This prospective, randomized, crossover manikin study was conducted at the Nagoya City University Hospital and the Nagoya City East Medical Centre from April 30, 2020 to May 11, 2020. The study protocol was reviewed and approved by the Nagoya City University Graduate School of Medical Sciences and Nagoya City University Hospital Institutional Review Board. After explaining the study flow to the participants by verbally and the video made for instruction, we obtained written consent for study participation. Patients were not involved in the study.
We recruited medical personnel who were working in the fields of anesthesia and intensive care without previous experience with aerosol boxes from five hospitals (Nagoya City University Hospital, Nagoya City East Medical Center, Kainan Hospital, Kariya Toyota General Hospital, and Aichi Children's Health and Medical Center) in Japan. Per standard recommendations, only experienced physicians will perform tracheal intubation for patients with COVID-19; therefore, our study only included personnel with > 2 years of dedicated anesthesia experience [1–8]. We designed and conducted this study not only as a research trial but also as a simulation training for clinical use in the operating room.
In this study, we compared the following three types of laryngoscopes: Macintosh direct laryngoscope (Macintosh), i-view, and C-MAC. The C-MAC is a video laryngoscope with an external display, while the i-view is a display-integrated, one-size-fits-all (equivalent to a size 4), single-use video laryngoscope. A reused size-3 blade was used for Macintosh, and a single-use size-3 Macintosh type blade was used for C-MAC. The AirSim Combo Bronchi X (TruCorp, Lurgan, Ireland) manikin was used for all procedures. A 7.0-mm tracheal tube with a stylet, angled by each participant, was used. Although some arrangements have been reported [16, 17], we created and used an acrylic box based on the original version of the report [10].
Before the main measurement using the box, participants were trained to familiarize themselves with the three laryngoscopes and the manikin used in the study. The training was conducted in the same manner as the main study, except in the following order (Macintosh, i-view, and C-MAC) without the box. All the training and main studies were conducted in an operating room at the two hospitals (Nagoya City University Hospital and Nagoya City East Medical Center). The manikin was placed on the operating table in the supine position under the box. Both the manikin and the box were fixed with tape on the table so that the top of the manikin's head was 10 cm away from the box. During training and the main test, the participating physician wore a long-sleeved gown, double gloves, a surgical mask, face shield or goggles and a surgical cap. The N95 mask was not used because they are in short supply and have limited influence on intubation procedures. Neither a covering hood nor a powered air-purifying respirator was used. The height of the operating table was adjusted per the participant. In the box, a laryngoscope was placed on the left side of the manikin, and a tray was placed on the right side, where a tracheal tube and a cuff syringe with 8 ml of air were prepared. Six l/min of oxygen was administered to the manikin with a facemask. The participants removed the mask, opened the manikin’s mouth and picked up the laryngoscope. Then the participants picked up a tracheal tube, performed tracheal intubation, removed the stylet and inflated air into the cuff of the tube by themselves, considering the minimal number of personnel in the operating room. The participants removed the outer glove on their right hand and grasped the reservoir bag for ventilation. The investigators (TN, YS) stood on the right side of the manikin, recording the intubation time with a stopwatch, and helped connect the anesthesia circuit to the tracheal tube. The intubation time was defined as the time between holding the laryngoscope and confirming the first expansion in both lungs. An intubation time > 60 s, esophageal intubation, or single-lung intubation were considered to indicate failure. After each procedure, the participant assessed the Cormack-Lehane grade and the subjective difficulty scale score of tracheal intubation (numeric rating scale 0–10, 0: no difficulty, 10: highest difficulty). In the training, participants performed tracheal intubation thrice using each laryngoscope, without the box. If the participants failed, an additional procedure was performed to achieve successful completion.
At least 2 h after the training, the participants began the main part of the study using the aerosol box. To compare the three laryngoscopes, we used a randomized crossover design by dividing the participants into six groups and testing them in the determined order (Figure 1). The investigator (TN) not participated in the study performed computer-generated randomization and allocated participants to the six groups. The allocation was blinded to the participant until immediately before preparing the determined laryngoscope. The participants performed tracheal intubation on the manikin with the box three times using each assigned laryngoscope. A washout period of at least 2 h was required before the next laryngoscope was used. The primary outcome was the intubation time. The secondary outcome included the success rate, Cormack-Lehane grade and subjective difficulty scale score.
Statistical Analysis
Based on a preliminary analysis performed by investigators who did not participate in the study, we estimated an intubation time of 25 s for C-MAC and 35 s for i-view. With an α error of 1.67% (adjusted for Bonferroni method), power of 90%, SD of 7, and correlation coefficient of 0.5 for 2-tailed statistical analysis, we arrived at a minimum sample size of 10 participants. However, we recruited as many eligible physicians as possible in the study because it also aimed to provide simulation training and increase the secondary endpoint estimate accuracy (success rate).
To compare the intubation time, Cormack-Lehane grade, and subjective difficulty scale score of the three laryngoscopes, we used the median values of the three measurements, considering the learning effect. The intubation time of i-view showed a non-normal distribution (Shapiro-Wilk test P<0.05); the intubation times of the three laryngoscopes are presented as the median [interquartile range (IQR)] values. The Cormack-Lehane grade and subjective difficulty scale score are also presented as median [IQR] values. We used the Friedman test to compare the three devices’ performances. If a significant difference was found, the Wilcoxon signed-rank test with Bonferroni adjustment was used for pairwise comparisons. We compared the proportion of participants who successfully performed all three procedures between the three different laryngoscopes using the Cochran Q test. We also compared outcomes with and without the aerosol box in each laryngoscope. All the statistical analyses were performed using R software (version 3.6.3, R Foundation for Statistical Computing, Vienna, Austria). A P value of <0.05 was considered statistically significant.