Introduction: Evidence regarding the efficacy of high-flow nasal cannula (HFNC) oxygenation for preoxygenation and apnoeic oxygenation is conflicting. Our objective is to evaluate whether HFNC oxygenation for preoxygenation and apnoeic oxygenation maintains higher oxygen saturation (SpO2) during rapid sequence intubation (RSI) in ED patients compared to usual care.
Methods: This was a multicentre, open-label, randomised controlled trial in adult ED patients requiring RSI. Patients were randomly assigned 1:1 to either intervention (HFNC oxygenation at 60L/min) group or control (non-rebreather mask for preoxygenation and nasal prongs of at least 15L/min oxygen flow for apnoeic oxygenation) group. Primary outcome was lowest SpO2 during the first intubation attempt. Secondary outcomes included incidence of SpO2 falling below 90% and safe apnoea time.
Results: One hundred and ninety patients were included, with 97 in the intervention and 93 in the control group. Median lowest SpO2 during the first intubation attempt was 100% in both groups. Incidence of SpO2 falling below 90% was lower in the intervention group (15.5%) compared to the control group (22.6%) (adjusted relative risk=0.68, 95% confidence interval [CI] 0.37–1.25). Post hoc quantile regression analysis showed that the first quartile of lowest SpO2 during the first intubation attempt was greater by 5.46% (95% CI 1.48–9.45, P=0.007) in the intervention group.
Conclusion: Use of HFNC for preoxygenation and apnoeic oxygenation, when compared to usual care, did not improve lowest SpO2 during the first intubation attempt but may prolong safe apnoea time.
Critically ill patients in the emergency department (ED) have shorter safe apnoea times due to physiological distress from decreased cardiac output, increased shunting and reduced pulmonary reserves.1 Hypoxia is a commonly encountered adverse event during rapid sequence intubation (RSI)2 and is associated with cardiac arrest, neurological injury and death.3 Therefore, there is significant interest in prolonging safe apnoea time without hypoxia or bag-valve-mask (BVM) ventilation,4,5 as the latter may cause aspiration risk in unfasted ED patients, unlike for most patients in intensive care units (ICUs). Existing strategies for preoxygenation include the use of high-flow face mask (HFFM),6 such as the non-rebreather mask (NRM) or via well-fitting BVM without additional positive pressure for those with higher degree of respiratory compromise. Apnoeic oxygenation is commonly delivered through application of standard nasal cannula
(NC) at 15L/min.7-10
Systematic reviews of apnoeic oxygenation in both ICU and ED settings suggest reduction in incidence of hypoxaemia.5,11 However, these reviews included heterogenous studies with different apnoeic oxygenation strategies (high-flow nasal cannula [HFNC] versus standard NC at 15L/min) and have differing controls (no intervention, BVM, non-invasive ventilation [NIV] or no control at all).5,11 Using HFNC for preoxygenation and apnoeic oxygenation in RSI may offer potential advantages over standard NC. These include better patient tolerance and generation of small amounts of positive end-expiratory pressure.12 HFNC oxygenation during intubation has been studied in randomised trials,13,14 and observational studies in ICUs15,16 and EDs.17,18 However, its role and effectiveness are still actively debated,19-21 with inconsistent evidence regarding its efficacy compared to NIV,22,23 HFFM23,24 or BVM,14 and whether it is best utilised in conjunction with NIV.25 Importantly, there has been no randomised controlled trial in the ED that evaluates the use of HFNC for preoxygenation and apnoeic oxygenation in RSI. Little is known about its efficacy in preventing hypoxia in ED patients during the peri-intubation period.
In our trial, we aimed to evaluate if HFNC use at 60L/min for preoxygenation and apnoeic oxygenation will maintain higher oxygen saturation (SpO2) for ED patients during RSI, compared to the usual care of preoxygenation with NRM and standard NC with at least 15L/min of oxygen flow for apnoeic oxygenation.
