- Email: [email protected]
Early experience with continuous cervical paravertebral block using a stimulating catheter
Early Experience With Continuous Cervical Paravertebral Block Using a Stimulating Catheter Andre´ P. Boezaart, M.D., Ph.D., Joe F. de Beer, MB.ChB., M.Med.(Orth), and Mercia L. Nell, R.N. Background and Objectives: This study reports our early experience with continuous cervical paravertebral block (CCPVB) using a stimulating catheter for the management of acute pain after shoulder surgery. Methods: This prospective observational study presents 256 CCPVB for pain relief after 14 different shoulder operations. Surgery was performed under general anesthesia and blocks were placed prior to induction of general anesthesia (n ⫽ 81 [32%]), after induction of general anesthesia (n ⫽ 116 [45%]), or postoperatively in the recovery room (n ⫽ 59 [23%]). A bolus dose of 30 mL of 0.5% ropivacaine was followed by an infusion of 0.1 mL/kg/h of 0.2% ropivacaine. Patient- or nurse-initiated bolus doses of 10 mL of the same drug were used for breakthrough pain and rescue analgesics were available. Postoperative pain, patient satisfaction, and motor function in different parts of the upper limb were evaluated immediately after surgery (time 0), and then 6, 12, 24, 48, 60 hours, and 14 days postoperatively. Results: An average of 2 (range 1-7) attempts were needed to advance the catheter while still stimulating the nerve. Average postoperative pain ranged from 0.27 ⫾ 1.04 cm to 0.78 ⫾ 1.56 cm (mean ⫾ SD) on a visual analog scale (VAS) (0-10 cm) for the first 48 hours and 3.8 ⫾ 2.1 cm and 3.5 ⫾ 2.4 cm at 60 hours and 14 days, respectively. Patient satisfaction on a VAS of 0 to 5 was 4.19 ⫾ 1.1, 4.28 ⫾ 1.01, and 4.69 ⫾ 1.05 at times 0, 6 hours, and 14 days, respectively. Motor function returned to normal in the fingers within 24 hours and in the shoulder within 60 hours. Complications included Horner’s syndrome (40%), dyspnea (8%), superficial skin infection (5%), posterior neck pain (22%), subclavian artery puncture (1%), contralateral epidural spread (4%), and 8% of the patients complained of an unpleasant “dead feeling” of the arm. Ninety-one percent of patients would request CCPVB again for future shoulder surgery. There was no evidence of nerve damage. Reg Anesth Pain Med 2003;28:406-413. Key Words: Paravertebral block, Continuous nerve block, Posterior approach, Interscalene block, Shoulder surgery, Stimulating catheter
W
e have previously described a technique for continuous cervical paravertebral block suitable for analgesia after shoulder surgery.1-4 This is a modification of the posterior approach to the brachial plexus originally described by Kappis,5 which was prompted by several factors. First was the requirement to provide adequate analgesia with min-
From the Department of Anesthesia, University of Iowa Hospitals and Clinics, Iowa City, IA (A.P.B.); Cape Shoulder Institute, Cape Town, South Africa (J.F.d.B); and Bristol Nuffield Hospital, Clifton Hill, Bristol (M.L.N.). Accepted for publication June 5, 2003. Dr. Boezaart initiated the concept of nerve stimulation via both the needle and catheter and the development of the StimuCath™. He also acts as consultant to Arrow International in the development of the StimuCath. Reprint requests: Andre´ P. Boezaart, M.D., Ph.D., 200 Hawkins Drive, 6-JCP, Iowa City, IA 52242-1079. E-mail: [email protected] © 2003 by the American Society of Regional Anesthesia and Pain Medicine. 1098-7339/03/2805-0009$30.00/0 doi:10.1016/S1098-7339(03)00221-9
406
imal motor block (often associated with continuous interscalene block) to allow patients to actively participate in physical therapy (especially patients undergoing arthroscopic capsulotomy for “frozen shoulder” or adhesive capsulitis). Second was to attempt to decrease the prevalence of an unpleasant “dead feeling” of the arm often associated with a dense sensory-motor block. Third was to address the need to perform continuous brachial plexus blocks postoperatively, or preoperatively in case of severely painful shoulder conditions without nerve stimulation. Fourth was to attempt to eliminate the posterior neck pain reported with the posterior approach. Finally was to decrease the incidence of neurologic complications reported with continuous interscalene block by Borgeat et al.6 as 14%, 7.9%, 3.9%, and 0.9% after 10 days, 1 month, 3 months and 6 months respectively. In a pilot study of 48 continuous paravertebral blocks using the landmarks first described by Kappis5 and later by Pippa et al,7 effective analgesia
Regional Anesthesia and Pain Medicine, Vol 28, No 5 (September–October), 2003: pp 406 –413
Continuous Cervical Paravertebral Brachial Plexus Block
•
Boezaart et al.
407
with minimal motor block was achieved, but most patients reported pain in the posterior aspect of the neck. This was probably related to the needle and catheter penetrating the tender extensor muscles of the neck. Modification of the continuous brachial plexus block was designed to avoid penetrating these muscles and prevent posterior neck pain. This article reports our early experience with continuous cervical paravertebral block.
Methods The institutional review board approved this study. Informed consent was obtained from 256 consecutive adult patients of either gender, American Society of Anesthesiologists Physical Class II or better classification or better, and scheduled for major shoulder surgery. All patients received a continuous cervical paravertebral block (CCPVB) for management of postoperative pain. The blocks were not used for surgical anesthesia. Patients with a history of central or peripheral neurologic disease, clinical evidence of pulmonary disease, psychiatric disease, or a history of narcotic dependence were excluded. Patients with allergies to meperidine, ropivacaine, or nonsteroidal anti-inflammatory drugs were also excluded. After explaining the procedure, including that there may be a higher risk of nerve damage when performing nerve blocks under general anesthesia, patients were given the choice of having a catheter for CCPVB placed either before or after induction of general anesthesia. If the patient had no preference, the catheter was placed before induction of general anesthesia. If the shoulder pathology was not clear and definitive corrective surgery was only done after diagnostic arthroscopy, blocks were placed postoperatively in the recovery room after the patient had emerged from general anesthesia. If no corrective surgery was performed, no blocks were placed. Needles and catheters for the continuous cervical paravertebral blocks were placed as previously described (Fig 1).1-4 All patients received a bolus of 30 mL of 0.5% ropivacaine (maximum 0.5 mL/kg) through the catheter immediately after catheter placement, at least one minute after a negative test dose of 2 mL of 2% lidocaine with 1:200,000 epinephrine and before the induction of general anesthesia. All operations were performed under general anesthesia. Standard intraoperative ASA monitoring was used. General anesthesia was induced with propofol (2 mg/kg) and fentanyl (2 g/kg) intravenously and maintained with a continuous intravenous infusion of propofol at (50 g/kg/min) and
Fig 1. Cervical paravertebral block. A 17- or 18-gauge sheathed Tuohy needle (StimuCath; Arrow Intl, Reading PA) enters the skin from the apex of the “V” formed by the trapezius and levator scapulae muscles on a line drawn from the spinous process of C6 to the suprasternal notch. The needle is advanced anterior, mesiad, and caudad, aiming for the suprasternal notch, until the pars intervertebralis or short transverse process of C6 is encountered. The needle, attached to a nerve stimulator and loss-of-resistance to air syringe, is now carefully “walked off” this bony part laterally, remaining medial to the posterior and middle scalene muscles until loss of resistance to air is encountered and motor response is evoked because of stimulation of the nerves when the space between the anterior scalene muscle and middle scalene muscle, where the roots of the brachial plexus are situated, is entered. The extensor muscles of the neck are thus avoided, and penetration of the vertebral artery and vein is not probable because of the protection by the pars intervertebralis and facet joints posterior. The phrenic nerve is situated on the belly of the anterior scalene muscle and the cervical sympathetic chain medial to it. The needle is now held steady in this position and a stimulating catheter is passed through it and advanced beyond the needle tip while stimulating the nerves via the “stimulating” catheter.4 The catheter is then securely tunneled to a convenient position.4
