Underuse of rapid sequence intubation outside emergency departments: preliminary, retrospective, observational study with emergency physician insights
Article information
Abstract
Objective
Rapid sequence intubation (RSI) involves the administration of induction agents and neuromuscular blockers before endotracheal intubation (ETI). However, RSI is underutilized outside emergency departments (EDs). We compared RSI adoption rates and ETI outcomes outside and within the EDs and investigated whether RSI adoption affected ETI outcomes outside the EDs.
Methods
This retrospective study included adults who underwent emergency ETI outside the operating room at one university hospital between March 2022 and February 2023. The exclusion criteria included cardiopulmonary resuscitation, intentional RSI avoidance, and tube exchange via an introducer. The primary outcome was first-pass success rate. Secondary outcomes were multiple attempts (≥3), prolonged ETI (>5 minutes), and complications. The association between RSI adoption and outcomes outside the ED was assessed using multivariate logistic regression.
Results
A total of 490 ETI cases was included: 286 occurred outside the ED and 204 within the ED. The mean age was 68.3±14.7 years and 290 were male. Cases outside the ED received fewer RSI attempts than cases within the ED (12.6% vs. 86.8%, P<0.001). The former showed fewer incidents of first-attempt success (62.2% vs. 88.2%, P<0.001), more numerous multiple attempts (11.5% vs. 2.0%), longer total time of ETI (8.4±8.3 minutes vs. 2.5±2.5 minutes, P<0.001), and more frequent complications (32.2% vs. 19.6%, P=0.003). However, multivariate logistic regression revealed no significant association between RSI adoption and these outcomes outside the ED: odds ratio, 1.74 (95% confidence interval [CI], 0.783–3.84), 0.167 (95% CI, 0.022–1.30), 1.04 (95% CI, 0.405–2.69), and 1.50 (95% CI, 0.664–3.40), respectively.
Conclusion
Outside the ED, RSI adoption was lower and ETI outcomes were poorer than those within the ED. However, no association was found between RSI adoption and ETI outcomes outside the ED.
INTRODUCTION
Approximately 1.6 million adults undergo endotracheal intubation (ETI) annually in the United States [1]. Emergency scenarios requiring ETI occur throughout hospitals. A rapid response team (RRT) found that the leading cause of cardiopulmonary arrest was respiratory arrest (42%), followed by cardiac arrest (18%), hypovolemic shock (13%), and septic shock (5%). These findings underscore the importance of successful emergency ETI [2].
Rapid sequence intubation (RSI) is the preferred method for ETI in emergencies [3,4]. By near-simultaneously administrating an induction agent (IA) and neuromuscular blocking agent (NMBA), RSI induces unconsciousness and paralysis, which requires rapid ETI [3]. This intervention is associated with a higher first-pass success rate (successful ETI on the first attempt) and a lower rate of complications, such as cardiac arrest, hypotension, hypoxemia, regurgitation, and esophageal intubation [5]. Introduced in the 1960s by anesthesiologists to reduce gastric regurgitation in the operating room, RSI has greatly evolved and has become a standard ETI method in emergency departments (EDs) [4–7].
However, minimal knowledge exists about its adoption rate outside the ED (except in a few intensive care units [ICUs]) and their consequences [8,9]. After recently becoming involved in an RRT, an author of this study, an emergency physician, observed that many doctors outside the ED do not adopt RSI during emergency ETI; they administer an IA only without an NMBA. Additionally, the authors observed that ETIs performed outside EDs that did not adopt RSI tended to require more time for the intervention and a greater number of attempts.
In this study, we investigated whether the adoption rate of RSI for administering both an IA and an NMBA before ETI and the outcome of emergency ETI differed outside and within the ED. Additionally, we investigated whether a relationship existed between RSI adoption and ETI outcomes outside the ED.
METHODS
Ethics statement
This study was approved by the Institutional Review Board of CHA Bundang Medical Center (No. 2023-09-021). The requirement for informed consent was waived due to the retrospective nature of the study.
Study design, setting, and participants
A retrospective cohort study was performed including all adults aged ≥18 years who had undergone emergency ETI outside the operating room in a university hospital from March 2022 to February 2023. The exclusion criteria were as follows: (1) full-scale or imminent cardiac arrest; (2) definite anatomically difficult airway in probable need of a surgical airway; (3) documented intentional avoidance of RSI for high-risk physiological conditions (hypoxemia, hypotension, severe metabolic acidosis, and right ventricular failure) or neuromuscular junction disorders for which RSI is optional or sometimes contraindicated; (4) ETI cases performed by anesthesiologists; and (5) exchange of the endotracheal tube via an intubating introducer [10–13].