Study design and setting
In this Pre- and Apnoeic high flow oxygenation for RApid sequence intubation in The Emergency department (Pre-AeRATE) study, an open-label, randomised, controlled trial, we enrolled adult patients who required RSI in EDs of the National University Hospital and Ng Teng Fong General Hospital in Singapore. Each ED receives over 110,000 attendances annually and performs 12–20 RSIs monthly. Pre-AeRATE was registered with ClinicalTrials.gov (NCT03396094). Ethics approval for waiver of consent at enrolment and delayed informed consent was obtained from the institutional ethics review board (DSRB reference number 2017/00348) in accordance with the Singapore regulations for clinical
trials under emergency situations.26
Subjects were randomised at 1:1 ratio into intervention and control groups, stratified by study site, with variable blocks of 4 and 6 via a web-based randomisation service generated by an independent statistician. Allocation concealment was maintained until completion of randomisation. Blinding of the ED team and patient was not possible due to the intervention. Clinicians in the admitting ICUs were blinded to the study allocation. The full protocol for this study was published elsewhere.27
Selection of participants
Inclusion criteria were patients aged ≥21 years requiring RSI due to any condition. The following were excluded: active “do-not-resuscitate” orders; crash, awake or delayed sequence intubations; requiring non-invasive positive pressure ventilation; cardiac arrest; suspicion or confirmed diagnosis of base of skull fractures or severe facial trauma that precluded placement of NC; pregnant women; and those incarcerated.
The intervention group using HFNC received 60L/min of warm and humidified oxygen at 37°C and fraction of inspired oxygen (FiO2) of more than 0.90 using the AIRVO 2 Humidifier with Integrated Flow Generator (Fisher & Paykel Healthcare Ltd, Auckland, New Zealand) during preoxygenation and apnoeic oxygenation phases. Control group was managed with usual care by preoxygenating using only NRM at flush rate, and then given at least 15L/min of non-humidified and nonheated oxygen from wall supply via NC for apnoeic oxygenation. Flush rate used for NRM preoxygenation reduces leak around the mask margins and is non-inferior to BVM, which is the other recommended modality.28
After ≥3 minutes of preoxygenation, induction medications were administered based on treating physician’s discretion, and apnoeic oxygenation commenced as per allocation. Intubation was attempted after 30–60 seconds, depending on the paralytic agent. End of intubation was defined as correct placement of endotracheal tube with confirmation using quantitative end-tidal carbon dioxide (ETCO2) monitoring.
Measurements and outcomes
Vital signs and airway features were assessed prior to preoxygenation. The primary endpoint was lowest SpO2 during the first intubation attempt, defined as time taken from administration of paralytic agent until quantitative ETCO2 was detected post-intubation if successful, or until the start of the second attempt if failed. The primary outcome analysis was restricted to the first intubation attempt as clinicians may deviate from initial treatment allocation based on their discretion after the first attempt. SpO2 was measured using the pulse oximeters, Philips Intellivue MP30 Patient Monitor (Royal Philips, Amsterdam, the Netherlands) and Zoll R Series defibrillator (Zoll Medical Corporation, Chelmsford, US) at 2 different areas in the upper extremities. A research coordinator, nurse or clinician not involved in the intubation recorded the lowest SpO2 and other variables collected during the attempts. Patients were monitored for peri-intubation adverse events (AEs) such as aspiration, arrhythmia and cardiac arrest during intubation or within 5 minutes after intubation. Main secondary outcomes were incidence of SpO2 falling below 90% and safe apnoea time during intubation (duration of apnoea where SpO2 remains ≥90% and censored at the time of successful intubation). Other secondary outcomes included number of intubation attempts, time (from induction) to successful intubation, peri-intubation AE, and various post-intubation clinical outcomes.
Sample size calculation
Based on our preliminary data (unpublished) and a previous study,24 we anticipated a standard deviation of 14% in the lowest SpO2. Enrolment of 184 patients (92 patients in each control and intervention groups) would provide statistical power of 80% (two-sided α of 0.05) to detect a 6% difference in lowest SpO2,15 allowing for a 5% dropout.
Analyses of baseline and efficacy data were performed with the intention-to-treat (ITT) population, stratified or adjusted for study site for outcome variables. Supplementary
analyses of efficacy data were performed with the perprotocol (PP) population, comprising randomised patients who received FiO2 of at least 0.70 during preoxygenation and apnoeic oxygenation with no major protocol deviation. This FiO2 value was chosen as it is the minimum acceptable level to be comparable with that delivered by NRM.29 Frequency and proportion of patients with any peri-intubation AE were summarised in the “as-treated” population, that is, according to actual approaches of preoxygenation and apnoeic oxygenation received.
