidal volume, end-tidal carbon dioxide and inspired and expired sevoflurane were noted every minute for 10 min from the beginning of induction (time points T1-T10). Bispectral index values were noted every 30 s. All data were also recorded immediately before and after orotracheal intubation. We noted the number of intubation attempts. Intubating conditions were evaluated by the Viby-Mogensen scale [8,9]. Ease of induction was evaluated by a visual analogue scale (VAS) ranging from 1 (most difficult induction possible) to 10 (easiest induction possible). Adverse events such as coughing, laryngospasm or movement were also recorded.

The sample size of our study was designed to evaluate an increase in ventilation in the PSV group. We found no study in adult patients concerning the extent of respiratory depression during sevoflurane anaesthesia. In paediatric patients, Brown and colleagues found that minute ventilation was reduced to 4.5 L m−2 min−1 during inhalation induction with sevoflurane [10]. We calculated that a sample size of 12 patients per group would reach an 80% power to detect a 10% increase in this ventilation with an α level of 0.05. Considering that respiratory depression might be different in adult patients, we decided to include at least 18 patients in each group. Normally distributed data were presented as means ± SD values and compared between groups with a t-test. Ordinal or quantitative variables that were not normally distributed were summarized as median (5th-95th centile) and were compared with U-tests. All tests were two-sided. We did not compare data within the groups, but looked for a significant difference between the groups regardless of measurement time. A P-value of 0.05 was considered as significant.


Eighteen patients were included in each group. One patient was removed from further analysis in Group 2 because the protocol was not followed. There was no difference among patients' characteristics between the two groups (Table 1). None of the patients presented criteria for predicted difficult intubation. During face mask ventilation no leak was clinically detected mainly because patients were not manually ventilated. This allowed the anaesthetist to hold the mask with both hands and to provide a better seal.

Table 1
Table 1:
Patient characteristics (mean ± SD or number of patients).

Expired tidal volume was greater in the PSV group during induction, except at time points T1, T2, T7, T9 and T10 (Fig. 1). Respiratory rate was not different between the two groups except at T8 (P = 0.046). End-tidal carbon dioxide was greater in the PSV group at T3, T5 and T7 and greater in the spontaneous ventilation group at T9 (Fig. 2). We also compared data before and immediately after tracheal intubation. While end-tidal carbon dioxide before intubation was not different between the groups, a significant difference was found after tracheal intubation. Median was 32 mmHg (32-38) in the PSV group vs. 38 mmHg (37-41) in the spontaneous group (P = 0.030). No difference was found in respiratory rate either before or after tracheal intubation (Fig. 3). Apnoea occurred in 12 patients (six in each group) after remifentanil infusion.

Figure 1.
Figure 1.:
Expired tidal volume from the beginning of the induction to the 10th minute. Results are given as median (error bars are 5th-95th centiles), * P < 0.05.
Figure 2.
Figure 2.:
End-tidal carbon dioxide from the beginning of the induction to the 10th minute. Results are given as median (error bars are 5th-95th centiles), * P < 0.05.
Figure 3.
Figure 3.:
Respiratory rate from the beginning of the induction to the 10th minute. Results are given as median (error bars are 5th-95th centiles), * P < 0.05.

Bispectral index values were lower in the PSV group except at T1, T2, T9 and T10 (Fig. 4). Inspired sevoflurane concentrations was similar in both groups during induction. Expired sevoflurane concentrations were lower in the PSV group at T1 (P = 0.011) and was higher in the PSV group at T8 (P = 0.012). Time to tracheal intubation was 7.19 ± 0.73 min in the PSV group vs. 7.50 ± 0.66 min in the spontaneous group (P = 0.181). In the PSV group, all patients were intubated at first attempt, whereas in the spontaneous group one patient required two attempts and another one required four attempts. The intubation score was ≥8 in 17 patients in the PSV group vs. 11 patients in the other group (P = 0.036) (Table 2). Two patients in this group had a score <5. Ease of induction was scored 10 in 13 patients in the PSV group and in nine patients in the spontaneous group. The lowest scores were 6 in the PSV group and 4 in the spontaneous group.

