Vascular access procedures, such as central venous and arterial catheterization, are commonly performed in the critical care setting. An estimated 5 million central venous catheters (CVCs) are placed annually in the United States¹ in a variety of settings, including critical care units, emergency departments, operating rooms, and in outpatient venues. The usual indications for CVC placement are to assist in hemodynamic monitoring; as a route for the administration of vasoactive medications, total parenteral nutrition, or other vascular irritants; and as a route for drawing blood. In addition, the Surviving Sepsis guidelines advocate measuring mixed venous oxygen saturation in the management of septic shock, which could ultimately lead to increased utilization of oximetric CVCs.

Arterial catheters are an important tool in the management of many intensive care unit (ICU) conditions, including shock, severe hypertension, and other circumstances, in which blood pressure monitoring is important. For a number of reasons, it seems that the role for arterial catheterization in the ICU may also increase. First, with the introduction of “minimally invasive” techniques now available to help estimate cardiac output, arterial catheter placement is becoming increasingly important for the management of selected patients with heart failure. Second, arterial catheterization can be used to assess the response to therapy in patients with pulmonary hypertension. Finally, there has been a significant amount of attention focused recently on respiratory variation of the peak arterial pressure as a means to predict fluid responsiveness in shock states.²

Peripherally inserted CVCs (PICCs) and peripherally inserted catheters sited in a midline position (midlines) have gained increased popularity as an alternative to CVCs in the care of selected patients because of their ease of insertion, longevity, and low rate of early complications. They have become an important component of the central venous access armamentarium. Peripheral intravenous (IV) catheters, long overlooked by the medical ultrasound community, can be placed with very high success rates even in very difficult to cannulate patients when performed under ultrasound guidance.

Vascular access is associated with a relatively low rate of serious complications.¹ However, an improved understanding of complications and why they occur may help the provider to reduce their risk. Complications associated with vascular access procedures are well described,¹ and can be categorized as patient or operator dependent (Table 27-1). Patient-dependent factors include body habitus, coagulopathy, and anatomic variation. Operator-dependent factors include the operator’s level of experience, time allotted to perform the procedure, and human factors, like fatigue and lack of ultrasound guidance.^3–5 The most common complications of CVC placement include accidental arterial puncture, failed placement, malposition of the catheter tip, hematoma, pneumothorax, and hemothorax, the frequency of which vary depending on the site of catheter insertion (Table 27-2). Arterial catheter placement can be complicated by venous puncture, multiple arterial punctures, significant hematoma, and failed placement. PICCs and midline placement are also associated with hematomas and arterial insertions. A common complication of PICC line placement is malposition of the catheter tip into the ipsilateral internal jugular (IJ) vein, or coiling in the subclavian vein or a thoracic branch, such as the thoracodorsal vein (Figure 27-1).

Complications from these procedures are associated with excess direct costs derived from prolonged hospital and ICU lengths of stay (LOS) and additional procedures, such as chest tube insertion or hematoma evacuation, to treat the complications. For example, a single episode of iatrogenic pneumothorax has an attributable LOS of 3–4 days.⁶ Indirect costs, such as additional provider time and patient suffering, are also important considerations.

There are several studies that assess the impact of ultrasonography in improving the success of vascular access procedures. In 1984, Legler et al. published a brief report describing the use of Doppler ultrasonography to locate the IJ vein for cannulation.⁷ Since then, three meta-analyses investigating the use of ultrasound for CVC placement or dialysis catheter placement,^8–10 several review articles, standardized procedure guidelines,^11,12 large case series,^13,14 and the SOAP-3 trial¹⁵ have been published. The body of evidence from these and other studies demonstrates that the use of two-dimensional (2D) ultrasound during central venous access is associated with fewer complications, fewer attempts before successful cannulation, shorter procedure times, and fewer failed procedures when compared with a landmark-based approach. As a result, the Agency for Healthcare Research and Quality and the British National Institute of Clinical Excellence have issued statements advocating the use of ultrasound guidance in central venous access procedures.^16,17 In addition, International Evidenced-based Recommendations on Ultrasound Guided Vascular Access have also been published.¹⁸

Despite these evidence-based guidelines, some providers continue to resist and use ultrasound only in potentially “difficult to cannulate” patients, such as the morbidly obese, or in cases of failed cannulation.¹⁹ Unfortunately, it is difficult to predict which patients will be difficult to cannulate, and the recognition of a failed attempt, as may arise from an occluded vessel, can only be viewed retrospectively after the failure has occurred and the patient has been adversely affected.²⁰ Some complications from CVC are considered preventable medical errors, which refers to either mistakes or poor outcomes that could potentially have been prevented or hospital-acquired conditions, which is a medical problem not present on admission.²¹ Although barriers to widespread adoption exist, such as a lack of training programs,²² ultrasound is a noninvasive tool that can help prevent complications and assists the operator in achieving optimal care for patients with less discomfort and fewer risks. Therefore, developing a strategy to overcome these barriers, both at the local and national level, is a worthwhile goal. It should be noted that even modest training programs have shown the potential to increase success rates and decrease complications.^23,24 Therefore, the consideration of ultrasound to improve safety in all central venous access procedures is recommended.

