First-Pass Success of Video Laryngoscope Compared With Direct Laryngoscope in Intubations Performed by Residents in the Emergency Department

Background: The video laryngoscope (VL) has been widely used for intubation in the emergency department (ED). However, their effectiveness remains controversial, particularly among airway management performed by residents in the ED. Methods: We aimed to examine whether the use of VL, compared to a direct laryngoscope (DL), was associated with higher first-attempt intubation success among intubations performed by residents in the ED. This is a secondary analysis of the data from a prospective, observational, multicentre study of 15 Japanese EDs from April 2012 through March 2020. We included all adult patients who underwent intubation with VL or DL by residents (postgraduate years ≤5) in the ED. The outcome measures were first-pass success and intubation-related adverse events (overall, major, and minor adverse events). To determine the association of VL use with each of the outcomes, we constructed logistic regression models with generalized estimating equations to account for patients clustering within the ED, adjusting for patient demographics, primary indications, intubation difficulty, and intubation methods. Results: Of 5,261 eligible patients who underwent an initial intubation attempt by residents, 1,858 (35%) patients were attempted with VL. Intubations performed with VL had a non-significantly higher first-pass success rate than those with DL (77% vs. 64%; unadjusted odds ratio (OR)=1.20; 95% CI=0.87-1.65; P=0.27). This association was significant after adjustment for potential confounders (adjusted OR, 1.33; 95% CI, 1.06-1.67; P=0.01). As for adverse events, the use of VL was associated with a lower rate of any (adjusted OR=0.67; 95% CI=0.51-0.86; P=0.002) and minor (adjusted OR=0.69; 95% CI=0.55-0.87; P=0.002) adverse events. Conclusion: The use of VL was associated with a higher first-attempt success rate and a lower rate of any adverse events compared to that with DL among intubations performed by residents in the EDs.


Introduction
Tracheal intubation is a critical intervention performed in the emergency department (ED).The firstattempt intubation success (first-pass success) rate in the ED remains suboptimal, ranging from 63% to 83%, as reported in studies using large registries [1][2][3].In addition, with the limited resources and an urgent situation, difficult intubations more frequently occurred in the ED than in the operating room setting (13% vs. 5.8%) [4,5].Regardless, physicians should achieve first-pass success in such settings as multiple intubation attempts lead to adverse events [6,7] and poor in-hospital outcomes [8].
In recent years, the video laryngoscope (VL) is more commonly used for intubation in critically ill patients in the ED [9].VL provides better glottic visualization and less oesophageal intubation than the direct laryngoscope (DL) [10][11][12].The utility of VL compared to DL in improving first-pass success rates and reducing adverse events was previously unclear; however, in recent times, there has been growing support for the effectiveness of VL [12][13][14][15].Nonetheless, the utility of VL in endotracheal intubation performed by resident physicians remains uncertain.In clinical education for residents, tracheal intubation is one of the most important procedures that residents should experience during their training program.Despite its importance, the effectiveness of VL on intubation performance among intubations performed by residents in the ED remains to be elucidated.
To address the knowledge gap in the literature, we used data from a prospective multicentre study to investigate whether the use of VL compared to the use of DL was associated with a higher first-pass success rate and lower adverse event rate of intubation by residents in EDs.A better understanding of this crucial issue in emergency airway management will inform physicians to develop optimal curricula for residency training.

Study design and setting
This is a secondary analysis of the data from a prospective, observational, multicentre study of emergency airway management, the second Japanese Emergency Airway Network (JEAN-2) study.The study design, setting, methods of measurement, and measured variables have been reported previously [16][17][18][19].In short, the JEAN-2 study was a consortium of 15 academic and community medical centers from different geographic regions across Japan.The participating institutions included 12 level-I and three level-II equivalent trauma centers with a median ED census of 26,692 patient visits per year (range=1,078-60,101 per year).All 15 EDs were staffed by emergency attending physicians and affiliated with emergency medicine residency programs.Intubations were performed by transitional-year residents (postgraduate years one and two), emergency medicine residents (both of which were at the discretion of attending physicians), attending emergency physicians, or other specialties (e.g., surgery, anesthesia, pediatrics).The institutional review board of each participating hospital approved the study protocol with a waiver of informed consent prior to data collection.

