Approval for the study was granted by the Shanghai Renji Hospital Ethics Committee (Ethical number: 2016[036]). Written informed consent was obtained from patients undergoing elective intra-thoracic surgeries requiring double lumen intubation. The present trial was registered at http://www.chictr.org.cn (registration number ChiCTR1900025718; principal investigator: Z.L.H.; date of registration: September 6, 2019).
Inclusion criteria were as follows: age 18–75 years old; ASA I–II, BMI<35 kg/m2, with Mallampati score of 1 or 2. All Mallampati scores were assigned by the same observer. Exclusion criteria were as follows: presence of any predictors of difficult intubation; Mallampati score >=3; inter-incisor distance <3 cm; thyromental distance <6 cm; neck extension <80°from neck flexion; cervical spine instability; history of difficult endotracheal intubation or difficult mask ventilation; and severe pulmonary ventilation dysfunction or risk of pulmonary aspiration.
Eligible patients were enrolled on the basis of the CONSORT Statement Extension for Randomized Controlled Trials of Nonpharmacological Trials, as displayed in the flowchart in Fig. 1. We randomly assigned 90 patients to GlideScope (Verathon Medical, Bothwell, UT, USA), C-MAC(D) (Karl Storz GmbHand Co.KG, Tuttlingen, Germany), or Macintosh groups. This was done using a closed envelope technique using a computer-generated block randomization method in blocks of 15. Before the study, the computerized randomization was performed and the allocation results were placed in individual numbered and sealed envelopes. The researcher responsible for recruitment blinded to the allocation result. After a patient was consented for the study, allocation was revealed. All endotracheal intubations were performed by five anesthesiologists with ten years’ working experience skilled in videolaryngoscopy.
Left‑side or right‑side 32Fr/35Fr Mallinckrodt™ DLTs (Mallinckrodt Medical, Athlone, Ireland) were selected for female patients and 35Fr/37Fr DLTs for male patients depending on whether their heights were below or above 155 cm for females and 165 cm for males. If the operation side was the left, right-side DLT was used; otherwise, the left-side DLT was used. To facilitate intubation, the distal 10–12 cm concavity of the DLT (with the stylet in situ) was molded along the blade convexity in each group. The tracheal and the bronchial cuffs of the DLT tubes were lubricated with sterile Surgilube.
No premedication was given before induction. Standard monitoring prior to induction included ECG, invasive arterial blood pressure, SpO2, and end-tidal carbon dioxide. After pre-oxygenation with 100% oxygen, anesthesia was induced with intravenous midazolam 0.05 mg.kg-1, propofol 1.5 mg.kg-1, fentanyl 5 μg.kg-1, and rocuronium 0.6 mg.kg-1. Ninety seconds after rocuronium administration, DLT intubation was performed using the allocated laryngoscope. The DLT was inserted with the distal concavity facing anteriorly until the bronchial lumen cuff passed the vocal cords. The stylet was then removed, and rotation was performed while tube was advanced. The left DLT rotated 90° counter‑clockwise, and the right DLT was rotated 90° clockwise to enter the respective mainstem bronchus. Hemodynamic changes were monitored during induction. If systolic blood pressure fell below 80 mmHg, ephedrine 5 mg was administrated intravenously. Atropine 0.5 mg was given for heart rate below 50 beats per minute. After the tip of the DLT was located in the targeted bronchus, the tracheal cuff was inflated and ventilation of the lungs started. Fiberoptic bronchoscopic assessment of adequate bronchial cuff placement was followed by DLT placement.
DLT insertion time was defined as from the time the laryngoscope passed the patient’s lips until three complete end-tidal carbon dioxide cycles were displayed on the monitor. Intubation success rate at the first attempt was recorded by the same observer. The difficulty of DLT insertion and delivery were assessed by the operator, using NRS ranging from 0 to 10. The NRS results were grouped as 0 = none, 1–3 = mild, 4–6 = moderate, and 7–10 = severe. C/L degrees were classified as four degrees (I, IIA, IIB, and III) and were assessed by the same operator. If the degree was not class I, external laryngeal pressure was provided by an assistant. The time taken for fiberoptic bronchoscopy was defined as the time from endobronchial intubation to placement confirmation using fiberoptic bronchoscopy. The operators examined blade surfaces for blood after removal. Hemodynamic parameters (mean arterial blood pressure and heart rate) were recorded 10 minutes before induction and 1, 3, and 5 minutes after intubation. After the assessment by fiberoptic bronchoscopy, the oral cavity, pharynx, larynx and teeth were examined for signs of laceration or bleeding by an independent investigator who was unaware of the type of laryngoscope used. One day after surgery, an independent investigator interviewed patients to assess the presence of sore throat and hoarseness of voice.
Statistical analysis
Based on previous studies [13,14], we determined that the mean intubation time for the GlideScope would be 45.6 s with a standard deviation of 10.7 s, and that of the C-MAC(D) would be 32.27 s with a standard deviation of 11.13 s [12]. Factoring possible drop-outs, we recruited 30 patients in each group, with an alpha value of 0.05 and a beta value of 0.2.
Data were expressed as median (interquartile range (IQR) [min–max]), mean ± SD, or absolute numbers, as required. Statistical analysis was performed using SPSS 13.0 (SPSS Inc., Chicago, IL). The Kruskal–Wallis test was used to analyze independent samples (the success rate at the first attempt, the times of intubation attempts, the DLT insertion time, the number of external laryngeal pressure applications, C/L degree, and NRS of DLT delivery and insertion). The Chi-square and Student–Newman–Keuls tests were used to analyze demographic data and the incidence of complications. For the analysis of hemodynamic response to intubation, a repeated-measures analysis of variance was used. Statistical significance was considered at P <0.05.