Publication

• Title: High-flow nasal cannula oxygen during endotracheal intubation in hypoxemic patients: a randomized controlled clinical trial
• Acronym: PREOXYFLOW
• Year: 2015
• Journal published in: Intensive Care Medicine
• Citation: Vourc’h M, Asfar P, Volteau C, Bachoumas K, Clavieras N, Egreteau PY, et al. High-flow nasal cannula oxygen during endotracheal intubation in hypoxemic patients: a randomized controlled clinical trial. Intensive Care Med. 2015;41(9):1538-1548.

Context & Rationale

• Background

  • Peri-intubation hypoxaemia and haemodynamic collapse are frequent during ICU endotracheal intubation, with clinically important morbidity and mortality risk.¹
  • Conventional preoxygenation with a high FiO₂ face mask is standard, but offers no oxygen delivery during laryngoscopy once the mask is removed.
  • High-flow nasal cannula (HFNC) can deliver heated, humidified oxygen at high flow rates and can remain in place during laryngoscopy (theoretical “peroxygenation” / apnoeic oxygenation).
  • Before PREOXYFLOW, ICU data supporting HFNC during intubation were mainly non-randomised and at risk of confounding, including a before–after study suggesting reduced desaturation in mild-to-moderate hypoxaemia.²
• Research Question/Hypothesis

  • In hypoxaemic ICU adults requiring endotracheal intubation, does HFNC (used for preoxygenation and left in situ during intubation) improve the lowest oxygen saturation during the intubation procedure versus a high FiO₂ face mask?
  • Hypothesis (as powered): HFNC would increase nadir SpO₂ during intubation by an absolute 6% compared with face mask preoxygenation.
• Why This Matters

  • Even brief desaturation episodes (e.g., SpO₂ <80%) during intubation are associated with downstream harms, including cardiac arrest and peri-intubation instability in critically ill patients.
  • If HFNC “peroxygenation” reduces nadir SpO₂ without increasing airway difficulty, it would represent a simple, scalable safety intervention for high-risk ICU intubations.

Design & Methods

• Research Question: Among hypoxaemic ICU adults requiring endotracheal intubation, does HFNC (continued during laryngoscopy) improve the lowest SpO₂ during the intubation procedure versus a high FiO₂ face mask?
• Study Type: Multicentre, randomised, open-label, controlled, parallel-group trial (six French ICUs); trial registration NCT01747109.
• Population:
  • Setting: ICU; adults intubated for acute hypoxaemic respiratory failure.
  • Key inclusion thresholds: PaO₂/FiO₂ <300 mmHg; respiratory rate ≥30/min; required FiO₂ ≥0.50 to achieve SpO₂ ≥90% within the hour before enrolment.
  • Key exclusions: Not reported (in full) in the manuscript text extract; examples include scenarios incompatible with protocolised rapid sequence intubation or situations requiring immediate airway control (Not reported).
• Intervention:
  • HFNC (Optiflow™) at 60 L/min, heated/humidified oxygen with FiO₂ 1.0 for 4 minutes preoxygenation.
  • HFNC left in place throughout induction and laryngoscopy until connection to mechanical ventilation (intended apnoeic oxygenation during the procedure).
• Comparison:
  • High FiO₂ face mask preoxygenation at 15 L/min oxygen flow for 4 minutes.
  • Mask removed after induction to permit laryngoscopy (no planned oxygen delivery during the apnoeic laryngoscopy interval).
• Blinding: Open-label; primary endpoint (lowest SpO₂) is physiologic and objectively recorded but performance bias and co-intervention bias remain plausible.
• Statistics: Powered to detect a 6% absolute increase in nadir SpO₂ (from 85% to 91%) with 80% power at a two-sided 5% alpha, assuming 2% attrition; required sample size 122, increased to 124 after early exclusions; analysis described as intention-to-treat but excluded 5 patients post-randomisation (modified ITT).
• Follow-Up Period: Clinical outcomes to day 28 (including ventilator-free days and 28-day mortality); SOFA recorded through day 5.