Transducer Selection

As described before (see Chapter 2), transducers come in a variety of frequencies, each with different properties and clinical applications. Two important concepts are important for ultrasonography in central venous access and need to be reviewed here. First, the relationship between ultrasound frequency and the depth of tissue penetration is an inverse relationship. This implies that low-frequency ultrasound (1–3 MHz) penetrates more deeply than high-frequency ultrasound (7–10 MHz). Second, the relationship between frequency and image detail, or resolution, is proportional. This means that low-frequency ultrasound has poorer resolution than high-frequency ultrasound. Therefore, high-frequency ultrasound provides a very detailed image of superficial structures, to a depth of approximately 5 cm, but cannot penetrate into deeper tissues. Alternatively, lower frequency ultrasound is capable of reaching into deeper structures, but provides a less detailed image. These relationships form the basis for transducer selection. For percutaneous vascular access, which is a procedure that is superficial, higher frequency transducers are ideal.

Modes

A-mode ultrasound has very few clinical applications and is not discussed further here. B-mode ultrasound creates recognizable 2D images. B-mode is the most common mode currently employed in diagnostic medical ultrasound. M-mode ultrasound uses information obtained with B-mode to create an image that demonstrates the movement of structures over time (Figure 27-2). The most common application of M-mode is to assess valve leaflet movement and wall motion in cardiac ultrasound.

Doppler mode also has several forms. The simplest produces no image; there is only an audible signal that varies in intensity with the velocity of the structure being studied (e.g., blood; Figure 27-3). Recently available ultrasound equipment uses Doppler in combination with B-mode to both create an image and give information about velocity (Figure 27-4). Color Doppler takes velocity information obtained by the Doppler shift and applies color to it. The Doppler is then superimposed on the B-mode image (Figure 27-5). Color Doppler is very commonly used in vascular applications, such as vascular access. The strength of the Doppler signal is related to the velocity of the target tissue (e.g., blood) and the angle of incidence. The best estimate of velocity occurs at an angle approaching zero (Figure 27-6). However, if the same vessel is imaged at 90°, there is no perceived motion of blood either toward or away from the transducer, and the Doppler signal fades. When the angle of incidence changes from one “side” of the 90° mark to the other “side,” the color of the blood within the target vessel changes (from red to blue). This is very important and a potential source of error when a beginner is becoming familiar with orientation and selecting a vessel for cannulation.

Techniques of Ultrasound Guidance

Ultrasound is not a substitute for a thorough knowledge of the landmark-based technique for central venous cannulation. Frequently, the beginner may focus on the image on the screen and be inattentive to anatomic landmarks and the position of the needle (Figure 27-7).

Ultrasound-guided procedures can be categorized as static or dynamic. Static guidance refers to the use of ultrasound to localize and mark a site on the skin to facilitate a subsequent percutaneous procedure, much like a traditional landmark-based approach. B-mode or Doppler ultrasound is used to locate the IJ vein, assess its patency, and mark a suitable site on the skin for cannulation. The cannulation itself is not performed with ultrasound. Dynamic guidance refers to performing the procedure in “real time,” with ultrasound imaging viewing the needle puncturing the vessel wall. For vascular access, static guidance appears to be inferior to dynamic, but still better than the landmark-based technique alone.¹² This is due to the time interval between marking with static guidance and the puncture, during which patients may move, or marks removed during skin preparation, both of which can lead to complications. Table 27-3 provides a comparison between static and dynamic guidance techniques. Dynamic guidance is more technically demanding because it requires significant eye–hand coordination. Another distinction is that cannulation can be performed either free hand or with needle guides. Recent data suggest that needle guides may help keep the needle tip in view which, theoretically, could lead to a reduced risk of mechanical complications.²⁵

Planes and Views

For our purposes, there are two planes to be considered: transverse and longitudinal, which refer to the orientation of the ultrasound transducer and the image to the vessel axis. A transverse view is a cross-section and provides the operator with information about structures that lay adjacent to the vessel of interest. For example, a cross-sectional view of the IJ vein will enable visualization of the adjacent common carotid artery and, perhaps, the vagus nerve, thyroid gland, and trachea (Figure 27-8).

A longitudinal view will depict structures anterior and posterior to the vessel of interest and may allow for visualization of the entire needle during cannulation, but does not allow simultaneous visualization of structures lateral to the vessel (Figure 27-9). All commonly utilized central venous and peripheral arterial sites can be visualized in either orientation.

Methods of Orientation

Orientation is probably the most important step to a successful procedure. Most transducers have an identifiable mark, known as a “notch,” on one side. This corresponds to a mark displayed on one side of the image, and allows right–left, or lateral, orientation (Figure 27-10). In rare instances, where the orientation is uncertain, a finger can be rubbed on one side of the transducer surface to produce an image and confirm the orientation (Figure 27-11).