Secondary Logo

Journal Logo

Preoxygenation by spontaneous breathing or noninvasive positive pressure ventilation with and without positive end-expiratory pressure

A randomised controlled trial

Hanouz, Jean-Luc; Lammens, Stéphane; Tasle, Marine; Lesage, Anne; Gérard, Jean-Louis; Plaud, Benoit

European Journal of Anaesthesiology (EJA): December 2015 - Volume 32 - Issue 12 - p 881–887
doi: 10.1097/EJA.0000000000000297

BACKGROUND In emergency situations requiring rapid airway control, shortening preoxygenate time is desirable.

OBJECTIVES The objective of this study is to compare the time to achieve an expired O2 fraction FeO2 of 90% (FeO2 90%) during preoxygenation with spontaneous breathing and positive pressure ventilation with and without positive end-expiratory pressure (PEEP).

DESIGN A randomised controlled trial.

SETTING Primary care in a university hospital in France from October 2006 to January 2008.

PATIENTS Adults patients scheduled for elective surgery. Exclusion criteria were rapid sequence induction, anticipated difficult airway management and refusal to provide consent.

INTERVENTION Patients were randomly allocated to preoxygenation with spontaneous breathing or positive pressure ventilation (positive inspiratory pressure: 12 cmH2O) without PEEP and with PEEP (positive inspiratory pressure: 12 cmH2O, PEEP: 6 cmH2O).

MAIN OUTCOME MEASURES Time to achieve an expired O2 fraction of 90% measured from positioning the face mask, and the time it took after endotracheal intubation for the SpO2 to fall to 93% (SpO2 93%) while the patient was apnoeic. Patient discomfort was recorded (visual analogue scale). Data are median (quartile 25th to 75).

RESULTS The time to achieve an FeO2 90% was shorter with positive pressure ventilation, with PEEP [140 (100 to 200) s] and without PEEP [153 (120 to 218) s], than with spontaneous breathing [190 (130 to 264) s; P = 0.002]. At 3 min, 47, 60 and 74% of patients achieved an FeO2 of 90% or more in the spontaneous breathing, positive pressure ventilation without and with PEEP groups, respectively (P = 0.01). Cox proportional-hazards regression showed that positive pressure ventilation with PEEP [hazard ratio 2.18; 95% confidence interval (95% CI) 1.42 to 3.36); P < 0.001] and without PEEP (hazard ratio 1.62; 95% CI 1.05 to 2.50; P = 0.03) were associated with a shorter time to an FeO2 90%. The time until SpO2 93% was not significantly different between spontaneous breathing [305 (263 to 383) s], positive pressure ventilation without PEEP [370 (300 to 450) s] and with PEEP [345 (245 to 435) s; P = 0.08]. The discomfort reported was 0 (0 to 18) mm and was comparable between groups (P = 0.22).

CONCLUSION Compared with spontaneous breathing, positive pressure ventilation with and without PEEP shortened preoxygenation time.

TRIAL REGISTRATION identifier: NCT02313766.

From the Department of Anaesthesia and Intensive Care Medicine, CHU de Caen (J-LH, AL, J-LG), Department of Anaesthesia, CHP Saint Martin, Caen (SL), Department of Anaesthesia, Hôpitaux du Leman, Thonon-les-Bains (MT), and Department of Anaesthesia and Intensive Care, CHU Saint-Louis Lariboisière Fernand Widal, Assistance Publique – Hôpitaux de Paris, Paris, France (BP)

Correspondence to Jean-Luc Hanouz, Pôle Réanimations Anesthésie SAMU (niveau 6), CHU de Caen, Av Côte de Nacre, 14033 Caen Cedex, France E-mail:

Published online 29 July 2015

Back to Top | Article Outline


Difficulty maintaining the airway and achieving adequate ventilation can lead to hypoxaemia during induction of anaesthesia.1 In emergency situations, such as caesarean section, and in the presence of acute respiratory, haemodynamic or neurological failure, the airway must be secured rapidly to ensure adequate oxygenation and ventilation. In these situations, the time to achieve preoxygenation is often prolonged because oxygen consumption is increased, cardiac output is decreased, and patient cooperation may not exist.2 In critically ill patients and during haemorrhagic shock, the time course of oxyhaemoglobin desaturation to below 85% has been shown to be dramatically short, only 20 to 30 s.3–53–53–5 Frequently, in urgent situations, intubation can be difficult, and the time necessary to complete intubation may be prolonged.6 Consequently, shortening the time needed to achieve efficient preoxygenation remains a critical issue in emergency situations.2

