Ultrasound-guided intermediate versus deep cervical plexus block for carotid endarterectomy: a randomized controlled study

Abstract

Background

Regional anesthesia is commonly preferred technique for carotid endarterectomy (CEA) as it allows neuromotor evaluation and provides more stable hemodynamics. Cervical plexus blocks (CPBs) including superficial, intermediate and deep, are used alone or in combination in carotid surgery. Ultrasound guidance provides better visualization with increased safety for CPBs. The aim of this study is to compare the anesthetic efficacy of deep and intermediate CPBs by assessing the number of patients requiring additional local anesthetic during CEA.

Methods

In this randomized, single-blind study, we enrolled 90 patients who underwent CEA. Patients who received either intermediate or deep CPB were divided into two groups: Group of Intermediate (GI) and Group of Deep (GD). The number of patients requiring additional lidocaine infiltration, time spent for block performance, extra number of needle insertions and orientations, assessment of cutaneous sensory loss in four peripheral cervical plexus nerve territories (lesser occipital, supraclavicular, transverse cervical and great auricular), pain scores, amount of additional lidocaine and complications were recorded. Patient and surgeon satisfaction were also evaluated.

Results

Nineteen out of 40 patients (47.5%) in GI and 10 out of 41 patients (24.5%) in GD needed additional lidocaine infiltration intraoperatively (p = 0.03). Time spent for intermediate block was 208.5 (50–332) s and for deep was 320 (90–780) s (p < 0.001). The blocks were performed successfully with single needle orientation 85% of GI and in 68% of GD (p = 0.11). The number of extra needle insertions was significantly higher in GD compared to GI (1 [1–4] and 1 [1–2], respectively; p = 0.04). All patients achieved adequate cutaneous sensory loss for surgery. Pain scores and amount of additional lidocaine were comparable. Patient satisfaction was significantly higher in GI than in GD (p = 0.03).

Conclusions

In this study we found that significantly fewer patients required additional lidocaine in deep CPB than in intermediate CPB with low initiation LA volumes under USG guidance.

Trial registration

The study was registered at ClinicalTrials.gov (registration number NCT05353218) on 24/04/2022.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12871-025-03460-w.

Introduction

Carotid endarterectomy (CEA) has been successfully performed both under general and regional anesthesia [1], with considerable experience being established for the later in the recent decades [2–5]. Regional anesthesia can be preferred according to center’s experience, patient’s inclination and clinical status [3, 6], as it allows neuromotor evaluations during the procedure and provides more stable hemodynamics in an awake patient [2–5].

Deep and superficial cervical plexus blocks have been successfully used for carotid surgery, whereas intermediate cervical plexus block (CPB) is a relatively new technique [7]. The nomenclature has been previously discussed and later clarified by subsequent trials which made a net distinction through fascia layers [7–11]. Initially, Telford and Stoneham suggested the name of “intermediate CPB” for the block performed in the cadaveric study of Pandit as the site of injection was “just below investing fascia”; underlining that superficial block would be associated with subcutaneous injection [7]. In the earliest study of intermediate CPB by passing investing fascia, Ramachnadran et al. compared superficial and intermediate CPB [9]. Moreover, Scimia et al. clearly defined the nomenclature for CPB according to target tissue: local anesthetic (LA) injection to subcutaneous tissue for superficial; between investing and prevertebral fasciae for intermediate and sub-prevertebral fascia for deep [12]. Recent anatomical review remarked that the distinction was probably more difficult in the absence of ultrasound (USG) guidance in historical studies [13].

Moreover, recent studies highlight the complexity of neck anatomy with variations of multiple fascial layers and anastomoses between nerve bundles in the carotid triangle [13–15]. The classification of CPB as deep, intermediate and superficial is perplexing and at times, these terms have been used erroneously due to this complex anatomy and blind techniques adopted for the initiation of these blocks. Currently, with the use of USG guided CPB instead of blind techniques, LA is administered to different layers of the cervical fascia blocking cervical nerves at different levels [5, 13, 16]. Although similar efficiency of these blocks at various injection levels in the past have been reported with older techniques utilizing high volume of LAs, much is yet unknown regarding the anatomical and functional information for the optimal spread and possible permeation of fascia in the cervical tissue [17]. Sensory block achieved or extended spread of block should be studied in detail, as it is the major determinant of clinical efficacy [17–19].

