Accepted Manuscript Apneic oxygenation reduces the incidence of hypoxemia during emergency intubation: A systematic review and meta-analysis

Ivan Pavlov, Sofia Medrano, Scott Weingart PII: DOI: Reference:

S0735-6757(17)30473-4 doi: 10.1016/j.ajem.2017.06.029 YAJEM 56757

To appear in: Received date: Revised date: Accepted date:

19 April 2017 ###REVISEDDATE### 13 June 2017

Please cite this article as: Ivan Pavlov, Sofia Medrano, Scott Weingart , Apneic oxygenation reduces the incidence of hypoxemia during emergency intubation: A systematic review and meta-analysis, (2017), doi: 10.1016/j.ajem.2017.06.029

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Title: Apneic Oxygenation Reduces the Incidence of Hypoxemia during Emergency Intubation: a Systematic Review and Meta-Analysis Running head: Apneic Oxygenation reduces hypoxemia

Authors and affiliations:

IP

T

Ivan Pavlov, M.D Department of Emergency Medicine Hôpital de Verdun

CR

Montréal (Québec) Canada Corresponding author: [email protected]

US

Sofia Medrano, M.D. Department of Biomedical Sciences Université de Montréal

M

ED

Scott Weingart, M.D. Division of Emergency Critical Care Stony Brook Hospital Stony Brook, NY, USA

AN

Montréal (Québec) Canada

PT

Keywords: airway management; oxygenation; apneic oxygenation, rapid sequence intubation; crash intubation.

CE

Meetings: This work has not been presented in any meetings, nor previously published in any form Funding: This work has not been supported by any grant or any other source of funding.

AC

Conflits of interest: None for all three authors. Word count: 2084 words Author contributions: IP conceived the study. Data were extracted and summarized by SM and IP. IP analyzed the data, and drafted the manuscript. All authors contributed substantially to its revision. IP takes responsibility for the paper as a whole.

ACCEPTED MANUSCRIPT Apneic Oxygenation Reduces the Incidence of Hypoxemia during Emergency Intubation: a Systematic Review and Meta-Analysis Introduction

T

Background

US

CR

IP

Emergency tracheal intubation is a lifesaving and time-sensitive procedure commonly performed on critically ill patients in the emergency department (ED), in intensive care units (ICU) and during prehospital transport. Failed intubations are up to 40 times more likely in the ED, as compared to elective intubations in the operating room[1], and hypoxemia is a common occurrence. At least one patient out of three experiences desaturation to less than 90%[2–4], and those patients are at risk for more severe hypoxemia and cardiac arrest.

M

AN

Critically ill patients have collapsed alveoli, high alveolar-arterial gradient and high metabolic requirements[5]. Their blood may arrive to the lungs poorly oxygenated, and a sizeable portion of it may pass directly through shunted alveoli without any reoxygenation. The high incidence of hypoxemia is thus hardly surprising, especially during the critical apneic phase that occurs between induction and the successful passing of the tracheal tube.

ED

Although a slight desaturation to 90% is not harmful, it represents the first step of the very steep slippery slope of the oxygen dissociation curve of hemoglobin[6,7].

PT

Clinically significant hypoxemia can thus be defined as any fall of SpO2 to a level at which any further drops of arterial partial pressure of oxygen will lead to near-exponential drops in SpO2. This is conventionally defined as SpO2<90[2,8–10], although the inflection point occurs more precisely at SpO2 93[4], which has led some authors to prefer this higher cut-off[11].

CE

Apneic Oxygenation

AC

During apnea, the differential rate between alveolar oxygen absorption and carbon dioxide excretion generates a negative pressure gradient, which results in aventilatory mass flow of gas from the upper respiratory tract into the lungs[12]. As long as the airway remains patent, any supplemental oxygen administrated through the nares will thus be delivered to the alveoli without any ventilation; this technique has been called apneic oxygenation (Ap-Ox). In healthy surgical patients, Ap-Ox can assure sufficient oxygenation for up to an hour, though it cannot prevent the progressive rise of PaCO2[13]. Ap-Ox has demonstrated a significant reduction in the incidence of clinically significant hypoxemia during elective intubation (including obese patients)[14–16], bronchoscopy[17] and pan-endoscopy[18]. However, there has not been a clear signal for intubations in the emergency department or the intensive care unit. Semler et al., in the FELLOW trial, were not able to demonstrate a benefit to

