Pilot Study to Evaluate the Accuracy of Ultrasonography in Confirming Endotracheal Tube Placement

Transcription

1 AIRWAY/BRIEF RESEARCH REPORT Pilot Study to Evaluate the Accuracy of Ultrasonography in Confirming Endotracheal Tube Placement Sandra L. Werner, MD, RDMS Charles E. Smith, MD Jessica R. Goldstein, MD Robert A. Jones, DO, RDMS Rita K. Cydulka, MD, MS From the Department of Emergency Medicine, MetroHealth Medical Center/Cleveland Clinic Foundation (Werner, Goldstein, Jones, Cydulka), and the Department of Anesthesia, MetroHealth Medical Center (Smith), Cleveland, OH. Study objective: Visualization of the vocal cords and end-tidal capnography are the usual standards in confirming endotracheal tube placement. Vocal cord visualization is, however, not always possible, and capnography is not 100% reliable and requires ventilation of the lungs to confirm placement. The goal of this study is to determine the accuracy of ultrasonography for detecting endotracheal tube placement into the trachea and esophagus in real time. Methods: This was a prospective, randomized, controlled study. Eligible patients were adults undergoing elective surgery requiring intubation. Exclusion criteria were a history of difficult intubation, abnormal airway anatomy, aspiration risk factors, and esophageal disease. Thirty-three patients were enrolled. After induction of anesthesia and neuromuscular blockade, the anesthesiologist placed the endotracheal tube in the trachea and esophagus in random order with direct laryngoscopy. During the intubations, a high-frequency, linear transducer was placed transversely on the neck at the suprasternal notch. Two emergency physicians, blinded to the order and performance of the intubations, independently recorded the location of the endotracheal tube according to the real-time ultrasonographic image. A 2-by-2 table was used to calculate sensitivity and specificity of the emergency physicians ability to detect placement of the endotracheal tube. Results: For each physician, the sensitivity for identifying the first intubation as tracheal was 100% (95% confidence interval [CI] 77% to 100%) with a specificity of 100% (95% CI 82% to 100%). One endotracheal tube was unintentionally placed twice in the esophagus, but both tube placements were identified as esophageal by the emergency physicians. Conclusion: In this pilot study, 2 emergency physicians experienced in ultrasonography accurately detected placement of endotracheal tubes during intubation with ultrasonography in select patients in the controlled environment of the operating room. Future studies should examine the use of ultrasonography to visualize endotracheal tube placement in real time by emergency physicians with less ultrasonographic training; use of the technique in the emergency department on a wider range of patients, including patients with difficult airways; and assessment of the utility of ultrasonography in confirmation of endotracheal tube position in already intubated patients. [Ann Emerg Med. 2007; 49:75-80.] /$-see front matter Copyright 2007 by the American College of Emergency Physicians. doi: /j.annemergmed INTRODUCTION Background Unrecognized esophageal intubation has a high rate of mortality and morbidity and is more likely to occur in critical care environments. Although numerous techniques have been described to confirm endotracheal tube placement, there is no perfect confirmation tool, and many methods are less reliable or unavailable in the emergency department (ED) or out-ofhospital environment. In addition, most methods require ventilation of the patient to confirm placement of the endotracheal tube. Detection of exhaled CO 2 by capnography, combined with direct visualization of the endotracheal tube Volume 49, NO. 1 : January 2007 Annals of Emergency Medicine 75

2 Ultrasonography for Confirming Endotracheal Tube Placement Editor s Capsule Summary What is already known on this topic Ultrasonographic visualization of real time tracheal tube passage in the neck has been described but has not been studied in live patients. What question this study addressed The authors tested the ability of 2 skilled ultrasonographers to determine tracheal or esophageal tube placement in 33 elective anesthesia patients in the operating room. Each patient underwent placement of the tube in both the trachea and esophagus in random order. A 5- to 10-MHz linear transducer was used to determine tracheal or esophageal location in real time as the tracheal tube passed through the neck. What this study adds to our knowledge Skilled ultrasonographers in a controlled operating room setting can consistently detect passage of a tracheal tube into either the trachea or esophagus in normal airways. How this might change clinical practice Ultrasonography offers a potential means of confirming tracheal tube placement without inadvertent ventilation of the stomach. This technique must be validated by a variety of operators, in emergency settings, and in patients who have difficult airways before it can be recommended as a means of tracheal tube confirmation. passing through the glottis, is the criterion standard in the operating room. However, in emergency intubations, a metaanalysis of the recent literature demonstrated capnography to be only 93% sensitive. 