MINERVA ANESTESIOLOGICA

Transcription

1 MINERVA ANESTESIOLOGICA EDIZIONI MINERVA MEDICA This provisional PDF corresponds to the article as it appeared upon acceptance. A copyedited and fully formatted version will be made available soon. The final version may contain major or minor changes. Real-time Ultrasound Guided Percutaneous Dilatational Tracheostomy -With and Without Bronchoscopic Control : An Observational Study Jose CHACKO, Gagan BRAR, Umesh KUMAR, Bhargav MUNDLAPUDI Minerva Anestesiol 2014 Jul 24 [Epub ahead of print] MINERVA ANESTESIOLOGICA Rivista di Anestesia, Rianimazione, Terapia Antalgica e Terapia Intensiva pissn eissn Article type: Original Paper The online version of this article is located at Subscription: Information about subscribing to Minerva Medica journals is online at: Reprints and permissions: For information about reprints and permissions send an to: journals.dept@minervamedica.it - journals2.dept@minervamedica.it - journals6.dept@minervamedica.it COPYRIGHT 2014 EDIZIONI MINERVA MEDICA

2 1 TITLE PAGE Title: Real time Ultrasound Guided Percutaneous Dilatational Tracheostomy -With and Without Bronchoscopic Control : An Observational Study List of authors: 1. Dr. Jose Chacko, Consultant, Critical Care, Manipal Hospital, Bangalore, India 2. Dr. Brar Gagan, Registrar, Critical Care, Manipal Hosptial, Bangalore, India 3. Mr. Kumar Umesh, Respiratory Therapist, Manipal Hospital, Bangalore, India 4. Dr. Bhargav Mundlapudi, Registrar, Critical Care, Manipal Hospital, Bangalore, India Corresponding author: Dr. Jose Chacko, Consultant, Critical Care, Manipal Hospital, Bangalore, India. Tel: , FAX: , chackojose@gmail.com Conflict of interest: None Source of funding: None

3 2 ABSTRACT Background. Real time ultrasound guidance may enable precise introducer needle and guidewire insertion during percutaneous dilatational tracheostomy (PDT). However, the safety of PDT done solely under ultrasound guidance has not been compared to additional bronchoscopic confirmation. We aimed to compare the efficacy, incidence of complications and procedural times of real time ultrasound guided PDTs done with and without bronchoscopic confirmation. Methods. We analysed all bedside PDTs performed over an 18 month period. On transverse view, the introducer needle was inserted under real time guidance into the chosen interspace. The interspace of guidewire entry was confirmed on longitudinal view. In the bronchoscope group, needle and guidewire position within the tracheal lumen were confirmed by bronchoscopy. Results. We performed 177 PDTs under ultrasound guidance during the study period 95 with and 82 without bronchoscopic confirmation. The introducer needle was inserted at the desired level on the first attempt in 81.4% of instances with no difference between groups (77.9% vs. 85.4%, p = 0.6). Introducer needle entry between the first and third tracheal cartilages did not differ between groups (90.5% vs. 92.7%, p = 0.6). Procedural times were longer in the bronchoscope group (13.9 min vs min, p < ). There were more episodes of fall in oxygen saturation below 90% in the bronchoscope group (16.8% vs. 3.7%, p = 0.006). Minor bleeding occurred in 6.2% of cases and did not differ between groups.

4 3 Conclusions. Ultrasound guidance for PDT may be equally safe, even without bronchoscopic confirmation; the incidence of hypoxic episodes may be less and the procedure can be performed in less time. Clinical trial registration. CTRI/2012/09/ Keywords: tracheostomy; ultrasound; bronchoscope; airway control; complications

5 4 Real-time Ultrasound Guided Percutaneous Tracheostomy -With and Without Bronchoscopic Control: An Observational Study INTRODUCTION Bedside percutaneous dilatational tracheostomy (PDT) is commonly performed under 1 bronchoscopic guidance., 2 Bronchoscopy enables direct visualisation of the introducer needle, guidewire and dilators within the tracheal lumen; the final position of the tracheostomy tube can also be confirmed. Several studies support the usefulness of bronchoscopy in preventing complications of 3 PDT., 4 However, other reports suggest a low incidence of complications even without 1. Veenith T, Ganeshamoorthy S, Standley T, Carter J, Young P. Intensive care unit tracheostomy: a snapshot of UK practice. Int Arch Med 2008; 1: Vargas M, Servillo G, Arditi E, Brunetti I, Pecunia L, Salami D, et al. Tracheostomy in Intensive Care Unit: a national survey in Italy. Minerva Anestesiol 2013; 79: Tomsic JP, Connolly MC, Joe VC, Wong DT. Evaluation of bronchoscopic assisted percutaneous tracheostomy. Am Surg 2006; 72: Berrouschot J, Oeken J, Steiniger L, Schneider D. Perioperative complications of percutaneous tracheostomy. Laryngoscope 1997; 107:

6 5 5 bronchoscopic guidance., 6 After the initial learning curve, bronchoscopic guidance may not add to the safety of PDT. 5, 7 In a retrospective study of 243 patients, there was no difference in the 8 incidence of complications with or without bronchoscopic guidance. Professional bodies in Intensive Care Medicine do not uniformly recommend routine bronchoscopy during PDT in their guidelines. 9, Díaz Regañón G, Miñambres E, Ruiz A, González Herrera S, Holanda Peña M, López Espadas F. Safety and complications of percutaneous tracheostomy in a cohort of 800 mixed ICU patients. Anaesthesia 2008; 11: Ahmed R, Rady SR, Mohammad Siddique JI, Iqbal M. Percutaneous tracheostomy in critically ill patients: 24 months experience at a tertiary care hospital in United Arab Emirates. Ann Thorac Med 201; 5: Cobean R, Beals M, Moss C, Bredenberg CE. Percutaneous dilatational tracheostomy. A safe, cost effective bedside procedure. Arch Surg. 1996; 131: Jackson LS, Davis JW, Kaups KL, Sue LP, Wolfe MM, Bilello JF, et al. Percutaneous tracheostomy: to bronch or not to bronch that is the question. J Trauma 2011; 71: Madsen KR, Guldager H, Rewers M, Weber SO, Købke Jacobsen K, Jensen R. Danish Society of Anesthesiology and Intensive Care Medicine. Guidelines for Percutaneous Dilatational Tracheostomy (PDT) from the Danish Society of Intensive Care Medicine (DSIT) and the Danish Society of Anesthesiology and Intensive Care Medicine (DASAIM). Dan Med Bull 2011; 58:C

7 6 11 Ultrasound scan of the neck offers excellent visualisation of airway anatomy. Pre procedural screening ultrasound has been used previously to identify prominent blood vessels 12 before PDT. Introducer needle insertion under real time ultrasound guidance was found to be 13 feasible in a pilot study. Real time ultrasound has also been used to guide PDT in critically ill, 14 obese patients. We have previously reported the use of real time ultrasound guidance to precisely insert the 15 introducer needle into a pre defined level in the trachea. In that study, we had used bronchoscopy. Burrel T, Sampson B. ANZICS, Percutaneous Dilatational Tracheostomy Consensus Statement, [ quality/saq resources] 11. Sustić A. Role of ultrasound in the airway management of critically ill patients. Crit Care Med 2007; 35: S Hatfield A, Bodenham A. Portable ultrasonic scanning of the anterior neck before percutaneous dilatational tracheostomy. Anaesthesia 1999; 54: Rajajee V, Fletcher JJ, Rochlen LR, Jacobs TL. Real time ultrasound guided percutaneous dilatational tracheostomy: a feasibility study. Crit Care 2011; 15:R Guinot PG, Zogheib E, Petiot S, Marienne JP, Guerin AM, Monet P, et al. Ultrasound guided percutaneous tracheostomy in critically ill obese patients. Crit Care 2012; 16: R Chacko J, Nikahat J, Gagan B, Umesh K, Ramanathan M. Real time ultrasound guided percutaneous dilatational tracheostomy. Intensive Care Med 2012; 38:

