Thoracic paravertebral block procedure guide

INTRODUCTION — The thoracic paravertebral block (TPVB) is a peripheral nerve block performed by injecting local anesthetic (LA) into the thoracic paravertebral space (TPVS). The TPVB targets spinal and sympathetic nerves, in order to produce an ipsilateral segmental somatic and sympathetic block. TPVB is used for anesthesia and analgesia for surgery of the chest and abdomen. This block requires a greater degree of technical skill than many other peripheral nerve blocks, due to the deeper location of the target, the narrow sonographic window, and the potential risks of pneumothorax and epidural or intrathecal spread (including total spinal). This topic will discuss the anatomy, ultrasound imaging, injection techniques for performing TPVB, and complications specific to TPVB.

General considerations common to all peripheral nerve blocks, including patient preparation and monitoring, use of aseptic technique, localization techniques, drug choices, contraindications, and complications, are discussed separately. (See "Overview of peripheral nerve blocks".)

ANATOMY — Each thoracic paravertebral space (TPVS) is a wedge shaped compartment that lies alongside vertebral column, in between the heads and necks of the ribs above and below. The anterior-posterior dimension of the TPVS is greatest medially (figure 1).

●Borders of each TPVS are as follows [1]:

•Medial – Intervertebral foramina, vertebral bodies, and intervertebral discs.

•Lateral – TPVS continues as the intercostal space; the border between the spaces is unclear.

•Anterior – Parietal pleura.

•Posterior (dorsal) – Ribs, transverse processes, superior costotransverse ligament (SCTL), and internal intercostal membrane (IIM).

The IIM is a continuation of the internal intercostal muscle, and it serves as the posterior border of TPVS lateral to the SCTL. The IIM fuses with the SCTL medially. The SCTL runs from the transverse process (TP) above to the rib below obliquely from cranial to caudad and also from lateral to medial. The ventral surface of the SCTL is at a depth similar to the ventral surface of the TP.

●Contents – The fat within the TPVS contains the dorsal and ventral rami of the intercostal (spinal) nerve and the sympathetic chain (which are the targets of a TPVB) and blood vessels.

●Communications – There are communications between ipsilateral TPVSs, between TPVSs on either side of the spine, and between the TPVS and other clinically relevant spaces [1,2].

•The TPVS communicates with the epidural space medially, the intercostal space laterally, and the cervical paravertebral space cranially (image 1).

•There is controversy over the caudal most limit of the TPVS, below T12. Some cadaver studies have found spread of dye into the low TPVS from injections below the arcuate ligament [3,4], while others have found no spread into the TPVS [5-7].

•The SCTL that forms most of the dorsal border of the TPVS has gaps (figure 2), which may allow spread of local anesthetic (LA) into the paravertebral space from more superficial injection (eg, with the midpoint transverse process to pleura [MTP] block or erector spinae [ESP] blocks).

CLINICAL ANATOMIC CORRELATIONS — The TPVB is used to provide anesthesia and analgesia for breast surgery, surgery that requires coverage of the axilla (eg, creation of dialysis access, axillary lymph node dissection), or for thoracoabdominal surgery as an alternative to thoracic epidural.

Important clinical issues related to anatomy include the following:

●The selected level for the block should be as close to the location for the dermatome for the incision or injury as possible (table 1). Three to five contiguous intercostal nerves may need to be blocked to provide complete coverage for one dermatome, due to the interconnections between spinal nerves and overlap in innervation of dermatomes. This can be accomplished with blocks at multiple levels, or with a larger volume single-injection. (See 'Single versus multiple injections' below.)

●Performing a TPVB more medially may theoretically reduce the risk of pleural puncture. However, placing a needle or stiff catheter more medially may increase the risk of neuraxial complications (eg, excessive epidural spread, accidental dural puncture with or without spinal block) as well as increase the chance of catheter migration anteriorly into the mediastinum (figure 3). Epidural spread may be more likely when using high volume and pressure during injection [2].