Since distribution of lowest SpO2 was highly skewed as observed in similar studies,,15,24 a stratified Mann-Whitney U test (namely the van Elteren test) was used to compare lowest SpO2 between treatment groups. Post hoc analysis with quantile regression of lowest SpO2 was performed with adjustment for covariates, selected backward stepwise into the final adjusted model.30
The Cochran-Mantel-Haenszel test was used to compare risk of SpO2 below 90% between groups. The common relative risk (RR) was estimated with its 95% confidence interval (CI) by the Mantel-Haenszel method. Safe apnoea time during intubation and time to successful intubation were compared using the stratified log-rank test. The hazard ratios (HR) of SpO2 falling below 90% and successful intubation were estimated from the Cox proportional hazards model. All analyses were performed using SAS 9.4 (SAS Institute, Cary, US) and P<0.05 indicated statistical significance.
From May 2018 to December 2019, a total of 518 patients were screened and 192 eligible patients were randomised, with 97 assigned to intervention group and 95 to control group. All intervention patients and 93 patients in the control group were included in ITT analysis (Fig. 1). In the control group, 2 patients were excluded from data analysis as 1 had a “do-not-resuscitate” order established after randomisation and the other was intubated in the operating room.
Fig 1. Flowchart illustrating patient enrolment, randomisation and treatment allocation.
ITT: intention-to-treat; NUH: National University Hospital; NTFGH: Ng Teng Fong General Hospital
*Refers to patients who were unconscious and apnoeic
Characteristics of study subjects
Overall, there was a predominance of males (124/190, 65.3%) with a mean age of 61.1 years (standard deviation [SD] 15.1) and mean estimated weight of 64.6kg (SD 14.0) (Table 1 and online Supplementary Table 1). The top 3 indications for intubation were non-traumatic intracranial, subarachnoid and subdural haemorrhages (57/190, 30.0%); shock states (28/190, 14.7%); and seizures (23/190, 12.1%).
Most patients were assessed to have no potential airway difficulty (121/190, 63.7%). A great proportion (176/190, 92.6%) were intubated using the C-MAC (Karl Storz, Tuttlingen, Germany) video laryngoscope; 20 patients (10.6%) had Grade 3 or 4 laryngeal view by the Cormack-Lehane classification. Most received NRM before randomisation (intervention 52/97 [53.6%] and control 44/93 [47.8%]; see online Supplementary Table 2) as part of prehospital care. Full data relating to the study procedure and treatment are summarised in the online Supplementary Tables 2 and 3.
Lowest SpO2 was maintained at 100% during the first intubation attempt in more than half of the patients in both groups. There was no statistical difference in the lowest SpO2 recorded during the first intubation attempt between the intervention (median SpO2 100%, interquartile range [IQR] 96.0–100) and control groups (median SpO2 100%, IQR 91.0–100) (Table 2 and Fig. 2). Comparison of lowest SpO2 between the 2 groups using the stratified Mann-Whitney U test provided P values of 0.138 and 0.061 in ITT and PP populations, respectively (Table 2). A post hoc quantile regression analysis of the first quartile of lowest SpO2 (adjusted for indication for intubation, potential airway difficulty and baseline SpO2) estimated a difference of 5.5% (95% CI 1.5–9.5, P=0.007) in first quartile between the 2 groups (online Supplementary Table 4). Outcomes were comparable between groups with respect to number of intubation attempts and time to successful intubation. Most patients were successfully intubated at first attempt (intervention group 80/97 [82.5%] versus control group 78/93 [83.9%]). Median time from induction to successful intubation was 3.0 minutes in the intervention group and 3.5 minutes in the control group (Table 2). Treatment in intervention group reduced the risk of SpO2 falling below 90% by at least 30% during the first intubation attempt and any attempt (adjusted RRs 0.55, P=0.084; 0.52, P=0.057 in first and any attempt in the PP population; and adjusted RRs 0.68, P=0.213; 0.65, P=0.156 in first and any attempt in the ITT population). Median safe apnoea time was prolonged by 3.0 minutes in the intervention group (10 minutes) compared to the control group (7 minutes), with HR of SpO2 falling below 90% = 0.57 (95% CI 0.28–1.12, P=0.104).