remifentanil (0.125 g/kg/min). Patients were mechanically ventilated to normocapnia with 60% oxygen in air through a laryngeal mask airway. Patients were positioned in the “beach chair” or lateral decubitis position depending on the preference of the surgeon. An infusion of ropivacaine (0.2%) was started immediately after initial bolus injection. The initial infusion rate was 0.1 mL/kg/h and continued into the postoperative period. The infusion rates were adjusted over the first 24 hours to obtain optimal analgesia and minimal motor block. If postoperative pain relief was not satisfactory and “breakthrough”
408
Regional Anesthesia and Pain Medicine Vol. 28 No. 5 September–October 2003
pain occurred, an additional bolus of 10 mL ropivacaine 0.2% was injected through the catheter by the attending nurse. This was permitted every 30 minutes, up to a maximum of 3 bolus injections per 6-hour period. If pain relief was still not effective, “rescue” enteral drugs, a combination of paracetamol 320 mg and codeine phosphate 8 mg 4 hourly were first given, and if this was not effective, parenteral drugs were administered. Parenteral analgesic drugs included 75 mg intramuscular diclofenac every 12 hours, or, if this was not effective, intramuscular meperidine (1 mg/kg) was administered every 4 hours. The use of intramuscular and oral analgesic usage was noted. The surgeon and anesthesiologist visited the patient twice daily to assess pain, to inspect the catheter site for signs of infection, and to note complications. Surgical pain was evaluated by the nursing staff on a visual analog scale of 0 cm to 10 cm (0 cm represented no pain and 10 cm represented the most severe pain imaginable). Pain was evaluated in the recovery room directly on emergence from general anesthesia (time 0) and thereafter at 6, 12, 24, 48, 60 hours, and 14 days after surgery. A homecare nurse visited patients who left the hospital earlier with catheters and infusions by elastometric pump in situ at these measurement intervals. Discharged patients were telephoned at these times if the catheters had been removed. Patient satisfaction with the analgesic regimen was assessed on a verbal scale of 0 to 5 (0 ⫽ extremely dissatisfied and 5 ⫽ extremely satisfied) at time zero in the recovery room, 6 hours and 2 weeks postoperatively. The average satisfaction scale and number of the patients who did not report a satisfaction scale of 5 of 5 were recorded. If signs of inflammation or infection were noted at the catheter site, the catheter was removed and the patient was placed on broad-spectrum oral antibiotic therapy. If bacteria were grown from the catheter exit site or tip, the antibiotic regimen was adjusted according to the antibiogram if necessary. Prophylactic antibiotics were not routinely used for surgery. If the continuous nerve block was no longer needed for acute pain relief, the infusion was discontinued for 3 hours and pain reassessed. If it were possible to manage the pain with oral analgesic drugs only, the catheter was removed. But, if parenteral opiates were required for effective pain relief, a bolus dose of 10 mL of 0.2% ropivacaine would be injected through the catheter and the continuous infusion resumed at its last effective infusion rate for an additional 24 hours. Infusions were continued until pain was manageable with oral analgesics but for no more than 7 days.
Table 1. Surgery Operation
n
%
Arthroscopic rotator cuff repair Open rotator cuff repair Arthroscopic subacromial decompression Arthroscopic Bankart repair Arthroscopic capsulotomy (frozen shoulder) SLAP repair Total shoulder arthroplasty Hemi shoulder arthroplasty Acromio-clavicular joint excision Laterjet stabilization procedure Open Bankart repair Subscapularis repair Arthroscopic sinovectomy Latissimus dorsi transfer Total
58 36 35 28 22 17 14 10 8 8 7 6 6 1 256
23 14 13 11 9 7 5 4 3 3 3 2 2 1 100
Motor function of the fingers (hand grip), wrist (flexion and extension), elbow (flexion and extension), and shoulder (flexion or abduction) was assessed at 0, 6, 12, 24, 48, 60 hours, and 14 days postoperatively on a scale of 0 to 5 with 0 representing no motor function and 5, normal motor function. If there was neck pain at the area where the catheter was placed, which could not be relieved with oral analgesics, nonsteroidal anti-inflammatory drugs, or application of a heating pad, the catheter was removed. All complications were noted. Horner’s syndrome, dyspnea, hematoma formation, and bilateral spread of the block were specifically noted. Two weeks after surgery, the patients were seen at the surgeon’s office or telephoned and specifically asked about residual anesthesia of the limb, burning sensation in the arm, neck pain, surgical pain, satisfaction with the analgesic regimen, and if the “dead feeling” of the limb bothered them during the infusion. Patients were asked if they would choose the same analgesic regimen for future shoulder operations. Data are reported as mean ⫾ standard deviation of the mean.
Results The average age of the patients entering this study was 55 ⫾ 14.5 years and included 92 men (35%) and 164 women (64%). Surgical procedures are listed in Table 1. The CCPVBs were performed in the recovery room after surgery for 59 patients (23%) using loss of resistance to air only for indication of correct needle placement. One hundred and sixteen patients (45%) requested that the CCPVB be performed after induction of general anesthesia, whereas 81 (32%) were placed before induction of general anesthesia. Vital signs in the recovery room and six hours postoperatively were stable.
Continuous Cervical Paravertebral Brachial Plexus Block Table 2. Postoperative Pain Pain (VAS 0-10) Time zero 6 hours postoperatively 12 hours postoperatively 24 hours postoperatively 48 hours postoperatively 60 hours postoperatively 2 weeks postoperatively
0.27 ⫾ 1.04 0.16 ⫾ 0.47 0.50 ⫾ 1.24 0.65 ⫾ 1.22 0.78 ⫾ 1.56 3.08 ⫾ 2.10 3.5 ⫾ 2.4
All patients received the same bolus injection (30 mL ropivacaine 0.5%) through the catheter after placement. The average infusion rate of ropivacaine 0.2% was 5.35 ⫾ 3.05 mL/h, with a minimum of 3 ml/h and a maximum of 10 mL/h. Data on postoperative pain are summarized in Table 2. Average postoperative pain was judged to be zero or very mild (visual analog scale [VAS] ⬍1) by the patients for the first 48 hours. However, they gave an average rating of 3.5 ⫾ 2.4 after 60 hours when the majority of infusions had been discontinued and catheters removed. One hundred twenty-eight patients required 1 additional bolus of 10 mL for “breakthrough” pain, 88 required 2, 54 required 3, and 39 required 4 or more bolus injections. Fiftyfour patients (21%) required 2 or more injections of diclofenac, whereas 36 (14%) required 2 or more injections of meperidine (mainly patients who received “open” procedures). The remaining patients (166) required either no analgesics or oral paracetamol-containing oral analgesic only. On average the catheters were removed after 41.65 ⫾ 14.69 hours of local anesthetic agent infusion (range 14120 hours). Patient satisfaction with the analgesic regimen was generally reported as excellent (5 on VAS of 0-5) except for 31 (12%) who rated the analgesic regimen as less than 5 (Table 3). The average satisfaction rating by these 31 patients was 2.33 ⫾ 2.08. All of these patients underwent “open” shoulder surgery. The average nerve stimulator current used for initial needle placement was 1.36 ⫾ 0.72 mA with a minimum of 0.3 and a maximum of 3 mA. Catheters were advanced, and muscle twitches remained unchanged at a mean nerve stimulator output of 1.06 ⫾ 0.52 mA (minimum 0.3 mA and
•
Boezaart et al.