In this study, we aimed to compare anesthetic efficacy and to assess loss of cutaneous pinprick sensation in USG guided deep and intermediate CPB in CEA. Our primary outcome was the number of patients who required additional lidocaine infiltration intraoperatively. Time spent for block performance, cutaneous pinprick sensation loss, intraoperative pain scores, amount of additional lidocaine, complications, satisfaction of the patients and the surgeon were also compared as secondary outcomes.

Methods

This prospective, randomized, single-blind study was initiated after local Clinical Research Ethics Committee approval (2021/923). The study was registered at ClinicalTrials.gov (registration number: NCT05353218) before the start of enrollment. This manuscript adhered to the Consolidated Standards of Reporting Trials (CONSORT) guidelines for randomized controlled studies.

Patients aged over 18 years who underwent CEA between June 2022 and April 2023 and who gave written informed consent to participate to the study were enrolled in this study. Patients who refused regional anesthesia, had local infection, known bleeding disorders and allergy of LAs were excluded from the study. Planned general anesthesia, previous ipsilateral surgery and radiotherapy in the block area were other exclusion criteria. Patients who were included to the study underwent training for cutaneous sensory loss assessment with pinprick test and pain evaluation with numeric rating score (NRS).

Patients were randomized by computer generated random numbers into two groups according to regional technique: Group of Intermediate CPB (GI) and Group of Deep CPB (GD). This randomization numbers in sealed opaque envelopes were only opened by the anesthesiologists who were going to perform the block. These anesthesiologists did not contribute to data collection. Staff anesthesiologist who managed the patient perioperatively, surgeon and observers were blinded to the regional technique until the end of analysis.

Cervical plexus blocks

All patients in the operating room were monitored with 5-lead electrocardiography and pulse oximetry following surgical site confirmation. Patients were sedated with iv 1 mg midazolam and iv 50 µcg fentanyl with nasal oxygen supplementation. An arterial cannula was placed into radial artery on the opposite side of surgery for monitoring invasive arterial blood pressure. Position was similar in all patients, supine and head rotated to contralateral side to perform blocks.

The cardiovascular anesthesiologists (ÖT or ZS) experienced in regional anesthesia, performed all blocks under USG guidance with a blunt peripheral block needle (50 mm, 22 gauge; Stimuplex^® A; B Braun, Melsungen, Germany). The time spent for performance of the block which was defined as the beginning from the insertion of USG probe and extending to the removal of needle from skin was recorded. Needle insertion was accepted as needle piercing the skin and extra number of needle insertions (the count of times the needle was withdrawn from the skin and reinserted more than once in GI and more than thrice in GD), and the number of patients requiring extra needle orientations (changes in needle orientation without complete withdrawal from the skin) were documented during the block application.

For intermediate CPBs, a high-frequency linear USG probe (5 to 13 MHz; GE Healthcare^®, Wauwatosa, Wis.) in transverse plane was placed on the posterior border of the sternocleidomastoid (SCM) muscle at C4 level. The needle was inserted in plane using anterior approach and advanced through the SCM muscle [10, 18]. After negative aspiration, the mixture of LA solution consisting of 10 mL 0.5% bupivacaine and 5 mL 2% lidocaine was injected below the investing fascia (superficial layer of the deep cervical fascia) beneath the SCM muscle (Fig. 1, Supplementary Material 1).

Fig. 1.

Image of intermediate CPB with USG guidance. Abbreviations:CA Carotid artery, CPB Cervical plexus block, IJV Internal jugular vein, LA Local anesthetic, SCM Sternocleidomastoid muscle, USG Ultrasound

For deep CPBs, a linear USG probe was placed in the transverse plane over the mastoid process and moved caudally along the posterior border of the SCM muscle. When transverse process of C2 was visualized in the USG window, the needle was introduced in plane using anterior approach towards to the hyperechoic shadow of the transverse process. After negative aspiration, 5 mL of the same LA solution was injected to a point 1–2 mm adjacent to the transverse process beneath the prevertebral fascia (deep layer of the deep cervical fascia) (Fig. 2, Supplementary Material 1). Practitioner shifted the probe caudally (along the posterior border of the SCM muscle) and applied the same procedure to C3 and C4 levels one by one injecting 5 mL LA to each of them [20].

Fig. 2.

