Rapid sequence intubation is the most common method of airway management in the emergency department (ED).1 Use of an induction agent and a neuromuscular-blocking agent results in transient apnea during the intubation attempt. If prolonged, apnea and resulting oxygen desaturation may lead to serious adverse events such as dysrhythmias, hypotension, and cardiac arrest.2 Although practice guidelines advocate preoxygenation to prevent oxygen desaturation, the adequacy of preoxygenation is not typically assessed in the ED.3, 4, 5
Editor’s Capsule Summary
What is already known on this topic
Preoxygenation is recommended before rapid sequence intubation.
What question this study addressed
Can end tidal expired oxygen levels (eto2) characterize preoxygenation before rapid sequence intubation?
What this study adds to our knowledge
In this series of 100 rapid sequence intubations, median eto2 levels before and after preoxygenation were 53% (interquartile range 43% to 65%) and 78% (interquartile range 64% to 86%). Of the 18 patients experiencing oxygen desaturation (SaO2 <90%), 14 (78%) had an eto2 level less than 85% before induction.
How this is relevant to clinical practice
eto2 level may be useful for guiding preoxygenation during rapid sequence intubation in the ED. However, formal validation is necessary before clinical implementation.
In the operating room, anesthesiologists have used gas analyzers to quantify and optimize preoxygenation.6, 7 The Difficult Airway Society suggests that when critically ill patients are intubated, preoxygenation should be performed until an eto2 level of greater than or equal to 85% is attained.5, 8 eto2 level has not been widely used in clinical practice in the ED.
We sought to describe our preliminary use of eto2 to assess preoxygenation during rapid sequence intubation in the ED.
We conducted a prospective observational cohort study at 2 urban, academic EDs in Sydney, Australia, and New York City. This study was approved with waiver of consent by the institutional review board and ethics boards at each institution.
The annual census of each ED was 175,000 patients (New York) and 80,000 patients (Sydney). The departments perform greater than 900 intubations a year combined (approximately 300 at the Sydney site and 600 at the New York site). Each department has emergency medicine trainees. The majority of intubations are performed by emergency medicine trainees under the direct supervision of an emergency medicine attending physician. Both sites practice rapid sequence intubation, with the specific technique used chosen by the emergency medicine attending physician.
Clinical airway management practices at each site called for a minimum of 3 minutes of preoxygenation with either a bag-valve-mask device or a nonrebreather mask before rapid sequence intubation. If a bag-valve-mask device was used, the mask seal was maintained by the operator, and assisted breaths were given at the discretion of the attending physician. Positive end-expiratory pressure could be provided if clinically warranted by means of a positive end-expiratory pressure valve connected to the bag-valve-mask device, and at pressures ranging from 1 to 20 cm H2O. For nonrebreather mask, the oxygen flow rate was selected by the emergency medicine attending physician and was set at either 15 L/min or flush rate (50 to 70 L/min at the New York site and 19 L/min at the Sydney site). The use of supplemental nasal cannula oxygen was chosen by the attending emergency physician and ranged from 15 L/min to flush rate.
We included all adult patients (≥18 years) undergoing rapid sequence intubation during October 2017 to February 2018. We excluded patients in cardiac arrest, receiving noninvasive ventilation before intubation, intubated in the out-of-hospital setting, or who underwent awake intubation.
To quantify eto2 level achieved with current preoxygenation practices, emergency physicians were blinded to the eto2 data collected during the procedure. Independent observers (research assistants, nurses, and residents) collected all eto2 measurements. The observers underwent training for the study and collected data in real time, using a standardized data collection tool.
Vital signs were obtained from the cardiac monitor (Philips IntelliVue; Philips, Andover, MA) in real time. eto2 level was measured by Phillips G5 Gas Analyzer (Philips) at the New York site and by a Philips G7 Gas Analyzer (Philips) at the Sydney site. The newer-generation G7 module is a more compact version of the G5 and facilitates similar eto2 measurements. Although not the focus of this study, the devices also measured FiO2. The analyzers use a single flow sensor to make measurements for FiO2 supplied and eto2 exhaled by gas sampling. SpO2 was measured through standard finger oximeters (Covidien Oximax; Covidien). Hypoxemia was defined as an SpO2 level less than 90%.
For patients undergoing bag-valve-mask preoxygenation, eto2 level was measured by side-stream gas sampling connected between the bag-valve-mask device and the mask. For patients receiving nonrebreather mask preoxygenation, side-stream sampling was used by means of a nasal prong gas sampler (AirLife etco2 Nasal Cannula at the New York site and CapnoEZY at the Sydney site). For patients receiving preoxygenation by nonrebreather mask plus nasal cannula oxygen, the New York site used a separate standard nasal cannula. The Sydney site used CapnoEZY nasal prongs. Waveform capnography was used to verify that ventilation was occurring during eto2 measurement recordings.
To validate the use of single-breath versus continuous eto2 measurements, we conducted a preliminary study using 4 healthy volunteers (attaining 20 measurements at each site). We compared continuous and single-breath eto2 measurements for both bag-valve-mask device and nonrebreather mask (n=10 measurements for each method). Each volunteer was preoxygenated for 3 minutes with bag-valve-mask device and nonrebreather mask. The eto2 was recorded at the end of 3 minutes of continuous analysis. At the conclusion of the 3 minutes, the subject held a single breath for 10 seconds and then exhaled. This single-breath measurement was then compared with the final measurement during the continuous eto2 measurement.
Preoxygenation interval was defined as the elapsed time from emergency medicine attending physician decision to intubate to the point of induction. Apnea time was defined as the elapsed time from induction to confirmation of tube placement by waveform capnography. All other data on baseline characteristics, pre- and postlaryngoscopy management, and clinical outcomes were collected from the medical record by study personnel. We tested interrater agreement for eto2 and SpO2 by using the first 5 cases.
The primary outcome was eto2 level, measured before preoxygenation and at induction. The secondary outcome was SpO2, measured before and immediately after intubation.
To validate the use of single-breath versus continuous eto2 in healthy volunteers, we used Pearson’s correlation coefficient and a Bland-Altman plot. For the clinical series, we analyzed the data with descriptive techniques. We determined median eto2 level before preoxygenation and at induction, and SpO2 before and after intubation, examining the full cohort, as well as stratifying by preoxygenation technique. We plotted eto2 and SpO2 measurements for each patient. We determined the proportion of patients achieving eto2 level greater than 85%. We also determined the proportion of patients who experienced oxygen desaturation (SpO2 <90%). For the initial 5 subjects, we assessed percentage agreement between raters. All analyses were performed with XLStat (version 2018.7; Addinsoft, New York, NY).
