- Email: [email protected]
Axillary brachial plexus block using peripheral nerve stimulator: A comparison between double- and triple-injection techniques
Original Articles
Axillary Brachial Plexus Block Using Peripheral Nerve Stimulator: A Comparison Between Double- and Triple-Injection Techniques Salvatore Sia, M.D., Antonella Lepri, M.D., and Paolo Ponzecchi, M.D. Background and Objectives: The multiple-injection technique for axillary block, in which the main 4 nerves of the plexus are located by a nerve stimulator and separately injected, has been shown to produce a high success rate. However, this technique may prove to be more difficult and time-consuming than other methods. Therefore, a simplified technique, with a reduced number of injections, might be desirable. A comparison between 2- and 3-injection techniques was made in the present double-blind study. Methods: One hundred patients were randomly allocated to 2 groups. In group 3N, the radial, median, and musculocutaneous nerves were located by a nerve stimulator and injections made. In group 2N, the radial and median nerves were located and injections made. Forty milliliters of local anesthetic was used. Results: A greater success rate for anesthetizing the musculocutaneous nerve was found in group 3N (98% v 80%; P ⬍ .005). No differences between the groups were found in the success rate for blocking the radial, median, and ulnar nerves. The rate of complete block (all the sensory areas distal to the elbow) was 90% in group 3N and 76% in group 2N. The time to perform the block was shorter in group 2N (5 ⫾ 1 v 6 ⫾ 1 minutes; P ⬍ .001). Conclusions: The 2-injection technique offers a success rate in blocking the 3 nerves innervating the hand similar to that obtained with the 3-injection technique. The latter approach should be considered when the musculocutaneous nerve distribution is involved in the surgical area. Reg Anesth Pain Med 2001;26:499-503. Key Words:
Anesthetic techniques, Regional, Brachial plexus, Axillary approach, Nerve stimulation.
T
he multiple-nerve stimulation technique for axillary block, in which the 4 distal nerves of the plexus are identified by a nerve stimulator and separate injections for each performed, has been shown to produce a high success rate and a rapid onset of block.1-5 However, because this technique requires more time to perform than other axillary approaches, a simplification of the technique might be desirable. We previously found that a 3-injection method in which the ulnar nerve was not purposely
See Editorial page 495
From the Department of Anesthesiology, Centro Traumatologico Ortopedico, Azienda Ospedaliera Careggi, Firenze, Italy. Accepted for publication April 19, 2001. Reprint requests: Salvatore Sia, M.D., Via Santelli, 41, 50134 Firenze, Italy. E-mail: [email protected] © 2001 by the American Society of Regional Anesthesia and Pain Medicine. 1098-7339/01/2606-0101$35.00/0 doi:10.1053/rapm.2001.25896
located provided a spread and a latency of sensory block equal to that obtained with a 4-injection technique.6 As fewer injections may be required to block the nerves in this region, this study was designed to compare the efficacy of the 2– and 3–stimulator-assisted injection techniques. Therefore, we compared the onset time and success rate of a 3-injection technique in which the radial, median, and musculocutaneous nerves were identified to a 2-injection technique in which the radial and the median nerves were identified.