8 7 to confirm guidewire position within the tracheal lumen. As we gained in experience, we began performing PDTs solely under real time ultrasound guidance. We hypothesised that PDTs performed under real time ultrasound guidance alone might be comparable in safety and efficacy to additional confirmation by bronchoscopy. In this retrospective observational study, we aimed to compare the efficacy, incidence of complications and time taken with real time ultrasound guided PDTs done with and without bronchoscopic confirmation. MATERIALS AND METHODS Design We collected data on consecutive bedside PDTs performed in our 27 bedded multi disciplinary intensive care unit (MICU) over an 18 month period. The study was approved by the Hospital Ethics Committee and a waiver of informed consent was granted in view of the observational design that did not require deviation from our established practice of using bronchoscopic or ultrasound guidance upon operator discretion. However, as per our protocol, written informed consent was obtained for the procedure. Patient anonymity was protected as information was abstracted from a database without unique identifiers. The study was registered with the Clinical Trials Registry India CTRI/2012/09/ Technique PDT was performed by introducer needle insertion under real time ultrasound guidance followed by bronchoscopic confirmation or solely under real time ultrasound guidance, at the discretion of the

9 8 operator. Briefly, after increasing F i O 2 to 1.0, anaesthesia was induced, followed by positioning and draping. Ultrasound scan of the neck was performed using a 12 MHz linear probe. In the long axis view, the endotracheal tube was visualised as a double line and withdrawn to the level of the cricoid or the first tracheal cartilage. If the tube was not clearly visualised on ultrasound, it was withdrawn under direct laryngoscopy till the cuff was seen just above the glottis. Alternatively, the endotracheal tube was removed and a laryngeal mask was inserted to maintain airway at the discretion of the operator. Identification of the level of entry The thyroid, cricoid and the tracheal cartilages were imaged in the short axis view by moving the probe in a caudal direction. The first tracheal cartilage was identified immediately below the cricoid cartilage (Fig. 1a). The probe was positioned transversely on the first or second tracheal cartilage and the introducer needle inserted immediately caudal to the probe at its midpoint as shown in Fig. 1b. This enabled entry between the first and second (T1 T2) or second and third (T2 T3) tracheal cartilages. The point of needle insertion was adjusted to prevent injury to any prominent blood vessels identified on colour Doppler imaging. A lower level was chosen if an enlarged thyroid gland was encountered between T1 and T3. The needle track was followed by soft tissue movement at its tip and indentation of the air mucosal interface as shown in Fig. 1c. A distinctive give way feel indicated entry into the tracheal lumen. Following introducer needle and guidewire insertion, the point of entry as seen on a clock face was visualised on the short axis view. The interspace of guidewire entry was confirmed on the long axis view as presented in Fig. 2a, b. In the bronchoscope assisted group, following guide wire insertion, fibreoptic bronchoscopy was performed, through the endotracheal tube or LMA to confirm position within the tracheal lumen. Tracheostomy was completed by the single stage dilator (SSD) technique or using the guide

10 9 wire dilating (GWD) forceps. A bronchoscope was available by the bedside in all cases if direct visualisation was felt necessary. Data collection We collected data on patient characteristics including age, sex, body mass index (BMI), underlying illness, day of tracheostomy, the APACHE II score, suspicion of C spine injury, type of operator (consultant vs trainee), method of airway maintenance during the procedure (LMA vs endotracheal tube), INR and platelet count on the day of the procedure, the PaO 2 /F i O 2 ratio immediately before the procedure, presence of a short neck or significant thyroid swelling as judged by the operator and the visibility of neck anatomy on ultrasound imaging. The number of attempts taken to pass the introducer needle, point of entry of guidewire into the trachea on a clock face and the interspace of entry were noted. Guidewire entry between eleven o clock and one o clock positions was described as median ; any entry lateral to this was considered paramedian. Needle to wire time was defined as the time taken from the first attempt at introducer needle insertion to successful passage of guidewire into the trachea. Total procedural time was the time interval between the first attempt at introducer needle insertion to correct placement of the tracheostomy tube as confirmed by capnography. Ease of ultrasonic visualisation of neck anatomy was graded by the operator as excellent, good or unsatisfactory. Complications encountered were recorded, including bleeding during or after the procedure, accidental extubation, cuff puncture by the introducer needle, subcutaneous emphysema, pneumothrorax or fall in oxygen saturation below 90% during the procedure. Major bleeding was defined as requiring transfusions or surgical intervention and minor bleeding as bleeding that could be arrested with pressure. Patients who were discharged from the

