Costoclavicular brachial plexus block for shoulder surgery: a narrative review

Article information

Korean J Anesthesiol. 2025;78(6):513-523

Publication date (electronic) : 2025 August 4

doi : https://doi.org/10.4097/kja.25422

Samita Pirotesak^1,2, Nazanin Fallah¹, Reef Alruqaie¹, Karoll Rodelo¹, Juan Francisco Asenjo¹, Julián Aliste^,1,3

¹Department of Anesthesiology, Montreal General Hospital, McGill University Health Center, McGill University, Montreal, QC, Canada

²Department of Anesthesiology, Siriraj Hospital, Mahidol University Faculty of Medicine, Bangkok, Thailand

³Department of Anesthesiology and Perioperative Medicine, University of Chile School of Medicine, Santiago, Chile

Corresponding author: Julián Aliste, M.D. Department of Anesthesiology, Montreal General Hospital, McGill University Health Center, McGill University, Office D10-165.2, 1650 Cedar Avenue, Montreal, Quebec H3G 1A4, Canada Tel: +1-514-934-1934 ext. 43795 Fax: +1-514-934-8249 Email: julian.aliste@mcgill.ca

Received 2025 May 20; Revised 2025 July 11; Accepted 2025 July 21.

Abstract

Introduction

The costoclavicular block (CCB), which was first described by Karmakar et al. [1] in 2015, represents a proximal approach for blocking the brachial plexus in the infraclavicular fossa. This technique targets the plexus cords where they cluster laterally to the incipient section of the axillary artery between the clavicle and the second rib. The CCB was initially expected to offer advantages such as targeting the plexus at a more superficial location, allowing better ultrasonographic visualization, involving easier needling, having reduced anesthetic volume requirements, and demonstrating a faster block onset.

In 2017, the same research group detailed the scanning method for and the dynamics of CCB, reporting a median onset time of 5 min for sensory and motor blockade of the distal nerves (musculocutaneous, radial, median, and ulnar) when using a single 20-ml injection of 0.5% ropivacaine [2]. However, subsequent studies have questioned whether the CCB offers meaningful advantages over lateral paracoracoid approaches, with reports of conflicting results regarding whether CCB use allowed a reduced anesthetic volume [3–7]. Therefore, although the CCB remains a viable alternative to lateral paracoracoid approaches, it does not necessarily surpass conventional lateral infraclavicular blocks for routine indications.

Recently, diaphragm-sparing anesthesia for shoulder surgery has resulted in renewed interest in the costoclavicular space (CCS). Some studies have suggested that injecting a local anesthetic at this level may provide adequate analgesia while preserving the ipsilateral diaphragm function [8–11]. In this review, we examine the evidence on the CCB available to date, focusing particularly on its role as a diaphragm-sparing technique for shoulder anesthesia.

Anatomy

Anatomy relevant to the costoclavicular block

The CCS is located between the posterior surface of the middle third of the clavicle and the anterior chest wall [1]. Within this space, the cords of the brachial plexus run parallel and lateral to the axillary vessels, passing between the clavicular head of the pectoralis major and the subclavius muscles anteriorly, and the serratus anterior muscle, overlying the chest wall, posteriorly (Fig. 1) [18].

Fig. 1.

Costoclavicular space (CCS). (A) Schematic representation of the anatomical structures in the CCS. The connective tissue septa separate the lateral cord from the posterior and medial cords of the brachial plexus. (B) Standard ultrasound image of the CCS used for the costoclavicular block. Most of the structures shown in A can also be identified in B. AA: axillary artery, AV: axillary vein, IC: intercostal muscles, L: lateral cord, M: medial cord, P: posterior cord, PMj: pectoral major muscle, SA: serratus anterior muscle, Sc: subclavius muscle.

The three cords each maintain a consistent orientation. The lateral cord is the most superficially located. The medial cord lies posterior to the lateral cord, but remains medial to the posterior cord (Fig. 1) [18].

A paraneural sheath of connective tissue encases the brachial plexus cords, separating them from the epimysium of the surrounding muscles [18,19]. Within this sheath, a connective tissue septum further compartmentalizes the plexus, isolating the lateral cord from the medial and posterior cords (Fig. 1A) [19].