maximum 3 mA). Needle placement was accurate after an average of 1.15 ⫾ 0.34 attempts (range 1-2), whereas catheter advancement with maintenance of the motor response was possible after an average of 2.56 ⫾ 1.2 attempts (range 1-7). The average depth at which the needle encountered the pars intervertebralis or transverse process was 5.70 ⫾ 1.14 cm (range 3-8 cm), and catheters were advanced 5.05 ⫾ 1.11 cm (range 3-6 cm) beyond the needle tip. Motor function was always better and returned earlier in the distal part of the extremity at the observation times after surgery. Full motor strength returned in all patients in the fingers and wrists after 48 hours, and at the elbows and shoulders after 60 hours (Fig 2, Table 4). Horner’s syndrome was present in 103 patients (40%) in the recovery room immediately after surgery. This decreased to 90 patients (35%) at 6 hours postoperatively, 32 (13%) after 12 hours, and 0 after 24 hours postoperatively. Hoarseness occurred in 25 (10%) of patients. Diaphragmatic function was not specifically assessed in this study, but 20 patients (8%) complained of dyspnea immediately postoperatively in the recovery room. In all cases, oxygen supplementation, placement of the patient in the Fowler’s position, and reassurance were adequate therapy. At 6 hours postoperatively dyspnea, occurred in 5 patients (2%) and in none after 12 hours postoperatively. Catheter sites showed signs of infection in 13 patients (5%) (redness [n ⫽ 10] or purulent discharge [n ⫽ 3]). Staphylococcus aureus was cultured from the catheter sites in 2 of these patients and Staphylococcus epidermidis in 7 patients. No bacteria could be cultured in the remaining 4 patients or from any of the catheter tips. The catheters were removed in all of these cases, and the patients were treated with broad-spectrum oral antibiotics. None of the patients developed any signs or symptoms of systemic or local infection. Fifty-six patients (22%) complained of posterior neck pain during the course of the continuous nerve block. This was effectively treated with oral or parenteral nonsteroidal anti-inflammatory drugs in 40 of the 56 patients (71%). In the remaining 16
Table 3. Patient Satisfaction
PACU (time zero) 6 hours postoperatively 2 weeks postoperatively
409
Satisfaction (0-5)
Satisfaction (if not 5/5)
Number of Patients Not 5/5 Satisfied (%)
4.19 ⫾ 1.10 4.28 ⫾ 1.01 4.69 ⫾ 1.05
3.09 ⫾ 0.83 3.01 ⫾ 0.72 2.33 ⫾ 2.08
45 (18%) 39 (15%) 31 (12%)
410
Regional Anesthesia and Pain Medicine Vol. 28 No. 5 September–October 2003
Fig 2. Motor function. Motor function at fingers, wrist, elbow, and shoulder directly postoperatively and then 6, 12, 24, 48, 60 hours, and 14 days postoperatively.
(29%), it was necessary to remove the catheters early to relieve neck pain. Three patients (1%) developed supraclavicular swelling due to hematoma formation shortly after placement of the catheters. Magnetic resonance imaging (MRI) studies revealed hematoma in the area of the subclavian artery in all 3 patients. In 2 of these cases, the nerve stimulator used in placement was later found to be defective. In the remaining patient, the catheter had been placed postoperatively without using a nerve stimulator and without encountering the pars intervertebralis or transverse process. The hematomas resolved without any further consequences in all patients within 7 days and the nerve blocks functioned well in all of them. Eleven patients (4%) had numbness in the contralateral hand. In 4 of these patients in whom contralateral numbness did not resolve within 6 hours, MRI and contrast studies were done, which showed epidural spread. All of the contralateral blocks resolved after the initial bolus injections dissipated. There were no further consequences from
continuous infusions in these patients, and all blocks were satisfactory. Twenty-one patients (8%) complained that the “dead feeling” of the limb was bothersome to them, whereas 10 (4%) had residual numbness in their hands 2 weeks after the surgery, which resolved completely after another week in all cases. No patient suffered nerve damage, and no neurologic complications were reported 3 weeks postoperatively. Of the 256 patients, 233 (91%) indicated that they would choose a similar analgesic regimen for future shoulder surgery, whereas 13 (5%) were not sure and the remaining 10 (4%) said they would prefer some other analgesic regimen. Of the latter 10 patients, six had surgery for shoulder replacement and the remaining four other “open” shoulder surgery.
Discussion Based on the low pain scores, infrequent use of adjuvant analgesics, high patient satisfaction rating,
Table 4. Motor Function
Time 0 6 hours postop 12 hours postop 24 hours postop 48 hours postop 60 hours postop 14 days postop
Fingers (0-5)
Wrist (0-5)
Elbow (0-5)
Shoulder (0-5)
2.88 ⫾ 1.82 3.27 ⫾ 1.56 4.27 ⫾ 1.28 4.59 ⫾ 0.80 5.00 ⫾ 0.00 5.00 ⫾ 0.00 5.00 ⫾ 0.00
1.96 ⫾ 1.84 2.73 ⫾ 1.73 3.37 ⫾ 1.66 4.35 ⫾ 0.86 5.00 ⫾ 0.00 5.00 ⫾ 0.00 5.00 ⫾ 0.00
1.04 ⫾ 1.51 1.38 ⫾ 1.58 2.62 ⫾ 1.68 3.65 ⫾ 1.17 4.22 ⫾ 1.09 5.00 ⫾ 0.00 5.00 ⫾ 0.00
0.35 ⫾ 1.06 0.73 ⫾ 1.31 1.20 ⫾ 1.12 2.41 ⫾ 1.70 3.44 ⫾ 1.42 5.00 ⫾ 0.00 5.00 ⫾ 0.00
Continuous Cervical Paravertebral Brachial Plexus Block
and the high number of patients who would prefer the same analgesic regimen for possible future surgery, CCPVB appears to provide satisfactory postoperative analgesia for patients undergoing shoulder surgery. However, the block may not be fully adequate in cases of shoulder arthroplasty, large rotator cuff repair surgery involving the subscapularis muscle, “open” acromioclavicular joint surgery, and major open stabilization procedures, since all the patients who were not satisfied with the analgesia underwent one of these “open” procedures. This finding may possibly be related to the CCPVB being performed with a 30-mL bolus and 3 to 10 mL/h rates of continuous local anesthetic infusion. This may not be enough to reach the origin or branches of the superficial cervical plexus nerves that provide sensory innervation to the skin and subcutaneous tissue around the shoulder. If so, a superficial cervical plexus block would be an easy adjuvant to CCPVB for major open shoulder surgery. CCVBP was also ineffective in the single case of latissimus dorsi muscle transfer surgery in the series, probably because the spread of local anesthetic was limited to the brachial plexus roots and did not reach the upper thoracic nerve roots. The upper thoracic nerve roots should probably be blocked additionally if this type of surgery is contemplated. The postoperative pain management protocol for continuous infusion and drug therapy may have been less than optimal. Patient-controlled bolus injections, as described by Borgeat et al.9,10 and Ilfeld et al.,11 may enhance the patient satisfaction. An average of 2 (1-7) attempts at catheter placement was required, indicating that even after successful needle placement, the catheter does not necessarily follow a perineural course or remain in the desired position. Although studies comparing stimulating catheters with nonstimulating catheters for this particular block are not yet available, at least 1 author has shown a clear advantage for stimulating catheters in other nerve blocks.12 We do not adhere to the concept of “volume distension” of the paravertebral space with saline before catheter placement as propagated by PhangDam et al.12 because saline in this space will dissipate electrical current and make nerve stimulation via the catheter impossible or will require much higher current (as PhangDam et al. noted). Loss of resistance to air may be less obvious in the cervical paravertebral block than during epidural needle placement. Catheter advancement was also associated with more resistance as compared with epidural catheter placement. The importance of preserving motor function has not been clearly defined by orthopedic surgeons
•
Boezaart et al.
