After determining the appropriate level(s) to be blocked based on the surgical procedure, the midline is marked, and a line parallel to the midline is marked 2.5 cm lateral on the side(s) to be blocked (Fig. 1B). After sterile preparation of the skin, subcutaneous infiltration of local anesthetic is performed along the paramedian line. Infiltration may be performed with chloroprocaine to minimize the risk of systemic local anesthetic toxicity while allowing infiltration of the entire area. The block needle is then introduced along this line at the level of the respective spinous processes and perpendicular to all skin planes, and it is advanced until its tip contacts the posterior surface of the transverse process. The depth to the transverse process varies with patient size and block level [47,48]. In an adult patient of normal body habitus, the needle should not be advanced past 4 cm without bony contact. If no bony contact is made, the needle should be reoriented cephalad or caudad to seek bony contact. After contact, the depth of contact is noted, the needle is partially withdrawn slightly, and oriented slightly more cephalad or caudad to “walk off” the transverse process. The needle is advanced 1 to 1.5 cm past the depth of bony contact, and local anesthetic is injected. Using a needle with depth markings on the shaft, or placing a “depth guard” on the needle may be helpful. A distinct “pop” may be felt as the needle tip enters the paravertebral space (through the costotransverse ligament), though this is not consistent.

It is important to not advance the needle more than 1.5 cm past the depth of bony contact, as this could lead to pleural puncture. The needle should not be redirected medially, as this can increase the risk of epidural or intrathecal injection [40]. Whether it is better to perform multiple small-volume injections or a single large-volume injection remains controversial. Multiple-injection techniques may be associated with an increased likelihood of block success [44,49], though each needle pass carries a risk of puncturing the pleura. Single-injection techniques may decrease this risk, although injecting a large volume of local anesthetic in a single paravertebral space may promote contralateral or epidural spread of anesthetic and produce more severe hemodynamic effects [50].

Surgical placement of catheters within the paravertebral space under direct vision [51], and video assistance during thoracoscopic procedures [52] have also been described.

The technique is identical to that described for the landmark-based approach, except that loss of resistance (LOR) is used to confirm positioning of the needle tip within the paravertebral space [17]. The LOR technique generally uses larger-gauge needles, and is especially well suited to placing catheters within the paravertebral space for continuous blocks [6]. A “pressure measurement technique” has been advocated as an objective and reproducible means to identify the TPVS [53]. This technique relies on measuring pressure changes during phases of the respiratory cycle, with a classic “pressure inversion” (pressure during expiration higher than during inspiration) indicating needle tip position within the TPVS. Fluoroscopy [42] may be used as a supplementary method of confirming correct needle or catheter placement within the TPVS.

The technique is essentially identical to that described for landmark-based and LOR techniques, with the addition of electrical neurostimulation as an additional confirmatory end point [54]. High stimulating currents (5 mA, 2 Hz, 0.3 ms pulse width) may be used initially, and contraction of the paraspinal muscles may be observed before the needle tip makes contact with the transverse process. Stimulation can be used to confirm positioning of the needle tip within the TPVS, as contractions of appropriate intercostal/abdominal muscles may be observed at low stimulating currents (0.4–0.6 mA). Because these contractions may be subtle and difficult to see (especially in obese patients) it may be helpful to have an assistant palpate the patient’s intercostal muscles during block placement. Slight movements of the needle tip within the TPVS may help elicit contractions by moving the needle tip closer to nerve roots. It is important that these movements should not be “in-and-out” as this could increase the risk or pleural puncture. Rather, “side-to-side,” “up-and-down,” or rotational maneuvers should be used. It is also helpful to use insulated needles designed for use with a nerve stimulator.