ACCEPTED MANUSCRIPT Ap-Ox in the intensive care unit[2], although it has been argued that the trial was severely underpowered and the control group was predominantly receiving positive-pressure breaths during the apneic period[19]. Despite animal models showing that shunted lungs do not preclude the benefit of Ap-Ox[20], the possibility remains that because of collapsed airways, pulmonary shunting and impaired membrane exchange, the sickest patients are the ones who cannot benefit from Ap-Ox, unless it is provided along with continuous positive airway pressure[21,22].

IP

T

Goals of This Investigation

CR

We conducted a systematic review and meta-analysis to determine the effect of Ap-Ox on the incidence of clinically significant hypoxemia during emergent endotracheal intubation.

US

Materials and Methods

AN

The details of the protocol for this systematic review were registered on PROSPERO (42017055771), and can be accessed at [http://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42017055771]. The metaanalysis results are reported in conformity with the PRISMA[23] and MOOSE[24] guidelines.

PT

ED

M

There's a paucity of studies of Ap-OX in the critical care literature, and we opted to include every study of Ap-Ox performed in an adult critical care setting (i.e. emergency department, intensive care unit, prehospital emergency transport), including studies where both preoxygenation and apneic oxygenation strategies were compared to standard care (i.e. no Ap-Ox). The only exclusion criterion was the absence of a control group. We did not exclude observational studies, given that their inclusion is statistically sound and would increase the power of the meta-analysis[25].

AC

CE

We searched the MEDLINE, EMBASE and Pubmed databases from inception to January 2017, without any restrictions of language. Details of the search strategy are available in the Supplementary Appendix, Fig E1. A search of Scopus, Google Scholar, ClinicalTrials.gov, as well as backward reference searching did not yield any supplementary references. Two reviewers (I.P. and S.M.) independently parsed through the abstracts of all articles identified by the initial search, and produced a list of potentially relevant articles. The full texts of these articles were then reviewed and a final list of studies to include in the meta-analysis was produced. No disagreement between the reviewers arose during this process. Data from the articles was abstracted independently by I.P. and S.M. The information included trial name, year of publication, details about trial design, number of patients in each study arm, the nature of the intervention and control therapies, main outcomes, including adverse effects and all stated and inferred limitations. Disagreements were resolved through consensus. The corresponding authors of six studies were contacted to obtain further clarification or supplemental data[2,10,11,21,26,27].

ACCEPTED MANUSCRIPT The risk of bias in randomized controlled trials was assessed with the Cochrane Collaboration's tool for assessing risk of bias[28] and with the RoBANS[29] for observational studies. Because less than 10 studies were included in the meta-analysis, we did not perform funnel plot testing to assess for reporting bias[28].

Data Analysis

IP

T

Since the main purpose of Ap-Ox is to reduce the incidence of hypoxemia, we chose the incidence of clinically significant hypoxemia as the main outcome of interest. The precise definition could vary between trials, as long as it corresponded to SpO 2 between 90 and 93%.

CR

Pre-planned secondary outcomes included: the incidence of increasingly severe levels of hypoxemia (SpO2<90%, SpO2<80%, SpO2<70%), and the incidence of death of any cause during the 28 days following endotracheal intubation.

AN

US

All the outcomes were computed in terms of relative risk, and according to both an intentionto-treat and per-protocol allocations, wherever applicable. Unless there was significant discrepancy between intention-to-treat and per protocol analyses, we reported only the results of the generally more conservative intention-to-treat analysis. We also planned a meta-analysis restricted to low bias studies, as reported in Tables E1 and E2.

M

The purpose of subgroup, secondary analyses was to ensure that if any effect of Ap-Ox were to be demonstrated on the primary outcome, it would remain quantitatively robust across all subgroup analyses, and would appear in both low and high quality studies[30].

CE

PT

ED

Outcomes were pooled with both fixed and random effects meta-analyses, using the Mantel– Haenszel method[31]. In all trials, the same intervention (Ap-Ox) was applied to similar populations (patients requiring emergency endotracheal intubation). Although the precise technique of Ap-Ox was different across trials, we expected a similar effect size of Ap-Ox accross all studies. Thus, we did not expect significant discrepancy between the results of fixed and random effect models. Heterogeneity was expressed as the I2 statistic.