11 10 hospital were followed up as outpatient at six months after discharge, specifically to look for long term complications related to tracheostomy. Statistical Analysis Data analysis was performed using Medcalc statistical software (version , Ostend, Belgium). Data are presented as number and percentage, mean ± SD or median with interquartile range (IQR). The Kolmogorov Smirnov test was used to test for normality of distribution of variables. The Student s t test was used if continuous variables were normally distributed; the Mann Whitney U test was employed otherwise. Categorical variables were tested using the Fisher s exact test. A probability value (p) of less than 0.05 was considered to be statistically significant. Baseline characteristics were compared between the group that had PDTs done under sole ultrasound guidance and those that had bronchoscopic confirmation. Additionally, we analysed data on variables that may have prompted operators to resort to bronchoscopic confirmation. These included the APACHE II score, Body Mass Index (BMI), suspected C spine injury, type of operator (consultant vs trainee), LMA use, INR > 1.5, platelet count < 50,000 per cu mm, PCO 2, P/F ratio, PEEP, short neck as judged by the operator, technique of PDT (GWD vs SSD), presence of significant thyroid swelling as judged by the operator and visibility of neck anatomy on ultrasound. Being an observational study, in which bronchoscope use carried out entirely at the discretion of the operator, we sought to adjust for possible imbalances in baseline variables that may have necessitated bronchoscopic confirmation. Univariate logistic regression analysis was performed with the variables mentioned above to predict bronchoscope use. Using a cut off P value of 0.25, a backward stepwise multivariate regression analysis was carried out with bronchoscope

12 11 use as the dependent variable and the above mentioned parameters as independent variables to identify any possible bias associated with bronchoscopic confirmation of PDT. The Hosmer Lemeshow test was used to evaluate goodness of fit of the model for agreement between observed and expected results, with P > 0.05 indicating a good fit. RESULTS PDT was successful in all except one patient, who had to undergo an open procedure due to excessive bleeding; however haemostasis was rapidly obtained without haemodynamic instability and no blood transfusions were required. We performed 177 PDTs over an 18 month period, after a median of six days (IQR: 5 8) of endotracheal intubation. Three patients underwent elective, open tracheostomy in two patients, access was restricted with visibility on ultrasound limited to the level of the first tracheal ring even after maximal neck extension. In the third instance, open surgical tracheostomy was combined with other operative procedures. Two PDTs were performed for emergency airway access under ultrasound guidance alone. A laryngeal mask was used in 17(9.6%) patients, while the endotracheal tube was withdrawn to maintain airway in the rest. Ninety five patients had bronchoscopic confirmation of guidewire position (US BR) while 82 patients underwent PDT solely under ultrasound guidance (US). Baseline characteristics did not differ between groups (Table 1). Delineation of neck anatomy by ultrasound was judged by the operator as excellent in 170 (96%) cases. Multivariate analysis did not reveal any predictor of

13 12 bronchoscope use with no variable being retained in the final model. Calibration using the Hosmer Lemeshow test revealed an adequate fit of the model (x 2 =6.64, P=0.58). Introducer needle entry characteristics Procedural characteristics are described in Table 2. The introducer needle was inserted into the trachea on the first attempt in 144 (81.4%) patients; there was no significant difference between groups. Entry was median in 90.4% of patients in the US group and 84.1% of patients in the US BR group, with no difference between groups. The level of entry was through the T1 T2 or T2 T3 interspace in 91.5% of cases, and was comparable between groups. Entry between T3 and T4 was encountered in 12 (6.7%) of cases and between T4 and T5 in 3 (1.7%) instances. Procedure times Total procedural time was significantly lower in the US group (13.9 ± vs 10.7 min, p < ). Needle to wire time was not significantly different between groups. Complications Thirty two complications were observed in the US BR group, compared to ten in the US group (Table 3). There were three accidental extubations during the procedure, all in the US BR group. We observed a significantly higher incidence of fall in oxygen saturation in the US BR group compared to the US group (16.8 vs 3.7%; p = 0.006). No major bleeding was encountered; the

14 13 incidence of minor bleeding did not differ between groups. Subcutaneous emphysema or pneumothorax were not observed during the study. We followed up 87 out of 102 patients who were discharged alive from hospital after undergoing PDT at six months post discharge 38 patients in the US and 49 in the US BR group. At this stage, five patients (4.8%) three in the US group and two in the US BR group had developed significant tracheal stenosis characterised by stridor and confirmed on bronchcoscopy, that required intervention. In three instances, we had to resort to bronchoscopic assistance, although the initial plan was to perform the procedure solely under ultrasound guidance. In two cases, bronchoscopy was done to confirm wire position within the tracheal lumen as the operator was unsure. The guidewire was seen to take a cranial course on ultrasound in another instance; this was confirmed by bronchoscopy and redirected caudally. DISCUSSION We compared real time ultrasound guided PDT with and without bronchoscopic confirmation in this retrospective observational study. Total procedural time was significantly shorter with a significantly lower incidence of fall in oxygen saturation in the ultrasound only group. Complications were few and did not differ between groups.