CHOOSING A BLOCK TECHNIQUE — TPVB can be performed by using surface landmarks or with ultrasound; ultrasound can be used continuously for real time ultrasound-guided block, intermittently during the block, or as a preprocedure scan to delineate anatomy.

Landmark approach versus use of ultrasound — We suggest using ultrasound guidance to improve success (accurately determine the level and location of transverse process [TP], guide the needle to the TP or thoracic paravertebral space [TPVS] depending on preferred TPVB approach, confirm needle tip location and local anesthetic [LA] deposition in the TPVS) and reduce complications (neuraxial blockade or pleural puncture) [1,8,9].

Evidence comparing ultrasound guidance with a landmark-based approach for TPVB is limited to quasi-experimental and non-experimental observational studies rather than randomized controlled trials. There is a learning curve for ultrasound use in regional anesthesia, and pleural puncture has been reported despite the use of ultrasound [10]. (See "Overview of peripheral nerve blocks", section on 'Ultrasound guidance'.)

In one trial, 72 females who underwent unilateral breast surgery were randomly assigned to have paravertebral block with multiple injections from T1 to T6 using ultrasound guidance versus anatomic landmarks [9]. Ultrasound guidance resulted in more successful blocks (94 versus 72 percent), more dermatomes blocked, and better pain control at two and four hours. The number of complications were similar, including two pleural punctures in the ultrasound group, and one in the landmark group.

Single versus multiple injections — TPVB can be performed using multiple injections at spaces within the desired dermatomal coverage, or using one or more larger single injections, expecting spread across multiple dermatomes. Multiple injections take longer, may be more uncomfortable for the patient, and could increase the risk of vascular and pleural puncture.

Literature comparing multiple versus single injections for TPVB is conflicting [11-13]. Most studies suggest that when the block is used for analgesia, single large volume injection is likely to be as effective as multiple injections. For some procedures, multiple injections may be required to ensure coverage of particular dermatomes, particularly for more cranial dermatomes (eg, T2 coverage for breast surgery). In studies that have used radiocontrast to evaluate the extent of spread after paravertebral injection, the ultimate extent of sensory block is usually slightly greater than the spread of contrast, perhaps due to diffusion on LA after initial injection.

ULTRASOUND-GUIDED OR ASSISTED BLOCKS — Real time continuous ultrasound-guided block is reasonable if the thoracic paravertebral space (TPVS) is easily visualized and if the clinician is experienced and technically proficient with ultrasound-guided blocks. If not, we use an intermittent ultrasound technique. For most patients, and for all continuous blocks, the authors use an intermittent ultrasound technique, whereby we use ultrasound in real time to insert the needle tip at the inferior aspect of the transverse process (TP), and again when injecting local anesthetic (LA). Others use continuous ultrasound guided block more routinely.

Ultrasound imaging — General principles of ultrasound imaging for TPVBs are discussed here. Specifics are discussed in the block techniques described below.

Ultrasound equipment — Use a high-frequency (6 to 13 MHz) linear transducer if the pleura is visualized at a depth <4 cm. Use a low-medium frequency (2 to 6 MHz) curvilinear transducer, preferably with a smaller footprint, when the pleura is >4 cm deep. Use a curved transducer with a small footprint for an in-plane approach with an axial transducer orientation. (See 'Transducer in axial orientation' below.)

For optimal visualization, use a large bore echogenic needle, either a block needle or an epidural needle. The needle and tubing should be meticulously cleared of air before performing the block. Inadvertent injection of air will distort further ultrasound imaging.

Scanning — Scanning can be performed with the transducer in axial, sagittal, or oblique orientation. There is no one correct orientation, and clinicians should use the orientation with which they are most comfortable. The following should be considered:

●The needle is best visualized when inserted in plane to the ultrasound beam.

●The pleura is best visualized with the transducer in an axial orientation.

●If the needle is inserted in plane (lateral to medial) with the transducer axial, or out of plane with the transducer in sagittal orientation, the needle will be aimed towards the neuraxis, potentially increasing the risk of epidural injection or dural puncture.