Two methods of preoxygenation that have been proposed are spontaneous breathing until the expired oxygen fraction (FeO2) reaches 90%, which usually takes 3 to 5 min, and eight deep breaths within 60 s, which requires the patient's cooperation.7 Nevertheless, it has been shown that inadequate preoxygenation (FeO2 <90%) is a frequent event during induction of anaesthesia.8

The latest generation of anaesthesia ventilators provide pressure support ventilation modes with trigger sensitivity and inspiratory flow comparable to those provided by ICU ventilators and can theoretically be used for preoxygenation.9 Preoxygenation using positive inspiratory pressure ventilation (PPV) associated with positive end-expiratory pressure (PEEP) has been shown to be feasible.5,10,115,10,115,10,11 It could be hypothesised that alveolar recruitment, increase in tidal volume and prevention of inwards leakage of air could shorten the time to achieve an FeO2 of 90% (FeO2 90%).

At the present time, the duration of preoxygenation has not been compared during spontaneous breathing and PPV with and without PEEP. Because this may be relevant in emergency situations, we tested the hypothesis that PPV with and without PEEP could shorten the time to reach FeO2 90% (primary endpoint). The secondary endpoint was to compare the time until SpO2 had fallen to 93% (SpO2 93%), as the absence of hypoxaemia during the apnoeic period required for endotracheal intubation remains the major goal of preoxygenation.

Back to Top | Article Outline

Materials and methods

This study was approved by the institutional review board (Comité de Protection des Personnes Nord Ouest III, CHU de Caen, Caen, France, authorisation 2005-039 delivered on 3 December 2005). All patients provided written informed consent at the end of the preanaesthesia consultation.

We recruited patients older than 18 years, who achieved American Society of Anesthesiologists (ASA) Physical Status classes I and II, and who were scheduled for surgery with general anaesthesia and oro-tracheal intubation. Exclusion criteria were rapid sequence induction, anticipated difficult mask ventilation (two or more factors among age >55 years, BMI >26 kg m−2, beard, history of snoring, lack of teeth), anticipated difficult intubation (history of difficult intubation, Mallampati class 3 and 4, thyromental distance <60 mm, interincisor distance <35 mm, limited cervical spine movement) and refusal to provide consent. Preoperative data recorded were age, height, weight, ASA physical status and history of asthma, current smoking (ex-smokers were considered after a 3-month period without smoking), number of pack-years. Obesity was defined as a BMI > 30 kg m-2.

On the day of surgery, each patient was randomly allocated (random number table, block sizes of 15) to one of the three preoxygenation methods: spontaneous breathing, noninvasive positive inspiratory pressure ventilation (12 cmH2O) without PEEP (PPV) or with PEEP at 6 cmH2O (PEEP). The positive inspiratory pressure support and PEEP were chosen according to previous data.5,11,125,11,125,11,12 After intravenous line placement and monitoring that included noninvasive blood pressure (measurement repeated every 3 min), continuous five-lead electrocardiography and peripheral O2 saturation (SpO2; IntelliVue MP70 Philips HealthCare, Amsterdam, The Netherlands), patients were preoxygenated through a face mask firmly applied and connected to the anaesthesia machine (Fabius GS; Dräger Medical SAS, Antony, France) delivering a fresh gas flow of 12 l min−1. The circle breathing system was flushed with the O2 bypass for 30 s before mask placement. The inspired O2 concentration was set at 100%, the inspiratory trigger sensitivity was set at −2l min−1, the adjustable pressure limiting valve was opened and maximal airway pressure was limited at 18 cmH2O. Inspired and expired fraction of O2 and CO2 were continuously measured (IntelliVue G5 Gas Module; Philips HealthCare, Amsterdam, The Netherlands) and displayed on the anaesthesia monitor. In all groups, the end of preoxygenation was defined by FeO2 90%.13,1413,14 At the end of preoxygenation, anaesthesia was induced with intravenous propofol 2.5 mg kg−1, alfentanil 40 μg kg−1 and succinylcholine 1 mg kg−1 followed by a continuous infusion of propofol 8 mg kg−1 h−1 to a target bispectral index between 40 and 50. The propofol and alfentanil stopped respiration in all groups. No face mask ventilation was provided before the orotracheal intubation, which was performed at the end of muscle fasciculation. The correct position of the tube (tracheal position above carena) was immediately checked using a fibrescope, and only if satisfactory, then atracurium 0.5 mg kg−1 was given. If tube misplacement occurred, the patient was excluded from the study and standard care was provided. The time from the end of intubation (before checking the endotracheal tube position with the fibrescope) until SpO2 had fallen to 93% was measured. The SpO2 and FeO2 were recorded every 10 s during the preoxygenation phase. SpO2 was recorded every 10 s during the apnoea phase.

