Background

The paramount importance of providing continuous chest compressions with the shortest possible interruption in the context of cardiopulmonary resuscitation (CPR) has already been proven to be prognostically relevant [1], is highlighted accordingly in the corresponding resuscitation guidelines [2], and undisputed. However, regarding airway management during CPR there has been much discussion in recent years. While several studies suggest tracheal intubation as the preferred option for airway management during CPR [3–6], randomised controlled trials found no benefit in relevant outcome parameters for patients with tracheal intubation compared to supraglottic airway management [7–9]. One reason for this might be interruptions of chest compressions during laryngoscopy and tracheal intubation [10–12]. At best, intubation should be performed with ongoing chest compressions, thereby eliminating the associated no-flow time.

In recent years, video laryngoscopy (VL) has become established in everyday elective anaesthesia as well as in emergency situations and in the preclinical setting. In comparison to direct laryngoscopy (DL), a significantly better view of the vocal cord level was demonstrated, even in the presence of difficult airway anatomy [13, 14]. These findings were also confirmed in everyday preclinical work [15]. In addition, an increased first-pass success (FPS) rate in emergency situations has been shown [16–19]. This effect was demonstrated in inexperienced users in particular [20–22]. Overall, the data situation remains inconclusive. However, we hypothesize that VL facilitates safe and expeditious tracheal intubation even under the aggravated conditions of ongoing chest compressions. To investigate this, in this randomized controlled study we recorded FPS rate and time to successful intubation with VL and compared them to DL in an adult manikin model with uninterrupted chest compressions.

Materials and methods

In December 2021, 28 anaesthesiologists and 28 anaesthesia nurses were recruited from the University Clinic for Anaesthesiology and Pain Medicine at the Inselspital in Bern, Switzerland. Participation was voluntary and free of charge. Written consent to use the data was obtained from all participants and they were assured that performance in the study scenario had no influence on the course result or any professional consequences. Subjects were blinded to the endpoints: they were only told that the study wanted to investigate whether endotracheal intubation could be performed "quickly and safely" with ongoing resuscitation. The participants’ clinical experience in anaesthesiology ranged from less than one to 34 years. All participants had extensive training in both DL and VL. Before the start of the study, the cantonal ethics committee of Bern confirmed that no authorization was required (Req-2021-01338, 11/22/2021). This study was designed as a pragmatic pilot study, and participants were recruited as part of an in-hospital training course. Thus, the size of the group, the profession of the subjects, and also their clinical experience were predetermined. A power analysis was deliberately omitted due to this setting. Nonetheless, effect sizes were statistically examined during the analysis.

Results

Results of all 28 anaesthetists and 28 anaesthesia nurses were included in the study. The participants’ experience in anaesthesia ranged from 0.75 to 34 years. The subjects were randomised into the two groups VL and DL, each with 28 participants (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Participants’ characteristics.

https://doi.org/10.1371/journal.pone.0281186.t001

FPS was 100% (95% CI: 87.9% - 100.0%) in the VL group and 67.8% (95% CI: 49.3% - 82.1%) in the DL group [group difference: 32.2% (95% CI: 17.8% - 50.8%), p<0.001, Table 2]. The median TTI was 27.5 seconds (IQR: 21.8–31.0 seconds) in the VL group and 30.0 seconds (IQR: 26.5–36.5 seconds) in the DL group (p = 0.019, Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Primary outcome: First-pass success presented with counts and percentages.

https://doi.org/10.1371/journal.pone.0281186.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Secondary outcome: Time to intubate–only in the case of FPS.

https://doi.org/10.1371/journal.pone.0281186.t003

We performed a multivariable logistic regression analysis of FPS rate to assess the influence of experience or occupational group. Here, experience was not a significant predictor (p = 0.6). Likewise, no significant difference between the occupational groups could be demonstrated (p = 0.4, Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Multivariable regression for the outcome first-pass success.

https://doi.org/10.1371/journal.pone.0281186.t004

Discussion

In our randomized manikin study, we were able to show that during ongoing chest compressions a higher FPS rate can be achieved and that the time until successful intubation is significantly shorter with VL compared to DL.