1 Capnography is also not widely available in EDs. Continuous end-tidal CO 2 (ETCO 2 ) monitoring was reported as being available by only 25% of respondents to a recent survey as part of the National Emergency Airway Registry Series. 2 In addition, detection of ETCO 2 by either capnography or colorimetric measurement depends on adequate pulmonary blood flow and is, therefore, unreliable in patients with cardiac arrest or those in a low-flow state, with a sensitivity of only 72% in these patients. 3 Cadaveric studies of ultrasonographic confirmation of endotracheal tube position have yielded promising results. In a randomized study of 140 intubations, 4 Ma et al noted that ultrasonography was 100% senstitive/97% specific in determining tracheal position of the tube during intubation. However, there may be significant tissue and physiologic differences between cadavers and live humans, and no controlled studies have been done to demonstrate the efficacy of ultrasonographic confirmation of tracheal intubation in humans. Werner et al Importance Ultrasonographic machines are available in many EDs and critical care areas and are becoming available in the out-ofhospital environment. Ultrasonographic images are not affected by very low pulmonary flow, as is capnography, and ultrasonographic detection of esophageal intubation can be made before ventilation of the patient. Earlier detection of esophageal intubation could prevent ventilation of the stomach and its associated complications, emesis and aspiration. If ultrasonography is shown to accurately and consistently detect esophageal intubations, applications could include its use in the ED, ICU, and out-of-hospital arenas, in which the most difficult conditions exist and the highest percentage of esophageal intubations occur. Goals of This Investigation The goal of this pilot study is to determine whether emergency physicians experienced in ultrasonography can accurately detect passage of the endotracheal tube into the trachea and esophagus in real time in select patients in a controlled environment to determine whether further study of this method in higher-risk patients and in the ED setting should be pursued. MATERIALS AND METHODS Study Design and Setting This was a prospective, randomized, controlled study conducted in the operating suite of an academic hospital during normal operating hours between October 2004 and March The study was approved by the institutional review board. Written consent was obtained from each subject. Selection of Participants The 2 physicians participating in the study had substantial ultrasonographic training during residency and had completed or partially completed an emergency medicine ultrasonographic fellowship but had not had any formal training specifically in airway ultrasonography. Eligible study subjects were patients 18 years and older who were undergoing elective surgery requiring tracheal intubation. A convenience sample of subjects was recruited during the preoperative evaluation before administration of sedative medications. Exclusion criteria were determined to minimize risk to participants in this pilot study involving intentional esophageal intubations and included a history of difficult intubation, abnormal airway anatomy (eg, Mallampati class 3 or 4, thyromental distance 6.5 cm, mouth opening 3 cm, cervical spine disease, anteriorly protruding incisors), aspiration risk factors or esophageal disease, or predicted difficult intubation at the discretion of the anesthesiologist. For each subject, after induction of anesthesia and neuromuscular blockade, the attending anesthesiologist (C.E.S.) placed the endotracheal tube once in the trachea and once in the esophagus in a random order with direct laryngoscopy. Standard endotracheal tubes sizes 7.0 and 8.0 were used for female and male patients, respectively. Use of a stylet and ventilation of the patient between intubations was at the discretion of the 76 Annals of Emergency Medicine Volume 49, NO. 1 : January 2007