Image of deep CPB with USG guidance. Abbreviations: CA Carotid artery, CPB Cervical plexus block, IJV Internal jugular vein, LA Local anesthetic, SCM sternocleidomastoid muscle, USG Ultrasound

Evaluation of cutaneous sensory loss

Cutaneous sensory loss was assessed at five-minute intervals up to the 20th minute using pinprick test in both groups. Sensory block assessment was performed in four areas according to the four peripheral cervical plexus nerve innervations (Supplementary Material 2). These areas were defined as:

1. - inferoposterior region of the ear, which is innervated by the lesser occipital nerve,

2. - superior region of the clavicle behind the SCM muscle, which is innervated by the supraclavicular nerve,

3. - region between the cricoid cartilage and the SCM muscle, which is innervated by the transverse cervical nerve,

4. - inferior region of the angle of the mandible which is innervated by the great auricular nerve.

Sensory block was assessed in three levels as follows: “no sensation (complete loss of sensation)’’, ‘‘feels touch (reduced sensation)’’ or ‘‘feels pain (no block)’. The first two answers (complete loss of sensation or reduced sensation) were accepted as adequate regional anesthesia for surgery at 20th minute, and surgery was started thereafter.

Intraoperative management

Sedation was achieved with 0.02 µg/kg/min remifentanil, and the level of sedation was evaluated using the Ramsay Sedation Scale. Intraoperative pain was measured using NRS with five-minute interval for the first 30 min of surgery and then at 15 min interval (NRS 0: no pain to NRS 10: the worst pain imaginable). If the patient reported NRS ≥ 4, an infiltration of 4 mL 1% lidocaine was administered by the surgeon into the surgical area. Additional lidocaine was permitted up to a maximum dose of 2 µg/kg. In cases where pain persisted despite reaching the maximum dose of lidocaine, the remifentanil infusion rate was planned to increase by 0.02 µg/kg/min. The number of patients who required additional lidocaine infiltration, amount of lidocaine required and the surgical areas (skin and subcutaneous tissues, carotid sheath, towards the mandible) where lidocaine was infiltrated, were noted.

Blood pressure was continuously monitored with invasive arterial measurements throughout the entire surgery. Intravenous vasoactive medications (esmolol, ephedrine, or norepinephrine) were administered to maintain systolic blood pressure between 140 mmHg and 180 mmHg. Hemodynamic instability was defined as either severe hypotension (systolic pressure < 100 mmHg) or severe hypertension (systolic pressure ≥ 180 mmHg), both of which required active management.

All carotid endarterectomies were performed by the same senior surgeon with patch angioplasty. The patient’s neurological status was monitored continuously throughout the procedure. Consciousness, orientation (to time, place, and person), and cooperation (ability to understand verbal commands and follow instructions) were assessed with verbal questions. Motor functions were evaluated with contralateral hand movements (squeezing a ball, lifting the arm). Any deterioration in cognitive status, speech, or motor ability prompted the immediate insertion of an intraluminal carotid shunt.

Postoperative follow-up

At the postoperative visit, patients and surgeons were asked to evaluate their satisfaction about the anesthesia method by using a 5-point Likert scale (1: very dissatisfied, 2: dissatisfied, 3: neutral, 4: satisfied, 5: very satisfied). Complications such as hematoma, hypoglossal nerve paralysis, hoarseness, facial paralysis, Horner’s syndrome, intravascular or intrathecal injection, need for general anesthesia conversion, if occurred, were recorded.

Statistical analysis

The sample size was calculated by a preliminary study in which 10 patients were included in each group. Five (50%) patients needed additional lidocaine in GI whereas two (20%) patients in GD. The estimated sample size was 36 patients per group with a significance level of 0.05 and a power of 0.8. Considering possible dropouts, a total of 90 patients were enrolled in the study.

The normality of data distribution was assessed by Kolmogorov–Simirnov test. Categorical data were analyzed with Chi-Square test and expressed as number (percentages). Quantitative data were presented as mean ± standard deviation, or median (minimum-maximum). Student t-test was used for normally distributed data and Mann–Whitney U test for non-normally distributed data. All statistical analysis was performed using IBM SPSS for Windows, version 21.0 (IBM Corp., Armonk, NY) and p < 0.05 was considered statistically significant.

Results

Ninety patients undergoing CEA between June 2022 and April 2023 were enrolled in the study. Nine patients were excluded from the study, as summarized in Fig. 3. Finally, a total of 81 patients were analyzed, including 40 patients in GI and 41 patients in GD.

