Original Contributions

Transcription

1 doi: /j.jemermed The Journal of Emergency Medicine, Vol. 42, No. 6, pp , 2012 Published by Elsevier Inc. Printed in the USA /$ - see front matter Original Contributions DIFFICULT AIRWAY MANAGEMENT IN THE EMERGENCY DEPARTMENT: GLIDESCOPE VIDEOLARYNGOSCOPY COMPARED TO DIRECT LARYNGOSCOPY Jarrod M. Mosier, MD, Uwe Stolz, PHD, MPH, Stephen Chiu, BA, and John C. Sakles, MD Department of Emergency Medicine, University of Arizona College of Medicine, Tucson, Arizona Reprint Address: Jarrod M. Mosier, MD, Department of Emergency Medicine, University of Arizona College of Medicine, 4233 N. Rio Cancion Dr. #119, Tucson, AZ 85718, Abstract Background: Videolaryngoscopy has become a popular method of intubation in the Emergency Department (ED), however, little research has compared this technique with direct laryngoscopy (DL). Objective: To compare the success rates of GlideScope (Verathon Inc., Bothell, WA) videolaryngoscopy (GVL) and DL in emergent airways with known difficult airway predictors (DAPs). Methods: We evaluated 772 consecutive ED intubations over a 23-month period. After each intubation, the physician completed a data collection form that included: demographics, DAPs, Cormack-Lehane view, optical clarity, lens contamination, and complications. DAPs included: cervical immobility, obesity, small mandible, large tongue, short neck, blood or vomit in the airway, tracheal edema, secretions, and facial or neck trauma. Primary outcome was first-attempt success rates. Multivariate logistic regression was performed to evaluate the odds of failure for DL compared to GVL. Results: Firstattempt success rate with DL was 68%, GVL 78% (Fisher s exact test, p = 0.001). Adjusted odds of success of GVL compared to DL on first attempt equals 2.20 (odds ratio [OR] 2.2, 95% confidence interval [CI] ). After statistically controlling for DAPs, GVL was more likely to succeed on first attempt than DL (OR 3.07, 95% CI ). Logistic regression of DAPs showed that the Prior Presentations: Mediterranean Emergency Medicine Congress V, Valencia, Spain, September 2009; Western Society of Academic Emergency Medicine, Sonoma, CA, March 2010; Society for Academic Emergency Medicine, Phoenix, AZ, June presence of blood, small mandible, obesity, and a large tongue were statistically significant risk factors for decreasing the odds of success with DL and increasing the odds of success of GVL. Conclusion: For difficult airways with the presence of blood or small mandible, or a large tongue or obesity, GVL had a higher success rate at first attempt than DL. Published by Elsevier Inc., Keywords endotracheal intubation; videolaryngoscopy; GlideScope; direct laryngoscopy; difficult airway Background INTRODUCTION The difficult airway is one in which the intubator predicts difficulty in securing an airway. Several anatomic and pathologic conditions have been identified that, if present, can reliably predict a difficult airway (Table 1). In contrast, a failed airway is one that is not predicted but is experienced after three failed attempts, or that has inability to maintain adequate oxygen saturations (1). The significance of a predicted difficult airway is that it may require alternative management strategies based on how well the patient can be oxygenated and ventilated through alternative means, and the physician s confidence in securing the airway with paralysis. If oxygenation cannot be maintained, the airway becomes a crash, or failed, airway. RECEIVED: 8 October 2010; FINAL SUBMISSION RECEIVED: 28 March 2011; ACCEPTED: 5 June

2 630 J. M. Mosier et al. Table 1. Difficult Airway Predictors Anatomic Obesity Large tongue Short neck Small mandible Cervical immobility Importance Modern direct laryngoscopy with the Macintosh laryngoscope was developed in the 1940s as a method of aligning the oral, pharyngeal, and laryngeal axes to provide the operator a direct view of the glottic inlet so an endotracheal tube could be inserted under visualization (2). Any factor that prohibits the aligning of these axes, or obscures the view of the vocal cords, creates a difficult or failed airway situation. Several devices exist for managing difficult airways, including flexible fiberoptic scopes, intubating laryngeal mask airways, optical and lighted stylets, and videolaryngoscopy. Ultimately, if these modalities fail or when adequate oxygenation cannot be maintained, surgical airway management is the next step (1). Videolaryngoscopy may provide superior intubating conditions compared to direct laryngoscopy that allows the operator to manage more difficult airways with rapid sequence intubation (RSI) (3 11). Platts-Mills et al. (2009) report the only published comparison between direct laryngoscopy and video laryngoscopy in the emergency department (ED) setting (12). They found an equivalent success rate between the devices with minimal to no previous experience with video laryngoscopy, suggesting it