Video laryngoscopy versus direct laryngoscopy for first-attempt tracheal intubation in the general ward

Tracheal intubation can be a hazardous procedure outside of the operating room, as it tends to be performed by inexperienced junior trainees and involves physiologically unstable patients. Accordingly, tracheal intubation outside of the operating room is associated with a higher rate of complications than the corresponding procedure inside the operating room [1–3]. Furthermore, a successful first-attempt intubation is important in the emergency setting [4] because related complications are associated with multiple intubation attempts [5, 6].

A video laryngoscope, defined as a laryngoscopic device to which a camera has been attached to the tip of the blade, could assist airway management by improving visualisation of the glottis in critically ill patients [7–10]. According to a few studies, video laryngoscopy yields a greater rate of first-attempt intubation success than direct laryngoscopy [7, 11]. However, a recent randomised clinical trial in an intensive care unit (ICU) setting found that video laryngoscopy did not yield higher first-attempt tracheal intubation success rate than direct laryngoscopy and was associated with higher rates of complications [12]. Therefore, the superior device for first-attempt intubation success in critically ill patients remains controversial. Furthermore, limited data are available regarding tracheal intubation in general ward settings [5, 13].

The objective of this study was to investigate which laryngoscopy mode is associated with higher first-attempt intubation success in a general ward. We hypothesised that tracheal intubation with the video laryngoscope would be associated with increased successful intubation on first attempt compared with the direct laryngoscopy.

During the study period, 1312 tracheal intubations were performed in general ward. A total of 354 patients were excluded which received tracheal intubation during cardiac arrest, 4 younger than 19 years, 4 which initially received tracheal intubation using a supraglottic airway device, 4 which records were unavailable. Because thirty-nine patients were intubated twice, and one patient was intubated three times, 41 cases also excluded. Among the 958 intubations, the initial laryngoscopy mode was video laryngoscopy in 493 (52%) and direct laryngoscopy in 465 (48%) (Fig. 1). As the alternative techniques, an i-gel and a laryngeal tube was used. Gum elastic bougie was used in four cases for 102 failures at first pass in the video laryngoscopy group and was not used for 192 failures in the direct laryngoscopy group. Tracheostomy was conducted in three patients. One tracheostomy was conducted after a failure of cricothyroidotomy. The other two tracheostomies were performed after failures of tracheal intubation because oxygen saturation was maintained at > 90%. One patient died after a failure of emergency cricothyroidotomy in a ‘cannot intubate, cannot oxygenate’ situation.

The overall first-attempt intubation success rate was 69%, with rates of 79% and 59% in the video laryngoscopy and direct laryngoscopy groups, respectively (Fig. 2). Among 192 unsuccessful tracheal intubation with direct laryngoscopy, second intubation attempt was performed by 90 (47%) video laryngoscopy. On the contrary, 102 unsuccessful tracheal intubation with video laryngoscopy changed to 7 (7%) direct laryngoscopy. Among inexperienced operators, the first-attempt intubation success rate was higher with video laryngoscopy (75%) than with direct laryngoscopy (52%, p < 0.001; Table 7). The incidence of difficult intubation was 294 (31%). The incidence of difficult intubation was higher in the direct laryngoscopy group (41%) than in the video laryngoscopy group (21%, p < 0.001; Table 3).

Variable	Total n = 958	Direct laryngoscopy n = 465	Video laryngoscopy n = 493	p
First-attempt success, n (%)	664 (69.3)	273 (58.7)	391 (79.3)	< 0.001
Difficult intubation*, n (%)	294 (30.7)	192 (41.3)	102 (20.7)	< 0.001
Cormack–Lehane grade†, n (%)				0.024
1	721 (76.8)	328 (72.6)	393 (80.7)
2	170 (18.1)	95 (21.0)	75 (15.4)
3	38 (4.0)	22 (4.9)	16 (3.3)
4	10 (1.1)	7 (1.5)	3 (0.6)
No. of intubation attempts	1 [1, 2]	1 [1, 2]	1 [1, 1]	< 0.001
Intubation duration‡ (min)	4 [3, 6)	4 [3, 7]	4 [3, 5]	< 0.001