To improve the chances of a favourable neurological outcome for patients after cardiac arrest, every effort has to be made not only to achieve the highest possible FPS [27], but also to optimize circulation during CPR with as little no-flow time as possible. Airway management during CPR has been a highly discussed and intensively investigated research area during the last years. Three studies suggest that tracheal intubation is associated with higher short- and long-term survival rates and better neurological recovery [4–6]. One large randomized controlled trial showed no significant difference in favourable functional outcome 30 days after OHCA between patients ventilated by a supraglottic airway compared with a tracheal tube [9]. Wang et al. [7] concluded from their PART trial that a strategy of initial laryngeal tube insertion was associated with significantly greater 72-hour survival compared with a strategy of initial tracheal intubation. However, this study has the significant limitation that the first pass success rate in patients who were tracheally intubated was only 51%. Therefore, a possible explanation for the inconclusive results of these studies on airway management might be the low success rates in tracheal intubation and interruptions of chest compressions for this procedure in contrast to positioning of an SGA. Future research on airway management during CPR should focus on how to improve FPS and reduce no-flow time for tracheal intubation.

According to our results, VL for airway management during CPR allows tracheal intubation with an excellent FPS even during uninterrupted chest compressions and performs significantly better than DL. The practicability and superiority of the VL—even with difficult airway anatomy, in emergency situations as well as complex preclinical situations—have been shown in several studies [16, 18, 19]. In addition, VL is also a suitable technique for inexperienced users to establish a safe airway in an emergency situation [28, 29]. This could even be demonstrated in beginners in the hospital setting [20–22]. Specifically, in a recent study of novice medical students, Keresztes and colleagues found that VL was superior to DL in terms of TTI in a normal airway and non-inferior in a difficult airway. In addition, a significantly smaller oesophageal false intubation rate could be observed in both cases [30].

Our regression analysis did not find an effect of either profession (physician, nurse) nor working experience in anaesthesia as a significant predictor of FPS. This result is interesting in that it could be postulated that users with many years of experience might be able to intubate more safely using DL. This could be assumed since these participants were regularly intubating patients using DL long before the introduction of VL. In addition to the proven advantage of VL for inexperienced users, it does not seem to be inferior for experienced practitioners either. Therefore, VL may provide further benefits in the prehospital setting, and—according to the finding of two recent studies—especially for providers with little experience in tracheal intubation [18, 19].

Although VL is becoming increasingly widespread it cannot be assumed that this technology is available everywhere. Nonetheless, for example, in German HEMS and the Swiss Air-Rescue (REGA) VL is used as the first-line device [13, 18]. Of course, the significantly higher costs for acquisition and maintenance compared to a conventional laryngoscope need to be noted. However, the trend here is towards falling prices and thus increasing availability [31].

Our study has clear limitations. Firstly, one issue is the use of a manikin that can only partially depict anatomical conditions. However, the benefit of using manikins is that we were able to establish standardized airway conditions for each participant. Secondly, due to the fact that the subjects were only present for one day, a cross-over design was not possible. In the setting of a cross-over design, the two intubations would have taken place within a short period of time and we then would expect a significant bias due to the short-term experience and training effect. This would not fulfil our goal of conducting a pragmatic trial simulating the setting of prehospital emergency care medicine where personnel are regularly confronted with unfamiliar situations for tracheal intubation. Thirdly, as this study was designed as a pragmatic pilot study, we could not complete a power analysis given the fixed population size. In this context, the small sample group must also be noted, despite the proven effect size. Fourthly, the clinical relevance of the absolute difference of 2.5 seconds in TTI can be questioned. For this reason, we chose FPS as the primary endpoint, which seems to be outcome relevant in CPR [27]. Fifthly, we did not include the experience with tracheal intubation as potential confounder for our multivariable regression model. However, this variable would very closely correlate with working experience in anaesthesia in our clinic at a university hospital. Therefore, we constrained to the working experience in anaesthesia as a simply and objectively measurable parameter.

Conclusions

In summary, the results of our study provide a good basis for a large-scaled clinical randomized controlled trial comparing VL and DL under ongoing chest compressions in CPR. We were able to show significantly higher FPS for VL under these circumstances as compared to DL. Airway management experience and professional group were not significant predictors of FPS.

Supporting information

Acknowledgments

The authors would like to thank the volunteers for participating in this study, as well as the Bern Simulation- and CPR-Centre (BeSiC) at the Bern University Hospital for providing facilities and equipment. We also would like to thank Jeannie Wurz, Medical Editor, and Darren Hight, PhD, Bern, Switzerland, for their editorial assistance.