Methods After written informed consent and institutional approval, 100 American Society of Anesthesiologists (ASA) physical status I and II patients, scheduled for elective surgery of the hand, wrist, or forearm under axillary brachial plexus block, were included in the study. All patients received fentanyl 1 g/kg and midazolam 20 g/kg intravenously (IV) 5 minutes before the block. The patients were randomly assigned to 1 of the following 2 groups:
Regional Anesthesia and Pain Medicine, Vol 26, No 6 (November–December), 2001: pp 499 –503
499
500
Regional Anesthesia and Pain Medicine Vol. 26 No. 6 November–December 2001
group 2N (n ⫽ 50) and group 3N (n ⫽ 50). The patients were placed supine, with the arm abducted 90° and bent at the elbow with the forearm supinated. The pulse of the axillary artery was palpated at the level of the major pectoral muscle crossing the axilla. A mixture of equal parts of 0.5% bupivacaine and 2% lidocaine was used. The subcutaneous tissue overlaying the artery was infiltrated with 4 mL of the solution before the block to anesthetize the medial cutaneous nerves of the arm and forearm. A 22-gauge 50-mm long, short-bevelled insulated needle (Stimuplex; Braun, Melsungen, Germany) was connected to the negative lead of the nerve stimulator (Stimuplex; Braun). The stimulating current was set to 0.5 mA and the stimulus frequency to 2 Hz. In group 2N, the needle was inserted superior (lateral) to the artery to identify the median nerve, and inferior (medial) to identify the radial nerve. Twenty milliliters of the solution was injected near each nerve. In group 3N, 10 mL of the solution was injected near the musculocutaneous nerve, 10 mL near the median nerve, and 20 mL near the radial nerve. All blocks were performed or supervised by the first author and assessed by an investigator unaware of group assignment. Pain associated with the injections was assessed after completion of the block using the visual analog scale (VAS). The time to perform primary block was defined as the time between the initial insertion of the needle to infiltrate the subcutaneous tissue overlying the artery and the removal of the insulated needle. The sensory onset time of the primary block was assessed in the areas supplied by 6 nerves (musculocutaneous, radial, median, ulnar, and medial cutaneous nerves of the arm and of the forearm) at 10, 20, and 30 minutes after the end of the initial block. Sensory loss was assessed via a 22-gauge needle and defined as analgesia (loss of pinprick) or anesthesia (loss of touch). The primary block was defined as complete when analgesia or anesthesia (surgical analgesia) was observed at 30 minutes in all the sensory areas below the elbow. After 30 minutes, in case of incomplete block, the unblocked nerve(s) implicated in the surgical site were blocked at the midhumeral, elbow, or wrist level and the sensory assessment continued 40 and 50 minutes after the primary block. Patients were declared ready for surgery when they had a complete block or, in case of incomplete block, surgical analgesia in all areas necessary for surgery. Latency time was considered as the time between the end of the primary block and the time when the patient was pronounced ready for surgery. Total anesthetic time was calculated by adding the times for block performance and latency. Primary block effectiveness was calculated as the
percentage number of patients in each group in which a complete block was obtained at 30 minutes. Primary block was also assessed at 30 minutes for each nerve separately. Secondary block effectiveness was calculated as percentage of patients in each group who had surgical analgesia after supplementary blocks. Motor block was assessed at 30 minutes and defined as complete (no movements against gravity), satisfactory (minor movements of the digits possible), or absent. Midazolam, in 1-mg increments, was administered IV to patients who requested sedation during surgery. Fentanyl, in 50-g increments, was administered IV in case of tourniquet pain. The incidence of adverse effects was noted. The incidence of acute nerve injury was evaluated at 48 hours. Neurological sequelae were recorded during the surgical follow-up visits at 10 and 30 days. The estimation of the necessary sample size was based on the results of our previous study,6 in which a 90% primary block effectiveness and a 97% rate of musculocutaneous nerve block were obtained using a 3-injection technique. A pilot study using a 2-injection technique was performed on 20 patients. An 80% primary block effectiveness was recorded and the musculocutaneous nerve was unblocked in all the cases of incomplete block. Assuming a primary block effectiveness of 90% against 80%, with an ␣ error of 5% and a  error of 20%, 195 patients would be required in each group. However, because we expected a difference between the groups only for the rate of the musculocutaneous nerve block and not for the rate of block of the other 3 nerves innervating the hand, we decided to choose as study outcome the frequency of blocking of the musculocutaneous nerve instead of the primary block effectiveness. Therefore, assuming a success rate of 97% against 80%, 47 patients in each group would be required to show significant difference at 80% power. Obviously, because only 50 patients in each group were enrolled in the study, the risk of  error for the primary block effectiveness cannot be excluded. Parametric variables were described as mean ⫾ SD and compared with Student’s t-test. Qualitative variables were described as number (percentage) in each category and analyzed with Fisher’s exact test. VAS score was described as median and ranges and analyzed with Mann-Whitney U-test. P ⬍ .05 was considered significant.
Results Demographics and type and duration of surgery were not significantly different between the groups (Table 1). The block characteristics are shown in
Axillary Block: Double or Triple Injection Table 1. Patient Characteristics and Type and Duration of Surgery
Age (yr) Sex (M/F) Weight (kg) Surgical site (forearm/wrist/hand) Duration of surgery (min) Tourniquet time (min)
Group 2N (n ⫽ 50)
Group 3N (n ⫽ 50)
43 ⫾ 10 29/21 65 ⫾ 12 8/10/32 66 ⫾ 22 52 ⫾ 19
40 ⫾ 14 26/24 62 ⫾ 11 12/10/28 72 ⫾ 26 60 ⫾ 20
NOTE. 2N, 2 injections; 3N, 3 injections. Data are expressed as mean ⫾ SD or n.