411
and physical therapists. In cases such as rotator cuff repair and shoulder arthroplasty, early motor function can be detrimental in shoulder surgery, especially if motor function is complete and the shoulder joint is free of pain. On the other hand, early return of motor function may be advantageous in cases of nonstructural surgery, such as capsulotomy for adhesive capsulitis (“frozen shoulder”). The continuous cervical paravertebral block provided ideal conditions for early return of motor function in hand, wrist, and elbow, whereas shoulder motor function returned later and simultaneously with the return of shoulder sensation. Early return of motor function added to patient comfort, and 92% of these patients did not complain of an uncomfortable “dead feeling” in the arm. Preservation or early return of motor function with this block is probably because the catheter is placed at the level of the roots of the brachial plexus. At this level, the posterior sensory fibers and anterior motor fibers have not yet combined. Because a posterior approach is used for this block, the sensory fibers are encountered first and relatively small infusion rates may spare the more “distant” motor fibers. The initial motor block gradually diminished from the inferior roots (hand, wrist, and elbow) upward toward the fifth cervical root (shoulder) where the catheter was initially placed, as the effects of the initial ropivacaine bolus wore off. This may also explain why some patients reported sensory electrical pulsations in the arm before a motor response was elicited and also why higher than customary current outputs were required to elicit a motor response in some cases. The complications observed with continuous cervical paravertebral block tended to occur in the immediate postoperative period while the effects of the main bolus dose of local anesthetic was present. Horner’s syndrome was common, whereas dyspnea was a rare complication. Although phrenic function was not specifically studied in this series, dyspnea was probably because of phrenic nerve paresis. Dyspnea resolved in all patients in whom it occurred within 6 to 12 hours and was alleviated by simple measures. Phrenic nerve paresis should theoretically have the same prevalence (20%) as after observed with continuous blocks via the modified lateral or longitudinal interscalene approach.8 Hoarseness from recurrent laryngeal nerve block mainly followed right-sided paravertebral block and resolved within 12 hours in all cases. Macintosh and Mushin13 remarked that the posterior approach to the brachial plexus as originally proposed by Kappis in 1912 was uniformly successful, but this block never gained popularity, probably because of the pain associated with it. We have
412
Regional Anesthesia and Pain Medicine Vol. 28 No. 5 September–October 2003
recently speculated that the pain may be associated with the penetration of the tender extensor muscles of the neck by the needle and catheter [3]. The continuous cervical paravertebral block was designed with this in mind, and the needle entry is accordingly from the apex of the “V” formed by the trapezius and levator scapulae muscles to avoid the penetration of the posterior extensor muscles of the neck. This may explain the low incidence of posterior neck pain reported by the patients in this series. Nadig et al.14 recently criticized the continuous paravertebral block that Vranken et al.15 performed for the management of pain in patients with terminal cancer. Apart from stressing the posterior neck pain associated with this technique, they also argued that nerve root penetration, subdural or extradural spread of local anesthetic agent, and bleeding because of needle placement could be problematic. In the present series, we did not observe nerve injury or subdural injection, although epidural spread was demonstrated by MRI in 4 patients and suspected in 7 more. Epidural spread was not clinically significant. Subdural injection is possible but was not encountered, perhaps because we used a relatively large 17-gauge blunt Tuohy needle and a bullet-tipped catheter. Dural penetration should not be more prevalent with this block than with the classical anterior interscalene approach. Subclavian artery penetration occurred in 3 patients. In 1 of the cases, a nerve stimulator was not used because of postoperative placement of the block and the bony structures were missed laterally. In the remaining 2 patients, the nerve stimulator was later found to be defective. A functioning nerve stimulator is thus essential because of the close anatomical proximity of the subclavian artery anterior and inferior to the brachial plexus at this level. The shoulder muscles will twitch when the brachial plexus is encountered, and this should occur before the subclavian artery is reached. That blood was not aspirated during arterial puncture was probably because of the stylet inside the needle. If the stylet was removed when the bony structures were encountered, and a loss-of-resistance-to-air syringe was applied to the needle, blood vessel penetration may have been apparent. Positioning the patient in the sitting position or with a 5° to 10° angle head-up tilt when the lateral decubitis position is used would probably cause collapse of the veins around this area, which could explain the absence of venous bleeding observed in the present study. Borgeat et al.5 recently reported a 14% incidence of paresthesia, dysesthesia, and arm pain not related to surgery 10 days after conventional continuous interscalene block, which decreased to 3.9%
and 0.9% after 3 and 6 months, respectively. We speculate that this might in part be because of the catheter placement and drug infusion where the brachial plexus passes over the first rib through the narrow space between the first rib and the clavicle. The more medial paravertebral approach avoids this area, which possibly explains the absence of temporary or permanent nerve damage in the present study when compared with the findings of Borgeat et al.5 Although some anesthesiologists avoid performing nerve blocks in patients during general anesthesia,16 our limited data suggest that placing the blocks in anesthetized patients diminished discomfort without adding risk. However, until definitive evidence to the contrary is presented, cervical paravertebral block should be performed in nonanesthetized patients whenever possible. Severely painful shoulder conditions, anxiety, and blocks in children are potential exceptions, but the patients or their guardians should be informed about the possible added risk of nerve damage when nerve blocks are performed under general anesthesia. Nerve damage unrelated to anesthetic technique or regional anesthesia has been reported in as many as 30% of arthroscopic shoulder operations,17 4% after shoulder arthroplasty,18 2.6% in rotator cuff repair,19 and 8.2% in patients who underwent surgery for anterior glenohumeral instability.20 All the surgical procedures in the present series were performed by the same shoulder surgeon (JFdB) experienced in arthroscopic, rotator cuff, arthroplasty, and instability surgery. It is therefore possible that the present results may not be generally applicable. Among the limitations of our study, the main one is that CCPVB was not compared with conventional continuous interscalene block. A study to compare these 2 blocks is underway. Another limitation is the lack of data on patient-controlled intravenous analgesia, which is probably a more objective measurement of acute pain. Determining phrenic nerve function or proprioception would have greatly enhanced our study. It would also have been valuable to evaluate CCPVB as a sole anesthetic for shoulder surgery, especially for arthroscopic shoulder surgery. The use of patient-controlled bolus injections as propagated by Borgeat et al9,10 and others11 would probably have added value to this study. In summary, we conclude that CCPVB appears to provide satisfactory analgesia for pain associated with shoulder surgery. If used for “open” shoulder surgery, larger volumes of local anesthetic agents should probably be used for the bolus injection and continuous infusions, and an additional superficial cervical plexus block should perhaps be performed. Our data show that CCPVB preserves motor func-
Continuous Cervical Paravertebral Brachial Plexus Block
tion of the distal parts of the arm or facilitates early return of such function. Complications occurred in the early postoperative period when the effects of the initial bolus were still present and the patients were still in the care of the anesthesiologist. Nerve damage or other serious complications were not observed in this limited series. Further research is required to compare this block with the conventional anterior approach to the continuous interscalene block and apply it for other major upper limb surgery or as sole anesthetic for shoulder and upper limb surgery.
9.
10.
11.
Acknowledgment
12.
The authors are indebted to Robert Koorn, MD, Richard W. Rosenquist, MD, and Steven C. Borene, MD, for their critical review of this manuscript.
13.
References
14.
1. Boezaart AP, de Beer JF. Continuous low cervical paravertebral block for shoulder surgery. Reg Anesth Pain Med 2001;26:68(abstr). 2. Boezaart AP. Continuous interscalene block for ambulatory shoulder surgery. Best Practice Res Clin Anaesth 2002;16:295-310. 3. Boezaart A P, Koorn R, Rosenquist RW. Paravertebral approach to the brachial plexus: an anatomic improvement in technique. Reg Anesth Pain Med 2003; 28:241-244. 4. Borene SC, Rosenquist RW, Koorn R, Haider N, Boezaart AP. An indication for continuous cervical paravertebral block (posterior approach to the interscalene space). Anesth Analg 2003;97:898-900. 5. Kappis M. Weitere Erfahrungen mit der Sympathektomie. Klin Wchnschr 1923;2:1441. 6. Borgeat A, Ekatodramis G, Kalberer F, Benz C. Acute and nonacute complications associated with interscalene block and shoulder surgery. Anesthesiology 2001;95:875-880. 7. Pippa P, Cominelli E, Marinelli C, Aito S. Brachial plexus block using the posterior approach. Eur J Anaesth 1990;7:411-420. 8. Boezaart AP, de Beer JF, du Toit C, van Rooyen K. A
15.
16.
17.
18.
19.
20.
•
Boezaart et al.