Immediately caudal to the CCS, the cephalic vein merges with the axillary vein, while slightly more distally, the thoracoacromial artery arises from the second portion of the axillary artery (Fig. 2) [18]. These vascular landmarks require careful consideration when selecting the approach for the needle.

Fig. 2.

Anatomical-technical aspects of costoclavicular block (CCB). The brachial plexus runs parallel and lateral to the axillary vessels, coursing between the clavicle and the chest wall. Caudad to the CCB site, the cephalic vein merges into the axillary vein, while slightly more distal, the thoracoacromial artery arises from the second portion of the axillary artery. A lateral-to-medial CCB with 20 ml of local anesthetic reliably blocks the brachial plexus cords and their branches, and via cranial spread of the anesthetic, it also targets the supraclavicular components of the brachial plexus, including the suprascapular nerve, while sparing the phrenic nerve.

Local injection of anesthetic into the CCS consistently anesthetizes the distal brachial plexus nerves, which is essential for distal extremity procedures. However, this space is also the point where key branches emerge: the axillary and subscapular nerves from the posterior cord, and the lateral pectoral nerve from the lateral cord [20]. Cadaveric studies have demonstrated that dye injections in this space effectively stain the cords and their branches, but also revealed cranial spread, with the dye reaching supraclavicular brachial plexus components [21]. Importantly, this cephalad spread may spare the PN and C5–C6 roots, potentially influencing their diaphragm-sparing profiles (Fig. 2) [21].

Costoclavicular block dynamics

Evidence from the literature indicates that the mechanism of action of the CCB involves the blockade of infraclavicular brachial plexus structures, along with that of some supraclavicular plexus components, in a volume-dependent manner. Blockade of the brachial plexus cords provides adequate anesthesia for procedures distal to the mid-humerus as well as analgesia for the proximal arm.

The minimal effective volume required for 90% success in achieving surgical anesthesia of the distal upper-extremity with a paracoracoid infraclavicular block may be up to 35 ml [22], while lower effective volumes have also been reported [23]. For the CCB, the minimal effective volume for achieving anesthesia in 90% of surgeries distal to the elbow ranges between 21 and 34 ml, depending on the criteria used to define success and the local anesthetic employed [4,6].

The volume required to block the brachial plexus effectively while sparing the PN when using paracoracoid or costoclavicular approaches has not been established in dose-finding studies to date. Studies using 35 ml of local anesthetic have reported a 9% incidence of hemidiaphragmatic paralysis (HDP) for both types of blocks [3], whereas studies using a volume of 20 ml for the CCB have shown a phrenic palsy rate between 0% and 7.5% [8,10,24]. However, 20-ml injectates delivered into the CCS have demonstrated complete SSN blockade in between 60% [8] and 87% [10] of cases at 30 min, even allowing successful shoulder surgery without the use of general anesthesia [10].

In a study of the retroclavicular approach to the infraclavicular fossa, a Bayesian analysis determined that the optimal volume of 0.5% ropivacaine for achieving adequate analgesia during shoulder surgery, with minimal PN risk (1%), was 25 ml [25]. Interestingly, within the 15–40-ml range, the study reported complete SSN blockade in 69%–96% of cases (with 96%–98% cases demonstrating partial-to-complete SSN blockade), complete subscapular nerve blockade in 98% of cases, and complete lateral pectoral nerve blockade in 96% of cases [25].

Certain aspects of the CCB technique remain unclear. Cadaveric studies have indicated that the plexus cords are separated by a septum at the level of the CCS (Fig. 1A) [19]; however, the spread of injectates within each of these compartments is yet to be fully characterized. Moreover, the ideal volume distribution across these compartments remains unclear.

A 2020 study by Layera et al. [26] found that dividing 34 ml of anesthetic into two injections, one delivered at the center of the cord and the other delivered between the medial cord and axillary artery, accelerated the onset of anesthesia for distal surgery. The double-injection approach resulted in more cases of hoarseness and Horner’s syndrome. However, the study did not systematically identify the septum before injection, leaving uncertainty regarding the actual proportion of injectates deposited in each compartment.