Table 2. The time to perform the block was significantly shorter in group 2N than in group 3N (5 ⫾ 1 v 6 ⫾ 1; P ⬍ .001). The primary block effectiveness was 90% in group 3N and 76% in group 2N (P ⫽ .054). A greater success rate for anesthetizing the musculocutaneous nerve was seen in group 3N (98% v 80%; P ⬍ .005). No differences were found for the rate of block of the 3 nerves innervating the hand (90% in both groups) (Table 3). Six group 2N and 4 group 3N patients underwent supplementary blocks. None of the patients experienced pain from the site of surgery; however, 2 patients in group 2N and 3 in group 3N reported tourniquet pain and fentanyl was administered to these 5 patients. Intraoperative sedation was requested by 6 patients in group 2N and 4 in group 3N. Immediate symptoms of intravascular injection were observed in 6 patients, 3 in each group; none of these patients required treatment. Unintentional elicitation of paresthesia was observed in 5 patients in group 2N and in 7 in group 3N. Neurologic complications were not observed.
Discussion The 4-stimulation technique, in which each distal nerve of the plexus is identified in the axilla and Table 2. Characteristics of the Block
Performance time (min) Onset time (min) Latency time (min) Total anesthetic time (min) Primary block effectiveness n (%) Secondary block effectiveness n (%) Motor block (complete/satisfactory) Venous puncture (n) VAS at block performance
Group 2N (n ⫽ 50)
Group 3N (n ⫽ 50)
5⫾1 16 ⫾ 8 21 ⫾ 10 26 ⫾ 11 38 (76) 50 (100) 35/15 5 1 (0-5)
6 ⫾ 1* 17 ⫾ 7 19 ⫾ 9 25 ⫾ 10 45 (90) 50 (100) 38/12 4 1 (0-5)
NOTE. 2N, 2 injections; 3N, 3 injections. Data are expressed as mean ⫾ SD, n (%), or median (range). Primary block effectiveness: percentage of patients in which a complete block was obtained at 30 minutes. Secondary block effectiveness: percentage of patients who had surgical analgesia after supplementary block. *P ⬍ .001.
•
Sia et al.
501
Table 3. Spread of Analgesia 30 Minutes After the Primary Block
All nerves All nerves but musculocutaneous Radial nerve Musculocutaneous nerve Median nerve Ulnar nerve Medial cutaneous arm nerve Medial cutaneous forearm nerve
Group 2N (n ⫽ 50)
Group 3N (n ⫽ 50)
38 (76) 45 (90) 46 (92) 40 (80) 48 (96) 49 (98) 40 (100) 40 (100)
45 (90) 45 (90) 47 (94) 49 (98)* 47 (94) 49 (98) 40 (100) 40 (100)
NOTE. Numbers of patients with percentages in parentheses. 2N, 2 injections; 3N, 3 injections. *P ⬍ .005.