AC

Although the median [2,21,32] or mean [26] lowest SpO2 during intubation was reported as the main outcome in all four randomized controlled trials, we decided against computing this descriptive statistic, for two important reasons. First, more than half of patients do not experience any hypoxemia during intubation [2–4], and the median SpO2 fails to inform us on the fate of precisely those sickest patients who could benefit from Ap-Ox[19]. Second, simple descriptive statistics, although useful when the data follow a normal distribution can easily become meaningless and, indeed, misleading, when distribution is skewed[33], which is indeed the case in all studies of Ap-Ox. In other words, it is more important to know the proportion of patients who experience hypoxemia during intubation, than the median or mean SpO2 of the whole group. All analyses were performed in R version 3.2.3[34], with the help of meta package[35].

ACCEPTED MANUSCRIPT Results

IP

T

After removal of duplicate records, the search strategy yielded a total of 369 references, 10 of which were reviewed in full (Fig 1). One of these studies compared two different strategies of Ap-Ox, without control group, and was therefore discarded[36]. The results of the nine other studies, including a total of 1953 patients, are summarized in Table 1. Four studies were randomized controlled trials and five were observational. One study[37] reported on a subset population of the second, larger, study[10] and was not included in the meta-analysis. Finally, one study[32] did not report the incidence of clinically significant hypoxemia and was used only for subgroup analyses, including the incidence of death.

CR

As reported in tables E1 and E2, six studies were judged to be at low risk of bias. The three others were at higher risk of bias, either because of lack of blinding and allocation concealment, or the lack of adjustment for confounding variables in observational studies.

AN

US

In three studies, Ap-Ox was administered by means of a high flow nasal cannula set at 50 to 60L/min, FiO2 100%[21,26,32]. All other studies used a conventional nasal cannula set at various flow rates (between 5 and > 15L/min in one study[10] and at 15L/min in the four remaining ones[2,11,27,38]).

ED

M

The incidence of clinically significant hypoxemia was assessed in seven studies[2,10,11,21,26,27,38], as summarized in Fig 2. It was defined as SpO2<93% in one study[11] and as SpO2<90 in the six remaining studies. The pooled absolute risk of clinically significant hypoxemia was 27.6% in the usual care group and 19.1% in Ap-Ox group. Ap-Ox reduced the relative risk of hypoxemia by 30% (95% confidence interval 0.59 to 0.82).

PT

This finding remained robust in sub-group analyses, with similar size effect across all levels of hypoxemia (Fig E2, E3 and E4 in Online Supplement). The benefit of Ap-Ox appeared as robust in analyses restricted to only higher quality studies (Fig E5, E6 and E7 in Online Supplement).

AC

CE

The outcome of death was available from three studies[2,21,32], and a pooled analysis (Fig 3) revealed an unexpected benefit which almost attained statistical significance. Fifty-seven (34.8%) patients treated with Ap-Ox died, compared to 69 (44.8%) treated with standard care (relative risk of death 0.77; 95% confidence interval 0.59 to 1.02).

Limitations

As summarized in Table 1, the studies included in this meta-analysis differed in the precise strategies of Ap-Ox (standard cannula vs high-flow nasal cannula) and in the reported main outcomes. Although we were able to obtain the data regarding our main outcome of interest, the incidence of clinically significant hypoxemia, from 7 studies out of 8, pooled analyses of secondary outcomes were less inclusive, with the outcome of death studied in only three studies. Three studies[21,26,32] evaluated different pre-oxygenation and Ap-Ox strategies. Although we limited our analyses of these studies to the period of Ap-Ox, it is obvious that differences in pre-

ACCEPTED MANUSCRIPT oxygenation could have influenced the outcomes. However, in those three studies, the differences in oxygenation appeared only during the apneic period, which suggest that despite different pre-oxygenation strategies, a similar level of oxygenation were achieved in both intervention and control groups.

T

More importantly, the standard care was contaminated with ventilation[2] or some form of ApOx[26,38]] in three studies. For instance, in the FELLOW trial[2], 77% of patients in the control group were ventilated just before and in-between laryngoscopy attempts. Any benefit of Ap-Ox will be diluted in a population that barely experiences apnea during the course of intubation.