Fig. 3.

Flowchart. Abbreviations: GD Group deep cervical plexus block, GI  Group intermediate cervical plexus block

Demographics and operative data were summarized in Table 1 and were similar between groups. All patients reached “complete loss of sensation” or “reduced sensation” as adequate anesthesia for surgery at 20th minute after the block performance and the surgery was started. Number of patients with adequate cutaneous sensory loss (adequate anesthesia: " complete loss of sensation” or “reduced sensation”) for surgery in all areas were similar between groups at all evaluation times till surgery. Detailed evaluation of dermatomal analgesia is depicted in Fig. 4.

Table 1.

Demographics and operative data

	GI (n = 40)	GD (n = 41)	p value
Age (years)	62.85 ± 11.14	66.82 ± 8.45	0.074
Gender
Female  Male	17 (42.5%) 23 (57.5%)	22 (53.6%) 19 (46.4%)	0.315
BMI (kg/m2)	33.99 ± 43.94	26.60 ± 2.83	0.286
ASA
ASA II  ASA III	17 (42.5%) 23 (57.5%)	19 (46.4%) 22 (53.6%)	0.82
Comorbidity
Hypertension  Diabetes mellitus  Coronary artery disease  Chronic obstructive lung disease	29 (72.5%) 13 (32.5%) 17 (42.5%) 0 (0%)	30 (73.1%) 15 (36.6%) 15 (36.6%) 2 (4.9%)	1 0.816 0.652 0.493
Anti-platelet therapy and anticoagulation
Acetyl salicylic acid  Acetyl salicylic acid with LMWH  Dual anti-platelet therapy  None	6 (15%) 25 (62.5%) 2 (5%) 5 (12.5%)	4 (9.7%) 26 (63.4%) 4 (9.7%) 6 (14.6%)	0.515 1 0.675 1
Contralateral stenosis	11 (27.5%)	12 (29.3%)	1
Side of surgery
Right  Left	19 (47.5%) 21 (52.5%)	22 (53.6%) 19 (46.4%)	0.579
Intraluminal carotid shunt	1 (2.5%)	3 (7.3%)	0.615
Intraoperative vasopressor requirement	9 (22.5%)	5 (12%)	0.22
Intraoperative antihypertensive requirement	16 (40%)	12 (29%)	0.31
Duration of surgery (min)	92.62 ± 19.67	97.34 ± 18.28	0.267
Total clamping time (min)	23.76 ± 4.99	22.30 ± 7.34	0.303

Data are expressed as mean ± standard deviation or number (percentage)

Abbreviations: ASA American Society of Anesthesiologists, BMI Body mass index, GD Group deep cervical plexus block, GI Group intermediate cervical plexus block, LMWH Low molecular weight heparin

Fig. 4.

Evaluation of cutaneous sensory loss in the innervation areas of the four peripheral cervical plexus nerves. Patients exhibiting “complete loss of sensation” or“reduced sensation” were included in the histogram to represent adequate regional anesthesia. The histogram depicts the temporal progression of sensory block comparison between the two study groups within the targeted dermatomal areas. Abbreviations: CPB Cervical plexus block

Time spent for block performance was 208.5 (50–332) s in GI and 320 (90–780) s in GD which showed statistically significant difference between groups (p < 0.001). The blocks were successfully performed with a single needle orientation in 34 (85%) patients in GI and 28 (68%) patients in GD, with no statistically significant difference between the groups (p = 0.11). The extra number of needle insertions (i.e. insertion of the needle more than once in GI and more than thrice in GD) was significantly higher in GD compared to GI (1 [1–4] and 1 [1–2], respectively; p = 0.04).

Pain scores throughout the surgery are summarized in Table 2. After start of the surgery, 19 patients (47.5%) in GI and 10 patients (24.5%) in GD needed additional lidocaine infiltration intraoperatively (p = 0.03), which was our primary outcome. The amount of additional lidocaine injected was similar between groups (5.05 ± 1.81 mL in GI, 6.4 ± 2.07 mL in GD; p = 0.08). In GI, additional lidocaine infiltration was administered in nine patients around the skin and subcutaneous tissues, in ten into the carotid sheath, and five towards the mandible. Seven, four and three patients required additional infiltration in GD at the same respective sites without statistical difference compared to GI (p = 0.53, 0.06, and 0.43, respectively). Sedation was maintained in all patients with the basal remifentanil infusion dose (0.02 µg/kg/min), and all patients had Ramsay Sedation Scale scores below three. None of the patients experienced persistent pain with additional lidocaine and there was no need to increase remifentanil dose.