Study participants
We included all adult patients (aged ≥18 years) who underwent tracheal intubation in the ED by postgraduate residents (defined as a doctor who has started working within five years after graduating from medical school, which included transitional-year residents, emergency medicine residents, and other specialty residents) from April 2012 through March 2020.We excluded patients intubated by other than VL and DL or by adjunctive devices, or those with missing data on the patient's age, body mass index (BMI), modified look-evaluate-Mallampati-obstruction-neck (LEMON), intubation devices, intubator's postgraduate year, or specialty.

Data collection and processing
The intubators filled out a standardized data collection form on each intubation.Collected information included the patient's demographics (age, sex, height, and weight), primary indication for intubation, components of modified LEMON criteria [20,21], all medications used for intubation, intubation methods (no medication, rapid sequence intubation (RSI), sedation without paralysis, and others), ED visit year, devices of intubation, intubator's level of training and specialty, intubation success or failure, the number of intubation attempts, vital signs (at pre-intubation, one minute and 30 minutes after intubation), and adverse events.The modified LEMON criteria were composed of external looks, the distance between the incisor teeth, the distance between the hyoid bone and the chin, airway obstruction, and neck immobility [20,21].The Japanese Emergency Medicine Network (JEMNet) coordinating center and site investigator at each ED reviewed the data forms.If information was missing or contained inconsistencies, the data form was returned to the intubator for completion or the site investigator asked the intubator for the airway management details.

Outcome measures
The primary outcome was first-pass success defined as successful intubation on the first attempt.We have not collected data on time for intubation.The secondary outcomes were intubation-related adverse events, including major and minor adverse events.Major adverse events were defined as cardiac arrest, dysrhythmia, hypotension (systolic blood pressure <90 mmHg), hypoxia (pulse oximetry saturation <90%), and oesophageal intubation with delayed recognition [6,9,17,18,22].Minor adverse events were defined as regurgitation, airway trauma, dental or lip trauma, mainstem bronchial intubation, and oesophageal intubation recognized immediately [6,9,18,22].

Statistical analysis
First, we compared the patients and airway management characteristics between VL and DL groups, using the Mann-Whitney U test and chi-square test as appropriate.Second, to investigate the association of devices (VL and DL) with each of the outcomes, we constructed the unadjusted and adjusted logistic regression models with generalized estimating equations (GEE) to account for patients clustering within the ED.In the multivariable models, we adjusted for age (<65, and ≥65 years old), sex, BMI (<25.0,25.0-29.9,and ≥30.0 kg/m 2 ), primary indication (medical indication, traumatic indication, and cardiac arrest), modified LEMON, premedication use, intubation methods (no medication, RSI, sedation without paralysis, and others), and ED visit year.We selected these potential confounders based on clinical plausibility and a priori knowledge [6,14,17].We also fit the unadjusted and adjusted multivariable logistic regression models with GEE in which VL was divided into C-MAC, McGrath, and other VLs (Airway Scope and GlideScope).
We conducted a series of sensitivity analyses to examine the robustness of our inference.First, we repeated the analyses for the outcomes with stratification by 1) BMI category (<25.0,25.0-29.9,and ≥30.0 kg/m 2 ), 2) primary indication (cardiac arrest and non-cardiac arrest), 3) modified LEMON (0 and ≥1), 4) intubation methods (RSI and non-RSI), and 5) intubator's specialty (transitional-year residents, emergency medicine residents, and other specialty residents).Lastly, we repeated the analysis by including intubations attempted by VL or DL with the use of a bougie.All P values of <0.05 were regarded as statistically significant.All statistical analyses were performed with STATA 14.1 (StataCorp, College Station, TX) and the EZR software (Saitama Medical Centre, Jichi Medical University, Saitama, Japan) -a graphical user interface for R (version 4.0.1;R Foundation for Statistical Computing, Vienna, Austria) [23].

Results
During the study period, there were 12,690 patients who underwent intubation in the ED.The JEAN-2 study recorded 12,346 patients (capture rate, 98.5%).After removing non-eligible patients (Figure 1), the current study included 5,261 patients who underwent an intubation attempt by resident physicians in the ED.Table 1 summarizes the baseline characteristics and airway management of patients.Overall, the median age was 72 years (IQR=59-81 years), and 38% were female.The VL and DL groups had some differences in the patient and airway management characteristics.For example, the VL group was younger and more likely to have a medical indication and to be intubated with RSI (P<0.05).