Key Results

This trial was not stopped early. Recruitment reached the revised target sample (124 randomised); modified intention-to-treat analysis included 119 patients (HFNC n=62; face mask n=57).

• HFNC did not improve nadir SpO₂ during intubation versus high FiO₂ face mask (median 91.5% vs 89.5%; p=0.44), despite remaining in place during laryngoscopy.
• Severe peri-intubation events were common in both groups: SpO₂ <80% occurred in 25.8% (HFNC) vs 22.8% (mask) and cardiovascular collapse in 38.7% vs 52.6%.
• A shorter duration of mechanical ventilation was observed with HFNC (median 6 vs 10 days; p=0.02), while other ICU outcomes (ventilator-free days, ICU length of stay, 28-day mortality) were not significantly different.

Internal Validity

• Randomisation and allocation: Centralised, computer-generated randomisation (block size 4) stratified by centre; allocation via an online system (concealment likely, but detailed safeguards Not reported).
• Post-randomisation exclusions: 5/124 randomised patients were excluded from the analysis set (2 withdrew consent, 2 did not meet inclusion criteria, 1 improved before intubation), resulting in a modified ITT population of 119; this introduces potential attrition/selection bias.
• Performance/detection bias: Open-label design; operators knew allocation, and co-interventions around induction/airway technique were not protocolised (risk of performance bias); primary endpoint (lowest SpO₂) is objective but still susceptible to procedural differences.
• Protocol adherence: 4-minute preoxygenation completed in 93.5% (HFNC) vs 94.7% (mask); failure to raise SpO₂ to 90% during preoxygenation occurred in 6.5% vs 3.5% (p=0.49).
• Baseline characteristics: Groups broadly similar for illness severity and oxygenation (SAPS II mean 54.5 vs 51.3; PaO₂/FiO₂ mean 120.2 vs 115.7 mmHg), but clinically important imbalances existed (age 64.9 vs 59.3 years; vasopressor support at inclusion 29% vs 19.3%; and “real emergency” intubations 19.3% vs 50.8%).
• Heterogeneity: Hypoxaemic ICU population (PaO₂/FiO₂ <300) with a high rate of peri-intubation complications; heterogeneity in urgency/operator experience and pre-intubation support likely diluted measurable treatment effect.
• Timing: Preoxygenation was standardised to 4 minutes; HFNC was continued through laryngoscopy until ventilator connection, aligning with the hypothesised “peroxygenation” mechanism.
• Dose: Clear separation in oxygen delivery strategy (HFNC 60 L/min vs face mask 15 L/min); however, achieved preoxygenation SpO₂ at end of preoxygenation was similar (97.1 ± 3.8 vs 96.3 ± 4.4).
• Separation of the variable of interest: HFNC remained in place during intubation whereas the face mask was removed after induction; despite this, nadir SpO₂ remained similar (91.5 [80–96] vs 89.5 [81–95]).
• Outcome assessment: Primary endpoint is physiologic and well-defined; secondary outcomes included objective ICU endpoints (ventilation duration, ICU length of stay, mortality).
• Statistical rigour: Sample size achieved as planned; however, powering assumed a large absolute difference (6%); multiple secondary outcomes were tested with no multiplicity control reported.

Conclusion on Internal Validity: Moderate—randomisation was centralised and the primary outcome objective, but open-label delivery, post-randomisation exclusions, and baseline imbalances (notably urgency of intubation) may have biased estimates and reduced sensitivity to detect true effects.

External Validity

• Population representativeness: Adult ICU patients with acute hypoxaemic respiratory failure (mean PaO₂/FiO₂ ~120 mmHg) reflect a common, high-risk real-world ICU intubation population.
• Important exclusions: Full exclusion list not reproduced in the extracted manuscript text (Not reported); generalisability to crash intubations and non-RSI contexts may be limited.
• Applicability: HFNC is widely available in high-income ICUs and feasible to deploy rapidly; effect may differ where non-invasive ventilation is routinely used as the preoxygenation comparator in severe hypoxaemia.
• Health-system transferability: Multicentre French ICU practice; findings likely translate to similar ICU environments, but operator mix, equipment, and airway management bundles may modify outcomes.

