Ultrasound of the neck and upper respiratory system has many potentially useful clinical applications.¹ Aside from vascular access (see Chapter 27), some of these indications include confirmation of satisfactory endotracheal tube (ETT) placement, evaluation of the larynx, guidance for percutaneous tracheostomy, evaluation of the paranasal sinuses, and assessment of vocal cord disorders. The data demonstrating improved outcomes by using ultrasound for imaging the upper airway remain scarce. However, there are important opportunities to improve care for the intensive care unit (ICU) patient that can be derived from its use.

Ultrasound use for the evaluation of the paranasal sinuses was recognized in Europe as technically feasible to confirm the presence of sinus disease as early as the 1960s.² However, widespread clinical application emerged only recently with the development of low-cost, high-quality bedside ultrasound imaging technology. Earlier studies established ultrasound as an alternative to computed tomography (CT) for the diagnosis of maxillary sinus disease and described the typical findings associated with sinusitis.^3,4 With improvements in imaging, more recent reports focused on improving the diagnostic accuracy of ultrasound by performing postural maneuvers.⁵ In 2006, Vargas et al. investigated the role of ultrasound for performing transnasal puncture of the maxillary sinus in intubated ICU patients. In patients suspected of having sinusitis, they found ultrasonographic evidence of maxillary sinusitis in 70% of patients, and of these, 93% had positive results from transnasal puncture, demonstrating the comparability of ultrasound to CT for the diagnosis and transnasal puncture of sinusitis.⁶ More recent studies investigated the characterization, or “staging,” of sinus fluid collections by noting the presence or absence of acoustic streaming in a model of sinusitis. More viscous collections (pus) are less likely to undergo acoustic streaming than less viscous collections.⁷

There are no studies that describe an improvement in ICU outcomes by using ultrasound instead of standard CT, even though a CT scan has more radiation, is more expensive, and requires the transportation of critically ill patients to and from the radiology department as well as the use of valuable critical care nursing time. There are important roles, however, for CT imaging of the sinuses that cannot be duplicated with ultrasonography. These include any planned surgical procedure involving the sinuses, suspected sinus trauma, and suspected malignant disease. This discussion focuses on the use of ultrasound for the evaluation of paranasal sinusitis.

Sinus disease is important to recognize in critically ill patients because it is a source of fever, which leads to costly diagnostic workups and empiric therapeutic regimens.^8,9 In addition, maxillary sinus disease is an independent risk factor for the development of nosocomial lung infections.¹⁰ Although not studied in any systematic fashion, it is also conceivable that undiagnosed sinusitis may lead to significant pain and agitation, resulting in the increased use of sedatives and analgesics, which could then delay extubation.

Technique

The anatomy of the paranasal sinuses is shown in Figure 17-1. The sinuses most amenable to ultrasonographic examination are the maxillary and frontal sinuses; however, most studies have been performed on the maxillary sinus. The maxillary sinus is contained within the maxilla, and is bordered by the orbital floor superiorly, the hard palate inferiorly, the nasal wall medially, and the zygoma laterally. In the normal state, the sinus is air filled, thus impairing the transmission of ultrasound energy. In this case, what is seen is the anterior wall only, with some artifact known as “acoustic shadowing” (Figure 17-2), which obscures all underlying structures; this is considered a negative study. When filled with fluid, ultrasound penetrates the anterior wall, “travels” through the fluid, and strikes the posterior or lateral walls and “reflects” back to the transducer, resulting in an image of the sinus cavity in its entirety (Figure 17-3). This is known as a “sinusogram,” which is a positive study. A partial sinusogram, where only the posterior wall or a side wall is seen, can occur due to the presence of an air–fluid level in the sinus or mucosal thickening. The patient’s position influences the appearance of the fluid in a partial sinusogram. In the supine position, fluid can “layer out” away from the anterior wall, resulting in either acoustic shadowing or a partial sinusogram. However, when placed in a semirecumbent or upright position, the fluid (if present) will follow gravity and cover the floor of the sinus, coming in contact with the anterior wall. When imaged, this results in either a partial or complete sinusogram, depending on how much fluid is present and on the transducer orientation or angulation (Figure 17-4).

For a sinus ultrasound, patients can be placed in a semirecumbent position. A 3–5 MHz cardiac probe with a small footprint is used. Proper transducer position is demonstrated in Figure 17-5. The horizontal plane is scanned first, angulating the probe cephalad (toward the orbital floor) and caudally (toward the floor of the sinus); this is followed by turning the transducer 90° and scanning from the medial to the lateral wall. The technique is then repeated on the alternate side. A complete ultrasound maxillary sinus scan, in contrast to CT scanning, can be performed in <60 seconds. If a complete sinusogram is seen, no further evaluation is necessary and the patient should be treated for sinusitis. If a partial sinusogram is obtained, postural maneuvers may help elucidate the cause: sinusitis versus mucosal thickening. Until proficiency with the ultrasound technique is mastered, bedside ultrasound can be correlated with CT scan.