3 Werner et al Ultrasonography for Confirming Endotracheal Tube Placement 5 pressure, and vocal cord position. A higher Intubation Difficulty Scale score indicates a more difficult intubation. The Intubation Difficulty Scale score for an intubation in a patient without a difficult airway, by a single operator, with a single attempt would be zero. There is no maximum because points are accrued for each additional attempt, operator, and technique. The emergency physicians independently recorded the location of each intubation on separate data collection sheets at the intubation. The data collection sheets were taken from the investigators and immediately placed in a separate envelope for later analysis by the anesthesiologist s assistant. Ultrasonographic images of each subject s trachea were recorded on compact flash card before intubation and immediately after each of the 2 study intubations and later reviewed. Twenty-five of the 33 cases were also videotaped. Figure 1. Demonstration of transducer placement on the anterior neck just superior to the suprasternal notch. anesthesiologist. During the intubations, a 5- to 10-MHz Sonosite Titan linear ultrasonography transducer (Sonosite, Bothell, WA) was placed transversely on the anterior neck just superior to the suprasternal notch by one of the 2 emergency physician investigators. Figure 1 demonstrates placement of the ultrasonographic transducer on the neck. The 2 emergency physicians independently recorded the location of the endotracheal tube for each intubation according to the real-time ultrasonographic image. The order of intubations was determined by a computer-generated random-order table and placed in a sealed envelope opened by the anesthesiologist just before induction of anesthesia. In addition to blinding of the intubation order, the emergency physicians were blinded to the actual performance of the intubations by a curtain hung over the patient s neck between the anesthesiologist and the emergency physicians. The transducer was applied to the patient only immediately before and during the actual intubations. Data Collection and Processing The age, height, weight, body mass index, intubation difficulty score, Mallampati score, and American Society of Anesthesiologists physical status were recorded by the anesthesiologist or his assistant for each subject on a data collection sheet. The Intubation Difficulty Scale score was recorded by the anesthesiologist after completion of the study intubations. The Intubation Difficulty Scale is a standard anesthesia scoring tool, with points given for the number of supplementary intubation attempts, number of supplementary operators, number of alternative intubation techniques, glottic exposure, lifting force, necessity of applied external laryngeal Primary Data Analysis A 2-by-2 table was used to calculate the sensitivity and specificity of the emergency physicians ability to confirm tracheal and esophageal tube placement. Stata Special Edition 8 (Stata Corporation, College Station, TX) was used for all calculations. Direct visualization of the location of the endotracheal tube by the anesthesiologist, together with the sustained presence of ETCO 2, was used as the criterion standard. RESULTS Twenty-six women and 7 men were enrolled in the study. The average age was years. The average height was inches. The average body mass index was kg/m 2. The median Mallampati score was 1 (range 1 to 3), and the median ASA physical status was 2 (range 1 to 3). For the initial intubation, the median Intubation Difficulty Scale score was 0 (range 0 to 5). The 2 physicians correctly identified all tracheal and esophageal intubations. For the first set of each pair of intubations, physician 1 identified 14 of 14 tracheal intubations and 19 of 19 esophageal intubations. For physician 1, the sensitivity for identifying the first intubation as tracheal was 100% (95% confidence interval [CI] 77% to 100%), and the specificity was 100% (95% CI 82% to 100%). The results for physician 2 were identical. Although the true CIs would likely have been more narrow had the data from the 2 physicians been combined, the small number of clusters and lack of variance in the outcome brings into question the consistency of combined estimates. Figures 2A and 2B demonstrate the findings of tracheal and esophageal intubations, respectively. One endotracheal tube was unintentionally placed twice in the esophagus, but both tube placements were correctly identified as esophageal by the emergency physicians. The misplaced tube was detected by the anesthesiologist with ventilation of the patient. In esophageal intubations, we found that the esophagus was usually to the left of the trachea (29/33), with 2 of the esophageal intubations being noted to the right of the trachea Volume 49, NO. 1 : January 2007 Annals of Emergency Medicine 77