The CCB reliably blocks the brachial plexus cords and provides adequate surgical anesthesia for distal upper limb procedures. While dividing the local anesthetic into two injections may accelerate anesthesia onset, the optimal volume per injection site has not yet been determined. Injectates of 20 ml have provided effective SSN blockade, often enabling shoulder surgery without the use of general anesthesia; however, the risk of PN paralysis remains. Future research should clarify the cranial spread of injectates on each side of the septum, its dependence on the injection volume, and an optimal strategy for ensuring appropriate volume distribution.

When approaches other than the interscalene block (ISB) are used to achieve surgical anesthesia for shoulder procedures, block success should be assessed preoperatively. A scale has recently been proposed to assess sensory blockade in the territories of the supraclavicular and axillary nerves, along with motor function of the axillary, suprascapular, lateral pectoral, and subscapular nerves [27] and some studies have already validated its effectiveness [25,28].

Technical aspects of the costoclavicular block relevant to shoulder surgery

The technical aspects of the CCB have been described well in the initial reports [1,2,18]. Its dynamics vary depending on the injection volume and site used [8,21,24,26]. However, planning a CCB for shoulder surgery requires additional considerations.

The benefits of continuous vs. single-dose blocks with adjuvants for shoulder surgery remain debated [29,30]. Despite this uncertainty, continuous blocks are often the method of choice for highly painful surgeries; however, evidence supporting the use of continuous CCB for both distal and proximal upper limb procedures remains scarce [31].

Shoulder surgery poses challenges to standard lateral-to-medial catheter placement, both preoperatively and postoperatively. A medial-to-lateral approach has been described for administering single-dose CCB in distal procedures; however, this approach may result in a higher incidence of vascular trauma [32]. In clinical settings in which other brachial plexus techniques are unsuitable and where prolonged analgesia is required, medial-to-lateral continuous CCB remains a viable option. However, as with the lateral-to-medial insertions, the catheter tip may migrate into the supraclavicular fossa [31], and the impact of this positioning on the PN remains unknown.

Regardless of the CCB injection volume used, this approach spares the cervical plexus. Therefore, when surgical anesthesia is required, an intermediate or superficial cervical plexus block must be administered. Alternatively, a more distal supraclavicular nerve block can be used.

While the CCB technique was originally described with the arm at 90º abduction and in external rotation, this position is often painful for patients undergoing shoulder surgery. In addition, achieving an optimal imaging view of the cords at this level is not always straightforward, and frequently necessitates dynamic adjustments involving different degrees of shoulder abduction and rotation. The shoulder rotation component displaces the humeral head, creating more space for lateral-to-medial needling. Alternatively, this effect can be achieved by placing a cushion between the scapulae.

Pharmacology of costoclavicular block

CCBs have been performed using long-acting local anesthetics, such as bupivacaine, levobupivacaine, and ropivacaine [1,2,6–8, 10,11,24], as well as with mixtures of short-acting and long-acting local anesthetics at varying concentrations [3,9,26,32]. Although some reports have described the use of adrenalized mixtures and adjuvants to prolong blockade effects, no trials have systematically evaluated their impact on the analgesic efficacy or diaphragm-sparing effects in shoulder surgery.

For distal extremity surgery, the reported onset times vary, based on local anesthetic selection. While ropivacaine 0.5% may take 20–45 min for anesthesia onset [6], adrenalized mixtures of bupivacaine 0.5% and lidocaine 2% take 10–22 min [3].

The only trial assessing shoulder surgical anesthesia with a CCB demonstrate sensory-motor blockade success within 25 min in 91% of cases, using 20 ml of 0.5% ropivacaine [10]. Additionally, a separate trial focusing on postoperative analgesia showed a 90% success rate at 30 min when using 20 ml of 0.5% levobupivacaine [8]. Another study reported a sensory block lasting 544 min and a motor block lasting 441 min when using 0.5% ropivacaine without perineural or intravenous adjuncts [10].

Some evidence of the success rates and PN compromise occurring with commonly used volumes in CCB for shoulder surgery is available, although only one randomized controlled trial (RCT) has systematically addressed this topic. Jo et al. [33] compared the use of 20 ml of 0.75% ropivacaine with that of 40 ml of 0.375% ropivacaine for CCBs in shoulder arthroscopy. Complete analgesia rates were similar: 23.3% (7/30) in the 20-ml group vs. 33.3% in the 40-ml group at 1 h postoperatively. The pain scores were also comparable: 3 (1–5) with 20 ml and 2 (0–4) with 40 ml. Notably, PN involvement differed, with no cases of paralysis and 16% of paresis in the 20-ml group, while the 40-ml group reported 13% paralysis cases and 33% paresis cases. Although the study did not report block duration or long-term pain levels, the time to first rescue analgesia was 6.9 h in the 20-ml group vs. 5.5 h in the 40-ml group (P = 0.6).