15 14 Fibreoptic bronchoscopy is used by most operators to guide PDT, 1,2, 16 although it is unclear whether this results in a lower incidence of complications. Besides requiring a skilled operator, damage to the bronchoscope can occur during introducer needle insertion 17 leading to costly 18 repairs. The use of ultrasound to guide percutaneous tracheostomy has not been extensively studied; real time guidance by tracking the tip of the introducer needle has been evaluated in a small series. 13 Ultrasound use allowed us good visualisation of the airway anatomy with clear delineation of the thyroid, cricoid and tracheal cartilages; this enabled entry at the chosen level in the trachea. Subsequently, we confirmed the level of guidewire entry on the long axis view by counting downwards from the cricoid cartilage. We entered the trachea between T1 and T3 in 91.5% of cases. Blind insertion has been shown to result in inadvertent high or low placement of the introducer 19 needle. A high tracheal stoma can predispose to stenosis due to proximity to the cricoid cartilage; 16. Kluge S, Baumann HJ, Maier C, Klose H, Meyer A, Nierhaus A, et al. Tracheostomy in the intensive care unit: a nationwide survey. Anesth Analg 2008; 107: Kollig E, Heydenreich U, Roetman B, Hopf F, Muhr G. Ultrasound and bronchoscopic controlled percutaneous tracheostomy on trauma ICU. Injury 2000; 31: Rozman A, Duh S, Petrinec Primozic M, Triller N. Flexible bronchoscope damage and repair costs in a bronchoscopy teaching unit. Respiration 2009; 77: Dexter JT. A cadaver study appraising the accuracy of blind placement of percutaneous tracheostomy. Anaesthesia 1995; 50:

16 15 while low insertion may injure blood vessels at the root of the neck, leading to life threatening 20 haemorrhage. In contrast to ultrasound, bronchoscopy may not reliably confirm the level of entry. Counting upwards from the carina is inaccurate as the number of tracheal cartilages is variable; 21 counting downwards may be difficult as withdrawing the endotracheal tube far enough to visualise 22 the cricoid cartilage may result in inadvertent extubation. We were successful in median insertion in 23 nearly 90% of cases as confirmed on ultrasound; paramedian insertion may lead to a lateral stoma that may lead to tracheal stenosis besides haemorrhage from damage to aberrant blood vessels. 20. McCormick B, Manara A R. Mortality from percutaneous dilatational tracheostomy. A report of three cases. Anaesthesia 2005; 60: Banninster L: Respiratory System. In: Gray H, Williams PL, eds. Gray s Anatomy: the Anatomical Basis of Medicine and Surgery. Edinburgh: Churchill Livingstone; 1995; Ambesh SP, Sinha PK, Tripathi M, Matreja P. Laryngeal mask airway vs endotracheal tube to facilitate bedside percutaneous tracheostomy in critically ill patients: a prospective comparative study. J Postgrad Med 2002; 48: Hutchinson RC, Mitchell RD. Life threatening complications from percutaneous dilational tracheostomy. Crit Care Med 1991; 19: Van Heurn LWE, Goei R, de Ploeg I, Ramsay G, Brink PRG. Late complications of percutaneous dilatational tracheostomy. Chest 1996; 110 : Shlugman D, Satya Krishna R, Loh L. Acute fatal haemorrhage during percutaneous dilatational tracheostomy. Br J Anaesth 2003; 90:

17 16 Our study suggests that introducer needle insertion using an out of plane approach with the probe placed in the transverse axis may improve the chances of median puncture compared to an in 26 plane approach. In our study, accidental extubation occurred during the procedure on three occasions; all associated with bronchoscope use. We encountered a significantly higher incidence of fall in oxygen saturation in the bronchoscope group. Hypoxic episodes are more common during 27 bronchoscope guided PDT., 28 Other complications were minor, relatively few, and did not differ between groups. Perforation of the posterior wall of the trachea is a serious complication of PDT. Ultrasonography cannot identify this complication, because the air column in the trachea reflects 29 ultrasound beams preventing visibility of objects passing through it. Bronchoscopic guidance may 26. Kleine Brueggeney M, Greif R, Ross S, Eichenberger U, Moriggl B, Arnold A, et al. Ultrasound guided percutaneous tracheal puncture: a computer tomographic controlled study in cadavers. Br J Anaesth 2011; 106: Agarwal A, Singh D. Is fibreoptic percutaneous tracheostomy in ICU a breakthrough. J Anaesthesiol Clin Pharmacol 2010; 26: Kost KM: Endoscopic percutaneous dilatational tracheotomy: a prospective evaluation of 500 consecutive cases. Laryngoscope 2005; 115: Schuler A. Image artifacts and pitfalls. In: Mathis G ed. Chest sonography. New York: Springer, 2008;

18 17 not reliably prevent injury to the posterior tracheal wall, with this complication escaping detection by the bronchoscopist in several reports. 8,29 30 We did not encounter any instance of posterior wall perforation in our study. We did not require extra personnel for the procedure a single operator controlled the ultrasound probe with the non dominant hand and the introducer needle with the dominant hand, while the airway was handled by a second operator. It is important to emphasise that direct visualisation by bronchoscopy may still be required on occasions in case of difficulty; bronchoscopic assistance may be appropriate till proficiency is gained with this technique. We carried out emergency PDTs on two occasions solely under ultrasound guidance. There are several reports of emergency PDTs;, the feasibility of ultrasound guidance for the rapid performance of PDTs in emergency situations may merit further investigation. We studied patients who had PDT performed solely under ultrasound guidance and compared it with those who had bronchoscopic confirmation following ultrasound guided introducer 30. Brander L, Takala J.Tracheal tear and tension pneumothorax complicating bronchoscopy guided percutaneous tracheostomy. Heart Lung 2006; 35: Davidson SB, Blostein PA, Walsh J, Maltz SB, VandenBerg SL. Percutaneous tracheostomy: a new approach to the emergency airway. J Trauma Acute Care Surg 2012; 73: S Ben Nun A, Altman E, Best LA. Emergency percutaneous tracheostomy in trauma patients: an early experience. Ann Thorac Surg 2004; 77:

19 18 needle and guidewire insertion. Besides requiring a skilled operator, bronchoscope use carries a 14, 33 recurrent expense incurred with cleaning, reprocessing and damage repair. 33. Mughal MM, Minai OA, Culver DA, Mehta A. Reprocessing the bronchoscope: the challenges. Semin Respir Crit Care Med 2004; 25:

20 19 Limitations of the study Our study is limited by its retrospective observational design. Bronchoscopic confirmation was entirely at the discretion of the operator; this may have led to possible bias with such confirmation being employed in cases that were perceived to be more difficult. Although no bias was identified on multivariate analysis, unidentified confounding variables may have influenced our results, although we consider this unlikely. Being an observational study, our technique of PDT was not uniform; airway control during the procedure was either by withdrawing the endotracheal tube or through an LMA. Besides, the procedure was carried out in a single centre by operators who had proficiency and experience with the use of ultrasound that may not be generalisable to other settings.

21 20 Prospective randomised controlled studies are required to compare the safety and efficacy of ultrasound guided PDT, performed with and without bronchoscopic control. It may also be important to compare cost effectiveness between ultrasound versus bronchoscopic guidance given that there is likely to a significant cost involved with re processing, repair and replacement of bronchoscopes over time. CONCLUSIONS We conclude that real time ultrasound guidance enables precise insertion of the introducer needle at the desired level in the trachea, in a median position during PDT. Ultrasound guidance may be equally safe, even without bronchoscopic confirmation. The incidence of hypoxic episodes may be less and the procedure can be carried out in less time. KEY MESSAGES PDTs may be performed safely under real time ultrasound guidance without bronchoscopic assistance Real time ultrasound guidance enables precise insertion of the introducer needle at the desired level, in the median position Hypoxic episodes may be significantly less when PDTs are performed solely under ultrasound guidance

22 21 Procedural times are significantly shorter when bronchoscopic assistance is not resorted to Author contributions JC: Conception and design, statistical analysis, drafting of the paper, final approval; GB: Data collection, drafting of the paper, proofing, final approval; BM: Conception and design, drafting of the paper, proofing, statistical analysis, final approval; UK: Data collection, statistical analysis, proofing, final approval