Figure 4.
Figure 4.:
Bispectral index from the beginning of the induction to the 10th minute. Results are given as median (error bars are 5th-95th centiles), * P < 0.05.
Table 2
Table 2:
Intubation scores.

The HR and MAP were not different between the two groups during the induction. MAP was never lower than 50 mmHg or higher than 120 mmHg. Two episodes of bradycardia <40 min−1 occurred (one in each group) and were treated by intravenous (i.v.) atropine (10 μg kg−1).

Concerning the patients who remained non-apnoeic during induction (12 in the PSV group and 11 in the spontaneous group), expired tidal volume was greater in the PSV group except at T1 (P = 0.154). Bispectral index was lower in the PSV group, except at T1, T2 and T3, although expired sevoflurane concentration was not significantly different (except at T1).


The purpose of this study was to determine whether PSV may compensate for the decrease in tidal volume during inhalation induction with sevoflurane, thereby providing better alveolar ventilation and a deeper anaesthetic level.

Expired tidal volume was increased when PSV was used. Expired tidal volume was greater in Group 1 during induction, except in the first 2 min when patients in both groups breathed spontaneously, and at T7, T9 and T10 when most of the patients were intubated and ventilated. The difference was therefore linked to the use of PSV in Group 1. End-tidal carbon dioxide was higher in Group 1 at T3, T5 and T6 although expired tidal volume was greater, probably because of a reduction in the dead space effect by PSV. Reduction of the dead space effect provides a more precise measurement of the expired alveolar gases, leading to a higher end-tidal carbon dioxide. After tracheal intubation, end-tidal carbon dioxide was lower in Group 1, confirming that ventilation was improved with PSV.

We chose to use PSV during inhalational induction with sevoflurane because it preserves the patient's spontaneous ventilation and decreases work of breathing. It could therefore reduce the incidence of atelectasis, which is increased by diaphragmatic activity inhibition [11]. Our results might have been more significant if PSV had been set to provide a 10 mL kg−1 expired tidal volume in Group 1. We chose rather to limit the pressure support to 15 cmH2O based on a preliminary study in our department showing that this level could maintain normocapnia during inhalational induction with sevoflurane (unpublished data). Moreover, several studies have proved that inspiratory pressure should be limited to 20 cmH2O in order to minimise the risk of gastric insufflation during mask ventilation in adult patients [5,12]. Our results might also have been more significant without the use of remifentanil, which increased the occurrence of apnoea and therefore the number of patients ventilated with PCV in both groups.

Concerning the depth of anaesthesia, it is noteworthy that bispectral index was lower in Group 1 during induction when patients were ventilated with PSV. It is well known that the level of anaesthesia is linked to alveolar ventilation during inhalation induction [13]. In the PSV group minute ventilation was greater and anaesthesia level was lower. This deeper anaesthesia level might have been due to an increase in alveolar distribution of sevoflurane. On the other hand, inspiratory and expiratory sevoflurane concentrations were not different between the groups. Inspiratory concentration was not expected to be different because the vapourizer settings were the same in both groups. Expiratory concentration might have been different, according to the lower bispectral index in the PSV group. However, the dead space effect during face mask ventilation may have reduced this difference. Moreover, end-tidal sevoflurane concentration does not reflect the cerebral concentration until the steady state is reached (T5). This might explain the dissociation between a significant difference in bispectral index level and a non-significant one in expiratory sevoflurane concentration. The deeper anaesthesia level provided better conditions for tracheal intubation in the PSV group. All patients in the PSV group and 15 patients in the spontaneous group were intubated at the first attempt. The ease of intubation was better in the PSV group (P = 0.036).

