Pre-oxygenation using high-flow nasal oxygen vs. tight facemask during rapid sequence induction

Sjöblom. Anaesthesia 2021 Feb 18; doi:10.1111/anae.15426

Clinical Question

  • In adult patients undergoing rapid sequence intubation for emergency surgery, does pre-oxygenation with high-flow nasal oxygen, compared with standard face-mask oxygen, reduce the incidence of oxygen desaturation?


  • The aim of rapid sequence induction (RSI) is to minimise the apnoeic period between the application of oxygen to the patient prior to and during induction and oxygenation via the endotracheal tube following intubation
  • Hypoxia (most commonly recognised by low oxygen saturations) is a recognised complication of rapid sequence induction, and endotracheal intubation in general
  • High flow nasal oxygen (HFNO) application has the potential advantage over facemask pre-oxygenation of being continually administered during the apnoea period
  • Apnoeic oxygenation makes physiologic sense and has been well-described in the medical literature but randomised controlled trials investigating this have only recruited small numbers of patients Pavlov systematic review 2017
  • Apnoeic oxygenation via nasal cannulae is recommended by the Faculty of Intensive Care Medicine in the Guidelines for the management of tracheal intubation in critically ill adults
  • It is currently unknown whether the continuous flow of oxygen during intubation prevents the incidence of hypoxia during RSI, either in the critical care or anaesthetic room setting
  • This study attempted to answer this clinical question in patients requiring an RSI for surgical intervention


  • Unblinded international prospective randomised controlled trial
  • Power calculation determined that a sample size of 326 patients would provide 80% power to detect a 7.5% absolute risk reduction in oxygen desaturations using HFNO with a type-1 error of 5%
  • The primary outcome was the number of patients who developed oxygen saturation <93% from the start of pre- oxygenation until 1 min after intubation as measured by peripheral pulse oximeter
  • Secondary outcomes were as follows:
    1. Differences between centres
    2. End-tidal gas concentrations (ETCO2 and ETO2) in the first breath after intubation
    3. Effect of office hours vs. on- call hours
    4. The number of patients with signs of gastric regurgitation
  • Patients were randomly allocated to either pre-oxygenation with high-flow nasal oxygen or with a tight- fitting face mask. This was done using sealed envelopes assigned in a 1:1 ratio in block sizes of 10


  • Six centres in Europe (five in Sweden and one in Switzerland) between March 2018 and February 2020


  • Inclusion: Adult patients undergoing elective surgery where RSI was planned
  • Exclusion:
    1. Basal metabolic index (BMI) > 35 kg.m2
    2. Pregnancy
    3. Need for non-invasive ventilation prior to anaesthesia
    4. Inability to reach oxygen saturations of >93% during pre-oxygentaion
  • 350 patients randomly allocated to either pre-oxygenation with HFNO or facemask
  • One patient excluded due to a protocol violation
  • Three patients were ventilated during the apnoea phase but were included in the results
  • 349 patients included in the final analysis
  • No significant difference in baseline or airway characteristics between groups
    • Facemask group had a higher prevalence of a  pulmonary comorbidity (18.3% vs. 13.2%)
    • Average BMI 25.1 in HFNO vs. 25.5 in facemask group
    • 19.5% HFNO had smoking history vs. 18.4% in facemask group


  • High-flow nasal oxygen
    • Administered using specifically designed nasal cannulae (Optiflow TM, Fisher and Paykal Healthcare, Auckland, New Zealand) with 30-50 l/min of heated and humidified oxygen
    • Patients could breathe with an open or closed mouth
    • Oxygen flow increased to 70 l/min once apnoea occurred and was administered continuously until the tracheal tube was in place


  • Tight-fitting facemask
    • 100% oxygen via a tight-fitting facemask with a fresh gas flow of 10 l/min delivered via a circle system

Management common to both groups

  • Standard monitoring applied
  • Rapid sequence induction performed according to local protocol at each hospital
  • Drugs and doses determined by the anaesthetist in charge
  • Pre-oxygenation conducted for a minimum of three minutes
  • Patients placed supine in a reverse Trendelenburg position
  • Chin lift and jaw thrust used during apnoea to maintain an open airway
  • No difference in anaesthesia drugs and dosages between groups
    • Propofol was used as the induction agent in 59% of patients
    • Thiopentone was used as the induction agent in 36.1% of patients
    • Suxamethonium was used as the paralytic agent in 67.0% of patients
    • Rocuronium was used as the paralytic agent in 32.7% of patients


  • Primary outcome: No difference was seen in oxygen desaturation incidence between the HFNO (5 patients; 2.9%) and facemask (six patients; 3.4%) groups (p=0.77)
  • No difference in lowest mean oxygenation saturations from the start of pre-oxygenation until one minute after intubation between the groups (HFNO mean 99.1% vs. facemask 99.0%)
  • Secondary outcomes:
  • No significant difference in desaturation incidence between centres
    • However, patients in Switzerland had higher rates of desaturation (7%) when compared with patients in Sweden (1.6%), p=0.009
    • Swiss patients also had longer apnoea (125.6 seconds vs. 102.2 seconds; p=0.004) and intubation times 64.2 seconds vs. 47.2 seconds; p=0.013) than Swedish patients
  • No difference in ETCO2 levels in the first breath after intubation between groups (4.64 vs. 4.56; p=0.33)
  • Higher levels of ETO2 in facemask group when compared with HFNO group (84.9 kpa vs. 76.7 kpa; p<0.001)
  • No significant difference in incidence of desaturation between  office hours (4.7%) and on-call hours (0.7%; p=0.06)
  • One incident of regurgitation was reported and this was in the HFNO group

Authors’ Conclusions

  • We have shown that high-flow nasal oxygen is a suitable method for pre- and peri-oxygention to maintain adequate oxygen levels during RSI and is an alternative to traditional facemask pre-oxygenation


  • This was a well-conducted study with an appropriate power calculation and randomisation strategy
  • There were few exclusion criteria so the study sample population was representative of the general population
  • Reported baseline characteristics were well-balanced between the two groups and the authors reported on demographic, comorbid and airway variables
  • Only one patient was excluded due to a protocol violation
  • Patients were included both in-hours and after-hours, improving generalisability
  • The secondary endpoints were all relevant


  • The power calculation was based on an anticipated oxygen desaturation incidence of 10% in the facemask group, but the observed desaturation incidence was only 3.4%. The study may have been underpowered to detect a difference in the primary outcome
  • The overall incidence of oxygen desaturation below 93% was low (3.1%)
  • The primary outcome, whilst reasonable, was not patient centred, and the clinical importance of an oxygen desaturation below 93% is unclear. Oxygen saturations below 85% were only observed in four patients
  • The reason for the difference between Switzerland and Sweden in oxygen desaturation rates, apnoea and intubation times is unclear but does raise the question of significant between-country difference in anaesthetic practice
  • No blinding of the intervention was possible, which may have introduced bias
  • No conclusions can be made about pregnant or obese patients as they were excluded from the study

The Bottom Line

  • This was a well-conducted study and showed no tangible benefit in either HFNO or facemask application for reducing oxygen desaturation in patients undergoing RSI performed by anaesthetists prior to surgery
  • The results are difficult to translate to my critical care practice but I remain unconvinced of the clinical benefits of apnoeic oxygenation via nasal cannulae during intubation
  • Users should continue to use the technique that they are most comfortable with until evidence suggests a change in practice

External Links


Summary author: Fraser Magee  @fraz65
Summary date: 31/03/21
Peer-review editor: @davidslessor



Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.