is at least as good as, if not superior to, direct laryngoscopy. Goals of this Investigation In this study we examined the comparison between direct laryngoscopy and GlideScope (Verathon Inc., Bothell, WA) videolaryngoscopy in the presence of known difficult airway predictors in the ED. Primary outcome was first-attempt success rates between the devices in the presence of each difficult airway predictor listed in Table 1. Study Design METHODS Pathologic Blood Vomit Airway edema Facial or neck trauma This was a 23-month retrospective review of prospectively collected data of all ED patients intubated between July 1, 2007 and May 31, A simple one-page data collection sheet was developed for the Continuous Quality Improvement database and was completed by the operator immediately after each intubation was performed. Structured data forms were cross-referenced to professional billing records to identify any missing data forms. If an intubation was identified without a completed form, the operator was sent a blank form for completion immediately. Study Setting The study was conducted at a tertiary urban university hospital with a Level I trauma center and annual ED census of approximately 60,000 patients. The ED is staffed full time with emergency medicine residents (postgraduate years 1 3) and emergency medicine attendings. Typical of academic EDs, airway management at the institution is the ultimate responsibility of the attending faculty emergency physician (EP), who determines which resident will perform the intubation and what technique/ device is used on a case-by-case basis. Intubators were allowed to choose what device and method they thought was appropriate for each intubation. If an initial intubation attempt was not successful, the supervising EP determined whether or not to switch to another device. Residents and attending EPs are familiarized with the devices as part of the residency curriculum, and a simulation laboratory is available for independent practice. Most often, however, experience is gained in the ED during the use of a device. Intubations are typically performed by emergency medicine residents; however, attending EPs typically complete the intubation if a resident is not successful. Rarely, paramedics, medical students, and offservice residents perform intubations. All trauma airways are handled by the emergency medicine residents, with attending EP supervision. On Tuesday mornings, attending EPs staff the ED without residents during protected didactic time. Occasionally, patients are intubated by attendings primarily during this protected time. For the GlideScope intubations, potential options were the GlideScope standard (reusable blade), GlideScope Cobalt (disposable blade), and GlideScope Ranger (portable GlideScope with reusable blade). For direct laryngoscopy, all blade types and sizes were available. Methods of intubation included RSI, oral intubation with sedation only, and oral intubation without the use of any medications. Selection of Participants All patients requiring intubation in the ED were entered into the database. Patients intubated before arrival by pre-hospital providers (emergency medical services) were excluded; however, pre-hospital failed airways were included. All intubation cases entered in the database during the study period that used a GlideScope video laryngoscope or traditional laryngoscopy were extracted for analysis.

3 Videolaryngoscopy vs. Direct Laryngoscopy in Difficult Airway 631 Methods of Measurement The registry data information forms were completed by the physicians who performed the intubation. Collected data forms were cross-referenced to professional billing records to identify any missing intubation forms. If an intubation form was missing, a data collection form was submitted to the operator for completion. Information collected included intubation indication, technique, outcome, medications used, demographics, and performance characteristics (Table 2). An attempt at intubation was defined as insertion of the laryngoscope blade into the patient s mouth, regardless of whether an attempt to pass a tracheal tube took place. First-attempt success was defined as the placement of a tracheal tube on the first attempt. Ultimate success was defined as tracheal intubation with the initial device selected regardless of the number of attempts. The data form included a list of potential factors that may make intubation more difficult (Table 1). Primary Data Analysis The structured data collection sheet included the following information: reason for intubation, diagnosis, presence of difficult airway predictors, Cormack-Lehane view, size and type of device used, type and number of complications, and the degree of lens contamination and fogging. Lens fogging was evaluated using a 10-point visual analog scale with 0 representing no fog, and 10 representing complete fogging. Data were analyzed using Stata version 11.0 (Stata- Corp, College