Just before leaving the postanaesthesia care unit, the patient was asked to evaluate the preoxygenation method with the following questions: How do you evaluate the preoxygenation phase on the scale (cursor placed on a 100 mm visual analogue scale) between very comfortable (0 mm) and very uncomfortable (100 mm)? Do you want to have the same preoxygenation method for another anaesthetic (yes/no)?

The primary endpoint of the study was the time for preoxygenation in seconds measured from face mask positioning to FeO2 90%.

The following data were also recorded and analysed as secondary endpoints: the duration of apnoea in seconds (from the end of tracheal intubation (but before checking the tube position with the fibrescope) to SpO2 93%) and the patient's comfort evaluated on a 0 to 100 mm visual analogue scale.

Back to Top | Article Outline

Statistical analysis

A 15% difference in the primary outcome, equivalent to a 30-s reduction in the time required to obtain 90% FeO2 between spontaneous breathing and PPV groups, was considered clinically relevant. On the basis of a preliminary measured spontaneous breathing preoxygenation time [180 ± 45 s (mean ± SD)in 20 patients], 48 patients per group were required to detect a 15% difference in preoxygenation time accepting a two-sided α error of 5% and β error of 10%. We planned to include 50 patients in each group.

Normality of quantitative data was tested using the Kolmogorov and Smirnov test. Equality of variance was tested by the Levene test if required. Data are reported as mean (SD) or median (quartile 25th to quartile 75th) as required. Quantitative data were compared between groups with one-way analysis of variance or Kruskal–Wallis test, as appropriate, followed by posthoc comparisons using the Student–Newman–Keuls method and the Conover Inman method, respectively. Qualitative data were compared with the χ2 test. Relationship between the dependent variables (time until FeO2 90% and time until SpO2 93%) and quantitative independent variables was tested with linear regression analysis. Cox proportional-hazards regression was used to examine the effect of groups and covariates on the time-to-event (time until FeO2 90% and time until SpO2 93%). Covariates were selected if the P value was less than 0.10 in the univariate analysis. Hazard ratios and their 95% confidence intervals (95% CI) were reported for groups and covariates. The model discrimination was evaluated by its concordance and the overall goodness of fit was evaluated by the likelihood ratio χ2 test.

All statistical tests were two-sided. Statistical analysis was performed using MedCalc Software v (MedCalc Software, Mariakerke, Blegium) and R 2.15.1: A Language and Environment for Statistical Computing (http:// with ‘survival’ package.

Back to Top | Article Outline


One hundred and fifty consecutive patients were included in the study from October 2006 to October 2007. One patient in the spontaneous breathing group and three patients in the PPV group were excluded (Fig. 1). Patients’ characteristics were similar between the groups (Table 1).

Fig. 1

Fig. 1

Table 1

Table 1

Back to Top | Article Outline

Effects of the ventilation mode on the duration of preoxygenation

The time from face mask positioning until FeO2 90% was shorter in the PEEP [140 (100 to 200) s] and PPV [153 (120 to 218) s] groups than the spontaneous breathing group [190 (130 to 264) s; P = 0.002]. At 3 min, 23 out of 49 (47%), 28 out of 47 (60%) and 37 out of 50 (74%) patients achieved at least 90% FeO2 in spontaneous breathing, PPV and PEEP groups, respectively (P = 0.01).