Because US guidance is a relatively new technique for aiding the placement of TPVBs, a standard technique does not yet exist. Several different techniques are possible. Each technique has specific advantages and risks that should be taken into consideration before deciding which one to use in a particular clinical situation. Because US imaging of the TPVS may be challenging and because there is a significant risk of pleural puncture or injection of anesthetic directly into a nerve root, the subarachnoid space, and even the spinal cord itself, this technique should not be attempted by practitioners without significant experience in performing US-guided procedures and TPVBs. For those who do have a sufficient amount of training and experience, use of ultrasonography can be helpful as an adjunct to traditional techniques, or as a stand-alone method of guiding block placement.

As an aid to traditional techniques, ultrasonography can be helpful for measuring the depth to the transverse process and pleura as well as planning the trajectory of the needle, which may be especially helpful in patients with abnormal anatomy. Jamieson and Mariano [55] published a case report of 2 TPVBs placed using a nerve-stimulator guided technique after a US prescan. Pusch and colleagues [56] reported a series of 22 patients who had TPVBs placed using a landmark-based technique after measuring the depth to the transverse process and pleura at the T4 level. These studies found that the US prescan reliably measured the depth to the transverse process. Hara and colleagues [57] reported a series of 25 patients with TPVBs placed at the T4 and T1 level using an LOR technique in combination with real-time US guidance, using ultrasonography to guide the needle to make contact with the transverse process. After contact was made, the needle was advanced without US visualization and LOR was used to confirm entry of the needle tip into the TPVS.

Several techniques have been described using ultrasonography as a stand-alone technique for placing single-injection and continuous TPVBs. US-guided procedures can be performed by imaging the target structure(s) in the long or short axis. Needle guidance can be performed using the in-plane or out-of-plane technique. Although any combination of transducer orientation and needle guidance technique can be used to perform a block, the in-plane needle approach is generally preferred for TPVBs as it allows more consistent identification of the needle’s tip. For the out-of-plane approach, practitioners should use surrogate methods such as hydrolocalization to locate the needle’s tip.

To obtain a long-axis view of the TPVS, the US transducer is placed in a sagittal orientation (parallel to the long axis of the spine, Fig. 2A) and moved medially and laterally to identify the spinous processes, the transverse processes, the ribs, and the pleura (Fig. 2B). By scanning medially and laterally, the transverse process and rib (Fig. 2C) can be identified, and the spatial relationship of the rib and transverse process can be delineated (Fig. 2D). This parasagittal view has the advantage of allowing simultaneous visualization of multiple levels. Use of a curved-array US transducer may provide a “wider” field of view (more levels). It may be difficult to use an in-plane needling technique with this view as the transducer’s “footprint” (50 mm) may make it difficult to pass the needle’s tip between the transverse processes (Fig. 2E). Use of a transducer with a smaller “footprint” (20–25 mm) may eliminate this problem, however. Alternatively, this technique can be used to guide the needle to contact the transverse process, and it can then be “walked off” using needle depth or LOR to confirm entry of the needle’s tip into the TPVS. This long-axis view can also be helpful to determine the extent of spread of local anesthetic within the TPVS. After the block is performed, the anesthetic fluid can be seen as a hypoechoic (dark) stripe superficial to the pleura. Scanning rostrally and caudally, the superior and inferior limits of local anesthetic spread can be determined.