Station, TX). Categorical data were analyzed using Fisher s exact test for proportions. Means with standard deviation were calculated for parametric data, and median plus interquartile range were calculated for non-parametric data. Analysis of variance and Kruskal-Wallis rank tests were used to compare multiple Table 2. Demographics DL GVL p Value Total 505 (65%) 267 (35%) Male 333 (66%) 190 (71%) NS Female 172 (34%) 77 (29%) NS Medical 294 (58%) 68 (25%) p = Trauma 211 (42%) 199 (75%) p = RSI 437 (87%) 235 (88%) NS Sedation only 4 (1%) 4 (2%) NS No Meds 63 (12%) 28 (10%) NS PGY (20%) 43 (16%) NS PGY (36%) 101 (38%) NS PGY (42%) 120 (45%) NS Attending 8 (2%) 3 (1%) NS DL = direct laryngoscopy; GVL = GlideScope videolaryngoscopy; RSI = rapid sequence intubation; PGY = post-graduate year. devices as appropriate. A significance level (alpha) of 0.05 was used for all overall analyses. This study was reviewed and approved by the University of Arizona Institutional Review Board. RESULTS There were 881 patients consecutively intubated in the ED during the study period. One hundred nine patients were excluded because direct laryngoscopy (DL) or Glide- Scope videolaryngoscopy (GVL) was not used, or because they were not the first device attempted. In 505 (65%) patients, DL was used as the first device, and in 267 (35%), GVL was used as the first device. This group comprised the study population (Table 2). The overall first-attempt success rate of DL was 347/505 (68%), and GVL was 208/267 (78%) (Fisher s exact test, p = 0.007). Table 3 shows the percentage of attempts with either blood/vomit or anatomic difficult airway predictors (DAPs) present on initial and rescue attempts for each device. Essentially, GVL was used on airways with more predicted DAPs than DL on first attempts, however rescue attempts showed no difference in the presence of DAPs. The odds ratios (ORs) of success with GVL over DL was 2.26 (95% CI ) on first attempts, and 2.08 (95% CI ) on rescue attempt with the presence of blood or vomit in the airway. In the presence of three or more anatomic DAPs, the ORs of success with GVL over DL was 2.72 (95% CI ) on first attempts, or 1.84 (95% CI ) on rescue attempts. A multivariable logistic regression analysis demonstrating the odds of failure with DL compared to GVL, controlling for various DAPs, is summarized in Table 4. The crude OR of failure of DL vs. GVL was 3.07 (95% CI ), and the adjusted OR of failure of DL vs. GVL, controlling for various DAPs (blood, small mandible, obesity, and large tongue) was 2.20 (95% CI ). The presence of blood (OR 2.79), small mandible (OR 2.93), obesity (OR 1.60), and a large tongue (OR 1.90) were all independent predictors of intubation failure Table 3. Percentage of First and Rescue Attempts with Difficult Airway Predictors Present First Attempt* Rescue Attempt** Presence of anatomic difficulty airway predictors GVL 59/267 (22%) 21/84 (25%) DL 43/505 (8.5%) 24/134 (18%) Presence of blood First Attempt*** Rescue Attempt**** or vomit GVL 119/267 (45%) 42/84 (50%) DL 166/505 (33%) 65/134 (49%) Fisher s exact test, *p = 0.002; **p = 0.231; ***p = 0.000; ****p =

4 632 J. M. Mosier et al. Table 4. Multivariate Regression Model for Intubation Failure Unadjusted OR (95% CI) in the final model. The other difficult airway predictors were not statistically significant in the final model and were not included in the final model. DISCUSSION Adjusted OR (95% CI) DL vs. GVL 3.07 ( ) 2.20 ( ) Presence of DAP: Blood 2.30 ( ) 2.79 ( ) Vomit 1.56 ( ) NS Short neck 2.20 ( ) NS Cervical immobility 1.05 ( ) NS Small mandible 2.77 ( ) 2.93 ( ) Obesity 1.95 ( ) 1.60 ( ) Airway edema 2.07 ( ) NS Facial trauma 1.12 ( ) NS Large tongue 2.53 ( ) 1.90 ( ) OR = odds ratio; CI = confidence interval; DL = direct laryngoscopy; GVL = GlideScope laryngoscopy; DAP = difficult airway predictor. Video laryngoscopy overcomes the pitfalls of DL by placing a micro video camera on the undersurface of the blade, allowing the operator to see around the anterior curvature of the supraglottic structures. This allows the person performing the intubation to visualize the airway anatomy by transporting the view from inside the mouth to a video monitor placed either on, or next to, the device. The indirect view of the airway obviates the need to align the oral, pharyngeal, and laryngeal axes to directly visualize passing a tracheal tube. There are several devices commercially available, including the GlideScope video laryngoscope system, Pentax airway scope (Pentax Medical Company, Montvale, NJ), McGrath (Aircraft Medical Limited, Edinburgh, UK), Res-Q-Scope (Res- Q-Tech, Fuquay Varina, NC), and C-MAC (Karl Storz GmbH & Co., Tuttlingen, Germany). Some devices (GlideScope, C-MAC) are equipped with an anti-fog mechanism that will provide better optical clarity. The most widely used device is the GlideScope video laryngoscope, which comes as either a portable unit (GlideScope Ranger) or