The Kaplan–Meier analysis for the time until 90% FeO2 is shown in Fig. 2. The Cox proportional-hazards regression showed that PEEP (hazard ratio 2.18; 95% CI 1.42 to 3.36; P < 0.001), PPV (hazard ratio 1.62; 95% CI 1.05–2.50; P = 0.03) was a time-independent covariate associated with a shorter time to FeO2 90%. The concordance of the Cox proportional-hazards regression model was 0.60, R2 = 0.11, and the Score test was 18.36 (P = 0.002).

Fig. 2

Fig. 2

Back to Top | Article Outline

Effects of ventilation mode on the time to desaturate

The time until SpO2 93% was similar between the groups [spontaneous breathing: 305 (263 to 383) s; PPV: 370 (300 to 450) s; PEEP: 345 (245 to 435) s; P = 0.08].

The time until SpO2 93% was shorter in obese vs. nonobese patients [200 (139 to 285) vs. 355 (278 to 430) s; P < 0.0001]. There was a significant relationship between the time until SpO2 93% and BMI (R2 = 0.28; P < 0.01).

The Kaplan–Meier analysis for the time until SpO2 93% is shown in Fig. 3. The Cox proportional-hazards regression showed that the preoxygenation method (PPV: hazard ratio 0.76; 95% CI 0.49 to 1.19 and PEEP: hazard ratio 0.82; 95% CI 0.53 to 1.25] did not modify the decrease in SpO2 during apnoea. The concordance of the Cox proportional-hazards regression model was 0.62, R2 = 0.11, and the Score test was 57.26 (P < 0.0001).

Fig. 3

Fig. 3

Back to Top | Article Outline

Patient evaluation of the preoxygenation technique

The discomfort of the preoxygenation phase was reported as low [0 (0 to 18)] on a 0 to 100 mm visual analogue scale and was similar between groups (Table 1). Ninety-seven percent of patients answered that they would accept the same preoxygenation method on another occasion.

Back to Top | Article Outline


The main result of this study is that compared with spontaneous breathing, preoxygenation with PPV with and without PEEP decreased the time to attain FeO2 90%. The time for SpO2 to drop to 93% was comparable across the groups.

In emergency situations, increased oxygen consumption, decreased functional residual capacity, decreased cardiac output, loss of cooperation and increased risk of unanticipated difficult airway can combine to dramatically increase the risk of severe hypoxaemia.2,52,5 Theoretical models have shown that not only is the rate of oxyhaemoglobin desaturation very sensitive to the pathophysiological disorders described above, but also to FeO2.3 The model anticipated that only 23 s was required to achieve marked oxyhaemoglobin desaturation to 85% in a scenario that included some pathophysiology.3 In an experimental study, the time for SpO2 to fall to less than 70% during apnoea was shorter with increasing levels of haemorrhage but could be markedly increased (seven-fold) by adequate preoxygenation.4 It has been reported that 46% of critically ill patients experienced critical desaturation (SpO2 <80%) during rapid sequence induction and intubation despite 3 min of preoxygenation.5 Consequently, reducing the time needed for adequate preoxygenation may be of paramount importance in emergency situations wherein the time available to safely manage the airway is much shorter than thought.

In patients with acute respiratory failure or hypoxaemia, preoxygenation through PPV with PEEP at 5 cmH2O before rapid sequence induction has been shown to maintain SpO2 and arterial partial pressure of O2 better than preoxygenation with bag-valve and mask.5 The time to desaturate could not be studied. It should be noted that assessing preoxygenation by SpO2 is flawed because it does not estimate oxygen stores.14 In morbidly obese patients scheduled for bariatric surgery, preoxygenation using PPV (pressure support: 8 to 10 cmH2O and PEEP: 6 cmH2O) was shown to accelerate the rise in FeO2 and to increase the proportion of patients with FeO2 at least 95% at the end of 5 min preoxygenation.11 The time to desaturate (time to SpO2 95%) was measured and arterial blood gas analysis performed in a small number of patients, but the findings cannot be extrapolated to nonobese patients, in part because of the reduced functional residual capacity in the obese. To our knowledge, only one study reported that 3-min oxygenation through PPV (pressure support at 4 to 6 cmH2O and PEEP at 4 to 6 cmH2O) shortened the time to reach SpO2 90% and increased the proportion of individuals with FeO2 at least 90%.10 However, the study enrolled only 20 healthy volunteers and the time to desaturate could not be studied. Healthy volunteers cannot be considered representative of patients during induction of anaesthesia.