The only study comparing different concentrations of local anesthetics for CCBs was conducted in the setting of forearm and hand surgery [34]. While the sensorimotor onset times did not differ significantly between 0.5% and 0.375% ropivacaine, the lower concentration resulted in a significantly shorter sensory block duration, with 0.375% ropivacaine lasting 455 min and 0.5% ropivacaine lasting 610 min.

Clinical validation of costoclavicular block in shoulder surgery

The CCB has been studied in various clinical scenarios, including both adult and pediatric patients, although most investigations have focused on forearm and hand surgery. In this section, we discuss the results of RCTs comparing the CCB to sham treatment, no block, or other blocks, specifically in shoulder surgeries.

A systematic approach was used to ensure inclusion of reliable trials. Prospective trial registration was verified, and major discrepancies between the registered and reported protocols were assessed. The bibliographies of all retained reports were manually searched for additional RCTs.

The final literature search in MEDLINE, EMBASE, and Google Scholar databases was conducted by two authors (SP and JA), covering the period from database inception to the last week of April 2025.

By using the search terms “costoclavicular block” and “shoulder surgery,” five RCTs comparing the CCB to other blocks were identified. Of these, four studies were retained for analysis (Table 1). Trials were excluded because of multiple retrospective protocol modifications and inconsistencies [35].

Table 1.

Randomized Trials Comparing the Use of Costoclavicular Block with Other Nerve Blocks in Shoulder Surgery

The Cochrane Database Risk-of-Bias Tool was used to assess the risk-of-bias in the selected RCTs. The domains evaluated included adequacy of sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, and other sources of bias (i.e., study design issues, early trial termination, and baseline imbalance in study groups). Risk-of-bias in these domains were categorized as “yes” (i.e., low risk-of-bias; green in Fig. 3) “no” (i.e., high risk-of-bias; red in Fig. 3) and “unclear” (i.e., unknown risk-of-bias; yellow in Fig. 3).

Fig. 3.

Risk-of-bias summary of randomized controlled trials pertaining to the use of the costoclavicular brachial plexus block in shoulder surgeries. = Low risk-of-bias, = Unclear risk-of-bias. = High risk-of-bias.

Among the four retained RCTs, only one studied CCBs under surgical anesthesia [10], whereas the other three evaluated CCBs in combination with general anesthesia [8,11,24].

In an overview of trial designs, Aliste et al. [8] compared the CCB to the ISB in 44 patients undergoing arthroscopic shoulder procedures. Patients received 20 ml of 0.5% levobupivacaine with epinephrine (5 μg/ml) along with an intermediate cervical plexus block (5 ml), intravenous dexamethasone, and standardized general anesthesia with postoperative analgesia. Jo et al. [24] compared the CCB to the superior trunk block (STB) in 70 patients undergoing arthroscopic rotator cuff repair, using 20 ml of 0.5% ropivacaine, which was divided into two injections for both techniques. In the CCB group, 10–15 ml anesthetic was deposited between the posterior and medial cords, followed by 5–10 ml deposited around the lateral cord. No cervical plexus blocks or adjuncts were administered, and anesthesia and postoperative analgesia were standardized in both groups. Lee et al. [11] compared the CCB to the ISB in 70 patients undergoing arthroscopic rotator cuff repair under standardized general anesthesia with postoperative analgesia, using 20 ml of 0.2% ropivacaine for the CCB and 10 ml of 0.2% ropivacaine for the ISB. Finally, Luo et al. [10] compared the CCB with the ISB in 212 patients undergoing elective arthroscopic shoulder surgery with propofol sedation, using 20 ml of 0.5% ropivacaine. Additionally, both groups received 3 ml of 0.5% ropivacaine in an ipsilateral supraclavicular nerve block to anesthetize the cutaneous area overlying the shoulder capsule.

As primary outcomes, Aliste et al. [8] and Jo et al. [24] assessed pain at 30 and 60 min after surgery, respectively. Lee et al. [11] focused on the incidence of HDP and Luo et al. [10] studied the success rate of SSN blockade.