Values are expressed as median (interquartile range) or n (%)

* Defined as more than two attempts of intubation

^†Reflects glottis visualisation, with a score range of 1 (good) to 4 (no glottis visualisation)

^‡Defined as the time between the infusion of induction medication and confirmation of endotracheal tube placement by capnography or chest radiography

In the power analysis referent to previous studies, we achieved a statistical power of 99–100% for our study populations (493 in the video laryngoscopy group and 465 in the direct laryngoscopy group). Therefore, our study has an adequate sample size and appropriate power. Hosmer–Lemeshow goodness-of-fit testing revealed that multivariate models were well fitted (χ² = 6.419, p = 0.600 for first-attempt intubation success; χ² = 1.666, p = 0.948 for intubation-related complications, respectively).

To our knowledge, this is the first study of the efficacy of video laryngoscopy when performed by non-anaesthesiologists in a general ward setting. The results of this study showed a higher first-attempt intubation success rate with video laryngoscopy than with direct laryngoscopy. However, there was no significant association of video laryngoscopy with lower intubation-related complications. Additionally, the number of intubation attempts was significantly lower in the video laryngoscopy group. Video laryngoscopy was also associated with improved visualisation of the glottis when compared with direct laryngoscopy.

In the general ward, tracheal intubation tends to be performed by less experienced operators and without capnography. Several studies have demonstrated that residents conducted 76–83% of first intubation attempts in critically ill patients [12, 19]. Bowles et al. reported that capnography was used in only 20% of general ward intubations [13]. In our hospital, all general ward tracheal intubations performed by inexperienced operators were supervised, involved capnography, and featured assistance by highly trained MET nurses. Accordingly, our study observed an overall first-attempt success rate of 69%, which is comparable to the rates of 56–75% reported for the ICU and emergency department (ED) [7, 9, 12, 20]. Furthermore, intubation-related complications were observed in 26% of the patients, and this was lower than the rates of 28–39% in the ICU and ED [1–3, 20–22].

Previous studies have reported an association of video laryngoscopy with an improved first-attempt success rate for difficult intubation cases; [10] in the ICU [7, 8, 18, 23, 24], ED [25, 26], and pre-hospital settings; [27] and in cases of in-hospital cardiac arrest [28]. However, recent randomised clinical trials (RCTs) [12, 20, 29] in ICU settings found that video laryngoscopy did not yield higher first-attempt tracheal intubation success rate than direct laryngoscopy. The MACMAN trial [12] emphasised that improved glottis visualisation with video laryngoscopy did not translate into an improved first-attempt success rate. The authors suggested that tracheal catheterisation under indirect vision was more difficult. However, our analysis suggests that first-attempt intubation success depends on complex interactions of the mode of laryngoscopy, level of operator experience, and prediction of difficult airways. In the MACMAN trial [12], 83% of tracheal intubations were performed by non-expert operators. Although another RCT by Janz et al. [20]. included tracheal intubations conducted by pulmonary and critical care medicine fellows, the fellows had each performed fewer than 5 tracheal intubations with video laryngoscopy during the study period and had a previous experience of only 10 tracheal intubations using video laryngoscopy. Also, in a single-centre pilot RCT by Griesdale et al. [29], video laryngoscopy did not yield an improved first-attempt success rate. However, that study included only 40 tracheal intubations conducted by novices (e.g. medical students or non-anaesthesiology residents), with an overall first-attempt success rate of 38%. Although there is an argument to use of video laryngoscopy could shorten the learning curve, a recent study of the GlideScope device demonstrated that 76 attempts would be required to master intubation whereas traditional cut-off for direct laryngoscopy is set at 50 attempts [30]. In other words, these RCTs were limited by the lack of sufficient time for the operators to acquire a reliable level of experience with video laryngoscopy.