Table 2.

Intraoperative pain scores

NRS	GI (n = 40)	GD (n = 41)	p value
0th min	0 (0–6)	0 (0–3)	0.554
5th min	0 (0–6)	0 (0–6)	0.864
10th min	0 (0–6)	0 (0–6)	0.233
15th min	1 (0–6)	0 (0–4)	0.125
20th min	0 (0–4)	0 (0–5)	0.923
25th min	1.5 (0–4)	1 (0–5)	0.825
30th min	0 (0–4)	0 (0–6)	0.952
45th min	0 (0–5)	0 (0–3)	0.950
60th min	0 (0–3)	0 (0–3)	0.971
75th min	0.5 (0–3)	0 (0–2)	0.09
90th min	0 (0–3)	0 (0–2)	0.447

Data are expressed as median (minimum-maximum)

Abbreviations: GD Group deep cervical plexus block, GI Group intermediate cervical plexus block, NRS Numeric rating scale

Hypoglossal nerve paralysis was observed in three patients (one in GI and two in GD). One patient experienced facial paralysis in GI. Hoarseness occurred in seven patients (three in GI and four in GD).

All complications recovered in 24 h. There was no statistical difference for complications between groups. Surgeon satisfaction was 5 (4-5) in GI and 5 (3-5) in GD and was similar between groups (p=0.25). Patient satisfaction was 5 (1-5) in GI and 4 (1-5) in GD and showed statically difference among groups (p=0.03). There was no need for general anesthesia conversion.

Discussion

This study demonstrated that deep CPB was associated with superior anesthetic efficacy as fewer patients required intraoperative additional lidocaine compared to patients receiving intermediate CPB during CEA, although both techniques provided adequate anesthesia.

Cervical plexus blocks are advantageous for carotid endarterectomies as they allow close neuromotor evaluations (2-5). A review from 2008 has advocated superficial cervical plexus block as the regional anesthetic technique of choice [21], although there is limited evidence relating to head-to-head comparisons for superficial, intermediate or deep blocks.

Studies comparing solely deep CPB with other techniques in CEA are specifically scarce (Table 3). One such study is by Stoneham et al., in which the time to first analgesic requirement and the number of patients requiring analgesic in the first 24 h is clearly decreased with deep CPB [22]. The second study that included a sole deep CPB group in chronological order was conducted by Rössel et al. in which they compared deep CPB with intermediate CPB and a combination of intermediate and perivascular injection [23]. Deep CPB was performed via a single injection at C4-5 level looking at LA spread with USG rather than blocking C2, C3 and C4 roots separately. They consequently reported deficiency in C2 cutaneous sensory loss with this technique. But even with this technique, though statistically insignificant due to the low number of patients recruited (n = 10 per group), deep CPB was better than intermediate CPB in terms of the number of patients who required additional LA.

Table 3.