a stationary unit, with both having the option of disposable single-use or reusable blades, and is available in pediatric and adult sizes. The blade has an exaggerated anterior curvature with the camera on the distal undersurface of the blade providing the view of the airway. It is advanced down the midline of the tongue into the vallecula instead of sweeping the tongue to the side, as done with direct laryngoscopy. Using the gear-stick technique, the blade compresses the hyoepiglottic ligament, lifting the epiglottis out of the way, allowing a view of the glottic aperture. A rigid stylet may be used that follows the curvature of the blade to facilitate tube advancement into the trachea. This becomes incredibly helpful as the curvature required to enter the glottic opening may lead to bending of the malleable standard stylet on the teeth and inability to advance the tube. Other devices have in-line tube guides (Pentax AWS, Res-Q-Scope) that allow passage of the endotracheal tube directly into the trachea through the device itself. The potential advantage is to eliminate the need for a stylet, and diminishes the difficulty with passing the tube into the trachea that can be experienced with the GlideScope. The C-MAC (Karl Storz) is designed to give the operator the advantages of both direct laryngoscopy and videolaryngoscopy. It is designed as a Macintosh 3 or 4 blade with a camera recessed on the undersurface of the blade. It can be used as both a direct laryngoscope using the Macintosh or Miller techniques, and a video laryngoscope. No studies have compared its performance with direct laryngoscopy outside of manikins, and one study compares the C-MAC with DL and GVL in manikins (13). Given that it can be used as both DL and a video laryngoscope, it provides an education advantage that the other scopes cannot yet provide. Several studies have evaluated the GlideScope use in both normal and difficult airway cases in the operating room, and simulated difficult airways on manikins. Those studied evaluated cervical rigidity or immobilization, tongue edema, airway obstruction, and obesity (3 8). There was improved glottic view with the GlideScope in those cases with tongue edema and cervical rigidity or immobilization, as well as patients with obesity and history of multiple failed intubation attempts. Few studies have compared GVL to DL. Several studies found improved views of the airway and higher success rates for patients with obesity, pharyngeal obstruction, cervical immobility, and tongue edema compared to Macintosh blade (14 20). Kim et al., in 2008, found an improved view of the airway, but longer time to intubation compared to the Macintosh blade in pediatric patients (21). Robitaille et al., in 2008, noted that, compared to DL with a Macintosh blade, GVL provided an improved view of the airway but did not minimize movement of the cervical spine (22). Platts-Mills et al. reported the first comparison of GlideScope to DL in the ED setting (12). They found that even with limited experience with the device, they had a success profile similar to DL (12). Our study demonstrates the first large-scale comparison of GVL to DL for difficult airway management in the ED. GVL seems to be superior to DL in our multivariate model controlling for various difficult airway predictors such as the presence of blood in the airway, obesity, small mandible, and a large tongue for first intubation attempts. The qualities that these difficult airway predictors all possess in common are obstruction of a direct line of sight with DL. The characteristics of GVL obviate the need

5 Videolaryngoscopy vs. Direct Laryngoscopy in Difficult Airway 633 for that direct line of sight; however, GVL characteristics are potentially susceptible to the difficulty of passing the endotracheal tube around the curvature to the vocal cords. Limitations This study is not without its significant limitations. Although this represents a large-scale study, it is retrospective in nature, and relies heavily on self-report with the standardized data collection sheets. Future research should consider a prospective study design with pre- and post-intubation evaluations of the difficulty of the airway as well as assessing inter-rater reliability for intubation characteristics. In addition, video recordings of all intubations should be obtained for independent evaluation of the airway difficulty. Lastly, given that the Cormack-Lehane grade of view scale was developed to determine adequate intubation conditions with DL, it may not be a suitable performance characteristic for videolaryngoscopy. Given that most failures with videolaryngoscopy are from inability to direct the endotracheal tube rather than inability to view the glottic opening, future research should investigate a modified score pertinent to video laryngoscopy. CONCLUSION GlideScope videolaryngoscopy seems to be superior to direct laryngoscopy in the presence of all difficult