413
new technique of continuous interscalene nerve block. Can J Anesth 1999;46:275-281. Borgeat A, Schappi B, Biasca N, Gerber C. Patientcontrolled analgesia after major shoulder surgery: Patient-controlled interscalene analgesia versus patient-controlled analgesia. Anesthesiology 1997;87: 1343-1347. Borgeat A, Tewes E, Biasca N, Gerber C. Patientcontrolled interscalene analgesia with ropivacaine after major shoulder surgery: PCIA vs. PCA. Br J Anaesth 1998;81:603-605. Ilfeld BM, Morey TE, Enneking K. Continuous infraclavicular brachial plexus block for postoperative pain control at home. Anesthesiology 2002;96:12971304. PhangDam C, Kick O, Collet T, Gouin F, Pinaud M. Continuous peripheral nerve block with stimulating catheters. Reg Anesth Pain Med 2003;28:83-88. Macintosh RR, Mushin WW. Brachial plexus. In: Macintosh RR, Mushin WW, eds. Local Analgesia, 4th ed. Edinburgh: E&S Livingston Ltd, 1967:8. Nadig M, Ekatodramis G, Borgeat A. Continuous brachial plexus block at the cervical level using a posterior approach in the management of neuropathic cancer pain. Reg Anesth Pain Med 2002;27:446. Vranken JH, van der Vegt MH, Zuurmond WWA, Pijl AJ, Dzolijk M. Continuous brachial plexus block at the cervical level using a posterior approach in the management of neuropathic cancer pain. Reg Anesth Pain Med 2001;26:572-575. Benumof JL. Permanent loss of cervical spinal cord function associated with interscalene block performed under general anesthesia. Anesthesiology 200; 93:1541-1544. McFarland EG, O’Neill OR, Hsu CY. Complications of shoulder arthroscopy. J South Orthop Assoc 1997;6: 190-196. Lynch NM, Cofield RH, Silber PL, Hermann RC. Neurologic complications after total shoulder arthroplasty. J Shoulder Elbow Surg 1996;5:53-61. Mansat P, Cofield RH, Kersten TE, Rowland CM. Complications of rotator cuff repair. Orthop Clin North Am 1997;28:205-213. Ho E, Cofield RH, Balm MR, Hattrup SJ, Rowland CS. Neurological complications of surgery for anterior shoulder instability. J Shoulder Elbow Surg 1999;8: 266-270.
W
e have previously described a technique for continuous cervical paravertebral block suitable for analgesia after shoulder surgery.1-4 This is a modification of the posterior approach to the brachial plexus originally described by Kappis,5 which was prompted by several factors. First was the requirement to provide adequate analgesia with min-
From the Department of Anesthesia, University of Iowa Hospitals and Clinics, Iowa City, IA (A.P.B.); Cape Shoulder Institute, Cape Town, South Africa (J.F.d.B); and Bristol Nuffield Hospital, Clifton Hill, Bristol (M.L.N.). Accepted for publication June 5, 2003. Dr. Boezaart initiated the concept of nerve stimulation via both the needle and catheter and the development of the StimuCath™. He also acts as consultant to Arrow International in the development of the StimuCath. Reprint requests: Andre´ P. Boezaart, M.D., Ph.D., 200 Hawkins Drive, 6-JCP, Iowa City, IA 52242-1079. E-mail: [email protected] © 2003 by the American Society of Regional Anesthesia and Pain Medicine. 1098-7339/03/2805-0009$30.00/0 doi:10.1016/S1098-7339(03)00221-9
406
imal motor block (often associated with continuous interscalene block) to allow patients to actively participate in physical therapy (especially patients undergoing arthroscopic capsulotomy for “frozen shoulder” or adhesive capsulitis). Second was to attempt to decrease the prevalence of an unpleasant “dead feeling” of the arm often associated with a dense sensory-motor block. Third was to address the need to perform continuous brachial plexus blocks postoperatively, or preoperatively in case of severely painful shoulder conditions without nerve stimulation. Fourth was to attempt to eliminate the posterior neck pain reported with the posterior approach. Finally was to decrease the incidence of neurologic complications reported with continuous interscalene block by Borgeat et al.6 as 14%, 7.9%, 3.9%, and 0.9% after 10 days, 1 month, 3 months and 6 months respectively. In a pilot study of 48 continuous paravertebral blocks using the landmarks first described by Kappis5 and later by Pippa et al,7 effective analgesia
Regional Anesthesia and Pain Medicine, Vol 28, No 5 (September–October), 2003: pp 406 –413
Continuous Cervical Paravertebral Brachial Plexus Block
•
Boezaart et al.
407
with minimal motor block was achieved, but most patients reported pain in the posterior aspect of the neck. This was probably related to the needle and catheter penetrating the tender extensor muscles of the neck. Modification of the continuous brachial plexus block was designed to avoid penetrating these muscles and prevent posterior neck pain. This article reports our early experience with continuous cervical paravertebral block.
Methods The institutional review board approved this study. Informed consent was obtained from 256 consecutive adult patients of either gender, American Society of Anesthesiologists Physical Class II or better classification or better, and scheduled for major shoulder surgery. All patients received a continuous cervical paravertebral block (CCPVB) for management of postoperative pain. The blocks were not used for surgical anesthesia. Patients with a history of central or peripheral neurologic disease, clinical evidence of pulmonary disease, psychiatric disease, or a history of narcotic dependence were excluded. Patients with allergies to meperidine, ropivacaine, or nonsteroidal anti-inflammatory drugs were also excluded. After explaining the procedure, including that there may be a higher risk of nerve damage when performing nerve blocks under general anesthesia, patients were given the choice of having a catheter for CCPVB placed either before or after induction of general anesthesia. If the patient had no preference, the catheter was placed before induction of general anesthesia. If the shoulder pathology was not clear and definitive corrective surgery was only done after diagnostic arthroscopy, blocks were placed postoperatively in the recovery room after the patient had emerged from general anesthesia. If no corrective surgery was performed, no blocks were placed. Needles and catheters for the continuous cervical paravertebral blocks were placed as previously described (Fig 1).1-4 All patients received a bolus of 30 mL of 0.5% ropivacaine (maximum 0.5 mL/kg) through the catheter immediately after catheter placement, at least one minute after a negative test dose of 2 mL of 2% lidocaine with 1:200,000 epinephrine and before the induction of general anesthesia. All operations were performed under general anesthesia. Standard intraoperative ASA monitoring was used. General anesthesia was induced with propofol (2 mg/kg) and fentanyl (2 g/kg) intravenously and maintained with a continuous intravenous infusion of propofol at (50 g/kg/min) and
Fig 1. Cervical paravertebral block. A 17- or 18-gauge sheathed Tuohy needle (StimuCath; Arrow Intl, Reading PA) enters the skin from the apex of the “V” formed by the trapezius and levator scapulae muscles on a line drawn from the spinous process of C6 to the suprasternal notch. The needle is advanced anterior, mesiad, and caudad, aiming for the suprasternal notch, until the pars intervertebralis or short transverse process of C6 is encountered. The needle, attached to a nerve stimulator and loss-of-resistance to air syringe, is now carefully “walked off” this bony part laterally, remaining medial to the posterior and middle scalene muscles until loss of resistance to air is encountered and motor response is evoked because of stimulation of the nerves when the space between the anterior scalene muscle and middle scalene muscle, where the roots of the brachial plexus are situated, is entered. The extensor muscles of the neck are thus avoided, and penetration of the vertebral artery and vein is not probable because of the protection by the pars intervertebralis and facet joints posterior. The phrenic nerve is situated on the belly of the anterior scalene muscle and the cervical sympathetic chain medial to it. The needle is now held steady in this position and a stimulating catheter is passed through it and advanced beyond the needle tip while stimulating the nerves via the “stimulating” catheter.4 The catheter is then securely tunneled to a convenient position.4
remifentanil (0.125 g/kg/min). Patients were mechanically ventilated to normocapnia with 60% oxygen in air through a laryngeal mask airway. Patients were positioned in the “beach chair” or lateral decubitis position depending on the preference of the surgeon. An infusion of ropivacaine (0.2%) was started immediately after initial bolus injection. The initial infusion rate was 0.1 mL/kg/h and continued into the postoperative period. The infusion rates were adjusted over the first 24 hours to obtain optimal analgesia and minimal motor block. If postoperative pain relief was not satisfactory and “breakthrough”
408
Regional Anesthesia and Pain Medicine Vol. 28 No. 5 September–October 2003
pain occurred, an additional bolus of 10 mL ropivacaine 0.2% was injected through the catheter by the attending nurse. This was permitted every 30 minutes, up to a maximum of 3 bolus injections per 6-hour period. If pain relief was still not effective, “rescue” enteral drugs, a combination of paracetamol 320 mg and codeine phosphate 8 mg 4 hourly were first given, and if this was not effective, parenteral drugs were administered. Parenteral analgesic drugs included 75 mg intramuscular diclofenac every 12 hours, or, if this was not effective, intramuscular meperidine (1 mg/kg) was administered every 4 hours. The use of intramuscular and oral analgesic usage was noted. The surgeon and anesthesiologist visited the patient twice daily to assess pain, to inspect the catheter site for signs of infection, and to note complications. Surgical pain was evaluated by the nursing staff on a visual analog scale of 0 cm to 10 cm (0 cm represented no pain and 10 cm represented the most severe pain imaginable). Pain was evaluated in the recovery room directly on emergence from general anesthesia (time 0) and thereafter at 6, 12, 24, 48, 60 hours, and 14 days after surgery. A homecare nurse visited patients who left the hospital earlier with catheters and infusions by elastometric pump in situ at these measurement intervals. Discharged patients were telephoned at these times if the catheters had been removed. Patient satisfaction with the analgesic regimen was assessed on a verbal scale of 0 to 5 (0 ⫽ extremely dissatisfied and 5 ⫽ extremely satisfied) at time zero in the recovery room, 6 hours and 2 weeks postoperatively. The average satisfaction scale and number of the patients who did not report a satisfaction scale of 5 of 5 were recorded. If signs of inflammation or infection were noted at the catheter site, the catheter was removed and the patient was placed on broad-spectrum oral antibiotic therapy. If bacteria were grown from the catheter exit site or tip, the antibiotic regimen was adjusted according to the antibiogram if necessary. Prophylactic antibiotics were not routinely used for surgery. If the continuous nerve block was no longer needed for acute pain relief, the infusion was discontinued for 3 hours and pain reassessed. If it were possible to manage the pain with oral analgesic drugs only, the catheter was removed. But, if parenteral opiates were required for effective pain relief, a bolus dose of 10 mL of 0.2% ropivacaine would be injected through the catheter and the continuous infusion resumed at its last effective infusion rate for an additional 24 hours. Infusions were continued until pain was manageable with oral analgesics but for no more than 7 days.