Count ribs — We count ribs with ultrasound rather than surface landmarks to more accurately identify the desired TP level, particularly when spinous processes (SPs) are not clearly palpable (movie 1). The ribs can also be counted by identifying the spinous prominens (C7), and counting caudally from there. At the levels T4 to T8, the spinous process of the vertebral body above correlates with the TP of the vertebral body one below.

●Place the transducer in sagittal orientation approximately 6 cm from midline. Start at the top to identify the first rib, or at the bottom to identify the 12^th rib.

●Slide the transducer along the ribs down or up, parallel to the spine, counting the ribs until you reach the desired level. Center the rib on the ultrasound screen.

●Slide the transducer medially along the rib until the TP can be seen as a distinct, more superficial bony structure on the screen.

●Center the TP on the ultrasound screen and draw a horizontal line to mark the TP. Label the rib.

Identify structures — While scanning, identify the following:

●Ribs, TP, lamina, and SP.

●The internal intercostal membrane (IIM), and the superior costotransverse ligament (SCTL) – Often difficult to visualize.

●Pleura – Crucial to visualize to avoid needle contact, and often well visualized laterally but poorly visualized medially due to the oblique angle relative to the transducer (figure 4 and figure 5).

●TPVS – Visualized as a dark space above the pleura with a width of approximately 3 to 5 mm.

Injection of fluid expands the TPVS and displaces the pleura anteriorly. NOTE: Anterior pleural deflection is a key sign of desired spread of LA.

●If visualization of the TPVS (or the needle tip) is poor, it may be safer to avoid a real time block with continuous ultrasound, and instead use an alternative technique, or an alternative block.

Block using continuous real time ultrasound — This technique involves real time use of ultrasound throughout the block procedure.

Patient positioning — The patient can be positioned prone, sitting, or in the lateral decubitus position with the side to be blocked up. We find the prone position easiest, but the best position often depends on the clinical circumstance.

Marking the skin

●Count the ribs, mark horizontal lines over the ribs and number them. Mark the desired paravertebral space (picture 1). (See 'Count ribs' above.)

●Draw a vertical (cranial-caudal) line along the midline (over SPs).

●Draw a vertical line over the tips of the TPs, to guide transducer placement and avoid mistaking ribs (more lateral) for TPs (figure 6).

Performing the block — We describe the block with the transducer in sagittal and axial orientations. We prefer a sagittal transducer orientation to avoid aiming the needle medially towards the neuraxis.

Transducer in sagittal orientation

●Place the transducer along the line over the tips of the TPs, approximately 2.5 cm lateral to the midline.

Identify the pleura. If necessary, slide the transducer laterally (figure 6) or tilt the transducer to aim the beam laterally to visualize the pleura.

●Insert the needle in plane to the transducer,0.5 to 1 cm from the caudad edge of the probe. Out of plane needle insertion with a sagittal transducer orientation is an advanced technique that should only be performed by experienced clinicians. With lateral to medial needle insertion, the tip is aimed at the neuraxis, and the pleura is not well visualized. With medial to lateral needle insertion, the TPVS is narrower (a smaller target), and the needle is aimed at the lung.

Note: It can be difficult to find a needle path shallow enough to visualize the needle, and yet steep enough to avoid the edge of the transducer and the tip of the TP. If this occurs, either slide the transducer cephalad to the point at which the caudad TP is just barely in view, or switch to a smaller footprint curvilinear transducer, or switch to an oblique transducer orientation. This concept is shown in a figure (image 2).

●Advance the needle to the TPVS. A "pop," may be felt as the needle passes through the SCTL. Some clinicians use loss of resistance as additional evidence that the needle tip is in the TPVS, though this endpoint is less reliable than it is for epidural placement.

●After negative aspiration, inject 1 mL of LA, visualizing ventral (anterior) displacement of the pleura as confirmation of correct placement of the needle tip in the TPVS.

●After negative aspiration, inject 3 to 5 mL of LA. If performing a high volume single-injection block, inject 20 mL LA in 5 mL increments, with gentle aspiration between injections.

Transducer in axial orientation — With axial transducer orientation, the block can be performed with in plane or out of plane needle insertion. For an in-plane approach with lateral to medial orientation, the target should be the lateral paravertebral space, to avoid neuraxial complications (image 3).