airway predictors in terms of success rate for first attempts, whereas it appears equivalent to DL in terms of success rate for rescue attempts. In the presence of blood in the airway, obesity, small mandible, or large tongue, GVL is superior to DL for successful intubation on first attempt. Overall, GVL seems to be superior to DL, with an adjusted OR for first-attempt intubation success of 2.2 (95% CI ). REFERENCES 1. Walls RM. Manual of emergency airway management. Philadelphia: Wolters Kluwer Health/Lippincott Williams & Wilkins; Macintosh RR. A new laryngoscope. Lancet 1943;241: Murrell GL, Sandberg KM, Murrell SA. GlideScope video laryngoscopes. Otolaryngol Head Neck Surg 2007;136: Manoach S, Paladino L. Manual in-line stabilization for acute airway management of suspected cervical spine injury: historical review and current questions. Ann Emerg Med 2007;50: Hirabayashi Y, Hakozaki T, Fujisawa K, et al. Use of a new videolaryngoscope (GlideScope) in patients with a difficult airway [Japanese]. Masui 2007;56: Benjamin FJ, Boon D, French RA. An evaluation of the GlideScope, a new video laryngoscope for difficult airways: a manikin study. Eur J Anaesthesiol 2006;23: Tan BH, Liu EH, Lim RT, et al. Ease of intubation with the Glide- Scope or Airway Scope by novice operators in simulated easy and difficult airways a manikin study. Anaesthesia 2009;64: Cooper RM. Use of a new videolaryngoscope (GlideScope) in the management of a difficult airway. Can J Anaesth 2003;50: Noppens RR, Mobus S, Heid F, Schmidtmann I, Werner C, Piepho T. Evaluation of the McGrath Series 5 videolaryngoscope after failed direct laryngoscopy. Anaesthesia 2010;76: Hirabayashi Y, Hakozaki T, Fujisawa K, et al. GlideScope videolaryngoscope: a clinical assessment of its performance in 200 consecutive patients [Japanese]. Masui 2007;56: Hirabayashi Y. Airway Scope versus Macintosh laryngoscope: a manikin study. Emerg Med J 2007;24: Platts-Mills TF, Campagne D, Chinnock B, et al. A comparison of GlideScope video laryngoscopy versus direct laryngoscopy intubation in the emergency department. Acad Emerg Med 2009; 16: McElwain J, Malik MA, Harte BH, et al. Comparison of the C-MAC videolaryngoscope with the Macintosh, Glidescope, and Airtraq laryngoscopes in easy and difficult laryngoscopy scenarios in manikins. Anaesthesia 2010;65: Savoldelli GL, Schiffer E, Abegg C, et al. Comparison of the Glidescope, the McGrath, the Airtraq and the Macintosh laryngoscopes in simulated difficult airways. Anaesthesia 2008;63: Powell L, Andrzejowski J, Taylor R, et al. Comparison of the performance of four laryngoscopes in a high-fidelity simulator using normal and difficult airway. Br J Anaesth 2009;103: Narang AT, Oldeg PF, Medzon R, et al. Comparison of intubation success of video laryngoscopy versus direct laryngoscopy in the difficult airway using high-fidelity simulation. Simul Healthc 2009;4: Malik MA, O Donoghue C, Carney J, et al. Comparison of the Glidescope, the Pentax AWS, and the Truview EVO2 with the Macintosh laryngoscope in experienced anaesthetists: a manikin study. Br J Anaesth 2009;102: Malik MA, Maharaj CH, Harte BH, et al. Comparison of Macintosh, Truview EVO2, Glidescope, and Airwayscope laryngoscope use in patients with cervical spine immobilization. Br J Anaesth 2008;101: Lim Y, Yeo SW. A comparison of the GlideScope with the Macintosh laryngoscope for tracheal intubation in patients with simulated difficult airway. Anaesth Intensive Care 2005;33: Kim HJ, Chung SP, Park IC, et al. Comparison of the GlideScope video laryngoscope and Macintosh laryngoscope in simulated tracheal intubation scenarios. Emerg Med J 2008;25: Kim JT, Na HS, Bae JY, et al. GlideScope video laryngoscope: a randomized clinical trial in 203 paediatric patients. Br J Anaesth 2008;101: Robitaille A, Williams SR, Tremblay MH, et al. Cervical spine motion during tracheal intubation with manual in-line stabilization: direct laryngoscopy versus GlideScope videolaryngoscopy. Anesth Analg 2008;106:935 41:table of contents.

6 634 J. M. Mosier et al. ARTICLE SUMMARY 1. Why is this topic important? Airway management is the paramount procedure performed by emergency providers, with the majority of their patients representing high-risk airways. Advances in airway management should be aggressively evaluated to improve patient outcomes and limit complications. 2. What does this study attempt to show? This study compares video laryngoscopy using the GlideScope (GVL) to direct laryngoscopy (DL) in the presence of known difficult airway predictors and attempts to show in which situations one technique is better than the other. 3. What are the key findings? In the presence of a small mandible, blood, obesity, or large tongue, GVL has higher odds of first-pass success than DL. 4. How is patient care impacted? Fewer attempts at intubation and lower complication rates have the potential to greatly improve emergency airway management in our emergency departments.