Incidences of ventilator-associated pneumonia and aspiration pneumonia were comparable between both groups (22.6% vs 22.7%). The treatment in the intervention group did not affect other clinical outcomes (Table 3). Seventeen (18.5%) patients in the intervention group and 13 (13.3%) patients in the control group experienced peri-intubation AE (Table 3).
In our cohort, SpO2 was maintained at 100% in more than 50% of patients in both groups. Use of HFNC compared with routine care did not show any statistically significant difference in oxygenation during the first and subsequent intubation attempts. These observations could be due to the result of better physiological reserves in our cohort, which largely comprised patients with neurological emergencies. However, in our post hoc analysis, patients having lowest SpO2 of below 95% with usual care were expected to have their lowest SpO2 improved by an average of 5% if they were managed with HFNC (P=0.032 from Mann-Whitney U test in this subgroup). This is clinically important as the oxygen dissociation curve has a steep gradient below 90% and patients can reach critical hypoxic state in a span of seconds.31 After adjustment for significant covariates using quantile regression analysis, the first quartile of lowest SpO2 achieved during the first intubation attempt was greater by 5.5% (95% CI 1.5–9.5, P=0.007) in the intervention group compared to the control group. This is consistent with a recent network meta-analysis, which showed that HFNC or NIV was better than conventional oxygen therapy for preoxygenation in ICU, at least in terms of lowest SpO2 achieved.32 Our study results add to the evidence by examining peri-intubation HFNC in ED patients.
An acceptable alternative strategy to optimise SpO2 for intubation is through positive pressure ventilation using NIV.4 Although NIV may be superior to conventional oxygenation therapy in preoxygenation, its use may be limited if patients are obtunded, have emesis or are unable to tolerate face masks.22,33 Unlike NIV, HFNC allows apnoeic oxygenation during intubation without additional maneuvers. Studies comparing HFNC versus NIV have yielded inconsistent results. In the largest study to-date (n=313), neither modality reduced risk of severe hypoxia during intubation,34 though there seems to be potential synergistic benefit in combining NIV with HFNC.25 Nevertheless, combining the use of both NIV and HFNC has its problems of inadequate mask seal due to obstruction from the HFNC nasal cannula, as well as increase in procedural complexity and healthcare costs due to using 2 oxygen delivery devices. This may not be adequate given current dearth in evidence for this oxygenation method.
Our study showed that HFNC was safe compared to our usual care. The intervention patients had lower median pre-intubation Glasgow Coma Scale (7 [IQR 6–14]) compared to controls (10 [IQR 6–14]). However, we did not observe significant increase in aspiration risk and other adverse events. The safety of HFNC use in RSI was replicated in other existing literature.14 Pending availability of more evidence, HFNC can be considered as a useful alternative in selected patients with contraindications to NIV, given its safety profile and better tolerance.
Apart from optimising preoxygenation, apnoeic oxygenation is a strategy that can prevent peri-intubation desaturation. Physiologically during apnoea, the much higher solubility of carbon dioxide in blood compared to oxygen allows oxygen to move more readily from alveoli into the bloodstream. This creates a sub-atmospheric alveolar pressure, allowing passive flow of oxygen from the pharynx to alveoli. Despite this rationale, previous randomised controlled trials in the ED and ICU did not demonstrate significant benefit in improving lowest SpO2 during intubation.<8,35 These negative findings could be a result of study procedural limitations. In both studies, oxygen flow at 15L/min was delivered through standard NC that are not designed for such flows. Additionally, a flow of at least 30L/min is necessary before positive airway pressures could be generated.36 By using an HFNC device that can deliver up to 60L/min of high-flow oxygenation and that is specifically designed for this purpose, we were able to overcome these limitations. We showed a lower risk of SpO2 falling below 90% in the intervention group at the first and any attempt, and safe apnoea time was prolonged by at least 3 minutes (Table 1). Despite lack of statistical significance, this could be clinically significant for 2 reasons. First, an SpO2 of 90% represents the beginning of the steep down-sloping gradient of the oxygen dissociation curve, whereby small changes in the arterial partial pressure of oxygen (PaO2) result in large changes in haemoglobin oxygen binding capacity.37 Second, an ill average-sized adult could reach an SpO2 of less than 60% in under 2 minutes once SpO2 drops to 90%.31 Hence, prolonging safe apnoea time is of utmost importance.