Fig. 2 Sonoanatomy of the TPVS in the long-axis view. (A) US transducer position for long-axis imaging of the TPVS. After determining the appropriate level(s) to be blocked, a high-frequency curved or linear-array transducer is placed vertically along a paramedian line 2 to 3 cm lateral to the midline. Other types of transducers may be helpful in specific clinical situations such as deeper blocks (low-frequency, wide-footprint, curved-array) or pediatrics (very high-frequency, small-footprint, linear-array “hockey sticks”). (B) Long-axis US image of the TPVS. Cephalad and caudad aspects of the US image are labeled. The top of the image is superficial (posterior) and the bottom deep (anterior). The transverse processes (TP) typically have hyperechoic flat posterior surfaces with acoustic shadowing below. The pleura (arrowhead) is below (anterior) and the deeper lung parenchyma can be seen. Note that the distance between the posterior surface of the transverse process and the pleura is approximately 1.5 cm (depth markings on right-hand side of image). Movement between the visceral and parietal pleura can be observed during inspiration and expiration (the “sliding lung sign”), allowing identification of the pleura (more difficult on still images than real-time). (C) By moving the US transducer laterally (maintaining parasagittal orientation) the ribs (R) can be demonstrated, and the pleura (arrowhead) and lung (L) can be seen below. Cephalad and caudad aspects of the image are labeled. The ribs typically have a rounded posterior surface with shadowing below, and are wider than the transverse processes. An intercostals artery is circled on the inferior aspect of the cephalad rib. Note that the distance from the superficial surface of the rib and the pleura is approximately 1 cm (depth markings on right-hand side of image). (D) Simultaneous long-axis US image of the transverse process (TP) and rib (R). Both are outlined. Cephalad and caudad aspects of the image are labeled. Note that the rib is anterior (deep) to the transverse process. The pleura is marked with arrows (arrowhead) and the lung (L) is seen below. (E) Curved-array US image of the TPVS in the “long-axis” view. Cephalad and caudad aspects of the image are labeled. Because the probe is narrow but allows a wide field of view, multiple levels can be imaged simultaneously and the needle can be directed almost vertically anterior (deep) using an in-plane technique to pass between the transverse processes (TP, outlined) and enter the TPVS (outlined). The ribs (R, outlined) can be seen below the transverse processes. The pleura (arrowhead) and lung parenchyma (L) can also be seen between the acoustic shadows created by the bony structures.

O’Riain and colleagues [58] described a similar parasagittal long-axis in-plane technique and reported block success in 8 of 9 patients undergoing breast surgery. All patients received a continuous TPVB placed at the T3 level on the operative side. The TPVS was expanded with saline and a catheter was then passed 3 cm past the needle’s tip. If there was no hemodynamic response to a standard test dose, 10 mL of 0.25% bupivacaine was injected through the catheter. The number of dermatomes blocked using this technique was not formally assessed. All patients had GA for the surgical procedure. No complications were reported. The investigators stated that needle visualization may be challenging using this approach due to the “acute angle the needle must take to enter between adjacent transverse processes.” Although the reported block success rate was high in this series, the small number of patients, the fact that blocks were not used for surgical anesthesia, and the lack of a control group (TPVBs performed without US guidance) make it difficult to make definitive statements regarding the efficacy of this approach. Confirmation with larger series and prospective comparisons with standard approaches are still needed.

To obtain a short-axis view of the TPVS, the transducer is placed in an axial orientation (perpendicular to the long axis of the spine, Fig. 3A). With the transducer in this orientation, the spinous processes, laminae, transverse processes, ribs, and pleura can be seen simultaneously (Fig. 3B). Because these structures cannot be seen simultaneously using the long-axis view, this type of imaging may allow better appreciation of the TPVS. The TPVS can be imaged using this view by scanning in a rostral or caudal orientation until the rib is not in view (Fig. 3C). The rib and pleura may appear similar on the US image, and must be distinguished from one another for the block to be performed safely. The rib can usually be identified by the characteristic acoustic dropout “shadowing” seen during US imaging of bone (hyperechoic surface, US waves unable to penetrate past the surface, see Fig. 3C). In contrast, the parenchyma of the lung can usually be visualized deep to the hyperechoic pleura. In addition, the “sliding lung sign” can help confirm the identity of the pleura. By having the patient take a deep breath, the sliding of the parietal and visceral pleura can be seen. After confirmation of the relevant anatomic relationships, the needle tip can then be positioned in the TVPS using an in-plane needling technique by inserting the needle from the lateral edge of the transducer (see Fig. 3A). This short-axis in-plane technique has the advantage of consistently allowing visualization of the needle as it is advanced toward the paravertebral space. However, this approach should be used with great caution, as failure to maintain visualization of the needle’s tip could allow the needle tip to pass through the intervertebral foramen into the central neuraxial space or the spinal cord itself.