In the studies of Aliste et al. [8], Luo et al. [10], and Lee et al. [11] postoperative analgesia and opioid requirements did not differ between groups, whereas in the study by Jo et al. [24], significantly lower pain levels were found at 1-h postoperatively in the STB group. However, Jo et al. reported that the pain intensity remained mild in both groups, and opioid consumption was similar in both groups in the first 48 h.

Although Aliste et al. [8] and Luo et al. [10] reported the absence of complete paralysis, Jo et al. [24] and Lee et al. [11] described incidence rates of complete paralysis of 9.7% and 5.9%, respectively. While Aliste et al. did not report partial paralysis rates, the other trials reported a partial HDP incidence of up to 16% [10,11,24].

The study by Luo et al. [10] deserves particular attention, because, in the context of patients with severe respiratory pathology requiring shoulder surgery, tracheal intubation and narcotic excess should be avoided. Their study proved that, in 106 patients, shoulder arthroscopy could be performed using a 20-ml CCB plus a 3-ml supraclavicular nerve block supplemented with propofol sedation, with a similar level of patient satisfaction as achieved with ISB.

In addition to its effective diaphragm-sparing properties, equivalent postoperative analgesia, and suitability for surgical anesthesia, the CCB offers additional noteworthy qualities. The CCB procedure duration is comparable to that of the ISB [8,10], with only a modest delay in onset [8]. Intraoperative analgesia is similar to that of the ISB under general anesthesia [8,11] or propofol sedation [10]. Finally, postoperative opioid usage in the case of the CCB does not differ significantly from that in other blocks, with comparable consumption over 24 [8,11,24] and 48 hours [11]. When reported, cumulative opioid rescue requirements were minimal [8], which underscores the opioid-sparing efficacy of the CCB.

Alternative diaphragm-sparing blocks

The mechanism of HDP following ISB is hypothesized to involve the rostral spread of the local anesthetic to the C3–C5 nerve roots or the direct migration of injectates to the PN, which courses along the anterior surface of the anterior scalene muscle [20]. Phrenic block is a well-known complication of the ISB, and leads to a 20%–30% reduction in respiratory function parameters [13].

Modification of the ISB technique, such as extrafascial injections, targeting distant sites within the plexus sheath, or using pharmacological alterations (different local anesthetics, reduced anesthetic concentrations, and lower injection volumes), has not reduced the high incidence of PN paralysis and the secondary risk of respiratory complications [20]. Consequently, other techniques have gained popularity in shoulder surgeries, to mitigate PN involvement. Among these, the anterior SSN and STB are noteworthy [15]. However, these techniques require use of lower-than-usual injectate volumes to prevent the local anesthetic from spreading to the PN [15]. A cadaveric study suggested that sparing the PN when using an anterior SSN block requires administration of an injectate volume of 4.2 ml [17]. Similarly, a clinical trial comparing 5 and 15 ml of injectate for the STB demonstrated incidences of mixed complete and partial phrenic paralysis of 14.3% in the 5-ml and 65.7% in the 15-ml group, although the postoperative analgesia was equivalent between the groups [36]. Regrettably, no clinical trials have compared these low-volume diaphragm-sparing techniques with standard blocks, and no studies to date have determined whether these low-volume techniques provide sufficient surgical anesthesia.

Studies using standard supraclavicular brachial plexus block have reported HDP rates of >50%. However, in 2009, a study using a posterolateral injection of 20 ml of local anesthetic to the plexus was described as being diaphragm-sparing in distal upper-extremity surgery, with complete absence of HDP [37].

In 2018, this modified supraclavicular block technique was tested as an analgesic alternative to ISB in arthroscopic shoulder surgery. The results showed similar analgesic efficacy, but found a 9% incidence of HDP in the supraclavicular group [38]. While this was initially considered to be insufficiently diaphragm-sparing, recent findings have indicated that the CCB produces HDP rates ranging from 0% to 9.7%, suggesting that reconsideration may be warranted.

One key advantage of supraclavicular block is the possibility of catheter placement both preoperatively and postoperatively. However, albeit diaphragm-sparing as a “single-shot” block, it can still result in PN involvement when provided as a continuous infusion, a topic that remains largely unexplored in shoulder surgery.

