Ultrasonographic guidance of thoracic drainage and biopsy procedures is an attractive alternative to computerized tomography (CT) or fluoroscopic guidance. While CT scanning is the standard for overall imaging of the chest in all cases of malignancy and many nonmalignant conditions, ultrasound may be used for procedure guidance. Ultrasound guidance eliminates further radiation exposure and is often less time consuming and more comfortable for the patient. The intensivist–sonographer, who engages in thoracic interventions, must possess the cognitive and manual skills required for pleural and lung sonography; and, in the case of anterior mediastinal biopsy, must be familiar with the ultrasound anatomy of the mediastinal organs. The ability to visualize inserted hardware and to recognize and interpret reverberation artifact associated with hardware is required for all but thoracentesis and some simple biopsy procedures.

Convex-array or sector-scanning probes with frequencies between 2 and 5 MHz (typically 3.5 MHz) are most suitable for thoracic sonography and are also the most versatile for thoracic procedure guidance.¹ Higher frequency transducers do have better near-field resolution at the expense of penetration depth. We do not recommend probes with a biopsy channel for thoracic interventions due to the common problem of ribcage interference with imaging. Instead, an approach combining imaging with traditional bony landmark detection, that is, finding the upper rib margin with a finder needle prior to insertion of other hardware, is generally preferred. As in general thoracic sonography, the mark on the probe is oriented cephalad and the corresponding mark on the screen is placed at the upper left of the image. Thus, the orientation of standard views, which is imaging in the longitudinal axis, is cephalad left and caudal right on the screen. However, during procedure planning and visualization of hardware, nonstandard imaging planes are routinely used.

Proper patient positioning is essential in interventional chest sonography. Free-flowing pleural effusions follow the gravitational gradient and collect in the most dependent part of the thoracic cavity. In the patient sitting upright, an effusion will collect in the inferior and posterior chest and is most easily accessed from a position behind the patient. The approach to positioning the critically ill patient varies with the size of the effusion, presence of obesity, number and type of support devices, and physiologic compromise, such as hemodynamic instability. Large effusions may be accessed with the patient in the supine position, which presents few problems during access. However, lateral access may be impossible in the very obese even with very large effusions. Adduction of the ipsilateral arm across the chest greatly improves lateral access and should uniformly be attempted. Posterior access may be facilitated by having the patient held in a sitting position but also by placing the patient at the very edge of the bed or even in the full lateral decubitus position. These positions require assistants to assure safety and prevent inadvertent movement of the patient. Occasionally, simply elevating the head of the bed allows lateral access even in cases where this was not possible in the fully supine position. Careful monitoring of endotracheal tubes and vascular access devices is required at all times during patient positioning.

Patient positioning for biopsy of lung or pleural lesions depends on the location of the lesion. The ability of the patient to comfortably maintain position for the duration of the procedure is essential for successful performance of the procedure and for maintaining sterility throughout. The overall guiding principle for positioning for sonographically guided procedures is the individualized approach that fully exploits the flexibility of sonographic imaging.²

By adding sonographic guidance, the success rate of the thoracentesis procedure increases and the complication rate decreases likely by eliminating inadvertent attempts to drain fluid when no fluid is present. Proficiency in diagnostic sonography of pleural effusion is therefore a prerequisite for the successful use of sonographically guided thoracentesis (see Chapter 18). The issues pertaining to sonographic guidance in particular, such as patient positioning and access site selection, are described here. The more common complications of thoracentesis include pneumothorax, pain, shortness of breath, cough, and vasovagal reactions. Other complications that have been described are re-expansion pulmonary edema, inadvertent liver or splenic injury, hemothorax, infection, subcutaneous emphysema, air embolism, and chest wall or subcutaneous hematoma. Of all the complications, sonographic guidance appears to result in lower rates of traumatic pneumothorax; with rates between 5–18% for clinically guided versus 1–5% for sonographically guided thoracentesis. Obviously, the likely reduction of traumatic pneumothorax is especially important in the mechanically ventilated patient, who is at much higher risk for tension pneumothorax than the spontaneously breathing patient.

In patients not receiving mechanical ventilation, the risk of pneumothorax associated with thoracentesis performed by a radiologist has been reported to be 2.7%³; in a surgical intensive care unit setting, the complication rate was found to be 2.4%.⁴ Ultrasound guidance for therapeutic thoracentesis appears to almost eliminate needle trauma as the immediate cause of postprocedure pneumothorax in spontaneously breathing patients. Pneumothorax in this setting has been reported to be associated with unexpandable lung and not laceration of the visceral pleura in almost all cases.⁵ The incidence of pneumothorax in mechanically ventilated patients has been reported by radiologists to be higher than in spontaneously breathing patients, with an overall rate of only 2%,⁶ but with a rate of 7% in intubated patients. However, Godwin and Sahn reported a similar risk of pneumothorax in patients receiving mechanical ventilation when compared with spontaneously breathing patients.⁷ Other studies have reported low pneumothorax rates in the mechanically ventilated patient.^8–11 Although no direct comparison between thoracentesis with versus without ultrasound guidance has been performed, Diacon et al. found that the use of ultrasonography for site location for thoracentesis was more accurate than standard physical examination.¹²

In addition to the apparent safety of sonographically guided thoracentesis in regard to pneumothorax, physician-performed bedside ultrasound guidance may obviate the need to transport the critically ill to interventional radiology, thus eliminating the indirect risks related to the transport of the critically ill.