Fig. 3 Sonoanatomy of the TPVS in the short-axis view. (A) US transducer position for short-axis US imaging of the TPVS. To perform an in-plane TPVB using this type of imaging, the needle is inserted from the lateral edge of the probe and directed medially. (B) Short-axis US image of the TPVS. Medial and lateral aspects of the Image are labeled. The posteromedial surface of the transverse process (TP) is well seen, but the anterior aspect is not due to acoustic shadowing. The pleura can be seen as a curved hyperechoic line with lung parenchyma beneath. Use of the sliding lung sign can be helpful to differentiate pleura from rib. To perform an in-plane TPVB using this view, the needle is directed from lateral to medial so that its tip is positioned within the TPVS (outlined). Constant visualization of the needle’s tip is essential to prevent injury to the pleura or entry into the spinal column through the intervertebral foramen (arrowheads). (C) Short-axis US image of the TPVS showing the transverse process (TP), lung (L) and rib (R) simultaneously. The pleura and lung can be distinguished from the rib as the parenchyma of the lung is visible, while structures deep to the rib’s surface are not well seen due to acoustic shadowing.

Luyet and colleagues [59], using the short-axis in-plane technique, placed styleted catheters in the TPVS of 20 cadavers. In all cases, the relevant anatomy and needle were easily visualized, and the catheters were easy to pass. However, during the subsequent anatomic dissection, a mixture of local anesthetic and contrast dye injected through 9 of the 20 catheters was seen in the epidural space, the pleural cavity, or the mediastinum. Although it is not clear how this could impact blocks performed in clinical situations, the high incidence of concerning patterns of local anesthetic spread in this study highlight that caution should be exercised using this technique.

Renes and colleagues [60] recently described a similar technique in a series of 36 patients undergoing various surgical procedures of the chest or abdomen. Each block was placed at a level deemed appropriate for the surgical procedure. Under real-time US guidance, the needle’s tip was positioned within the TVPS. Fifteen milliliters of 0.75% ropivacaine was then injected through the needle and a catheter passed 2 cm past the needle’s tip. Placement of the catheter was confirmed by injecting an additional 5 mL of 0.75% ropivacaine with US visualization of spread within the TPVS. A blinded observer assessed block success, and catheter placement was confirmed radiographically by injecting contrast dye through the catheter tip. Renes and colleagues reported a block success rate of 100% with reduced cold sensation in a median of 6 dermatomal segments, and all catheters were radiographically well -positioned. One patient had evidence of epidural spread of contrast dye, but had no sensory changes on the contralateral side. All patients in this series reported adequate pain relief, and no complications were reported. No control group was included for comparison.

When we (the authors of this article) use this short-axis, in-plane technique to place a catheter in the TPVS, we inject saline (3–5 mL) through the needle to create a “pocket” within the TVPS for the catheter to pass into. Because of the high rate of apparently malpositioned styleted catheters in the aforementioned cadaver study by Luyet and colleagues, we do not use catheters with stylets. We prefer to use wire-reinforced flexible single-orifice epidural catheters (such as the Arrow Flextip Plus: Teleflex Medical, Research Triangle Park, NC. USA). Occasionally we will use a styleted stimulating peripheral nerve catheter (such as the Arrow StimuCath: Teleflex Medical), as these catheters may be easier than the epidural catheters to visualize with ultrasonography. If we are using one of these stimulating catheters, we remove the stylet before passing the catheter past the needle’s tip. Once we have passed the catheter 3 to 4 cm past the needle tip (we do not recommend passing the catheter more than this to minimize the risk of the catheter advancing into the epidural space), we inject saline (3–5 mL) through the catheter using real-time US visualization to confirm proper spread of fluid within the TPVS. We prefer to confirm appropriate spread of fluid injected through the needle and catheter with saline in order to avoid potential inadvertent subarachnoid or epidural injection of local anesthetic. We inject local anesthetic under low pressure, slowly and incrementally, to decrease the risk of epidural spread and minimize hemodynamic consequences if the catheter is malpositioned within the epidural or subarachnoid space.