CR

IP

Despite these limitations, we did not find any significant heterogeneity between studies; this is reassuring as to the overall reliability of our reported outcomes.

Discussion

AN

US

We found that Ap-Ox reduces the incidence of clinically significant hypoxemia by 30% during emergency intubation in critically ill patients. The effect size of Ap-Ox was consistent in all subgroup analyses, and across progressively more severe levels of hypoxemia. The findings of each individual study, including the negative FELLOW trial[2], were consistent with this reported benefit of Ap-Ox.

ED

M

The reported efficacy of Ap-Ox in the prevention of hypoxemia is in accord with the results of studies done in the operating room[13–15], is confirmed by animal models of respiratory failure[20], and is consistent with basic physiology[8,9,12].

PT

The results of this meta-analysis are thus robust and should inform the clinical practice of emergency intubation. A protocol for clinical use has been previously published and its recommendations are consistent with the findings of this meta-analysis[9].

CE

We did not expect to find a reduction of mortality with the use of Ap-Ox. This finding was a non-statistically significant trend and should be interpreted as preliminary and hypothesisgenerating, though the confidence interval is bordering on conventional statistical significance. It certainly merits further study.

AC

If this finding is true, we can hypothesize two possible mechanisms. First, Ap-Ox is associated with a higher rate of successful intubation on the first attempt[10], and this has been clearly associated with a reduced risk of adverse effects, such as critical desaturation, or aspiration due to repeated use of bag-valve-mask reoxygenation[39]. Second, hypoxemia demands reoxygenation attempts that may be accompanied by iatrogenic hyperventilation by the operator with resulting hypocapnia, which has been associated with higher mortality risk[40]. In conclusion, there is strong evidence that apneic oxygenation reduces the incidence of hypoxemia during emergency intubation in a critical care setting. These findings support the inclusion of apneic oxygenation in everyday clinical practice.

ACCEPTED MANUSCRIPT Conflicts of interest None for all authors Funding None for all authors Acknowledgments

IP

T

We would like to thank all investigators who provided supplementary data not reported in their original papers.

CR

References

US

[1] Cook T, MacDougall-Davis S. Complications and failure of airway management. British Journal of Anaesthesia 2012;109:i68–85.

AN

[2] Semler MW, Janz DR, Lentz RJ, Matthews DT, Norman BC, Assad TR, et al. Randomized trial of apneic oxygenation during endotracheal intubation of the critically ill. American Journal of Respiratory and Critical Care Medicine 2016;193:273–80.

M

[3] Gebremedhn EG, Mesele D, Aemero D, Alemu E. The incidence of oxygen desaturation during rapid sequence induction and intubation. World Journal of Emergency Medicine 2014;5:279.

PT

ED

[4] Davis DP, Hwang JQ, Dunford JV. Rate of decline in oxygen saturation at various pulse oximetry values with prehospital rapid sequence intubation. Prehospital Emergency Care 2008;12:46–51.

CE

[5] Mosier JM, Hypes CD, Sakles JC. Understanding preoxygenation and apneic oxygenation during intubation in the critically ill. Intensive Care Medicine 2016:1–3. [6] Severinghaus JW. Simple, accurate equations for human blood O2 dissociation computations. Journal of Applied Physiology 1979;46:599–602.

AC

[7] Bohr C, Hasselbalch K, Krogh A. Ueber einen in biologischer Beziehung wichtigen Einfluss, den die Kohlensaurespannung des Blutes auf dessen Sauerstoffbindung ubt. Archiv F Physiol 1904;16:40l. [8] Weingart SD. Preoxygenation, reoxygenation, and delayed sequence intubation in the emergency department. The Journal of Emergency Medicine 2011;40:661–7. [9] Weingart SD, Levitan RM. Preoxygenation and prevention of desaturation during emergency airway management. Annals of Emergency Medicine 2012;59:165–75. [10] Sakles JC, Mosier JM, Patanwala AE, Arcaris B, Dicken JM. First pass success without hypoxemia is increased with the use of apneic oxygenation during rapid sequence intubation in the emergency department. Academic Emergency Medicine 2016.