Because the pulmonary and critical care physician is already familiar with the basic procedure, thoracentesis is ideally suited for the initial adoption of sonographic guidance for thoracic interventions. While diagnostic sonography of pleural effusion is concerned with the detection and characterization of pleural fluid and the size of the effusion, sonographic guidance has the selection of a suitable access site and avoidance of organ puncture as the primary goals.

For ultrasound-guided thoracentesis to be performed safely and successfully, particular attention must be directed toward patient and operator positioning relative to the ultrasound machine to allow unencumbered use of the device without compromising sterility. The procedure field needs to be free of monitoring and support devices. Determination of a suitable access site requires the demonstration of pleural fluid immediately adjacent to the parietal pleura and sufficient distance from organs throughout the respiratory cycle. The diaphragm, liver, or spleen should be identified unequivocally. This is necessary so as not to confuse the curvilinear line of Morrison’s pouch (hepatorenal recess), located between the liver and kidney, with the diaphragm (Figure 24-1 and Video 24-1). On the left, a curvilinear line may also be seen between spleen and kidney, and this line may also be mistaken for the diaphragm by the inexperienced sonographer. Misidentification of these lines can result in inadvertent hepatic or splenic injury. After marking the access site, position of the mark should once again be sonographically verified. Sonography allows the measurement of the distances between the parietal pleura and the various underlying organs. Provided that the patient does not change position between the ultrasound examination and needle insertion and does not cough during the procedure, these distance measurements are reliable and are not subject to compression artifact as are measurements of chest wall thickness.^2,13–15 As the depth of the parietal pleural surface is indicated by pleural fluid return when the needle is advanced under aspiration, the measured distances between parietal pleural surface and underlying organs are measures of how much further and at which angle the needle may be safely advanced. An important caveat is related to possible occlusion of the needle lumen by clot or the solid contents of a complex effusion. This pitfall may be avoided by remembering that fluid return is expected approximately within 5 mm after passing the upper margin of the inferior rib. The needle lumen may also be cleared by injecting small amounts of local anesthetic if occlusion is suspected.

We do not routinely use real-time visualization of the thoracentesis needle during insertion. Real-time visualization adds complexity due to the need for a sterile sleeve, and may interfere with maintaining the proper needle insertion angle.¹⁶

After achieving proper local anesthesia and pleural fluid return, the needle is withdrawn and a small incision is made to allow insertion of the thoracentesis catheter if large volume thoracentesis is performed. Care must be taken that no air is inadvertently introduced during catheter insertion and connection to the drainage system.¹⁷ In case only a diagnostic sample is required, we withdraw the needle into the rib interspace until fluid return stops. We then exchange the syringe containing the local anesthetic with a clean syringe and reinsert the needle to the same depth at which pleural fluid was obtained before and begin aspiration of the sample. Anterior lung sliding may be documented prior to the procedure, and its continued presence after the procedure will reliably exclude an immediate postprocedure pneumothorax.

Indications for sonographically guided tube thoracostomy include complicated parapneumonic effusion, empyema, malignant pleural effusion, and pneumothorax. The type and size of the chest tube is dictated by the underlying condition. Large bore tubes are typically inserted for acute hemothorax or for pneumothorax and bronchopleural fistula in the mechanically ventilated patient; whereas small-bore pigtail catheters are the most versatile of chest tubes and are suitable for most other conditions. For chronic outpatient management of malignant pleural effusion, tunneled catheters may be used. Regardless of the type of tube used, the principles of ultrasound guidance are similar to those of simple ultrasound-guided thoracentesis. We routinely image a guide wire in real time, if applicable, to ascertain proper placement prior to use of dilators or catheter insertion (Figure 24-2 and Video 24-2). In order to image hardware, the transducer is rotated along its long axis to bring the wire or catheter into the sonographic image plane. The tip of a typical J-type guide wire can easily be seen with ultrasonography.

As the drainage devices remain in place, consideration has to be given to patient comfort whenever feasible and the device preferably should be inserted more laterally than during typical thoracentesis. However, this is not always possible and occasionally a pigtail catheter must be inserted posteriorly.

The use of sonography in catheter placement for pneumothorax is limited, as the distance between parietal and visceral pleura cannot be visualized in the presence of intrapleural air. However, sonography is useful in order to exclude underlying organs touching the parietal pleura immediately at the intended insertion site. We recommend CT guidance for pneumothorax in a complex pleural space due to adhesions from previous pleural injury or disease.

Ultrasound-guided needle biopsy is suitable for peripheral lung lesions, anterior mediastinal masses, and lesions of the pleura itself. However, any lesion must abut an accessible area of the parietal pleura and must be readily visualized by sonography (Figure 24-3 and Video 24-3). Critical to the success of the biopsy is operator proficiency in chest ultrasonography and the ability to correlate CT images with ultrasound findings. CT imaging is essential for procedure planning in order to characterize the lesion and document the extent and local topography, such as proximity of the heart or other structures (Figure 24-4 and Video 24-4). However, sonographic guidance is preferred to CT guidance of device insertion due to lack of radiation exposure and better patient comfort. Accessibility of the lesion with ultrasound guidance is determined by imaging of the lesion and confirmation of a clear needle path without intervening air, bone, organs, or vasculature. The angle of needle entry and penetration depth is then determined. Determination of penetration depth is susceptible to skin compression artifact, and this requires employment of strategies similar to the ones described above (see discussion on thoracentesis).