●Center the transducer over the line that was drawn over the tips of the TPs. Slide the probe caudad or cephalad until the transverse process and rib disappear so that the lamina and pleura are well visualized. The probe will usually be aligned with the intercostal space.

●Insert the needle as follows:

•In plane – Insert the needle from lateral to medial, 2 to 3 cm lateral to the edge of the transducer to allow for a shallow needle trajectory (movie 2).

•Out of plane – Insert the needle from caudal to cranial, 2 to 3 cm caudal to the midpoint of the edge of the transducer at a shallow angle. Tilt the ultrasound transducer beam caudally to improve needle visualization (movie 3).

●Advance the needle to the TPVS. A "pop," may be felt when the needle passes through the intercostal membrane.

●Inject 1 mL of LA, visualizing ventral (anterior) displacement of the pleura as confirmation of correct placement of the needle tip in the TPVS.

●After negative aspiration, inject 3 to 5 mL of LA. If performing a high volume single-injection block, inject 25 ml LA in 5 mL increments, with gentle aspiration between injections (movie 2).

Block with intermittent ultrasound — For this technique, ultrasound is used in real time to insert the needle tip at the inferior aspect of the TP, and again when injecting LA. The needle is advanced from the TP into the TPVS without ultrasound guidance. The authors use this technique for most single-injection blocks, and for all continuous blocks. We describe this technique briefly here (movie 4).

●Use a needle with centimeter markings.

●With the transducer in a sagittal orientation, insert the needle in plane in a caudal to cranial direction, placing the tip on the inferior aspect of the TP, as when performing an erector spinae plane (ESP) block. (See "Erector spinae plane block procedure guide".)

●Estimate the change in needle angulation and depth required to place the needle in the TPVS (figure 7).

●Set the transducer aside. Retract the needle to subcutaneous tissue, reinsert and walk off the TP at the estimated angle to the predetermined depth.

●Obtain ultrasound visualization again, including the needle tip. After negative aspiration, inject 1 to 2 mL of LA, visualizing anterior pleural displacement. Inject for the block as above. (See 'Performing the block' above.)

Block with preprocedure ultrasound — For this technique, ultrasound is used to locate the TP (figure 8), and to determine the depth to the TP, and pleura.

The block itself is performed without ultrasound, as shown in a figure (figure 9).

LANDMARK-BASED BLOCK

Landmarks — The spinous processes (SP) are surface landmarks used for performing landmark-based block, and are palpable in most patients of normal size. The tip of the transverse process (TP) is approximately 2.5 cm lateral to the SP of the vertebra above, and the thoracic paravertebral space (TPVS) is approximately 1 to 1.5 cm deeper than the dorsal surface of the TP (figure 10).

Patient positioning and marking — For purely landmark-based block, the desired level is determined by counting spinous processes rather than ribs. Counting starts with the most prominent SP (spinous prominens) at the base of the neck, which is typically C7. Using surface landmarks, such as using the inferior border of the scapula as a surrogate marker for the T7 level, can result in a block that is two to three levels away from the intended site [14].

●Position the patient sitting, with the neck flexed and shoulders relaxed.

●Draw a horizontal line at the superior aspect of the desired SP, remembering that the tip of the SP is at the level of the TP of the vertebra below [15]. For example, the TP of T4 is lateral to the SP of T3.

●Draw a vertical line on the patient's skin, 2.5 cm lateral to the desired SP (ie, the SP immediately cephalad to the desired TP).

●The needle insertion site is at the intersection of the horizontal and vertical lines.

Performing the block — For single injections, we use the type of block needle that would be used for peripheral nerve blocks. We use a stimulating needle if we'll use nerve stimulation as an endpoint. If using loss of resistance as an endpoint, we use a Tuohy needle. We use a needle with centimeter markings to more easily determine depths.

The block is performed by initially placing the needle tip on the TP, and then stepping off the TP and redirecting the needle caudally into the paravertebral space.