Table 1. Surgery Operation
n
%
Arthroscopic rotator cuff repair Open rotator cuff repair Arthroscopic subacromial decompression Arthroscopic Bankart repair Arthroscopic capsulotomy (frozen shoulder) SLAP repair Total shoulder arthroplasty Hemi shoulder arthroplasty Acromio-clavicular joint excision Laterjet stabilization procedure Open Bankart repair Subscapularis repair Arthroscopic sinovectomy Latissimus dorsi transfer Total
58 36 35 28 22 17 14 10 8 8 7 6 6 1 256
23 14 13 11 9 7 5 4 3 3 3 2 2 1 100
Motor function of the fingers (hand grip), wrist (flexion and extension), elbow (flexion and extension), and shoulder (flexion or abduction) was assessed at 0, 6, 12, 24, 48, 60 hours, and 14 days postoperatively on a scale of 0 to 5 with 0 representing no motor function and 5, normal motor function. If there was neck pain at the area where the catheter was placed, which could not be relieved with oral analgesics, nonsteroidal anti-inflammatory drugs, or application of a heating pad, the catheter was removed. All complications were noted. Horner’s syndrome, dyspnea, hematoma formation, and bilateral spread of the block were specifically noted. Two weeks after surgery, the patients were seen at the surgeon’s office or telephoned and specifically asked about residual anesthesia of the limb, burning sensation in the arm, neck pain, surgical pain, satisfaction with the analgesic regimen, and if the “dead feeling” of the limb bothered them during the infusion. Patients were asked if they would choose the same analgesic regimen for future shoulder operations. Data are reported as mean ⫾ standard deviation of the mean.
Results The average age of the patients entering this study was 55 ⫾ 14.5 years and included 92 men (35%) and 164 women (64%). Surgical procedures are listed in Table 1. The CCPVBs were performed in the recovery room after surgery for 59 patients (23%) using loss of resistance to air only for indication of correct needle placement. One hundred and sixteen patients (45%) requested that the CCPVB be performed after induction of general anesthesia, whereas 81 (32%) were placed before induction of general anesthesia. Vital signs in the recovery room and six hours postoperatively were stable.
Continuous Cervical Paravertebral Brachial Plexus Block Table 2. Postoperative Pain Pain (VAS 0-10) Time zero 6 hours postoperatively 12 hours postoperatively 24 hours postoperatively 48 hours postoperatively 60 hours postoperatively 2 weeks postoperatively
0.27 ⫾ 1.04 0.16 ⫾ 0.47 0.50 ⫾ 1.24 0.65 ⫾ 1.22 0.78 ⫾ 1.56 3.08 ⫾ 2.10 3.5 ⫾ 2.4
All patients received the same bolus injection (30 mL ropivacaine 0.5%) through the catheter after placement. The average infusion rate of ropivacaine 0.2% was 5.35 ⫾ 3.05 mL/h, with a minimum of 3 ml/h and a maximum of 10 mL/h. Data on postoperative pain are summarized in Table 2. Average postoperative pain was judged to be zero or very mild (visual analog scale [VAS] ⬍1) by the patients for the first 48 hours. However, they gave an average rating of 3.5 ⫾ 2.4 after 60 hours when the majority of infusions had been discontinued and catheters removed. One hundred twenty-eight patients required 1 additional bolus of 10 mL for “breakthrough” pain, 88 required 2, 54 required 3, and 39 required 4 or more bolus injections. Fiftyfour patients (21%) required 2 or more injections of diclofenac, whereas 36 (14%) required 2 or more injections of meperidine (mainly patients who received “open” procedures). The remaining patients (166) required either no analgesics or oral paracetamol-containing oral analgesic only. On average the catheters were removed after 41.65 ⫾ 14.69 hours of local anesthetic agent infusion (range 14120 hours). Patient satisfaction with the analgesic regimen was generally reported as excellent (5 on VAS of 0-5) except for 31 (12%) who rated the analgesic regimen as less than 5 (Table 3). The average satisfaction rating by these 31 patients was 2.33 ⫾ 2.08. All of these patients underwent “open” shoulder surgery. The average nerve stimulator current used for initial needle placement was 1.36 ⫾ 0.72 mA with a minimum of 0.3 and a maximum of 3 mA. Catheters were advanced, and muscle twitches remained unchanged at a mean nerve stimulator output of 1.06 ⫾ 0.52 mA (minimum 0.3 mA and
•
Boezaart et al.
maximum 3 mA). Needle placement was accurate after an average of 1.15 ⫾ 0.34 attempts (range 1-2), whereas catheter advancement with maintenance of the motor response was possible after an average of 2.56 ⫾ 1.2 attempts (range 1-7). The average depth at which the needle encountered the pars intervertebralis or transverse process was 5.70 ⫾ 1.14 cm (range 3-8 cm), and catheters were advanced 5.05 ⫾ 1.11 cm (range 3-6 cm) beyond the needle tip. Motor function was always better and returned earlier in the distal part of the extremity at the observation times after surgery. Full motor strength returned in all patients in the fingers and wrists after 48 hours, and at the elbows and shoulders after 60 hours (Fig 2, Table 4). Horner’s syndrome was present in 103 patients (40%) in the recovery room immediately after surgery. This decreased to 90 patients (35%) at 6 hours postoperatively, 32 (13%) after 12 hours, and 0 after 24 hours postoperatively. Hoarseness occurred in 25 (10%) of patients. Diaphragmatic function was not specifically assessed in this study, but 20 patients (8%) complained of dyspnea immediately postoperatively in the recovery room. In all cases, oxygen supplementation, placement of the patient in the Fowler’s position, and reassurance were adequate therapy. At 6 hours postoperatively dyspnea, occurred in 5 patients (2%) and in none after 12 hours postoperatively. Catheter sites showed signs of infection in 13 patients (5%) (redness [n ⫽ 10] or purulent discharge [n ⫽ 3]). Staphylococcus aureus was cultured from the catheter sites in 2 of these patients and Staphylococcus epidermidis in 7 patients. No bacteria could be cultured in the remaining 4 patients or from any of the catheter tips. The catheters were removed in all of these cases, and the patients were treated with broad-spectrum oral antibiotics. None of the patients developed any signs or symptoms of systemic or local infection. Fifty-six patients (22%) complained of posterior neck pain during the course of the continuous nerve block. This was effectively treated with oral or parenteral nonsteroidal anti-inflammatory drugs in 40 of the 56 patients (71%). In the remaining 16
Table 3. Patient Satisfaction
PACU (time zero) 6 hours postoperatively 2 weeks postoperatively
409
Satisfaction (0-5)
Satisfaction (if not 5/5)
Number of Patients Not 5/5 Satisfied (%)
4.19 ⫾ 1.10 4.28 ⫾ 1.01 4.69 ⫾ 1.05
3.09 ⫾ 0.83 3.01 ⫾ 0.72 2.33 ⫾ 2.08
45 (18%) 39 (15%) 31 (12%)
410
Regional Anesthesia and Pain Medicine Vol. 28 No. 5 September–October 2003
Fig 2. Motor function. Motor function at fingers, wrist, elbow, and shoulder directly postoperatively and then 6, 12, 24, 48, 60 hours, and 14 days postoperatively.