After comparing the RSI adoption rate and outcomes between the “within ED” and “outside ED” groups, we focused on the “outside ED” group to determine whether RSI adoption was related to each outcome. During the study period, ETI was performed not by the RRT, but by junior doctors responsible for the patient. Most patients in the within ED group were treated by emergency medicine residents. For all patients in the outside ED group, ETI was initially performed by physicians or surgeons (usually residents) other than emergency medicine physicians, although sometimes those physicians were called to assist in difficult cases.
Variables and measurements
The demographic information of the subjects in this study were age, sex, and body mass index (BMI) [6,14]. Clinical information was collected regarding the indication for ETI and high-risk physiology for performing RSI, including hypoxemia based on oxygen saturation <90%, systolic blood pressure <90 mmHg, severe metabolic acidosis with pH <7.10, or right ventricular failure [11–13]. Oxygen saturation was measured using pulse oximetry or arterial blood gas analysis (ABGA). To identify metabolic acidosis and right ventricular failure, ABGA and echocardiography that had been performed within 1 hour and 5 years from the time of ETI, respectively, were noted. Right ventricular failure was considered present if reported and pulmonary hypertension was severe.
For ETI factors, the location of ETI, the performer’s specialties and level of ETI experience, and the use of video (vs. direct) laryngoscopy were included [6,15]. The location of the ETI was classified as within or outside ED (ICU or general ward). Specialty doctors were categorized as emergency medicine doctors, physicians, and surgeons. The level of ETI experience was graded as basic (<10), intermediate (10–60), and advanced (>60) based on a review of the literature and subsequent discussion among the clinician-authors. These grades approximately corresponded to the average cumulative ETI experience of emergency medicine residents in this hospital at the end of their first semester of years 1, 2, and later, respectively [6,16,17].
RSI was considered to have been adopted only when IA and NMBA had been administered, either as a bolus or as continuous infusion, before the first ETI attempt. The primary outcome of the ETI was first-pass success. Secondary outcomes were multiple attempts (≥3), prolonged ETI (total time of ETI >5 minutes, median and mode values) (Supplementary Fig. 1), and complications. An ETI attempt was defined as insertion of the laryngoscope through the mouth. We primarily used nursing charts to document the total ETI time and number of attempts, which were verified by examining medical charge specifications. When discrepancies arose, the higher number of attempts was used. Successful placement of an endotracheal tube was confirmed by end-tidal CO2 monitoring or chest radiography without deteriorating oxygenation for >10 minutes [10]. The total time of ETI was defined as the time from the first attempt until the time of successful endotracheal tube placement and fixation. Complications included ETI failure, cardiac arrest, hypotension (defined as any decrease in systolic pressure to <90 mmHg or a decrease by even 1 mmHg in cases of preexisting hypotension following ETI), hypoxemia (defined as any decrease in oxygen saturation, as determined by pulse oximetry, to <90%, or a decrease by even 1% in cases of preexisting hypoxemia following ETI), arrhythmia, regurgitation, and esophageal intubation with delayed recognition (based on end-tidal CO2, chest radiography, or clinical deterioration) [5,6]. Considering the duration of RSI medications, adverse events were regarded as complications of ETI if they occurred within 20 minutes of the ETI attempt. For cases receiving vecuronium or rocuronium, the criteria increased to ≤40 minutes [3,18].
All measurements were obtained by retrospectively reviewing electronic medical records, including doctors’ notes, nursing charts, and medical charge specifications, in which the time of each process had been documented. The first author reviewed the medical records and conducted the measurements for each variable. In instances of uncertainty regarding the measurement of any variable, the discrepancies were resolved through consultation with another author.
Statistical analysis and sample size calculation
Demographic, clinical, and ETI information (especially adoption of RSI) and outcomes were compared according to the location of ETI (outside ED group vs. within ED group). We focused on the difference in adoption rates of RSI and outcomes of ETI between the two groups. To determine whether the adoption of RSI was related to outcomes of ETI outside the ED, we used the outside ED group only for further analyses. Variables that showed a possible relationship with each outcome in the univariate comparison (P<0.15) were incorporated into the multivariate logistic regression analysis. Age and BMI were categorized as ≥70 years or below and ≥25 kg/m2 or below, respectively [19]. For the 14 records with missing values of BMI, average values of the same sex were substituted: 22.6 kg/m2 for men and 23.1 kg/m2 for women.