Studies comparing deep CPB with other CPBs in CEA

	Author	Study design	Groups	Results
Sole DCPB comparisons	Stoneham et al. (18) 1998	RCT (n = 40)	*DCPB vs. *SCPB (All groups: 20 mL bupivacaine 0.375%) (DCPB at C4)	*Primary outcome: Amount of supplemental lidocaine similar *Time to 1 st analgesic requirement and # of patients requiring analgesic in the 1 st 24 h ↓ in DCPB
	Rössel et al. (19) 2019	RCT pilot study (n = 30)	*DCPB (20mL ropivacaine 0.5%) vs. *ICPB (20mL ropivacaine 0.5%) vs. *ICPB (20mL ropivacaine 0.5%) + perivascular infiltration (10mL ropivacaine 0.3%) (DCPB at C4-5)	*Primary outcome: Plasma concentration of ropivacaine DCPB > ICPB + perivascular infiltration > ICPB *Amount of supplemental lidocaine and # of patients requiring supplemental lidocaine ICPB > DCPB > ICPB + perivascular infiltration *Duration of block performance DCPB > ICPB + perivascular infiltration > ICPB
DCPB in combination comparisons	Pandit et al. (23) 2000	RCT (n = 40)	*DCPB (10mL bupivacaine 0.375%) + SCPB (20mL bupivacaine 0.375%) vs. *SCPB (30mL bupivacaine 0.375%) (DCPB at C4)	*Primary outcome: Amount of supplemental lidocaine similar *Intraoperative VAS similar *Time to 1 st analgesic requirement and # of patients requiring analgesic in the 1 st 24 h similar
	De Souza et al. (20) 2005	RCT (n = 125)	*DCPB (10mL bupivacaine 0.375%) + SCPB (20mL bupivacaine 0.375%) vs. *SCPB (30mL bupivacaine 0.375%) (DCPB at C4)	*Primary outcome: Amount of supplemental lidocaine similar *# of patients requiring supplemental lidocaine similar
	Sait Kavaklı et al. (21) 2016	RCT (n = 47)	*DCPB (15mL bupivacaine 0.5%) + SCPB (10mL bupivacaine 0.5%) vs. *ICPB (25mL bupivacaine 0.5%) (DCPB at C2, C3 and C4)	*Primary outcome: Amount of supplemental lidocaine ↑ in ICPB *Intraoperative VAS and # of patients requiring supplemental lidocaine ↑ in ICPB *Time to 1 st analgesic requirement similar
	Opperer et al. (12) 2022	RCT, observer blinded (n = 45)	*DCPB vs. *ICPB vs. *SCPB (All groups: 20mL ropivacaine 0.5%) (All groups: +10mL prilocaine 0.1% subcutaneous infiltration) (DCPB at C2, C3 and C4)	*Primary outcome: Measurements of ipsilateral diaphragmatic movement SCPB > ICPB > DCPB * Amount of supplemental prilocaine SCPB > DCPB > ICPB *Postoperative VAS similar

The USG guided blocks are annotated in italic, whereas landmark guided blocks (blind technique) are underlined

Abbreviations: CEA Carotid endarterectomy, CPB Cervical plexus block, DCPB Deep cervical plexus block, ICPB Intermediate cervical plexus block, RCT Randomized controlled trial, SCPB Superficial cervical plexus block, USG Ultrasound, VAS Visual analog scale, # Number

When randomized controlled studies reporting the use of deep CPB in combination with other blocks are investigated, one study that compared superficial to combined (deep and superficial) block reported similar number of patients requiring LA supplementation in both groups [24]. However, this study reported unusually high percentages of patients requiring LA (81.96% in superficial group and 89.06% in combined group), possibly due to blind techniques used. Another study that utilized deep CPB in combination with superficial CPB compared it to “intermediate CPB” which consists of three different area injections, namely posterolateral to the carotid bifurcation, beneath the SCM muscle and subcutaneous infiltration along the incision line [25]. This study concluded that deep and superficial CPB combined is superior to multi-injection intermediate block in terms of number of patients requiring lidocaine as well as the amount of supplemental lidocaine. We can speculate that, when subcutaneous infiltration in their “intermediate group” is taken into account as a modified superficial block, their results are comparable to ours. The last study of deep CPB compared all three blocks (superficial, intermediate and deep) but also included injection of 10 mL of prilocaine 1% to surgical incision line before surgery. The primary outcome of that study was diaphragmatic dysfunction rather than block efficacy [17]. The number of patients who required LA supplementation was not given in the study. Although the authors of this study reported a statistically significant difference between groups in the need for LA supplementation volume, the difference is between the superficial and intermediate groups rather than intermediate and deep CPB as the study is underpowered to detect a difference between intermediate and deep groups.

To our knowledge, this is also the first study that records cutaneous sensory loss across four nerve territories- namely the lesser occipital, supraclavicular, transverse cervical and great auricular nerves- following intermediate and deep CPBs in CEA patients. Calderon et al., in their prospective observational study, have evaluated lesser occipital, supraclavicular and transverse cervical nerve territories for intermediate CPB with anterior approach similar to ours [18]. However, greater auricular nerve territory should also be checked as this nerve runs superior at the upper portion of CEA incision and when incision is elongated towards mandible due to the location of bifurcation of carotid artery or thrombi, unblocked region can result in patient dissatisfaction [26].