Pre-AeRATE has several strengths. First, this is the only randomised controlled trial to-date evaluating HFNC use for RSI in ED, in Singapore. Second, to our knowledge, this is also the largest trial (190 patients analysed) comparing HFNC with standard oxygen therapy during the peri-intubation period in any clinical setting, in Singapore.13,14,24,38,39 Third, we conducted a post hoc analysis employing a quantile regression analysis as an endpoint due to the skewed distribution in such studies and in our cohort (Fig. 2), where a minority would experience much greater desaturation.40 This allowed us to generate a hypothesis on the usefulness of HFNC on these patients in extremis. Despite similar decrease in median SpO2 between the intervention and control arms, safe apnoeic time was noted to be prolonged by HFNC use. Fourth, this pragmatic trial reflects real-world conditions, and shows the plausibility of using HFNC for a time-sensitive procedure in ED and in patients with high aspiration risk, such as trauma with bloody airways and depressed consciousness.
Fig. 2. Lowest SpO2 during first intubation attempt.
SpO2: oxygen saturation; Q1: 25th percentile; Q3: 75th percentile
This trial has its limitations. First, the time-sensitive nature of RSI precluded measurements of PaO2 in our protocol, thereby rendering stratification of severity based on hypoxaemia impossible. SpO2 was used as an endpoint instead, as it is a ubiquitous non-invasive clinical parameter that allows better external validity. Second, it was impracticable to blind patients and ED clinical staff due to the nature of the intervention. However, objective measurements of SpO2 by 2 separate devices and blinding of the ICU teams to study allocation made biasing of results unlikely. Third, consecutive recruitment was not possible due to clinicians focusing on resuscitation in the fast-paced ED environment.
Fourth, although we included patients requiring RSI due to any indication, our eventual cohort comprised 42.1% of neurological conditions (intracranial haemorrhage and seizures) and may under-represent those with predominantly cardiorespiratory compromise, who could provide greater physiological insight. As this trial was designed to be pragmatic, we did not specify strict inclusion criteria due to indications for intubation. The observation of more than 50% of patients in each treatment group maintaining SpO2 at 100% during the first intubation attempt may be attributed to better cardiorespiratory reserves and low risk of desaturation in patients with predominant neurological conditions. By performing post hoc regression analysis of the first quartile of lowest SpO2, we were able to evaluate the treatment effect on patients in physiological extremis who had higher desaturation risk (compared to the general study population). Although the intervention group comprised a slightly higher proportion of neurological conditions by chance, this difference was not statistically significant. Lastly, given our primary endpoint was to evaluate the lowest SpO2 achieved, the study was not powered to make meaningful conclusions regarding patient-oriented outcomes such as mortality or length of ICU stay. Future studies should focus on such endpoints, other long-term outcomes and evaluate the cost-effectiveness of using HFNC during the peri-intubation period.
We found that HFNC use for preoxygenation and apnoeic oxygenation, when compared to usual care, did not show improvement in median lowest SpO2 achieved during the first intubation attempt. However, such HFNC use may prolong safe apnoea time. Our study showed that patients without neurological indications for intubation were likely to desaturate faster, or have a more challenging intubation, and may benefit from the
longer apnoea time that HFNC provides.
Dr Mui Teng Chua received the Clinician Scientist Individual Research Grant, New Investigator Grant from National Medical Research Council, Singapore for the conduct of this investigator-initiated trial. The grant was paid directly to the institution for conduct of this trial. The study funder has no role or authority in study design; collection, management, analysis, and interpretation of data; writing of the report; or the decision to submit the report for publication. Dr Matthew Edward Cove reports consultancy fees from Baxter and Medtronic paid directly to the National University Hospital, Singapore. Fisher & Paykel Healthcare did not participate in the design of the study or the decision to submit this article for publication.