Some use the lamina instead of the TP as the point of bony contact, and then advance the needle laterally off the lamina into the paravertebral space. We do not use such a paralaminar approach, as it targets the medial TPVS and may increase the potential for neuraxial injection. In addition, catheter placement using the paralaminar approach has resulted in placing the catheter tip too far anteriorly, near the vertebral body and sympathetic trunk [16].

●Infiltrate the skin and soft tissue at the injection site with 1% lidocaine, 1 to 3 mL, using a 25 gauge needle.

●Insert the needle at the marked site, perpendicular to the skin in all planes.

●Advance the needle until the tip contacts bone, aiming for the TP. The average depth of TP is approximately 2.5 cm from the skin, though in our experience the range is quite wide (1 to 8 cm depending on body habitus). The depth to the TP can be estimated quickly and safely by inserting the needle 1.5 cm lateral to the midline, to contact the lamina with the needle tip.

Note: The TPs in the upper thorax are wider and deeper than in the midthorax, and locating bone may require needle angle or insertion site adjustment. Start with a shallow needle depth.

●Once the needle contacts bone, note the depth. Withdraw the needle without removing it from the skin, and redirect it 15 degrees caudad, inserting the tip 1 cm deeper than the depth to the TP. Cephalad needle redirection off the TP can puncture the pleura (figure 11). Do not angle the needle in a medial (toward neuraxis) or lateral (toward pleura) direction.

●A pop may be felt as the needle passes through the superior costotransverse ligament (SCTL). However, loss of resistance during needle pass through SCTL is more subtle than during epidural placement and not as reliable.

●After negative aspiration, inject 3 to 5 mL of local anesthetic (LA). For single-injection high volume block, inject 20 mL in 5 mL increments, with gentle aspiration between injections. If there is resistance to injection advance the needle slightly forward until it is easy to inject, but no more than 1.5 cm past depth of TP.

LOCAL ANESTHETIC CHOICE AND DOSE — Local anesthetics (LAs) are chosen according to the goal of the block (surgical anesthesia or analgesia) and the desired duration of the effect of the block (table 2). LAs for peripheral nerve blocks and the use of adjuvant drugs are discussed in more detail separately. (See "Overview of peripheral nerve blocks", section on 'Drugs'.).

●Choice of LA – Usual options for TPVB are as follows:

•Surgical anesthesia only – 2% lidocaine or 1.5% mepivacaine or ropivacaine 0.5 to 0.75%, or bupivacaine 0.5%

•Postoperative analgesia – 0.25 to 0.5% bupivacaine or 0.5% ropivacaine. In our experience, cutaneous sensory block after bupivacaine or ropivacaine lasts on average six to eight hours. Analgesic duration may last longer [17].

•Surgical anesthesia and postoperative analgesia, either:

-For rapid onset – Equal volumes of a short-acting LA (2% lidocaine or 1.5% mepivacaine) plus a long-acting LA (0.5% bupivacaine or 1% ropivacaine). Note that mixing LAs results in onset and duration that are both intermediate between the two agents [18].

-When rapid onset is not required (ie, 30 minutes for onset is acceptable) – 0.25 to 0.5% bupivacaine or 0.5% ropivacaine.

For bilateral block that requires coverage for multiple dermatomes (eg bilateral breast surgery), a lower concentration of LA may be required to stay within recommended maximum doses.

●Volume of LA

•For multiple single level injections – 3 to 5 mL per injection

•For large volume single-injection block, or two ipsilateral large volume injections – 15 to 20 mL total spreads on average three to five continuous levels [11,12,19]

●Block onset – Dermatomal distribution of the block should be mapped 20 to 30 minutes after injection to ensure that the block is adequate.

CONTINUOUS BLOCK — Continuous paravertebral block catheters may be used to provide prolonged postoperative analgesia beyond the duration provided by a single-injection block.

●Catheter placement technique – The technique for placing the catheter is similar to the technique used for single-injection block.

•Use a 19 or 20 gauge Tuohy needle for this technique; insert the catheter through the needle.

•Use a soft-tipped catheter to reduce the risk of puncturing the pleura or blood vessels, or having the catheter migrate outside the paravertebral space.