23 22 Table legends: Table 1 Baseline characteristics. Data are expressed as mean (± standard deviation), median (interquartile range), or number and percentage (%). SSD, Single step dilator; GWD, guidewire dilating forceps; OR (CI), Odds Ratio (confidence interval). Table 2 Procedural characteristics. Data are expressed as mean ± standard deviation or number and percentage. Table 3 Complications. Data are expressed as number and percentage (%) Figure legends: Figure 1 Short axis (transverse) views on ultrasound. (A) ISTH = thyroid isthmus; asterisks = tracheal cartilage. (B) Needle insertion in the short axis. (C) arrow = Introducer needle indenting the tracheal ring Figure 2 Long axis (longitudinal) views on ultrasound. (A) Probe placed to view guidewire. (B) CR cricoid cartilage; T1, T2 first and second tracheal cartilages; arrow air mucosal interface. The guidewire can be seen entering the trachea between T1 and T2

24 Table 1. Baseline characteristics. Characteristic No Bronchoscope (82) Bronchoscope (95) OR (CI) for bronchoscope P value Age, years 57.3 ± ± Male/Female 54/41 48/ Technique SSD/GWD 88/7 79/ Univariate regression analysis with independent variables BMI 23.6 (3) 23.5 (1.9) 1.04 ( ) 0.35 APACHE II 16 (8) 15 (6.5) 1.01 ( ) 0.61 Suspected 2 (2.4%) 3 (3.2%) 1.3 ( ) 0.77 C-spine injury Trainee 43 (52.4%) 40 (42.1%) 0.93 ( ) 0.80 performed LMA 7 (8.5%) 10 (10.5%) 1.22 ( ) 0.73 INR >1.5 5 (6.1%) 7 (7.4%) 1.26 ( ) 0.65 Platelets 12 (14.6%) 17 (17.9%) 1.0 ( ) 0.87 <50,000/cu mm PCO 2, mm Hg 32.5 (8) 33 (5) 1.16 ( ) 0.85 P/F ratio 209 (40) 208 (37) 1.0 ( ) 0.45 PEEP >5 cm 3 (3.7%) 4 (4.2%) 1.27 ( ) 0.56 H 2 O Short neck 3 (3.7%) 3 (3.2%) 0.86 ( ) 0.85 SSD 79 (96.3%) 89 (93.7%) 0.48 ( ) 0.30 Thyroid 7 (8.5%) 8 (8.4%) 0.99 ( ) 0.98 enlargement Sub-optimal visibility on ultrasound 3 (3.7%) 4 (4.3%) 1.16 ( ) 0.85 Data are expressed as mean (± standard deviation), median (interquartile range), or number and percentage (%). SSD, Single step dilator; GWD, guidewire dilating forceps; OR (CI), Odds Ratio (confidence interval).

25 Table 2. Procedural characteristics. Bronchoscope No Bronchoscope p value (n = 95) (n = 82) One attempt 74 (77.9%) 70 (85.4%) 0.25 Two attempts 16 (16.8%) 4 (4.9%) More than two 5 (5.3%) 8 (9.8%) attempts Median entry 86 (90.4%) 69 (84.1%) 0.25 Good visibility 91 (95.8%) 79 (96.3%) 1.00 Entry at T1 T2 or 86 (90.5%) 76 (92.7%) 0.60 T2 T3 Needle to wire time, 13.6 ± ± seconds Total procedure time, minutes 13.9 ± ± (8.2) Data are expressed as mean ± standard deviation or number and percentage.

26 Table 3. Complications. Data are expressed as number and percentage (%) Bronchoscope No bronchoscope Total p value n = 95(%) n = 82(%) Bleed during 3 (3.2) 1 (1.2) 4 (2.3) 0.62 procedure Bleed after 4 (4.2) 3 (3.7) 7 (4) 1.00 procedure Accidental 3 (3.2) 0 (0) 3 (1.7) 0.25 extubation during procedure Cuff puncture 2 (2.1) 1 (1.2) 3 (1.7) 1.00 Saturation below 16 (16.8) 3 (3.7) 19 (10.7) % Damage to 1 (1.1) Not applicable bronchoscope Tracheal stenosis 3 (3.2) 2 (2.4) 5 (2.8) 1.00

27

28

29

30

31