This study is funded by the Clinician Scientist Individual Research Grant, New Investigator Grant from National Medical Research Council, Singapore. The study funder has no role or authority in the study design; collection, management, analysis and interpretation of data; writing of the report; or the decision to submit the report for publication.
ClinicalTrials.gov (NCT03396094). Registered on 10 January 2018.
- Law JA, Broemling N, Cooper RM, et al. The difficult airway with recommendations for management – Part 2 – The anticipated difficult airway. Can J Anaesth 2013;60:1119-38.
2. Chan GWH, Chai CY, Teo JSY, et al. Emergency airway management in a Singapore centre: A registry study. Ann Acad Med Singap 2021;50:42-51.
3. Mort TC. The incidence and risk factors for cardiac arrest during emergency tracheal intubation: a justification for incorporating the ASA guidelines in the remote location. J Clin Anesth 2004;16:508-16.
4. Weingart SD, Levitan RM. Preoxygenation and prevention of desaturation during emergency airway management. Ann Emerg Med 2012;59:165-75.e1.
5. Oliveira JE Silva L, Cabrera D, Barrionuevo P, et al. Effectiveness of apneic oxygenation during intubation: A systematic review and meta-analysis. Ann Emerg Med 2017;70:483-94.e11.
6. Gleason JM, Christian BR, Barton ED. Nasal cannula apneic oxygenation prevents desaturation during endotracheal intubation: An integrative literature review. West J Emerg Med 2018;19:403–11.
7. Thorpe CM, Gauntlett IS. Arterial oxygen saturation during induction of anaesthesia. Anaesthesia 1990;45:1012-5.
8. Semler MW, Janz DR, Lentz RJ, et al. Randomized trial of apneic oxygenation during endotracheal intubation of the critically ill. Am J Respir Crti Care Med 2016;193:273-80.
9. Ramachandran SK, Cosnowski A, Shanks A, et al. Apneic oxygenation during prolonged laryngoscopy in obese patients: a randomized, controlled trial of nasal oxygen administration. J Clin Anesth 2010;22:164-8.
10. Sakles JC, Mosier JM, Patanwala AE, et al. Apneic oxygenation is associated with a reduction in the incidence of hypoxemia during the RSI of patients with intracranial hemorrhage in the emergency department. Intern Emerg Med 2016;11:983-92.
11. Pavlov I, Medrano S, Weingart S. Apneic oxygenation reduces the incidence of hypoxemia during emergency intubation: a systematic review and meta-analysis. Am J Emerg Med 2017;35:1184-9.
12. Groves N, Tobin A. High flow nasal oxygen generates positive airway pressure in adult volunteers. Aust Crit Care 2007;20:126-31.
13. Simon M, Wachs C, Braune S, et al. High-flow nasal cannula versus bag-valve-mask for preoxygenation before intubation in subjects with hypoxemic respiratory failure. Respir Care 2016;61:1160-7.
14. Guitton C, Ehrmann S, Volteau C, et al. Nasal high-flow preoxygenation for endotracheal intubation in the critically ill patient: a randomized clinical trial. Intensive Care Med 2019;45:447-58.
15. Miguel-Montanes R, Hajage D, Messika J, et al. Use of highflow nasal cannula oxygen therapy to prevent desaturation during tracheal intubation of intensive care patients with mild-to-moderate hypoxemia. Crit Care Med 2015;43:574-83.
16. Besnier E, Guernon K, Bubenheim M, et al. Pre-oxygenation with high-flow nasal cannula oxygen therapy and non-invasive ventilation for intubation in the intensive care unit. Intensive Care Med 2016;42:1291-2.
17. Doyle AJ, Stolady D, Mariyaselvam M, et al. Preoxygenation and apneic oxygenation using Transnasal Humidified Rapid-Insufflation Ventilatory Exchange for emergency intubation. J Crit Care 2016;36:8-12.
18. Kim TH, Hwang SO, Cha YS, et al. The utility of noninvasive nasal positive pressure ventilators for optimizing oxygenation during rapid sequence intubation. Am J Emerg Med 2016;34:1627-30.
19. Ricard JD, Gregoretti C. Nasal high-flow preoxygenation for endotracheal intubation in the critically ill patient? Pro. Intensive Care Med 2019;45:529-31.