(29%), it was necessary to remove the catheters early to relieve neck pain. Three patients (1%) developed supraclavicular swelling due to hematoma formation shortly after placement of the catheters. Magnetic resonance imaging (MRI) studies revealed hematoma in the area of the subclavian artery in all 3 patients. In 2 of these cases, the nerve stimulator used in placement was later found to be defective. In the remaining patient, the catheter had been placed postoperatively without using a nerve stimulator and without encountering the pars intervertebralis or transverse process. The hematomas resolved without any further consequences in all patients within 7 days and the nerve blocks functioned well in all of them. Eleven patients (4%) had numbness in the contralateral hand. In 4 of these patients in whom contralateral numbness did not resolve within 6 hours, MRI and contrast studies were done, which showed epidural spread. All of the contralateral blocks resolved after the initial bolus injections dissipated. There were no further consequences from
continuous infusions in these patients, and all blocks were satisfactory. Twenty-one patients (8%) complained that the “dead feeling” of the limb was bothersome to them, whereas 10 (4%) had residual numbness in their hands 2 weeks after the surgery, which resolved completely after another week in all cases. No patient suffered nerve damage, and no neurologic complications were reported 3 weeks postoperatively. Of the 256 patients, 233 (91%) indicated that they would choose a similar analgesic regimen for future shoulder surgery, whereas 13 (5%) were not sure and the remaining 10 (4%) said they would prefer some other analgesic regimen. Of the latter 10 patients, six had surgery for shoulder replacement and the remaining four other “open” shoulder surgery.
Discussion Based on the low pain scores, infrequent use of adjuvant analgesics, high patient satisfaction rating,
Table 4. Motor Function
Time 0 6 hours postop 12 hours postop 24 hours postop 48 hours postop 60 hours postop 14 days postop
Fingers (0-5)
Wrist (0-5)
Elbow (0-5)
Shoulder (0-5)
2.88 ⫾ 1.82 3.27 ⫾ 1.56 4.27 ⫾ 1.28 4.59 ⫾ 0.80 5.00 ⫾ 0.00 5.00 ⫾ 0.00 5.00 ⫾ 0.00
1.96 ⫾ 1.84 2.73 ⫾ 1.73 3.37 ⫾ 1.66 4.35 ⫾ 0.86 5.00 ⫾ 0.00 5.00 ⫾ 0.00 5.00 ⫾ 0.00
1.04 ⫾ 1.51 1.38 ⫾ 1.58 2.62 ⫾ 1.68 3.65 ⫾ 1.17 4.22 ⫾ 1.09 5.00 ⫾ 0.00 5.00 ⫾ 0.00
0.35 ⫾ 1.06 0.73 ⫾ 1.31 1.20 ⫾ 1.12 2.41 ⫾ 1.70 3.44 ⫾ 1.42 5.00 ⫾ 0.00 5.00 ⫾ 0.00
Continuous Cervical Paravertebral Brachial Plexus Block
and the high number of patients who would prefer the same analgesic regimen for possible future surgery, CCPVB appears to provide satisfactory postoperative analgesia for patients undergoing shoulder surgery. However, the block may not be fully adequate in cases of shoulder arthroplasty, large rotator cuff repair surgery involving the subscapularis muscle, “open” acromioclavicular joint surgery, and major open stabilization procedures, since all the patients who were not satisfied with the analgesia underwent one of these “open” procedures. This finding may possibly be related to the CCPVB being performed with a 30-mL bolus and 3 to 10 mL/h rates of continuous local anesthetic infusion. This may not be enough to reach the origin or branches of the superficial cervical plexus nerves that provide sensory innervation to the skin and subcutaneous tissue around the shoulder. If so, a superficial cervical plexus block would be an easy adjuvant to CCPVB for major open shoulder surgery. CCVBP was also ineffective in the single case of latissimus dorsi muscle transfer surgery in the series, probably because the spread of local anesthetic was limited to the brachial plexus roots and did not reach the upper thoracic nerve roots. The upper thoracic nerve roots should probably be blocked additionally if this type of surgery is contemplated. The postoperative pain management protocol for continuous infusion and drug therapy may have been less than optimal. Patient-controlled bolus injections, as described by Borgeat et al.9,10 and Ilfeld et al.,11 may enhance the patient satisfaction. An average of 2 (1-7) attempts at catheter placement was required, indicating that even after successful needle placement, the catheter does not necessarily follow a perineural course or remain in the desired position. Although studies comparing stimulating catheters with nonstimulating catheters for this particular block are not yet available, at least 1 author has shown a clear advantage for stimulating catheters in other nerve blocks.12 We do not adhere to the concept of “volume distension” of the paravertebral space with saline before catheter placement as propagated by PhangDam et al.12 because saline in this space will dissipate electrical current and make nerve stimulation via the catheter impossible or will require much higher current (as PhangDam et al. noted). Loss of resistance to air may be less obvious in the cervical paravertebral block than during epidural needle placement. Catheter advancement was also associated with more resistance as compared with epidural catheter placement. The importance of preserving motor function has not been clearly defined by orthopedic surgeons
•
Boezaart et al.
411
and physical therapists. In cases such as rotator cuff repair and shoulder arthroplasty, early motor function can be detrimental in shoulder surgery, especially if motor function is complete and the shoulder joint is free of pain. On the other hand, early return of motor function may be advantageous in cases of nonstructural surgery, such as capsulotomy for adhesive capsulitis (“frozen shoulder”). The continuous cervical paravertebral block provided ideal conditions for early return of motor function in hand, wrist, and elbow, whereas shoulder motor function returned later and simultaneously with the return of shoulder sensation. Early return of motor function added to patient comfort, and 92% of these patients did not complain of an uncomfortable “dead feeling” in the arm. Preservation or early return of motor function with this block is probably because the catheter is placed at the level of the roots of the brachial plexus. At this level, the posterior sensory fibers and anterior motor fibers have not yet combined. Because a posterior approach is used for this block, the sensory fibers are encountered first and relatively small infusion rates may spare the more “distant” motor fibers. The initial motor block gradually diminished from the inferior roots (hand, wrist, and elbow) upward toward the fifth cervical root (shoulder) where the catheter was initially placed, as the effects of the initial ropivacaine bolus wore off. This may also explain why some patients reported sensory electrical pulsations in the arm before a motor response was elicited and also why higher than customary current outputs were required to elicit a motor response in some cases. The complications observed with continuous cervical paravertebral block tended to occur in the immediate postoperative period while the effects of the main bolus dose of local anesthetic was present. Horner’s syndrome was common, whereas dyspnea was a rare complication. Although phrenic function was not specifically studied in this series, dyspnea was probably because of phrenic nerve paresis. Dyspnea resolved in all patients in whom it occurred within 6 to 12 hours and was alleviated by simple measures. Phrenic nerve paresis should theoretically have the same prevalence (20%) as after observed with continuous blocks via the modified lateral or longitudinal interscalene approach.8 Hoarseness from recurrent laryngeal nerve block mainly followed right-sided paravertebral block and resolved within 12 hours in all cases. Macintosh and Mushin13 remarked that the posterior approach to the brachial plexus as originally proposed by Kappis in 1912 was uniformly successful, but this block never gained popularity, probably because of the pain associated with it. We have
412
Regional Anesthesia and Pain Medicine Vol. 28 No. 5 September–October 2003
