Videolaryngoscopes for Double Lumen Tube Intubations

Several regional surveys among thoracic anesthesiologists showed that the double-lumen endobronchial tubes (DLT) is still the first choice for lung separation.[1-3] The introduction of videolaryngoscopes (VL) enables the use of the DLTs instead of a bronchial blockers for lung separation in patients with difficult airway.[4-5]

There are numerous VLs, including, stylets, channeled and traditional VLs. The GlideScope®-assisted DLT insertion (Verathon Inc., Bothell, WA, USA), has been associated with variable intubation durations according to the experience of the intubating operators, despite superior vocal cord views.[6-7] The channeled VL, like as the Airtraq® (Prodol Limited, Viscaya, Spain) and the standard non-channeled blade of the King Vision™ (KVL™) (King Systems, Indianapolis, IN, USA), may offer additional benefits for DLT intubation in patients with limited mouth opening or restricted neck movement, [8-10] in whom the use of traditional VL like as the Glidescope® could be difficult. This is because of the larger outer diameter, the distal curvature and the increased rigidity of the DLT.[11]

Of note, the longer intubation durations with the use of different VL-assisted DLT insertion could be shortened with building up the operator's experience.

Comparing the effects of the three studied VLs on the time to DLT intubation in humans has not yet been studied.

We hypothesize that time to successful DLT intubation using the channeled VL, namely, the Airtraq® and KVL™, will be shorter than with the Macintosh (MAC) laryngoscope and GlideScope® VL when used by non-expert anesthesiologists. We will compare the effects of the MAC, GlideScope®, Airtraq® and KVL™ on the time to DLT intubation, laryngoscopic view, ease of intubation, number of intubation attempts, and number of optimization maneuvers in patients undergoing thoracic procedures using DLTs for one-lung ventilation (OLV).

Standard monitoring, and bispectral index (BIS), or state and response entropy based depth of anesthesia, (GE Healthcare, Helsinki, Finland) will be applied to all patients. Neuromuscular blockade will be measured with a train of four stimulation (TOF) of the ulnar nerve. Invasive blood pressure monitoring will be achieved by cannulation of the radial artery. After preoxygenation, general anesthesia will be induced with propofol 1.5 to 3mg kg-1 and fentanyl 2-3 µg kg-1 or remifentanil 0.05-0.2 ug Kg-1 min-1, titrated to achieve a BIS value < 60 or a state entropy values of less than 50 and a difference between response and state entropy of less than 10. Rocuronium (0.6 mg kg-1) will be given to facilitate intubation with an appropriate sized left-sided DLT.

In all groups, during the insertion of the DLT, exercise caution to avoid damage to the tracheal cuff by the upper teeth during its passage through the mouth opening. After DLT intubation, the tracheal cuff will be inflated and ventilation of the lungs commenced. Then, the correct position of its tip will be confirmed with a fiberoptic bronchoscope.

The first intubation attempt will be considered a failure if the trachea is not successfully intubated within 150 s or if the peripheral oxygen saturation (SpO2) decreases by 5%. Following an initial tracheal intubation failure, the participants will be allowed to use any maneuver and device they would normally use to navigate the DLT into the trachea.

Demographic data will be recorded including grade, experience with the VLs, and previous DLT intubation experience with the VLs.

All data, with the exception of the difficulty of device use score, will be recorded by blinded investigators.

A pilot study showed that the mean time to DLT intubation using the GlideScope® VL was 72 seconds with a standard deviation of 45.2 seconds. An a priori power analysis indicated that a sample size of 32 participants was sufficiently large to detect a 50% difference in the time to DLT intubation, during the use of the channeled VL, a type-I error of 0.008 (0.05/6 possible comparisons) and a power of 90%. We will add more patients (10%) for a final sample size of 35 participants to compensate dropping out during the study.

Data will be tested for normality using Kolmogorov-Smirnov test. Categorical data will be analyzed using the Fisher's exact test. Repeated-measures analysis of variance (ANOVA) will be used for continuous parametric variables, and the differences will be corrected by the post hoc Bonferroni test. Kruskal-Wallis test will be used for nonparametric values, and post hoc pairwise comparisons will be performed using the Wilcoxon rank-sum t test Continuous data will be presented as mean (standard deviation, SD), ordinal data will be presented as median (inter-quartile range), and categorical data will be presented as number (%). A value of p< 0.05 will be considered to be statistically significant.