•Place the needle using one of the techniques described above. We place the needle with the bevel facing slightly medially to increase the likelihood of vertical spread to adjacent paravertebral spaces.

•While stabilizing the needle, insert the catheter through the needle and advance approximately 4 to 5 cm beyond the needle tip. Shorter catheter advancement can be used to minimize the risk of having the catheter tip migrate out of the TPVS, but there may be a higher risk of catheter dislodgement.

•After negative aspiration, inject a test dose (3 mL 1.5% lidocaine with epinephrine). Visualize anterior displacement of the pleura, and observe the patient for signs of intrathecal or intravascular injection.

•Remove the needle without withdrawing the catheter.

•Inject 20 mL of local anesthetic (LA) in 5 mL increments, with gentle aspiration between injections.

•Check the level of the sensory block (testing both sides to rule out epidural spread) and observe the patient for hemodynamic changes suggestive of neuraxial spread.

For continuous TPVB, if extensive coverage is required (more than four dermatomes) we recommend placing catheters no further than four paravertebral spaces apart. If we place more than one catheter, we bolus with 15 ml LA per catheter.

●Positioning, draping, and securing the catheter – Important technical aspects to consider when placing a catheter for continuous TPVB include:

•Placing a perineural catheter takes longer than a single-injection block and sufficient time must be allocated prior to surgery (budget 15 to 30 minutes).

•Use meticulous sterile technique when placing a catheter, with a large sterile field, a sterile ultrasound sheath, and with personnel wearing a hat, mask, sterile gloves and preferably a gown.

•Secure the catheter well to prevent leakage and/or dislodgement.

-We apply surgical glue at the catheter insertion site, primarily to prevent leakage, which can disrupt dressing integrity, lead to catheter dislodgement, and to help fix the catheter in place.

-We apply liquid adhesive (eg, tincture of benzoin or similar) around the insertion site.

-We use a catheter fixation device. Avoid placing such devices within the surgical field or where a tourniquet will be placed.

•Coil the excess catheter and fix it to the skin with a transparent adhesive dressing. Ensure the insertion site, fixating device, and catheter are well covered and visible.

●Infusion drug dose – After injecting a bolus of LA as described above, we start high-volume infusions using programmed intermittent bolus to improve LA distribution. We usually use programmed boluses of 15 to 20 mL of ropivacaine 0.2% (or other long-acting LAs in equipotent doses) every three to four hours, with patient-controlled boluses of 5 to 10 mL per hour.

SIDE EFFECTS, COMPLICATIONS, AND CONTRAINDICATIONS

●Side effects and complications – The most common side effects of TPVB may be mild hypotension (as a result of sympathetic block) [20], mild signs and symptoms of local anesthetic systemic toxicity (LAST), and musculoskeletal pain after block resolution, though the incidence of such side effects is not known.

Complications of TPVB occur in 2.6 to 5 percent of blocks when a blind, anatomic approach is used [21-23]. Reported complications include Horner syndrome, pneumothorax, vascular puncture, and epidural or intrathecal spread. The risk of pneumothorax increases with multiple levels of injection [24], and has been reported to occur in 0.8 percent of cases using a landmark technique with nerve stimulation [25]. Ultrasound guidance may reduce the incidence of accidental pleural puncture [26]. This was suggested by a single institution retrospective review of over 1400 TPVB using ultrasound to guide a single axial, in-plane injection, which reported no cases of pleural puncture or symptomatic pneumothorax [27].

The incidence of LAST is likely similar to intercostal nerve blocks, since the TPVS communicates with the highly vascular intercostal space, and relatively high volumes of local anesthetic (LA) may be used for TPVB. However, the reported incidence of LAST after TPVB is low. In a retrospective review of over 300 continuous TPVBs placed for rib fractures, there were two cases of possible LAST, one of which was felt to be unlikely [28]. In the single institution review described above, there were two cases of possible LAST in over 1400 blocks [27]. (See "Local anesthetic systemic toxicity", section on 'Block site'.)