FIGURE 1: Study flow
During the study period, the JEAN-2 study recorded a total of 12,346 patients who underwent emergency airway management at one of the 15 emergency departments (capture rate, 98.5%).Of these, 5,261 patients were eligible for the current analysis.
Abbreviation: JEAN, Japanese Emergency Airway Network * We excluded fiberoptic intubation, lighted-style-assisted tracheal intubation, nasal intubation, cricothyrotomy, the use of supraglottic devices, and the use of bougie.

Sensitivity analysis
The associations between VL use and higher first-pass success were also observed in the sensitivity analyses (Table 3).For example, the use of VL was associated with a significantly higher first-pass success rate in the groups of BMI <25.0 kg/m   The associations between VL use and lower rates of intubation-related adverse events were also observed in the sensitivity analyses (Tables 5-7).For example, the use of VL was associated with a significantly lower rate of any adverse events in the groups of BMI <25.0 kg/m 2 (adjusted OR=0.64; 95% CI=0.50-0.83;P=0.001) and transitional-year residents (adjusted OR=0.44; 95% CI=0.31-0.61;P<0.001).Similarly, this association remained significant after stratification by primary indication, modified LEMON, and RSI use (all P<0.05).Lastly, the use of VL (including the use of bougie) was also associated with a significantly lower adverse event rate (adjusted OR=0.62; 95% CI=0.50-0.78;P<0.001; Table 4).

Discussion
Based on the prospective multicentre data of 5,261 ED patients intubated by resident physicians, intubation with VL achieved a significantly higher first-pass success rate than that with DL.In particular, intubation with C-MAC had the highest rate.In addition, we also found that the use of VL was associated with significantly fewer overall adverse events.The inference was consistent across the different statistical assumptions and most patient subgroups.
Few studies have reported the effectiveness of VL in intubations by residents.Within the sparse literature, a systematic review has reported that VL used outside the operating room setting by novices and trainees was associated with a higher first-pass success rate compared to DL [12].In an ED setting, a single-center study has shown that VL use was associated with a higher first-pass success rate compared to DL [14].Our findings are consistent with these previous studies.In contrast, research has reported that the association of VL and first-pass success rate differs by type of VL [13].For example, a secondary analysis of a prospective multicentre cohort study (the National Emergency Airway Registry-Ⅲ) has reported that only C-MAC was associated with a higher first-attempt success rate, while overall VL did not have a significantly higher or lower rate of intubation success [13].The apparent differences between the studies may be attributable to the differences in the study design, target population, data collection methods, analysis, or any combination of these factors.The current large prospective study builds on these earlier reports and extends them by demonstrating a robust relationship between VL use with a higher first-pass success rate and a lower adverse event rate.
There are several possible explanations for the observed relationship of VL use with first-pass success.First, the use of VL contributes to better glottic visualization [11].Critically ill patients have various difficult intubation factors (e.g., difficult airways and inadequate preoxygenation) [24].Improved glottic visualization by VL may have enabled residents to intubate such patients easily.In addition, the advantage of C-MAC is the use of a standard Macintosh-type blade, which may be familiar to physicians [25].Second, the dynamic supervision by attending physicians through the shared VL monitor may have also contributed to superior intubation outcomes [26].Our data also demonstrated that the use of VL was not associated with a lower rate of major intubation-related adverse events but associated with any and minor adverse events, consistent with previous reports showing the association between the use of VL and reduced oesophageal intubations [10,12].These findings can be explained, at least partially, by the improved intubation maneuvers with the use of DL because novice physicians use a conventional DL by applying extra force on the oral structures [27,28], thereby leading to airway trauma.Furthermore, the better glottic visualization with VL may have reduced the oesophageal intubation.Moreover, a higher first-pass success rate with VL likely led to fewer adverse events as multiple intubation attempts constitute a major risk factor for adverse events [6].Lastly, these potential mechanisms are not mutually exclusive.Notwithstanding the complexity of the mechanisms, the observed superiority of VL in ED intubations supports the use of VL as the first device of choice for resident physicians in the ED.