Future research

Use of ultrasound-guided diaphragm-sparing regional anesthesia has gained significant popularity in shoulder surgery over the past decade. However, several critical research questions remain unanswered, highlighting the need for further studies in this field.

Although diaphragm-sparing techniques, such as the CCB and the STB, can provide adequate analgesia, they lack the versatility and reliability of ISB, which remains the gold standard in shoulder surgery. The ISB is preferred because of its precise targeting of the required neural structures and its distance from the surgical site, allowing the use of single-dose or continuous blocks, both preoperatively and postoperatively.

Most patients who receive an ISB develop complete or partial HDP, but tolerate the decrease in ventilatory function well, even in ambulatory cases. This raises a crucial research question: how can patients who require alternatives to ISB be accurately identified? While extreme conditions, such as severe respiratory pathology or extreme obesity, make diaphragm-sparing approaches necessary, the risk of ventilatory insufficiency remains unclear in some patients. Thus, choosing alternative blocks may compromise regional analgesia and can increase reliance on narcotics, with potential side effects.

The literature increasingly supports the use of ultrasound as a valuable perioperative tool for assessing diaphragmatic function [39]. The methodologies that have been propose include assessing diaphragmatic excursion in M-mode, intercostal diaphragmatic excursion, and diaphragmatic thickening at the apposition zone.

Although the concepts of partial and complete paralysis are widely accepted, their definition remain arbitrary, and no study to date has determined the impact of varying degrees of paralysis on respiratory function. Furthermore, the correlation of preoperative baseline ipsilateral and contralateral diaphragm functions with HDP tolerance require investigation.

Regarding the CCB, future research should explore the optimal local anesthetic volumes for double-injection techniques, the ideal injection volume in each compartment for shoulder procedures, and the variations in proximal spread based on the injection site and volume.

Additionally, further investigations should address alternative strategies for continuous analgesia, focusing on block efficacy and HDP prevention, and optimization of SSN blockade without incurring PN spread, which remains a key challenge for ensuring successful analgesia/anesthesia with CCB. Combining low-volume CCBs and low-volume anterior SSN blocks represents a potential research avenue, as do comparisons of the CCB with the retroclavicular approach, incorporating its recently defined ideal volume [25].

Conclusions

In patients with respiratory pathologies requiring shoulder surgery, minimizing HDP is crucial when performing regional blocks for both anesthesia and analgesia. In the search for a more efficient alternative to the infraclavicular paracoracoid approach for distal upper-extremity procedures, a more proximal plexus approach within the CCS has been proposed. Although the CCB does not enhance the efficiency of infraclavicular block, it exhibits favorable dynamics for shoulder blockade.

The nerves originating from the brachial plexus cords that innervate the shoulder joint are effectively covered by the CCB. Additionally, the CCB demonstrates reasonable supraclavicular spread, allowing suprascapular nerve blockade. Therefore, the CCB provides adequate postoperative and intraoperative analgesia, with a lower incidence of phrenic paralysis than observed with alternative blocks.

From a practical standpoint, and based on the available evidence, the CCB technique for shoulder procedures involves a single-shot block using a lateral-to-medial, ultrasound-guided needle approach, with a total local anesthetic volume of 20 ml. Owing to the presence of septae that separate the brachial plexus cords from each other, this volume may need to be divided into two separate injections to ensure adequate spread. Shoulder pathology may limit abduction and rotation, resulting in suboptimal visualization of the standard anatomical landmarks. The most suitable indication for this technique may be patients with a known respiratory risk who are undergoing shoulder surgery, where most of the severe pain can be reasonably controlled during the first 24 hours. Finally, when surgical anesthesia is the goal, a superficial or intermediate cervical plexus block should be included and a thorough preoperative assessment of block efficacy should be conducted.

Ongoing studies should continue to validate these encouraging findings, while exploring technical modifications that could optimize the success rate of anesthesia and analgesia during shoulder surgery and could further minimize PN involvement.

Additionally, future research should focus on identifying effective and safe continuous regional analgesia using CCB, while maintaining diaphragm preservation. Investigations into alternative injection techniques, volume adjustments, and combined approaches may further improve shoulder surgery outcomes, while reducing respiratory complications.

Article information Continued