Conclusion on External Validity: Generally good for hypoxaemic ICU intubations in resourced settings, but limited for settings where NIV is standard preoxygenation in severe hypoxaemia or where immediate “crash” intubations predominate.

Strengths & Limitations

• Strengths: Pragmatic multicentre ICU randomised design; objective primary endpoint (nadir SpO₂); clear, protocolised oxygen delivery strategies with mechanistic separation (HFNC maintained during laryngoscopy vs mask removed); clinically relevant secondary outcomes.
• Limitations: Open-label delivery; modified ITT with post-randomisation exclusions; comparator was high FiO₂ face mask rather than NIV (potentially suboptimal in severe hypoxaemia); baseline imbalances in urgency and airway difficulty predictors; primary endpoint may not fully capture clinically meaningful “time-below-threshold” hypoxaemia burden.

Interpretation & Why It Matters

• Clinical implication

  HFNC (60 L/min) used as both preoxygenation and “peroxygenation” did not reduce nadir SpO₂ compared with a high FiO₂ face mask, in a cohort with substantial baseline hypoxaemia and high complication rates.
• Mechanistic inference

  Any apnoeic oxygenation benefit from HFNC was insufficient to overcome the dominant determinants of desaturation in ICU intubation (shunt physiology, limited oxygen reserve, haemodynamic instability) in this population.
• Practice relevance

  PREOXYFLOW helped recalibrate expectations: HFNC alone should not be assumed to prevent severe hypoxaemia during ICU intubation, particularly in moderate-to-severe hypoxaemic respiratory failure.

Controversies & Subsequent Evidence

• Comparator choice (mask vs NIV): Editorial critique emphasised that HFNC generates limited PEEP and may be an inadequate standalone strategy in severe hypoxaemia, arguing that NIV is a more biologically plausible comparator for maximising preoxygenation in shunt-dominant physiology.³
• Signal versus noise in secondary outcomes: The reduction in duration of mechanical ventilation (6 vs 10 days) occurred without statistically significant differences in ventilator-free days, ICU length of stay, or mortality, raising concern for chance findings amid multiple secondary comparisons.
• Non-randomised precursor evidence: A before–after ICU study suggested HFNC might reduce desaturation in mild-to-moderate hypoxaemia, but PREOXYFLOW did not reproduce a benefit in a more hypoxaemic cohort with high baseline complication rates.²
• Combination strategies (NIV + HFNC): The blinded single-centre OPTINIV trial tested NIV preoxygenation combined with HFNC apnoeic oxygenation (vs NIV alone) and demonstrated improved oxygenation during intubation, supporting the concept that positive-pressure preoxygenation and apnoeic oxygenation may be complementary rather than competing approaches.⁴
• NIV versus HFNC in hypoxaemic respiratory failure: In FLORALI-2, NIV preoxygenation did not reduce severe hypoxaemia overall compared with HFNC plus apnoeic oxygenation, but a pre-specified subgroup with PaO₂/FiO₂ ≤200 mmHg showed less severe hypoxaemia with NIV, reinforcing the likelihood of effect modification by baseline severity.⁵
• Pragmatic contemporary evidence favouring NIV versus oxygen mask: The multicentre PREOXI trial (ED/ICU) found lower hypoxaemia incidence with NIV preoxygenation compared with an oxygen mask, updating the evidence base towards positive-pressure preoxygenation when feasible.⁶
• Guideline synthesis: Major airway management guidance for critically ill adults emphasises primacy of oxygenation and supports escalation to non-invasive respiratory support strategies (including NIV and HFNC) tailored to physiology and operator expertise.⁷⁸⁹
• Evidence integration: Recent comparative syntheses (including network meta-analysis) increasingly frame preoxygenation as a strategy bundle (positioning, positive pressure, apnoeic oxygenation, and post-induction ventilation decisions) with likely effect modification by baseline hypoxaemia severity.¹⁰