ACCEPTED MANUSCRIPT [11] Wimalasena Y, Burns B, Reid C, Ware S, Habig K. Apneic oxygenation was associated with decreased desaturation rates during rapid sequence intubation by an Australian helicopter emergency medicine service. Annals of Emergency Medicine 2015;65:371–6. [12] Bartlett R, Brubach H, Specht H. Demonstration of aventilatory mass flow during ventilation and apnea in man. Journal of Applied Physiology 1959;14:97–101.

T

[13] Frumin MJ, Epstein RM, Cohen G. Apneic oxygenation in man. The Journal of the American Society of Anesthesiologists 1959;20:789–98.

CR

IP

[14] Lee S-C. Improvement of gas exchange by apneic oxygenation with nasal prong during fiberoptic intubation in fully relaxed patients. Journal of Korean Medical Science 1998;13:582– 6.

US

[15] Ramachandran SK, Cosnowski A, Shanks A, Turner CR. Apneic oxygenation during prolonged laryngoscopy in obese patients: A randomized, controlled trial of nasal oxygen administration. Journal of Clinical Anesthesia 2010;22:164–8.

AN

[16] Taha S, Siddik-Sayyid S, El-Khatib M, Dagher C, Hakki M, Baraka A. Nasopharyngeal oxygen insufflation following pre-oxygenation using the four deep breath technique. Anaesthesia 2006;61:427–30.

ED

M

[17] Pathak V, Welsby I, Mahmood K, Wahidi M, MacIntyre N, Shofer S. Ventilation and anesthetic approaches for rigid bronchoscopy. Annals of the American Thoracic Society 2014;11:628–34. [18] Rudlof B, Hohenhorst W. Use of apneic oxygenation for the performance of panendoscopy. Otolaryngology–Head and Neck Surgery 2013;149:235–9.

PT

[19] Pavlov I. Apneic oxygenation has not been disproven. Am J Respir Crit Care Med 2016;193:1316.

AC

CE

[20] Engström J, Hedenstierna G, Larsson A. Pharyngeal oxygen administration increases the time to serious desaturation at intubation in acute lung injury: An experimental study. Critical Care 2010;14:R93. [21] Jaber 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 Medicine 2016;42:1877–87. [22] Braun U, W. H. Dauer der Preoxygenation bei Patienten mit regelrechter und gestörter Lungenfunktion. Anaesthesist 1980;29:125–31. [23] Moher D, Liberati A, Tetzlaff J, Altman Douglas G, et al. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med 2009;6:e1000097.

ACCEPTED MANUSCRIPT [24] Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: A proposal for reporting. Jama 2000;283:2008–12. [25] Shrier I, Boivin J-F, Steele RJ, Platt RW, Furlan A, Kakuma R, et al. Should meta-analyses of interventions include observational studies in addition to randomized controlled trials? A critical examination of underlying principles. American Journal of Epidemiology 2007;166:1203– 9.

IP

T

[26] Simon M, Wachs C, Braune S, Heer G de, Frings D, Kluge S. High-flow nasal cannula versus bag-valve-mask for preoxygenation before intubation in subjects with hypoxemic respiratory failure. Respiratory Care 2016;61:1160–7.

CR

[27] Riyapan S, Lubin J. Apneic oxygenation may not prevent severe hypoxemia during rapid sequence intubation: A retrospective helicopter emergency medical service study. Air Medical Journal 2016;35:365–8.

US

[28] Higgins J, Green S. Cochrane handbook for systematic reviews of interventions version 5.1. 0 [updated march 2011]. The Cochrane Collaboration 2011.

AN

[29] Park J, Lee Y, Seo H, Jang B, Son H, Kim S, et al. Risk of bias assessment tool for nonrandomized studies (RoBANS): Development and validation of a new instrument. In:. Proceedings of the 19th cochrane colloquium, 2011, pp. 19–22.

ED

M

[30] Yusuf S, Wittes J, Probstfield J, Tyroler HA. Analysis and interpretation of treatment effects in subgroups of patients in randomized clinical trials. JAMA 1991;266:93–8. [31] Greenland S, Robins JM. Estimation of a common effect parameter from sparse follow-up data. Biometrics 1985:55–68.