The reported incidence of failed block varies widely, from zero to as high as 19 percent [11,25,29], and may depend on the block technique.

Unintentional bilateral block may occur, likely as a result of neuraxial spread when injection is medially within the TPVS.

●Contraindications – Relative contraindications specific to TPVB include a history of thoracic surgery or infection that may have caused scarring of the TPVS, which may result in less reliable spread of LA or higher risk of pleural puncture.

Most clinicians consider anticoagulation a relative contraindication to TPVB, due to the deep and noncompressible location of the paravertebral space. Deep blocks in non-compressible locations may increase the risk of hematoma. As a result, guidelines from the American Society of Regional Anesthesia and Pain Medicine recommend following the same guidelines for anticoagulation for TPVB that would be used for neuraxial anesthesia [30] These guidelines are discussed in detail separately. (See "Neuraxial anesthesia/analgesia techniques in the patient receiving anticoagulant or antiplatelet medication".)

Other contraindications are similar to those that apply to any regional anesthesia technique (eg, allergy to LAs, patient refusal, infection at the block site). (See "Overview of peripheral nerve blocks", section on 'Contraindications'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Local and regional anesthesia".)

SUMMARY AND RECOMMENDATIONS

●Anatomy

•Each thoracic paravertebral space (TPVS) is a wedge shaped compartment that lies alongside vertebral column, in between the heads and necks of the ribs above and below. The anterior boundary of the TPVS is the pleura. The posterior border consists of ribs, transverse processes, superior costotransverse ligament (SCTL), and internal intercostal membrane (IIM) (figure 1). (See 'Anatomy' above.)

•Performing the block medially may reduce the risk of pleural puncture, but may increase the risk of dural puncture, intrathecal injection, or excessive epidural spread (figure 3). (See 'Clinical anatomic correlations' above.)

●Choice of block technique – We suggest using ultrasound guidance for thoracic paravertebral block (TPVB) rather than a landmark-based technique, to improve block success and reduce complications. (See 'Choosing a block technique' above.)

●Ultrasound-guided block – Block technique is described briefly here, and in more detail above. For ultrasound-guided block, the pleura must be visualized to avoid puncture and guide needle placement. Anterior pleural displacement is a key sign of desired spread of local anesthetic (LA) . (See 'Ultrasound-guided or assisted blocks' above.)

•Continuous real time ultrasound – The block is performed with continuous ultrasound visualization, with the transducer centered over the tips of transverse processes (movie 2). (See 'Block using continuous real time ultrasound' above.)

•Intermittent ultrasound – Ultrasound is used to insert the needle tip at the inferior edge of the tip of the transverse process (TP) and to estimate the angle and depth required to place the tip into the TPVS. The needle is then advanced without ultrasound, after which ultrasound is used again to confirm correct needle tip placement and injection of LA (movie 4). (See 'Block with intermittent ultrasound' above.)

●Landmark-based block – The block is performed by initially placing the needle tip on the TP, and then stepping off the TP and redirecting the needle caudally into the paravertebral space. (See 'Landmark-based block' above.)

●LA choice and dose (See 'Local anesthetic choice and dose' above.)

•For multiple single-injection blocks, 3 to 5 mL of LA are typically used per injection. For large volume single-injection block, 15 to 20 mL are used.

•LA are chosen according to the goal for the block and desired duration (table 2). For bilateral block that requires coverage for multiple dermatomes (eg bilateral breast surgery), a lower concentration of LA may be required to stay within recommended maximum doses

●Continuous block – Continuous block is performed as for single-injection block, using a Tuohy needle with a catheter inserted through the needle into the TPVS. (See 'Continuous block' above.)

●Side effects and complications (See 'Side effects, complications, and contraindications' above.)

•TPVB may cause hypotension (due to sympathetic block), mild local anesthetic systemic toxicity (LAST), and musculoskeletal pain after block resolution.

•Reported complications include pneumothorax, Horner syndrome, neuraxial spread, and LAST.

•Most clinicians consider anticoagulation a relative contraindication to TPVB, due to the deep and noncompressible location of the TPVS.