Summary

• In hypoxaemic ICU adults (mean PaO₂/FiO₂ ~120 mmHg), HFNC (60 L/min) used through laryngoscopy did not improve nadir SpO₂ versus high FiO₂ face mask (91.5% vs 89.5%; p=0.44).
• Peri-intubation complications were frequent in both groups (any complication 58.1% vs 68.4%), with severe desaturation (SpO₂ <80%) occurring in ~23–26%.
• HFNC did not significantly change cardiovascular collapse, ICU length of stay, ventilator-free days at day 28, or 28-day mortality.
• A shorter duration of mechanical ventilation was observed with HFNC (6 vs 10 days; p=0.02), but this isolated secondary signal should be interpreted cautiously.
• Subsequent trials suggest that positive-pressure preoxygenation (NIV), and combination strategies (NIV + HFNC), may be more effective than HFNC alone in moderate-to-severe hypoxaemia.

Overall Takeaway

PREOXYFLOW is a landmark ICU airway trial because it rigorously tested a widely adopted physiologic idea—HFNC “peroxygenation” during intubation—and showed no reduction in nadir oxygen saturation or severe peri-intubation complications versus a high FiO₂ face mask. It shifted the field from assumption-driven adoption of HFNC alone towards a more physiology-matched approach, later supported by trials favouring NIV (and combination strategies) in patients with more severe hypoxaemia.

Bibliography

• 1Jaber S, Amraoui J, Lefrant JY, Arich C, Cohendy R, Landreau L, et al. Clinical practice and risk factors for immediate complications of endotracheal intubation in the intensive care unit: a prospective, multiple-center study. Crit Care Med. 2006;34(9):2355-2361. DOI
• 2Miguel-Montanes R, Hajage D, Messika J, Bertrand F, Gaudry S, Rafat C, et al. Use of high-flow nasal cannula oxygen therapy to prevent desaturation during tracheal intubation of intensive care patients with mild-to-moderate hypoxemia. Crit Care Med. 2015;43(3):574-583. DOI
• 3Demoule A, Rello J. High flow oxygen cannula: the other side of the moon. Intensive Care Med. 2015;41:1673-1675. DOI
• 4Jaber S, Monnin M, Girard M, Conseil M, Cisse M, Carr J, et al. Apnoeic oxygenation via high-flow nasal cannula oxygen combined with non-invasive ventilation preoxygenation for intubation in hypoxaemic patients in the intensive care unit: the single-centre, blinded, randomised controlled OPTINIV trial. Intensive Care Med. 2016;42(12):1877-1887. DOI
• 5Frat JP, Ricard JD, Quenot JP, Pichon N, Demoule A, Forel JM, et al. Non-invasive ventilation versus high-flow nasal cannula oxygen therapy with apnoeic oxygenation for preoxygenation before intubation of patients with acute hypoxaemic respiratory failure: a randomised, multicentre, open-label trial. Lancet Respir Med. 2019;7(4):303-312. DOI
• 6Gibbs KW, Semler MW, Driver BE, et al; PREOXI Investigators and the Pragmatic Critical Care Research Group. Noninvasive Ventilation for Preoxygenation during Emergency Intubation. N Engl J Med. 2024;390(23):2165-2177. DOI
• 7Higgs A, McGrath BA, Goddard C, Rangasami J, Suntharalingam G, Gale R, et al. Guidelines for the management of tracheal intubation in critically ill adults. Br J Anaesth. 2018;120(2):323-352. DOI
• 8Quintard H, l’Her E, Pottecher J, Adnet F, Constantin JM, De Jong A, et al. Experts’ guidelines of intubation and extubation of the ICU patient. Ann Intensive Care. 2019;9:13. DOI
• 9Acquisto NM, Mosier JM, Bittner EA, Cleveland KC, Diaconu CC, English D, et al. Rapid sequence intubation of the critically ill adult: Society of Critical Care Medicine clinical practice guidelines. Crit Care Med. 2023;51(10):1411-1430. DOI
• 10Pitre T, Orde S, Pechlivanoglou P, et al. Non-invasive respiratory support strategies for preoxygenation before emergency tracheal intubation: a network meta-analysis. Lancet Respir Med. 2025;13(7):593-605. DOI