One may question the need for a deeper block as success has been reported with the combination of superficial and intermediate blocks with relatively fewer complications [4]. However, the only meta-analysis that compared the complication frequency of deep CPB (alone or in combination) with superficial/intermediate CPB was in fact performed a decade earlier by Pandit et al. [4]. This meta-analysis contained five randomized controlled trials, and only one of them was a direct comparison of deep and superficial CPB [22]. The rest of the evidence was mostly observational studies or case reports of deep block compared to general anesthesia or combined block efficacy. One problem with this meta-analysis is that deep blocks used in the evidence were performed with blind techniques and paresthesia was actively sought out in order to elicit a successful block [22, 27]. Deep CPB was also stated to have a higher conversion ratio to general anesthesia, which the authors have attributed to technical difficulty of deep CPB indicating problems associated with blind technique. Also, in contrast to USG guided blocks of the recent years, both studies looking at deep CPB in comparison to superficial CPB in randomized controlled trials, performed deep CPB at a single root (C4) which may affect drug spread [22, 27]. Last but not the least, although deep CPB was stated to have higher number of serious complications, the absolute incidence was relatively low (0.25%). This meta-analysis also confirmed the scarcity of randomized controlled trials of solely deep CPB using USG techniques in block efficacy for CEA.

Another different aspect of our study is that, compared to the studies that contain sole deep CPB, we utilized a relatively low volume for initiation of the block (20 mL vs. 15 mL respectively) [17, 22, 23]. We used the same volume for both groups to avoid dose related bias. This may also explain the higher number of patients needing additional LA in the intermediate group as Pandit et al. defined intermediate block where injectate is administered deeper than the investing fascia of the neck but superficial to the prevertebral fascia. Thus, the volume of LA plays an important role in drug spread in the intermediate block as the recognized mechanism of action for intermediate CPB is mostly diffusion of LA via permeable prevertebral fascia [4, 8, 13].

In this study, we did not encounter any serious systemic or block related complications possibly due to USG guidance, but we encountered paralysis of the hypoglossal and facial nerves, and hoarseness due to block consequences. Similar complications were reported in a higher frequency in the USG guided studies for hoarseness [14–16, 18, 25], hypoglossal nerve paralysis [2, 14] and fascial nerve paralysis [2, 15–18, 25].

Ease of block is one of the key factors in the selection of a regional anesthesia technique which can be assessed by time spent for the block performance. The time required to perform deep CPB was significantly longer than intermediate CPB which is not surprising due to the greater number of needle punctures (three versus one, respectively). Although the procedural time for intermediate CPB has been rarely reported, our results are consistent with previous studies [16, 18]. Lower rates of patient satisfaction in deep CPB in our study may also be explained by three needle insertions as both CPBs were associated with low pain scores throughout surgery. Meanwhile, surgeon satisfaction was similar for both blocks.

This study has some limitations. Firstly, patients were not blinded to study group allocations as they were awake and aware of needle puncture numbers. But the anesthesiologists who performed the pinprick test and intraoperative follow-up, and the surgeon were blinded to group allocations. Secondly, pain assessment and cutaneous sensory block examination was limited to the intraoperative period and not extended to postoperative course as the primary outcome of this study was to compare the efficacy of two cervical plexus blocks. Finally, we could not show superiority of any blocks in terms of safety as this study was not designed to detect this issue, and the complication rate was quite low.

Conclusions

In this study with the use of USG and low initiation LA volumes, we found that significantly fewer patients required additional lidocaine in deep CPB with three injections than in intermediate CPB during CEA. Further focused anatomical trials would be warranted.

Abbreviations

CEA

Carotid endarterectomy

CPB

Cervical plexus block

GD

Group of Deep Cervical plexus block

GI

Group of Intermediate Cervical plexus block

LA

Local anesthetic

NRS

Numeric rating scale

SCM

Sternocleidomastoid

USG

Ultrasound

Authors’ contributions

ÖT and NS participated in the study design. NC, EED and DA collected the data. ÖT and NS analyzed the data. ÖT, NS and DA wrote the manuscript. ZS reviewed the manuscript. All authors have read and approved final version of the manuscript.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the principles of the Declaration of Helsinki. Ethical approval was obtained from Istanbul University Istanbul Faculty of Medicine Clinical Research Ethics Committee (approval number: 2021/923), and written informed consent was obtained from all participants prior to their enrollment in the study.

Consent for publication

Not Applicable.

Competing interests

The authors declare no competing interests.