The RSI adoption rate was assumed to be 20% for the outside ED group and 80% for the within ED group. When the power and α error were set to 80% and 0.05, respectively, the minimum sample size for each group was calculated as 16. With first-pass success, the primary outcome, the rate was assumed to be 60% for the outside ED group and 80% for the within ED group, and the sample size for each group was 120.
To execute multiple logistic regression analyses, a maximum of 13 independent variables (age, sex, BMI, indication for ETI, four physiological high-risk factors, location of the ETI, performer’s specialty, performer’s level of ETI experience, use of video laryngoscope, and adoption of RSI) required ≥130 outcome events, conservatively [20]. Assuming a first-pass success rate of approximately 60%, 325 ETI procedures performed outside the ED were required [21].
As a subgroup analysis, we compared the ETI characteristics according to the location (ICU vs. general ward). Chi-square and t-tests were used for categorical and continuous variables, respectively. All statistical analyses were performed using IBM SPSS ver. 29 (IBM Corp). Statistical significance was set at P<0.05. Data were presented as mean±standard deviation.
RESULTS
Among the 798 adult cases that had undergone emergency ETI outside the operating room, we excluded 280 definite and 8 impending CPR cases, 3 cases of a definitely anatomically difficult airway in probable need of a surgical airway (2 tongue cancers and 1 tracheal cancer), 1 intentional avoidance (myasthenia gravis) case, 5 ETI cases performed by anesthesiologists, and 11 cases of exchange of endotracheal tubes via an intubating introducer. Therefore, 490 ETI cases (age, 68.3±14.7 years; BMI, 22.8±4.5 kg/m2; 290 men, 59.2%) were initially included: 286 in the outside ED group (age, 68.4±13.8 years; BMI, 22.4 ± 4.1 kg/m2; 160 men, 55.9%; 254 ICU and 32 general ward cases) and 204 in the within ED group (age, 68.3±15.9 years; BMI, 23.5 ± 5.0 kg/m2; 130 men, 63.7%) (Fig. 1).
Study flowchart. ETI, endotracheal intubation; OR, operating room; RSI, rapid sequence intubation; ED, emergency department.
The RSI adoption rate was significantly lower in the outside ED group compared to the within ED group (12.6% vs. 86.8%, P<0.001). The outside ED group showed worse outcomes compared to the within ED group: a lower first-pass success rate (62.2% vs. 88.2%, P<0.001), a greater number of multiple attempts (11.5% vs. 2.0%, P<0.001), a greater total time of ETI (8.4±8.3 minutes vs. 2.5±2.5 minutes, P<0.001), and more frequent complications (32.2% vs. 19.6%, P=0.003). They also showed differences in BMI, indications for ETI, high-risk physiology, doctor’s specialties, level of ETI experience, and use of a video laryngoscope (Table 1).
Characteristics of ETI cases according to the location of ETI (n=490)
We then focused on the outside ED group only to study if adoption of RSI was related to the outcomes of ETI performed outside the ED. In the univariate analysis, the adoption of RSI outside the ED was associated with patients aged <70 years, higher BMI, indications for ETI, lower burden of hypoxemia, surgeon (vs. physician), and use of video laryngoscopy. However, the ETI outcomes did not reveal a significant association with RSI adoption in the univariate analysis (Table 2).
Characteristics of ETI cases performed outside the emergency department according to RSI adoption (n=286)
A first-pass success, the primary outcome, did not show significant associations with ETI characteristics in the univariate analysis except for other outcomes that could not have been “predictors” (Table 3). To identify the predictors of first-pass success using multiple logistic regression analysis, we incorporated the adoption of RSI (the key focus of the study), age, and indications for ETI (P<0.15). First-pass success was not associated with RSI adoption but only with age (Table 4).