Ultrasonography has recently been shown to be helpful in the in vivo analysis of vocal cord function. Clinical utility of this technique includes the pre- and/or postsurgical evaluation of vocal cord function at the time of thyroid or complex thoracic surgery,¹¹ assessment for the presence/absence of vocal cord paralysis,¹² or identification of vocal cord dysfunction syndrome. One recent study shows a reasonable correlation between findings of ultrasonography when compared to laryngoscopy in cases of vocal cord paresis.¹³ In addition, ultrasound has been shown to be a useful adjunct to endoscopy in vocal cord injection procedures, typically performed to improve phonation in cases of unilateral vocal cord paralysis.¹⁴

The technique of vocal cord visualization is fairly straightforward, but the operator should expect a learning curve. Using a high-resolution linear array transducer (6–13 mHz) usually works best. Position the patient supine with the neck slightly extended. Palpate the external anatomy, including thyroid and cricoids cartilages. Place the transducer transversely over the midportion of the thyroid cartilage perpendicular to the axis of the trachea. Move the transducer axially along the thyroid cartilage until the glottis structures are visualized; identifying the arytenoids first, which are bright structures and, therefore, usually easy to identify, helps orient the operator. Slight angulation of the probe cranially or caudally may improve visualization. Once the cords are visualized, have the patient “hum” to assess cord function in real time (Video 17-1).

In 1987, Raphael and Conard used B-mode 2D transtracheal ultrasound to assess the capability of ultrasound to visualize and confirm ETT placement.¹⁵ In this study, the investigators were primarily interested in verifying the intratracheal placement in patients already known to have successful tracheal intubations. The study was not attempting to identify esophageal intubation or any other malposition. The authors suggested that ETT cuffs be filled with saline to reduce the acoustic impedance of the air-filled ETT balloon, which would improve ultrasound transmission (similar to having a full bladder during abdominal ultrasonography). The contrast between the air-filled trachea and saline-filled balloon allows the position of the ETT to be identified more easily. It was also suggested that a longitudinal view, combined with a slight to-and-fro motion, could improve visualization. They concluded that this technique was beneficial for certain patient populations like pregnant women or patients receiving frequent chest radiographs to monitor ETT position. Building on this data, a recent study demonstrated that novice sonographers, after only a 50-minute training session, could accurately identify the intratracheal position of a saline-filled ETT cuff,¹⁶ underscoring the relative accessibility of these techniques to most practitioners.

A blinded, prospective study of 40 patients undergoing elective surgery was performed to identify esophageal intubation using 2D ultrasound.¹⁷ The authors identified esophageal intubations with a sensitivity of 100% and correctly identified all five esophageal intubations, and 34 out of 35 tracheal intubations. Two additional studies, using both live patients and cadavers, confirmed the high sensitivity and specificity of 2D ultrasound to evaluate ETT position.^18,19 In addition, improved sensitivity and specificity could be achieved by using a dynamic approach (visualization of the tube during placement) as compared with a static approach (confirmation of placement after the fact).¹⁹ Multiple studies have also concluded that tracheal ultrasound is as effective as traditional capnography for determining ETT position, and that the technique is faster than auscultation/capnography, even in obese patients.²⁰

Endotracheal tube malposition in the right mainstem bronchus can also be identified by using bilateral pleural ultrasound.^21–24 The parietal–visceral pleural interface can be easily identified with a high-frequency probe (Figure 17-6). This interface, known as the visceral–parietal pleural interface (VPPI), has a characteristic “shimmering” appearance during lung ventilation. The two pleural surfaces can be seen to slide past one another, which is responsible for producing the shimmering effect. If the ETT is positioned in a mainstem bronchus, the sliding or shimmering will either be greatly reduced or absent on the contralateral side. This assumes that there is no anatomic airway obstruction, such as an obstructing tumor, causing the reduced or absent pleural shimmer, which could lead to a false-positive test and inappropriate repositioning of the ETT. This approach was confirmed by Weaver et al. using cadavers.²⁵ The sensitivity for identifying esophageal intubation was 95–100% and the sensitivity for a right mainstem intubation versus a tracheal intubation was lower, at 70–75%, using the sliding lung sign.

Other applications for upper airway ultrasound, aside from verification of ETT position, include appropriate ETT size selection, confirmation of laryngeal mask airways, the prediction of a difficult airway, and the prediction of postextubation stridor. Lakhal et al. studied whether or not appropriate selection of ETT size could be informed by using translaryngeal ultrasound. After measuring the diameter of the airway in the subglottic region by translaryngeal ultrasound, the investigators then compared their measurement to those obtained by using magnetic resonance imaging and found a strong correlation between the two measurements. The authors suggested that ultrasound could accurately gauge the diameter of the subglottic airway and could guide selection of an appropriately sized ETT.²⁶ Gupta et al. published a study comparing ultrasound evaluation of laryngeal mask airway (LMA) position compared to endoscopy in patients undergoing elective intubation for same-day surgery procedures. The authors concluded that ultrasound accurately assessed LMA position, and that this technique could potentially replace endoscopy in this patient population.²⁷