CE

PT

[32] Vourc’h M, Asfar P, Volteau C, Bachoumas K, Clavieras N, Egreteau P-Y, et al. High-flow nasal cannula oxygen during endotracheal intubation in hypoxemic patients: A randomized controlled clinical trial. Intensive Care Medicine 2015;41:1538–48. [33] Anscombe FJ. Graphs in statistical analysis. The American Statistician 1973;27:17–21.

AC

[34] R Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2015. [35] Schwarzer G. Meta: An R package for meta-analysis. R News 2007;7:40–5. [36] Miguel-Montanes R, Hajage D, Messika J, Bertrand F, Gaudry S, Rafat C, et al. Use of highflow nasal cannula oxygen therapy to prevent desaturation during tracheal intubation of intensive care patients with mild-to-moderate hypoxemia*. Critical Care Medicine 2015;43:574–83. [37] Sakles JC, Mosier JM, Patanwala AE, Dicken JM. Apneic oxygenation is associated with a reduction in the incidence of hypoxemia during the RSI of patients with intracranial hemorrhage in the emergency department. Internal and Emergency Medicine 2016:1–10.

ACCEPTED MANUSCRIPT [38] Dyett J, Moser M, Tobin A. Prospective observational study of emergency airway management in the critical care environment of a tertiary hospital in Melbourne. Anaesth Intensive Care 2015;43:577–86. [39] Sakles JC, Chiu S, Mosier J, Walker C, Stolz U. The importance of first pass success when performing orotracheal intubation in the emergency department. Academic Emergency Medicine 2013;20:71–8.

AC

CE

PT

ED

M

AN

US

CR

IP

T

[40] Davis DP, Dunford JV, Poste JC, Ochs M, Holbrook T, Fortlage D, et al. The impact of hypoxia and hyperventilation on outcome after paramedic rapid sequence intubation of severely head-injured patients. Journal of Trauma and Acute Care Surgery 2004;57:1–10.

ACCEPTED MANUSCRIPT Figure captions.

Fig 1. PRISMA flow diagram of studies included in the meta-analysis.

IP

T

Fig 2. Apneic oxygenation is associated with a significantly lower incidence of clinically significant hypoxemia during endotracheal intubation.

CR

Fig 3. A pooled analysis of the three studies which reported the outcome of death at 28 days, apneic oxygenation was associated with a trend towards lower mortality. This finding borders, but does not achieve, conventional statistical significance.

US

Fig E1. The search strategy in Medline. Similar strategies were used in other databases.

AN

Fig E2. In a sub-group analysis of the incidence of SpO2 < 90%, apneic oxygenation was associated with a significantly lower risk of desaturation.

ED

M

Fig E3. A sub-group analysis restricted to the incidence of SpO2 < 80% shows a trend toward the benefit of apneic oxygenation, consistent with the effect size of the main outcome reported in Fig 2.

CE

PT

Fig E4. A sub-group analysis restricted to the incidence of SpO2 < 70% shows a trend toward the benefit of apneic oxygenation, consistent with the effect size of the main outcome reported in Fig 2.

AC

Fig E5. A sub-group analysis restricted to higher quality studies shows significantly lower incidence of clinically significant hypoxemia, just in the overall analysis reported in Fig 2.

Fig E6. A sub-group analysis of the incidence of SpO2 < 90%, restricted to higher quality studies, shows a significantly lower risk of desaturation, just as in Fig E2, which reports findings from all studies.

Fig E7. A sub-group analysis of the incidence of SpO2 < 80%, restricted to higher quality studies, shows a significantly lower risk of desaturation, just as in Fig E2, which reports findings from all studies.

PT

ED

M

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

AC

CE

Fig. 1

AC

CE

PT

ED

M

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT Table 1. Summary of included studies of apneic oxygenation Study design

Setting

FELLOW Semler et al. 2 2015

Randomized open-label ICU trial

OPTINIV Jaber et al. 21 2016

Randomized doubleblind controlled trial

Interventio n 77 patients. Ap-Ox with NC at 15 L/min

Control Results

Limitations

73 patients . No Ap-Ox

No statistical difference in median lowest SpO2, in the incidence of SpO2 < 90%, SpO2 < 80%, or decrease of SpO2 > 3%