Characteristics of ETI cases performed outside the emergency department according to first attempt result (n=286)
Univariate and multivariate logistic regression analysis to reveal the predictors of first-pass success outside the emergency department
Similar analyses were performed for the secondary outcomes. In the univariate analysis, multiple attempts were not significantly associated with any ETI characteristics except other outcomes (Supplementary Table 1). Among the candidate predictors (P<0.15 in the univariate analysis) incorporated into the multiple logistic regression analysis, RSI adoption showed high collinearity with NMBA administration (r=0.915, P<0.001), as the former almost entirely depended on the latter. Among these two variables, we selected RSI adoption, which was the focus of this study, while avoiding NMBA administration [20]. Multiple attempts revealed association with obesity. RSI adoption showed an equivocal relationship (P=0.087) (Supplementary Table 2). Similarly, prolonged ETI (>5 minutes) was negatively associated with advanced ETI experience. However, prolonged ETI did not show an association with RSI adoption (Supplementary Tables 3, 4).
For multivariate analysis to reveal the predictors of complications, we incorporated other outcomes—first-pass success, multiple attempts, and prolonged ETI—as these would influence ETI complications. Complications were associated with hypotension only, and the adoption of RSI was not associated with complications (Supplementary Tables 5, 6). In the subgroup analysis, patients in the ICU and general ward showed no differences in ETI characteristics except the indication for ETI (Supplementary Table 7).
DISCUSSION
The findings of this study demonstrate that RSI is less frequently adopted outside the ED during emergency ETI, and that these ETI outcomes are worse than those of ETI performed within the ED, where RSI is adopted more frequently. However, no significant relationship was found between RSI adoption and ETI outcomes when we analyzed ETI cases performed outside the ED.
To the best of our knowledge, this is the first study to investigate the adoption rate of RSI outside the ED (12.6% in this study) during emergency ETI as well as the outcomes of emergency ETI in terms of first-pass success, multiple attempts, prolonged attempts, and complications compared with those performed in the ED. This study is also original in its examination of relationships between RSI adoption and ETI outcomes outside the ED.
RSI is a well-established and preferred emergency ETI method [3,4,21]. RSI was introduced in the 1960s by anesthesiologists to reduce gastric regurgitation in the operating room during elective operations. It has evolved greatly to become a standard method of emergency ETI in the ED since the late 1990s [4,16]. Moreover, some intensivists working in ICUs have recently shown interest in RSI, focusing on NMBA [8,9]. However, the overall adoption rate of RSI outside the EDs (ICUs and general wards) and its consequences had not been investigated.
As this study revealed, doctors outside the ED underuse RSI compared with those within the ED (12.6% vs. 86.8%). In addition to the cases of relatively rare definite anatomically difficult airway in probable need of surgical airway, which were excluded from this study (3 out of 1,014, 0.3%), high-risk physiologies (hypoxemia, hypotension, severe metabolic acidosis, and right ventricular failure) are associated with poor ETI outcomes, such as hypoxemia, cardiovascular collapse, or arrest [11,18,22]. Doctors outside the ED may be concerned that these complications could result from RSI using popular IA such as midazolam and propofol to produce hypotension [3]. In this study, RSI was negatively associated with hypoxemia, a high-risk physiological condition, and hypotension did show association with ETI complications (Table 2 and Supplementary Table 6). However, the lower adoption of RSI outside the ED compared to that within the ED was primarily due to the significantly lower usage of NMBAs (14.7% vs. 87.7%) rather than induction agents (85.0% vs. 95.6%). We believe that fear of a “cannot ventilate, cannot intubate” situation can lead doctors outside the ED to avoid administering NMBAs [23].
Patients with ETI performed outside the ED also showed worse outcomes than those with ETI performed within the ED, including fewer first-attempt successes, a greater number of multiple attempts, more prolonged procedures, and a greater number of complications. However, use of RSI was not significantly associated with any of the ETI outcomes.
Factors other than RSI may have led to poor ETI outcomes outside the ED. Many studies have suggested that human factors have the greatest impact on ETI outcomes. Performance by non–emergency medicine doctors and less ETI experience, which characterized the outside ED group (Table 1), are predictors of first-pass failure [24]. Video laryngoscopy, a well-known predictor of first-pass success, was less frequently used outside the ED (Table 1) [15,17]. General wards outside an ED even tend to be less equipped for RSI, which suggests not only an equipment problem, but also weak teamwork, which is essential for successful ETI [25,26]. Patients outside the ED more frequently have respiratory failure, shock, and airway problems as ETI indications (Table 1), which are well-known risk factors for poor ETI outcomes [11,18,22]. Despite adjusting for these factors in multiple logistic regression analyses, these differences may explain the poorer ETI outcomes outside the ED.