4 Ultrasonography for Confirming Endotracheal Tube Placement Werner et al Figure 2. A, Ultrasonographic image of tracheal intubation. Only the shadowing posterior to the tracheal rings is visualized. The esophagus is not visualized. B, Ultrasonographic image of esophageal intubation. Shadowing is seen posterior to the tracheal rings and in the left paratracheal space where the esophagus has been opened by the endotracheal tube. and 2 immediately posterior, making them significantly more difficult to detect because motion posterior to rather than adjacent to the trachea was more difficult to appreciate. LIMITATIONS This study has several limitations that may overestimate its accuracy in the emergency medicine setting. First, because we used a dichotomized study design for each set of intubations, if placement of the first tube was confidently observed, the location of the other intubation could be assumed. Second, because we sought to examine the feasibility of this technique, our study was limited to low-risk patients and conducted in a highly controlled environment. The study results, therefore, may not be generalizable to the ED setting or patient 78 Annals of Emergency Medicine Volume 49, NO. 1 : January 2007

5 Werner et al population. In addition, we examined only the ability to detect endotracheal tube position during actual intubations. We did not examine its use in static, postintubation tube placement. Finally, the emergency physicians determining endotracheal tube placement have had substantial training in ultrasonography. DISCUSSION Our study demonstrates that emergency physicians with ultrasonographic training, in a controlled environment, can differentiate tracheal versus esophageal tube passage in real time before ventilation. Use of ultrasonography to confirm endotracheal intubation is attractive for several reasons. Ultrasonographic machines are available in many EDs and critical care areas. Ultrasonographic images are not affected by very low pulmonary flow, as is capnography. Unlike capnography, ultrasonographic detection of esophageal intubation can be made before bag-valve-mask tube ventilation. Earlier detection could prevent ventilation of the stomach and its associated complications, emesis and aspiration. In addition, ultrasonography can potentially be used in the out-of-hospital arena, in which the most difficult conditions exist and the highest percentages of esophageal intubations occur. Ultrasonography can indirectly assess tracheal intubation by visualizing ventilation, either by detecting sliding of the pleura 6-8 or movement of the diaphragm. However, both of these techniques require ventilation of the patient to determine the position of the endotracheal tube and may be affected by the presence of a pneumothorax. 5 Ma et al studied real-time detection of tracheal versus esophageal intubations in a cadaveric study and found that ultrasonography was 100% sensitive and 97% specific in determining tracheal position of the tube during real-time intubations. As in our study, they used a linear, high-frequency transducer in the short or transverse axis. However, the transducer was positioned just below the cricothyroid membrane, whereas we placed the transducer slightly more caudally, just superior to the suprasternal notch at the level of the thyroid gland. We chose this position to avoid placing unintentional pressure on the cricothyroid membrane during the intubation and because an abstract published by Schmitt et al 9 demonstrated increased accuracy in detection of endotracheal tube placement with the transducer positioned at the suprasternal notch. In addition, the transverse view allowed us to visualize the esophageal and tracheal areas during the intubation, whereas the sagittal view is limited to a long-axis view of either the trachea or one of the paratracheal spaces. Ultrasonographic imaging of the endotracheal tube at the neck in already intubated patients has been evaluated and 10 described by Drescher et al. They described the esophageal intubation as an anechoic circle that appears posterior to the thyroid tissue adjacent to the trachea as the tube enters the esophagus. (See Figure 2A for an example.) The intubated trachea was described as containing a hyperechoic or comet sign, and the normal trachea was described as having a periodic Ultrasonography for Confirming Endotracheal Tube Placement resonance artifact, although neither of these findings was uniformly present. We did not specifically look for the comet tail sign or periodic resonance during the live intubations. However, on review of the recorded intubations, whereas we did see movement in the trachea or esophagus, we did not see with any regularity the comet sign described in endotracheal intubations or the periodic resonance signs in unintubated tracheas. As noted above, in esophageal intubations, we found that the esophagus was to the left of the trachea in 29 of 33 cases, to the right of the trachea in 2, and immediately posterior in the