The present study showed that preoxygenation with PPV (pressure support: 12 cmH2O with and without PEEP at 6 cmH2O) decreased the time to FeO2 90% compared with spontaneous breathing. These findings confirm those reported in obese patients and extend the results to unselected ASA I and II patients.11 Although the clinical relevance of the difference (40 to 50 s) may be questionable, it might be crucial in sicker patients and in emergency cases. The mechanisms by which preoxygenation time is shortened might be increased tidal volume and minute ventilation resulting in a more efficient wash out of nitrogen,3,133,13 an increase in end expiratory lung volume through recruitment of collapsed alveoli15 and prevention of inward air leakage diluting inspired oxygen.16 Although it was a secondary endpoint of the study, the time to desaturate to SpO2 93% was not significantly different between the groups. As 90% FeO2 was the end point for preoxygenation, this result highlights the comparable efficacy of preoxygenation in all groups. The time to desaturate depends mainly on oxygen reserve (i.e. functional residual capacity) and consumption, and that should be similar between the groups (Table 1). Nevertheless, there was a trend towards a longer time to SpO2 93% in the PPV groups, and it has been suggested that continuous positive airway pressure (6 cmH2O) applied during preoxygenation will prolong the time to desaturate compared with zero end-expiratory pressure.17 Specifically designed studies should examine the effect of preoxygenation with pressure support ventilation modes on desaturation time in healthy individuals.

Despite a high level of positive pressure support (12 cmH2O), reported discomfort in the preoxygenation ventilation mode was low with 89% reporting a visual analogue discomfort score less than 30 mm; this was similar between the groups (Table 1). Finally, 97% would agree to have the same ventilation mode on another occasion. In obese patients, the tolerance of PPV with PEEP was reported as good and no different from spontaneous breathing through a face mask.11 It has also been shown that few patients experience significant discomfort from the mask during preoxygenation and that anaesthetists overestimate the degree of patient discomfort.18

Our results should be interpreted with the knowledge that preoxygenation with an inspired O2 concentration of 100% has been shown to promote pulmonary atelectasis.19 But PEEP during induction of anaesthesia has also been shown to prevent pulmonary atelectasis in the obese,12 and PPV has been shown to increase end expiratory lung volume and pulmonary function after intubation.15 Taken together, these results support the use of PPV and PEEP during preoxygenation to decrease the early pulmonary atelectasis observed after intubation. This issue requires further study.

Our results require cautious interpretation. First, the study was not powered to detect a difference in the time to desaturate, which remains the major goal of preoxygenation. However, decreasing the time to FeO2 90% is also important in emergency situations in order to minimise the risk of critical hypoxia. Second, we did not record respiratory rate and tidal volume during preoxygenation, but it is clearly established that PPV results in lower respiratory rate, higher tidal volume and increased end-expiratory lung volume.15 Third, we did not measure preoperative functional residual capacity, a major determinant of O2 reserve at the end of the preoxygenation. Nevertheless, the characteristics of the population studied strongly suggest that the functional residual capacity should be unaltered and equally distributed across the groups (Table 1). Fourth, we did not examine side effects of PPV such as gastric distension and pulmonary aspiration during endotracheal intubation. In morbidly obese patients, it has been suggested that PPV during preoxygenation results in moderate gastric distension. However, using anaesthesia ventilators, peak airway pressure can be limited in contrast to manual ventilation. The latter has been shown to promote gastric distension.20 At the present time, no study has been designed to examine the effect of PPV on pulmonary aspiration. Fifth, preoxygenation was performed in a supine position rather than a head-up position that has been shown to increase functional residual capacity and thus O2 reserves after preoxygenation.13 However, head-up is not the routine position for preoxygenation in nonobese patients.

In conclusion, we have shown that, compared with spontaneous breathing, preoxygenation with PPV with and without PEEP decreased the time to FeO2 90%. Importantly, the time to desaturate was comparable between the groups.