Second, the broad definition of RSI in this study (administering an IA and an NMBA before ETI) may have contributed to the lack of a relationship between RSI adoption and ETI outcomes outside the ED. Our RSI definition only included “paralysis and induction,” one of the seven P’s of RSI: preparation and planning, preoxygenation, pretreatment, paralysis and induction, protection and positioning, placement with proof, and post-intubation management [14,25–27]. While we ensured paralysis and induction, placement with proof, and post-intubation management, we could not confirm the completion of the other P’s, which require experienced personnel, equipment, and teamwork [17,24–26]. Adhering to the strictest RSI definition, less achievable in this retrospective study, might have shown a relationship between lower RSI adoption and worse ETI outcomes outside the ED, similar to those within the ED [21,24,25].
This study has several limitations. First, as this was a single-center retrospective observational study, we could not generalize our conclusions. Second, this retrospective study using a chart review had the potential for bias [28]. For example, the number of attempts and the total time of ETI based on a retrospective review, primarily on nursing charts, may have been less accurate than the prospective collection of data. Similarly, right ventricular failure, which is a high-risk physiology, and complications such as arrhythmia, regurgitation, and esophageal intubation with delayed recognition might have been missed in some cases owing to the lack of echocardiography, electrocardiography monitoring, and less aggressive detection of possible complications in this retrospective study. A multicenter, prospective cohort study focusing on all seven P’s of RSI and recording the variables at the time of ETI would render more valid and generalizable conclusions. Third, clinicians use factors such as head and neck pathology, reduced mouth opening, limited neck extension, obstructive sleep apnea, morbid obesity, and progressive airway compromise, along with mnemonics such as the LEMON criteria, to identify difficult airways [14,29]. Owing to the retrospective design, our definition of “definite anatomically difficult airway likely requiring a surgical airway” was too narrow to identify all potential difficult airways in advance. Last, we recruited fewer ETI cases outside the ED than the target (286 vs. 325). However, the number of candidate variables to be incorporated into multiple logistic regression analysis to predict the four outcomes was 3 to 7 (for complications), and the number of outcomes exceeded 10-fold the number required for each outcome to satisfy the basic condition [20].
In conclusion, RSI is less frequently performed outside the ED, and ETI outcomes are worse for those performed outside the ED than those performed in the ED. However, in this preliminary retrospective observational study, we did not find an association between RSI adoption and outcomes of ETI performed outside the ED. Multicenter prospective cohort studies are required to obtain valid and generalizable conclusions.
Notes
Author contributions
Conceptualization: all authors; Data curation: SYP; Formal analysis: SBC; Methodology: all authors; Supervision: SBC; Writing–original draft: SBC; Writing–review & editing: all authors. All authors read and approved the final manuscript.
Conflicts of interest
The authors have no conflict of interest to declare.
Funding
The authors received no financial support for this study.
Data availability
Data analyzed in this study are available from the corresponding author upon reasonable request.
SUPPLEMENTARY MATERIAL
Supplementary materials are available from https://doi.org/10.15441/ceem.24.227.
Supplementary Table 1.
Characteristics of ETI cases according to multiplicity of ETI attempts outside the emergency department (n=286)
Supplementary Table 2.
Univariate and multivariate logistic regression analysis to reveal the predictors of multiple ETI attempts outside the emergency department
Supplementary Table 3.
Characteristics of ETI cases according to total time of ETI performed outside the emergency department (n=286)
Supplementary Table 4.
Univariate and multivariate logistic regression analysis to reveal the predictors of prolonged ETI performed outside the emergency department
Supplementary Table 5.
Characteristics of ETI cases performed outside the emergency department according to complication (n=286)
Supplementary Table 6.
Univariate and multivariate logistic regression analysis to reveal the predictors of complication of ETI performed outside the emergency department
Supplementary Table 7.
Characteristics of ETI performed outside the emergency department according to specific location (n=286)
Supplementary Fig. 1.
Total time of ETI.
References
Article information Continued
Notes
Capsule Summary
What is already known
Rapid sequence intubation (RSI) is a standard practice for endotracheal intubation (ETI) in emergency departments (EDs) and involves administration of induction agents and neuromuscular blockers before ETI.
What is new in the current study
Outside EDs, RSI adoption is lower and ETI outcomes are poorer compared to those within the EDs. This preliminary study found no significant association between RSI adoption and ETI outcomes outside the EDs.