There is a concern of video laryngoscopy is associated with severe complications. The MACMAN trial demonstrated that video laryngoscopy was associated with severe life-threatening complications and higher incidence of severe desaturation [12]. The results of our study did not show that video laryngoscopy reduced intubation-related complications, and video laryngoscopy was associated with more severe desaturation than direct laryngoscopy. However in our study, the incidence rates of a pre-intubation SpO₂ < 80% and predicted difficult airway were higher in the video laryngoscopy group. This may affect the rate of intubation-related complications (Additional file 3: Table S3).

In this study, hospital mortality was higher in the video laryngoscopy group than in the direct laryngoscopy group. However, meta-analysis of the 12 RCTs showed that in-hospital mortality was not significantly different between video laryngoscopy and direct laryngoscopy [31]. We think many factors such as disease severity, organ dysfunction, underlying disease and comorbidities also contribute to mortality.

There were similar complications associated with the video laryngoscope such as the insertion of a styletted endotracheal tube through the right palatopharyngeal arch [32, 33]. In addition, even small amounts of oropharyngeal blood or vomitus can easily contaminated the lens of the video laryngoscope [34]. The GlideScope with a hyper-angulated blade is not the best device to learn intubation skills and required stylet mandatory. We cannot guarantee whether other types of video laryngoscopes would yield similar results.

This study had several limitations. First, this was a retrospective analysis, and the operators chose the laryngoscopy mode according to their preference. Thus, the device used for the first intubation attempt was not randomly assigned. Accordingly, the level of operator experience may have influence the device success rates. However, we performed an adjustment for differences in baseline characteristics of patients using a propensity-score matching to reduce the effect of confounding factors. As we described in Additional file 2: Tables S2 and Additional file 4: Table S4, the results of the propensity-score matching showed that the use of video laryngoscopy is associated with increased odds of successful intubation on first attempt compared with the direct laryngoscopy. Second, we did not demonstrate the Mallampati score and the percentage of glottis opening (POGO) score. Unfortunately, these variables were not routinely recorded in our intubation data collection sheet. Third, the experienced operators in our study may differ from those in other studies [7, 12, 20]. In previous studies, the authors defined an expert operator as a critical care medicine attending physician or anaesthesiologist. However, it is unlikely that every tracheal intubation in real-world general ward settings is performed under the supervision of such experts. In our hospital, an anaesthesiologist is not always available for tracheal intubation outside of the operating room. A rescue airway team, which includes an anaesthesiologist, is activated only in ‘cannot intubate, cannot oxygenate’ situations. In addition, the results of our analysis showed that the first-attempt success rate of experienced operators was sufficiently high to allow them to supervise inexperienced operators. Therefore, stratification of the level of operator experience according to board certification seems appropriate. Fourth, only 21% of patients received paralytics, which is associated with improved first-attempt success, improved Cormack–Lehane grade, and decreased procedure-related complications in critically ill patients [35]. Low rate of neuromuscular blockade use could have decreased the first-attempt success rate, and the video laryngoscopy group seems to be an awake intubation group which require expertise. However, in this study the first-intubation success rate was lower in direct laryngoscopy group, which usage of neuromuscular blockers was higher. In addition, after propensity-score matching, the first-attempt success rate was greater in video laryngoscopy group than direct laryngoscopy group. Therefore, further study is needed for this issue.

A strength of our study is the enrolment of a relatively large number of general ward patients who underwent tracheal intubation; accordingly, the results of our study will provide information to support the selection of intubation devices in a general ward setting.

In this study, we observed a higher first-attempt intubation success rate with video laryngoscopy than with direct laryngoscopy in a general ward setting. However, video laryngoscopy did not reduce intubation-related complications. Furthers trials on best way to perform intubation in the emergency settings are required.