20. Chanques G, Jaber S. Nasal high-flow preoxygenation for endotracheal intubation in the critically ill patient? Maybe. Intensive Care Med 2019;45:532-4.
21. Hanouz JL, Gérard JL, Fischer MO. Nasal high-flow preoxygenation for endotracheal intubation in the critically ill patient? Con. Intensive Care Med 2019;45:526-8.
22. Vourc’h M, Baud G, Feuillet F, et al. High-flow nasal cannulae versus non-invasive ventilation for preoxygenation of obese patients: the PREOPTIPOP randomized trial. EClinicalMedicine 2019;13:112–9.
23. Bailly A, Ricard JD, Le Thuaut A, et al. Compared efficacy of four preoxygenation methods for intubation in the ICU: retrospective analysis of McGrath Mac videolaryngoscope versus Macintosh laryngoscope (MACMAN) trial data. Crit Care Med 2019;47:e340-8.
24. Vourc’h M, Asfar P, Volteau C, et al. High-flow nasal cannula oxygen during endotracheal intubation in hypoxemic patients: a randomized controlled clinical trial. Intensive Care Med 2015;41:1538-48.
25. Jaber S, Monnin M, Girard M, 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;42:1877-87.
26. Munn MW. The Legislative Framework Governing Clinical Trials in Singapore. Asia Pacific Biotech News 2006;10:1210-5. Available at: http://www.asiabiotech.com/10/1021/1210_1215.pdf. Accessed on 23 May 2018.
27. Chua MT, Khan FA, Ng WM, et al. Pre- and Apnoeic high flow oxygenation for RApid sequence intubation in the Emergency department (Pre-AeRATE): Study protocol for a multicentre,
randomised controlled trial. Trials 2019;20:1-9.
28. Driver BE, Prekker ME, Kornas RL, et al. Flush rate oxygen for emergency airway preoxygenation. Ann Emerg Med 2017;69:1-6.
29. Pruitt W, Jacobs M. Breathing lessons: basics of oxygen therapy. Nursing 2003;33:43-5.
30. Koenker RW, Hallock KF. Quantile regression: An introduction. J Econ Perspect 2001;15:143-56.
31. Benumof JL, Dagg R, Benumof R. Critical hemoglobin desaturation will occur before return to unparalyzed state from 1 mg/kg succinylcholine. Anesthesiology 1997;87:979-82.
32. Fong KM, Au SY, Ng GWY. Preoxygenation before intubation in adult patients with acute hypoxemic respiratory failure: a network meta-analysis of randomized trials. Crit Care 2019;23:1-12.
33. Chua MT, Kuan WS. The use of high-flow nasal cannula in acute decompensated heart failure: ready for prime time yet? J Emerg Crit Care Med 2017;1:22.
34. Frat JP, Ricard JD, Quenot JP, 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;7:303-12.
35. Caputo N, Azan B, Domingues R, et al. Emergency department use of apneic oxygenation versus usual care during rapid sequence intubation: a randomized controlled trial (The ENDAO Trial). Acad Emerg Med 2017;24:1387-94.
36. Ritchie JE, Williams AB, Gerard C, et al. Evaluation of a humidified nasal high-flow oxygen system, using oxygraphy, capnography and measurement of upper airway pressures. Anaesth Intensive Care 2011;39:1103-10.
37. Weingart SD. Preoxygenation, reoxygenation, and delayed sequence intubation in the emergency department. J Emerg Med 2011;40:661-7.
38. Mir F, Patel A, Iqbal R, et al. A randomised controlled trial comparing transnasal humidified rapid insufflation ventilatory exchange (THRIVE) pre-oxygenation with facemask preoxygenation in patients undergoing rapid sequence induction of anaesthesia. Anaesthesia 2017;72:439-43.
39. Lodenius, Piehl J, Östlund A, et al. Transnasal humidified rapid-insufflation ventilatory exchange (THRIVE) vs. facemask breathing pre-oxygenation for rapid sequence induction in adults: a prospective randomised non-blinded clinical trial. Anaesthesia 2018;73:564-71.
40. Yu K, Lu Z, Stander J. Quantile regression: applications and current research areas. J R Stat Soc Series B Stat Methodol 2003;52:331-50.