A. Control group contaminate d with 56 patients (77%) ventilated between induction and laryngoscopy B. Trial underpowere d to detect a reduction in the incidence of severe desaturation

25 patients. Pre-Ox with NIV and HFNC (60L/min; FiO2 100%) and Ap-Ox with HFNC

24 patients . PreOx with NIV alone and no Ap-Ox

Significantl y higher lowest SpO2 (100% vs 96%; p=0.029%) and significantl y lower incidence of SpO2 < 80% (0% vs 21%, p=0.05%) in the interventio n group

Different Pre-Ox and Ap-Ox strategies in each arms. However, the difference in SpO2 appeared during the apneic phase, which support the efficacy of the Ap-Ox part of the intervention

62 patients. Pre-Ox and Ap-Ox with HFNC at 60L/min, FiO2 100%

57 patients . PreOx with highflow facial mask at 15L/mi n. No Ap-Ox

Median lowest SpO2 91.5% vs 89.5% in control group. (p=0.44)

Different Pre-Ox and Ap-Ox strategies in each arm.

AC

CE

PT

ED

M

ICU

AN

US

CR

IP

T

Trial

PREOXYFLO W Vourc’h et 32 al. 2015

Randomized controlled trial

ICU

ACCEPTED MANUSCRIPT

Randomized controlled trial

ICU

Sakles et al. 10 2016a

Single center observational casecontrol study

ED

Sakles et al. 35 2016b

Single center observational casecontrol study

20 patients. Pre-Ox and Ap-Ox with HFNC at 50L/min, FiO2 100%

20 patients . PreOx and partial Ap-Ox with BVM

Mean lowest SpO2 89% vs 86% in control group (p=0.56)

1. Contaminate d control group: BVM left in place during the apneic phase until the laryngoscopy attempt. 2. Differences in SpO2 appear before the intubation attempts. 380 255 FPS-H: 1. patients. patients 82% vs Observation Ap-Ox with . No 69% in the al study. NC at Ap-Ox control 2. Ap-Ox various flow group. with NC at rates (5, Adjusted various flows 10, 15 and OR 2.2, rates, the >15 L/min) (95% CI lowest of 1.5—3.3) them probably not effective. 72 patients. 55 Incidence Subset of the Ap-Ox with patients of previous NC at . No SpO2<90 observationa various flow Ap-Ox during the l study, rates(5, 10, intubation restricted to 15, >15 L 7% vs patients with /min) 29% in the intracranial control hemorrhage group. adjusted OR 0.13(95 % CI 0.03— 0.53)

AC

CE

PT

ED

M

AN

US

CR

IP

T

Simon et al. 25 2016

Riaypan et al. 26 2016

Single helicopter service retrospective case-control study

ED

Helicopter 29 patients. emergenc Ap-Ox with y service NC at 15L/min

64 patients . No Ap-Ox

Incidence of SpO2<90 during intubation 17.2% vs. 21.9% in the control group

Small, retrospective , database study

ACCEPTED MANUSCRIPT (p=0.78)

Prospective observational study

AN

US AC

CE

PT

ED

M

Emergenc y intubation s in the ICU, ED, and on the wards of a tertiary care hospital

47 patients. Ap-Ox with NC at 15 L/min

Incidence Before-after of study. SpO2<93 % during intubation 16.5% vs 22.6% in the control group. Crude OR 0.68 (95% CI 0.47— 0.98)

T

Dyett, et al. 36 2015

310 patients . No Ap-Ox

IP

Retrospectiveuncontroll Helicopter 418 ed before-after study emergenc patients . y service Ap-Ox with NC at 15L/min

CR

Wimalasena, 11 et al. 2015

92 patients , among whom 29 receive d partial Ap-Ox with BVM or NIV and 63 receive d no Ap-OX

Incidence of SpO2 <93% during intubation 12.8% vs 19.6% (p=0.32)

Small, observationa l study. Control group contaminate d with patients who received apneic oxygenation and/or ventilation between induction and laryngoscopy

List of abbreviations, in order of appearance : ICU, intensive care unit; Ap-Ox, apneic oxygenation; NIV, noninvasive ventilation; HFNC, high flow nasal cannula; Pre-Ox, preoxygenation; BVM, bag valve mask; NC, nasal cannula.