Limitations
This study has several potential limitations.First, the passive surveillance system adopted in this study may have led to self-reporting bias, resulting in a possible overestimation of the first-attempt success rate and underestimation of the adverse event rate.Regardless, we believe that a capture rate of 98.5% with standardized data collection ensured the quality of the study.Second, we excluded patients intubated with adjunctive devices (e.g., bougie).The use of a bougie may have resulted in a higher rate of successful initial intubation regardless of VL or DL [29,30].Indeed, our sensitivity analysis demonstrated a consistent inference.Third, with the increased use of VL over the years in Japan [9], there were some differences in patient characteristics between DL and VL groups.While we adjusted patent characteristics and ED visit years, there were potential unmeasured confounding factors (such as the actual skill and experience of each intubator).Fourth, the associations between VL use and higher first-pass success among transitional residents and emergency medicine residents were not significant probably due to the limited sample size in the subgroup analysis.Fifth, although time to intubation may be associated with adverse event rates, it is unclear because we have not collected time to intubation.Lastly, our study sample consisted of academic EDs in Japan.While it is tempting to dismiss the generalizability of our inferences, the observed relationship is clinically plausible [12][13][14][15].

TABLE 3 : Adjusted odds ratio for first-pass success, according to subgroups
Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, primary indications, modified LEMON, premedication use, and ED visit year as well as accounting for within-ED clustering.||Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, primary indications, modified LEMON, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.
Abbreviations: CI, confidence interval; DL, direct laryngoscope; OR, odds ratio; PGY, postgraduate year; RSI, rapid sequence intubation; VL, video laryngoscope * Multivariable logistic regression model with generalized estimating equations (GEE) adjusting for age, sex, primary indications, modified LEMON, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.† Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, modified LEMON, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.‡ Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, primary indications, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.§

TABLE 4 : Unadjusted and adjusted associations of video laryngoscope that include bougie use with first-pass success and adverse events in the emergency department
Abbreviations: CI, confidence interval; DL, direct laryngoscope; OR, odds ratio; VL, video laryngoscope; C-MAC, cipher-based message authentication codes Multivariable logistic regression model with generalized estimating equations (GEE) adjusting for age, sex, body mass index, primary indications, modified LEMON, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.* Other video laryngoscopes include Airway Scope and GlideScope.† Major adverse events include cardiac arrest, dysrhythmia, hypotension, hypoxia, and oesophageal intubation with delayed recognition.‡ Minor adverse events include regurgitation, airway trauma, dental or lip trauma, mainstem bronchial intubation, and oesophageal intubation recognized immediately.

TABLE 5 : Adjusted associations between video laryngoscope and any intubation-related adverse events, according to subgroups
Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, primary indications, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.§Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, primary indications, modified LEMON, premedication use, and ED visit year as well as accounting for within-ED clustering.||Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, primary indications, modified LEMON, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.
Abbreviations: CI, confidence interval; DL, direct laryngoscope; OR, odds ratio; PGY, postgraduate year; RSI, rapid sequence intubation; VL, video laryngoscope * Multivariable logistic regression model with generalized estimating equations (GEE) adjusting for age, sex, primary indications, modified LEMON, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.† Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, modified LEMON, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.‡ 2023 Sugaya et al.Cureus 15(10): e47563.DOI 10.7759/cureus.475639 of 14 P value

TABLE 6 : Adjusted associations between video laryngoscope and major intubation-related adverse events, according to subgroups
Multivariable logistic regression model with generalized estimating equations (GEE) adjusting for age, sex, primary indications, modified LEMON, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.†Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, modified LEMON, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.‡Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, primary indications, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.§Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, primary indications, modified LEMON, premedication use, and ED visit year as well as accounting for within-ED clustering.||Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, primary indications, modified LEMON, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.

TABLE 7 : Adjusted associations between video laryngoscope and minor intubation-related adverse events, according to subgroups
Multivariable logistic regression model with generalized estimating equations (GEE) adjusting for age, sex, primary indications, modified LEMON, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.†Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, modified LEMON, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.‡Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, primary indications, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering.§Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, primary indications, modified LEMON, premedication use, and ED visit year as well as accounting for within-ED clustering.||Multivariable logistic regression model with GEE adjusting for age, sex, body mass index, primary indications, modified LEMON, premedication use, intubation methods, and ED visit year as well as accounting for within-ED clustering. *