remaining 2. The position of the esophagus is an interesting 11 finding, given the results of a study by Smith et which al, examined by magnetic resonance imaging the effects of cricoid pressure on the position of the esophagus relative to the trachea at the level of the cricoid cartilage. They found that the esophagus was laterally displaced in 53% of the necks without and 90% of the necks with cricoid pressure, with leftward displacement of the esophagus in 2 of 3 cases. Although we did not measure the pressure exerted on the neck by the ultrasonographic transducer, it is possible that even a small amount of pressure on the trachea at the level of the suprasternal notch caused lateral displacement of the esophagus. If so, it raises the question of whether the ultrasonographic transducer placement may have a compressive effect on the trachea or any impact on laryngeal view. However, it is also possible that the esophagus travels the trachea more laterally as it moves inferiorly from the level of the cricoid cartilage toward the suprasternal notch, accounting for our finding of a relatively high proportion of laterally displaced esophagi. Future studies should examine the use of ultrasonography as a method for real-time assessment of endotracheal tube placement by emergency physicians with only basic ultrasonographic training, whether use of the technique has an impact on laryngeal view, and the feasibility of using the technique in the ED, ICU, and out-of-hospital environment with a wider range of patients. In conclusion, our study demonstrated that emergency physicians experienced in ultrasonography can accurately detect passage of the endotracheal tube into the trachea and esophagus in real time in a controlled environment with ultrasonography. The authors wish to thank Ramachandra Avula, MD, for his assistance in this study and Joshua Howland Sarver-Tamayo, MD, PhD, for his thoughtful review of the manuscript and helpful suggestions for statistical analysis. Supervising editor: Richard M. Levitan, MD Author contributions: SLW, CES, RAJ, and RKC developed the study protocol. SLW, CES, JRG, and RAJ conducted data collection. SLW and RKC conducted statistical analysis. SLW drafted the article. CES, JRG, RAJ, and RKC were responsible for article review and revision. SLW takes responsibility for the paper as a whole. Volume 49, NO. 1 : January 2007 Annals of Emergency Medicine 79

6 Ultrasonography for Confirming Endotracheal Tube Placement Werner et al Funding and support: The authors report this study did not receive any outside funding or support. Publication dates: Received for publication November 18, Revisions received March 28, 2006; May 2, 2006; and June 8, Accepted for publication July 6, Available online October 2, Abstract presented as a poster at the American College of Emergency Physicians annual meeting, September 2005, Washington, DC. Reprints not available from the authors. Address for correspondence: Sandra L. Werner, MD, RDMS, Department of Emergency Medicine, MetroHealth Medical Center, 2500 MetroHealth Drive, Cleveland, OH 44109; ; fax ; swerner@metrohealth.org. REFERENCES 1. Li J. Capnography alone is imperfect for endotracheal tube placement confirmation. J Emerg Med. 2001;20: Deiorio NM. Continuous end-tidal carbon dioxide monitoring for confirmation of endotracheal tube placement is neither widely available nor consistently applied by emergency physicians. Emerg Med J. 2005;22: MacLeod BA. Verification of endotracheal tube placement with colorimetric end-tidal CO 2 detection. Ann Emerg Med. 1991;20: Ma G, Hayden SR, Chan TC, et al. Using ultrasound to visualize and confirm endotracheal intubation [abstract]. Acad Emerg Med. 1999;6: Adnet F. The Intubation Difficulty Scale (IDS). Anesthesiology. 1997;87: Chun R, Kirkpatrick AW, Sirois M, et al. Where s the tube? evaluation of hand-held ultrasound in confirming endotracheal tube placement. Prehosp Emerg Care. 2004;19: Hsieh KS, Lee CL, Lin CC, et al. Secondary confirmation of endotracheal tube position by ultrasound. Crit Care Med. 2004; 32(Suppl):S374-S Weaver B, Lyon M, Blaivas M. Confirmation of endotracheal tube placement after intubation using the ultrasound lung sliding sign. Acad Emerg Med. 2005;8: Schmitt JM, Ma G, Hayden SR, et al. Suprasternal versus cricothyroid ultrasound probe position in the confirmation of endotracheal tube placement by bedside ultrasound [abstract]. Acad Emerg Med. 2000;7: Drescher MJ, Conard FU, Schamban NE. Identification and description of esophageal intubation using ultrasound. Acad Emerg Med. 2000;7: Smith KJ, Dobranowski J, Yip G, et al. Cricoid pressure displaces the esophagus: an observational study using magnetic resonance imaging. Anesthesiology. 2003;99: Annals of Emergency Medicine Volume 49, NO. 1 : January 2007