Back to Top | Article Outline

Acknowledgements relating to this article

Assistance with the study: none.

Financial support and sponsorship: this work was supported by CHU de Caen, Caen France.

Conflicts of interest: none.

Presentation: none.

Back to Top | Article Outline


1. Cheney FW, Posner KL, Lee LA, et al. Trends in anesthesia-related death and brain damage. A closed claims analysis. Anesthesiology 2006; 105:1081–1086.
2. Weingart SD, Levitan RM. Preoxygenation and prevention of desaturation during emergency airway management. Ann Emerg Med 2012; 59:165–175.
3. Farmery AD, Roe PG. A model to describe the rate of oxyhaemoglobin desaturation during apnoea. Br J Anaesth 1996; 76:284–291.
4. Pehböck D, Wenzel V, Voelckel W, et al. Effects of preoxygenation on desaturation time during hemorrhagic shock in pigs. Anesthesiology 2010; 113:593–599.
5. Baillard C, Fosse JP, Sebbane M, et al. Noninvasive ventilation improves preoxygenation before intubation of hypoxic patients. Am J Respir Crit Care Med 2006; 174:171–177.
6. Schwartz DE, Matthay MA, Cohen NH. Death and other complications of emergency airway management in critically ill adults. Anesthesiology 1995; 82:367–376.
7. Baraka AS, Taha SK, Aouad MT, et al. Preoxygenation: comparison of maximal breathing and tidal volume breathing techniques. Anesthesiology 1999; 91:612–616.
8. Baillard C, Depret F, Levy V, et al. Incidence and prediction of inadequate preoxygenation before induction of anaesthesia. Ann Fr Anesth Reanim 2014; 33:e55–e58.
9. Jaber S, Tassaux D, Sebbane M, et al. Performance characteristics of five new anesthesia ventilators and four intensive care ventilators in pressure-support mode: a comparative bench study. Anesthesiology 2006; 105:944–952.
10. Tanoubi I, Drolet P, Fortier LP, Donati F. Inspiratory support versus spontaneous breathing during preoxygenation in healthy subjects: a randomized, double blind, cross-over trial. Ann Fr Anesth Reanim 2010; 29:198–203.
11. Delay JM, Sebbane M, Jung B, et al. The effectiveness of noninvasive positive pressure ventilation to enhance preoxygenation in morbidly obese patients: a randomized controlled study. Anesth Analg 2008; 107:1707–1713.
12. Coussa M, Proietti S, Schnyder P, et al. Prevention of atelectasis formation during the induction of general anesthesia in morbidly obese patients. Anesth Analg 2004; 98:1491–1495.
13. Tanoubi I, Drolet P, Donati F. Optimizing preoxygenation in adults. Can J Anesth 2009; 56:449–466.
14. Campbell IT, Beatty PC. Monitoring preoxygenation. Br J Anaesth 1994; 72:3–4.
15. Futier E, Constantin JM, Pelosi P, et al. Noninvasive ventilation and alveolar recruitment maneuver improve respiratory function during and after intubation of morbidly obese patients: a randomized controlled study. Anesthesiology 2011; 114:1354–1363.
16. Gagnon C, Fortier LP, Donati F. When a leak is unavoidable, preoxygenation is equally ineffective with vital capacity or tidal volume breathing. Can J Anaesth 2006; 53:86–91.
17. Herriger A, Frascarolo P, Spahn DR, Magnusson L. The effect of positive airway pressure during preoxygenation and induction of anaesthesia upon duration of nonhypoxic apnoea. Anaesthesia 2004; 59:243–247.
18. Schlack W, Heck Z, Lorenz C. Mask tolerance and preoxygenation: a problem for anesthesiologists but not for patients. Anesthesiology 2001; 94:546.
19. Edmark L, Kostova-Aherdan K, Enlund M, Hedenstierna G. Optimal oxygen concentration during induction of general anesthesia. Anesthesiology 2003; 98:28–33.
20. Ho-Tai LM, Devitt JH, Noel AG, O’Donnell MP. Gas leak and gastric insufflation during controlled ventilation: face mask versus laryngeal mask airway. Can J Anaesth 1998; 45:206–211.
© 2015 European Society of Anaesthesiology