recently speculated that the pain may be associated with the penetration of the tender extensor muscles of the neck by the needle and catheter [3]. The continuous cervical paravertebral block was designed with this in mind, and the needle entry is accordingly from the apex of the “V” formed by the trapezius and levator scapulae muscles to avoid the penetration of the posterior extensor muscles of the neck. This may explain the low incidence of posterior neck pain reported by the patients in this series. Nadig et al.14 recently criticized the continuous paravertebral block that Vranken et al.15 performed for the management of pain in patients with terminal cancer. Apart from stressing the posterior neck pain associated with this technique, they also argued that nerve root penetration, subdural or extradural spread of local anesthetic agent, and bleeding because of needle placement could be problematic. In the present series, we did not observe nerve injury or subdural injection, although epidural spread was demonstrated by MRI in 4 patients and suspected in 7 more. Epidural spread was not clinically significant. Subdural injection is possible but was not encountered, perhaps because we used a relatively large 17-gauge blunt Tuohy needle and a bullet-tipped catheter. Dural penetration should not be more prevalent with this block than with the classical anterior interscalene approach. Subclavian artery penetration occurred in 3 patients. In 1 of the cases, a nerve stimulator was not used because of postoperative placement of the block and the bony structures were missed laterally. In the remaining 2 patients, the nerve stimulator was later found to be defective. A functioning nerve stimulator is thus essential because of the close anatomical proximity of the subclavian artery anterior and inferior to the brachial plexus at this level. The shoulder muscles will twitch when the brachial plexus is encountered, and this should occur before the subclavian artery is reached. That blood was not aspirated during arterial puncture was probably because of the stylet inside the needle. If the stylet was removed when the bony structures were encountered, and a loss-of-resistance-to-air syringe was applied to the needle, blood vessel penetration may have been apparent. Positioning the patient in the sitting position or with a 5° to 10° angle head-up tilt when the lateral decubitis position is used would probably cause collapse of the veins around this area, which could explain the absence of venous bleeding observed in the present study. Borgeat et al.5 recently reported a 14% incidence of paresthesia, dysesthesia, and arm pain not related to surgery 10 days after conventional continuous interscalene block, which decreased to 3.9%
and 0.9% after 3 and 6 months, respectively. We speculate that this might in part be because of the catheter placement and drug infusion where the brachial plexus passes over the first rib through the narrow space between the first rib and the clavicle. The more medial paravertebral approach avoids this area, which possibly explains the absence of temporary or permanent nerve damage in the present study when compared with the findings of Borgeat et al.5 Although some anesthesiologists avoid performing nerve blocks in patients during general anesthesia,16 our limited data suggest that placing the blocks in anesthetized patients diminished discomfort without adding risk. However, until definitive evidence to the contrary is presented, cervical paravertebral block should be performed in nonanesthetized patients whenever possible. Severely painful shoulder conditions, anxiety, and blocks in children are potential exceptions, but the patients or their guardians should be informed about the possible added risk of nerve damage when nerve blocks are performed under general anesthesia. Nerve damage unrelated to anesthetic technique or regional anesthesia has been reported in as many as 30% of arthroscopic shoulder operations,17 4% after shoulder arthroplasty,18 2.6% in rotator cuff repair,19 and 8.2% in patients who underwent surgery for anterior glenohumeral instability.20 All the surgical procedures in the present series were performed by the same shoulder surgeon (JFdB) experienced in arthroscopic, rotator cuff, arthroplasty, and instability surgery. It is therefore possible that the present results may not be generally applicable. Among the limitations of our study, the main one is that CCPVB was not compared with conventional continuous interscalene block. A study to compare these 2 blocks is underway. Another limitation is the lack of data on patient-controlled intravenous analgesia, which is probably a more objective measurement of acute pain. Determining phrenic nerve function or proprioception would have greatly enhanced our study. It would also have been valuable to evaluate CCPVB as a sole anesthetic for shoulder surgery, especially for arthroscopic shoulder surgery. The use of patient-controlled bolus injections as propagated by Borgeat et al9,10 and others11 would probably have added value to this study. In summary, we conclude that CCPVB appears to provide satisfactory analgesia for pain associated with shoulder surgery. If used for “open” shoulder surgery, larger volumes of local anesthetic agents should probably be used for the bolus injection and continuous infusions, and an additional superficial cervical plexus block should perhaps be performed. Our data show that CCPVB preserves motor func-
Continuous Cervical Paravertebral Brachial Plexus Block
tion of the distal parts of the arm or facilitates early return of such function. Complications occurred in the early postoperative period when the effects of the initial bolus were still present and the patients were still in the care of the anesthesiologist. Nerve damage or other serious complications were not observed in this limited series. Further research is required to compare this block with the conventional anterior approach to the continuous interscalene block and apply it for other major upper limb surgery or as sole anesthetic for shoulder and upper limb surgery.
9.
10.
11.
Acknowledgment
12.
The authors are indebted to Robert Koorn, MD, Richard W. Rosenquist, MD, and Steven C. Borene, MD, for their critical review of this manuscript.
13.
References
14.
1. Boezaart AP, de Beer JF. Continuous low cervical paravertebral block for shoulder surgery. Reg Anesth Pain Med 2001;26:68(abstr). 2. Boezaart AP. Continuous interscalene block for ambulatory shoulder surgery. Best Practice Res Clin Anaesth 2002;16:295-310. 3. Boezaart A P, Koorn R, Rosenquist RW. Paravertebral approach to the brachial plexus: an anatomic improvement in technique. Reg Anesth Pain Med 2003; 28:241-244. 4. Borene SC, Rosenquist RW, Koorn R, Haider N, Boezaart AP. An indication for continuous cervical paravertebral block (posterior approach to the interscalene space). Anesth Analg 2003;97:898-900. 5. Kappis M. Weitere Erfahrungen mit der Sympathektomie. Klin Wchnschr 1923;2:1441. 6. Borgeat A, Ekatodramis G, Kalberer F, Benz C. Acute and nonacute complications associated with interscalene block and shoulder surgery. Anesthesiology 2001;95:875-880. 7. Pippa P, Cominelli E, Marinelli C, Aito S. Brachial plexus block using the posterior approach. Eur J Anaesth 1990;7:411-420. 8. Boezaart AP, de Beer JF, du Toit C, van Rooyen K. A
15.
16.
17.
18.
19.
20.
•
Boezaart et al.
413
new technique of continuous interscalene nerve block. Can J Anesth 1999;46:275-281. Borgeat A, Schappi B, Biasca N, Gerber C. Patientcontrolled analgesia after major shoulder surgery: Patient-controlled interscalene analgesia versus patient-controlled analgesia. Anesthesiology 1997;87: 1343-1347. Borgeat A, Tewes E, Biasca N, Gerber C. Patientcontrolled interscalene analgesia with ropivacaine after major shoulder surgery: PCIA vs. PCA. Br J Anaesth 1998;81:603-605. Ilfeld BM, Morey TE, Enneking K. Continuous infraclavicular brachial plexus block for postoperative pain control at home. Anesthesiology 2002;96:12971304. PhangDam C, Kick O, Collet T, Gouin F, Pinaud M. Continuous peripheral nerve block with stimulating catheters. Reg Anesth Pain Med 2003;28:83-88. Macintosh RR, Mushin WW. Brachial plexus. In: Macintosh RR, Mushin WW, eds. Local Analgesia, 4th ed. Edinburgh: E&S Livingston Ltd, 1967:8. Nadig M, Ekatodramis G, Borgeat A. Continuous brachial plexus block at the cervical level using a posterior approach in the management of neuropathic cancer pain. Reg Anesth Pain Med 2002;27:446. Vranken JH, van der Vegt MH, Zuurmond WWA, Pijl AJ, Dzolijk M. Continuous brachial plexus block at the cervical level using a posterior approach in the management of neuropathic cancer pain. Reg Anesth Pain Med 2001;26:572-575. Benumof JL. Permanent loss of cervical spinal cord function associated with interscalene block performed under general anesthesia. Anesthesiology 200; 93:1541-1544. McFarland EG, O’Neill OR, Hsu CY. Complications of shoulder arthroscopy. J South Orthop Assoc 1997;6: 190-196. Lynch NM, Cofield RH, Silber PL, Hermann RC. Neurologic complications after total shoulder arthroplasty. J Shoulder Elbow Surg 1996;5:53-61. Mansat P, Cofield RH, Kersten TE, Rowland CM. Complications of rotator cuff repair. Orthop Clin North Am 1997;28:205-213. Ho E, Cofield RH, Balm MR, Hattrup SJ, Rowland CS. Neurological complications of surgery for anterior shoulder instability. J Shoulder Elbow Surg 1999;8: 266-270.