The tolerance of inhalation induction was good. No movement, agitation or movement occurred. Patient-to-ventilator adaptation was good in the PSV group although the inspiratory trigger was set at its lowest level. The automatic switch from PSV to pressure controlled ventilation in the apnoeic patients in Group 1 was well tolerated. Haemodynamic tolerance was good and no hypotension occurred. Even though our patients were young and healthy, inhalation induction with sevoflurane is known to provide good haemodynamic tolerance, even during vascular surgery [14] or in elderly patients [15]. Two episodes of bradycardia (one in each group) were efficiently treated by i.v. atropine.

Even though the anaesthetists who performed the study were not blinded to the respiratory mode because activation of PSV was perceptible, the risk of bias was low as comparison between the two groups was performed from measured data and not from subjective evaluation.

In conclusion, our results show that during inhalational induction with sevoflurane and remifentanil, pressure support ventilation can compensate for the respiratory depression due to these drugs. It allows better distribution of sevoflurane, leading to a deeper anaesthetic level and better intubating conditions. The tolerance of the technique is good.


1. Yamaguchi S, Egawa H, Mishio M, Okuda Y, Kitajima T. Bispectral index monitoring during vital capacity rapid inhalation induction technique with sevoflurane. J Clin Anesth 2003; 15: 24-28.
2. Doi M, Ikeda K. Respiratory effects of sevoflurane. Anesth Analg 1987; 66: 241-244.
3. Ma D, Chakrabarti MK, Whitwam JG. The combined effects of sevoflurane and remifentanil on central respiratory activity and nociceptive cardiovascular responses in anesthetized rabbits. Anesth Analg 1999; 89: 453-461.
4. Eger EI. New inhaled anesthetics. Anesthesiology 1994; 80: 906-922.
5. Von Goedecke A, Voelckel WG, Wenzel V et al. Mechanical versus manual ventilation via a face mask during the induction of anesthesia: a prospective, randomised, crossover study. Anesth Analg 2004; 98: 260-263.
6. Christie JM, Smith RA. Pressure support ventilation decreases inspiratory work of breathing during general anesthesia and spontaneous ventilation. Anesth Analg 1992; 75: 167-171.
7. Langeron O, Masso E, Huraux C et al. Prediction of difficult mask ventilation. Anesthesiology 2000; 92: 1229-1236.
8. Viby-Mogensen J, Engbaek J, Eriksson LI et al. Good clinical research practice (GCRP) in pharmacodynamic studies of neuromuscular blocking agents. Acta Anaesthesiol Scand 1996; 40: 59-74.
9. Koscielniak-Nielsen ZJ, Horn A, Sztuk F, Eriksen K, Theil Skovgaard L, Viby-Mogensen J. Timing of tracheal intubation: monitoring the orbicularis oculi, the abductor pollicis or use a stopwatch? Eur J Anaesthesiol 1996; 13: 130-135.
10. Brown K, Aun C, Stocks J, Jackson E, Mackersie A, Hatch D. A comparison of the respiratory effects of sevoflurane and halothane in infants and young children. Anesthesiology 1998; 89: 86-92.
11. Rothen HU, Sporre B, Engberg G, Wegenius G, Hedenstierna G. Re-expansion of atelectasis during general anesthesia: a computed tomography study. Br J Anaesth 1993; 71: 788-795.
12. Weiler N, Heinrichs W. Assessment of pulmonary mechanics and gastric inflation pressure during mask ventilation. Prehosp Disas Med 1995; 10: 101-105.
13. Nathan N, Bazin JE, Cros AM. Induction par inhalation. Ann Fr Anesth Réanim 2004; 23: 884-889.
14. Godet G, Watremez C, El Kettani C, Soriano C, Coriat P. A comparison of sevoflurane, target controlled infusion of propofol and propofol/isoflurane anesthesia in patients undergoing carotid surgery: a quality of anesthesia and recovery profile. Anesth Analg 2001; 93: 560-565.
15. Yamaguchi S, Ikeda T, Wake K, Okuda Y, Kitajima T. A sevoflurane induction with gradual reduction of concentration is well tolerated in elderly patients. Can J Anesth 2003; 50: 26-31.


© 2005 European Society of Anaesthesiology