injected with a quarter of the total local anesthetic mixture, provides a fast onset and a high success rate.1-5 However, this technique might prove to be more time-consuming than other approaches. The nerves are clustered closely together; therefore, selective block of individual nerves might be difficult to achieve because local anesthetic solution may permeate to adjacent nerves after earlier injections. Moreover, the withdrawal and redirection of the stimulating needle to elicit 4 different muscular twitches could increase patient discomfort. Therefore, a simplified technique with a reduced number of injections might be desirable. In a previous study, we found that a 3-injection method in which the ulnar nerve was not purposely located and 20 mL of solution was injected near the radial nerve provides a spread and a latency of sensory block equal to that obtained with a 4-injection technique.6 To further reduce the number of injections, we decided to study the effectiveness of a 2-injection technique. The 2 nerves located in group 2N were not chosen neither randomly (e.g., the 2 nerves located first) nor on the basis of the surgical site. We planned to inject with 20 mL of solution both the radial and median nerves. The radial nerve was chosen on the basis of our previous results.6 The injection near the median nerve was made to promote spread to the musculocutaneous nerve, which is distant from the remaining brachial plexus at this level. The results of Yamamoto et al.7 supported this choice. These investigators found that, when using a single-injection technique, the most reliable block of the musculocutaneous nerve was achieved when anesthetic was injected to the area supplied by the median nerve compared with the ulnar or radial nerves. The possibility that a different combination of blocked nerves or a different distribution of the solution might be equally or even more effective cannot be excluded. An axillary block is considered successful by most physicians when analgesia is present in all areas
502
Regional Anesthesia and Pain Medicine Vol. 26 No. 6 November–December 2001
subject to surgical interventions. This definition is sufficient clinically, but may imply a falsely high success rate and makes comparison difficult.2 Therefore, to standardize the criteria of success, we considered successful (complete block) only the primary blocks including all nerves to the forearm. To further standardize, a time of 30 minutes was arbitrarily chosen for the evaluation of the primary block, even if a delay might have extended the area of analgesia. In the present study, there was a higher but not statistically significantly improved primary block effectiveness in group 3N (90% v 76%; P ⫽ .054). This result was related to a greater success rate for anesthetizing the musculocutaneous nerve (97% v 80%; P ⬍ .005), while no differences between the groups in success rate for anesthetizing the 3 nerves innervating the hand were found. In fact, when the musculocutaneous nerve is not considered in the statistical analysis, the rate of successful block is 90% in both groups. The success rate for anesthetizing the median, ulnar, and radial nerves obtained in both groups was similar to that reported in other studies using a 4-nerve stimulation approach.1-5 The success rate for anesthetizing the musculocutaneous nerve in studies using a 4-injection approach1-5 is similar to that recorded in group 3N. Few studies have reported on the effectiveness of a different number of injections and on the importance of a separate location of the musculocutaneous nerve when using a multiple-injection technique with a nerve stimulator. The results of these studies are conflicting. Baranowski and Pither,8 using a technique in which 1 to 3 nerves were located, found that the more nerves detected, the higher the success rate of the block. They also found that the musculocutaneous nerve was blocked in 67% of the cases without a specific stimulation of the nerve, while the block of the nerve was unsuccessful in the 4 cases in which the nerve was identified and injected. Lavoie et al.9 found that stimulation of the musculocutaneous nerve and one other nerve innervating the surgical site gives the same success rate (no supplemental block needed) as stimulation of all 4 major nerves of the plexus. However, these results were not confirmed by Bouaziz et al.,10 who found that a complete block (all 4 nerves) was obtained in only 54% of patients in which the musculocutaneous plus another nerve were located and injected at the axilla, compared with 88% of patients in which all 4 nerves of the plexus were located at the midhumeral level. The results obtained by Bouaziz et al.10 in the 2-injection group (musculocutaneous plus another nerve) are similar to those obtained with a single-injection technique1,8,11 and might suggest, in our opinion, that
at least 2 injections near the brachial plexus sheath are needed to obtain a satisfactory rate of complete block. Inberg et al.,11 using a 2-injection technique in which the median plus radial or ulnar nerves (depending on the site of surgery) were located and injected, obtained a primary success rate (no need for supplementary block) of 92% at 40 minutes and an 88% success rate for the musculocutaneous nerve block. They stated that the incidence of incomplete block of the musculocutaneous nerve could be minimized with a 2-injection technique. Unlike these investigators, we believe that a selective location and injection of the musculocutaneous nerve permits a higher success rate for the block of this nerve. A recent study by Coventry et al.12 compared 2 neurostimulation techniques using 30 mL of local anesthetic solution. In one group, the musculocutaneous and median nerves were identified and injected, while in the other group the musculocutaneous, median, and radial nerves were used. Complete sensory block (all the nerves to the forearm) occurred in 53% and in 97% of patients in the 2-injection and 3-injection groups, respectively (P ⬍ .001). The results of this study are in agreement with ours. A single injection near the brachial plexus, even if associated with a selective block of the musculocutaneous nerve, does not produce a reliable brachial plexus block. Specific ulnar nerve localization and injection is not essential for the block of the plexus. A 3-injection technique in which the musculocutaneous, median, and radial nerves are located and injected offers a high success rate for block of all the nerves innervating the forearm and the hand. The time to perform the block was significantly shorter in group 2N than in group 3N (5 ⫾ 1 v 6 ⫾ 1 minutes; P ⬍ .001). However, a difference of 1 minute in the mean performance time could be considered clinically unimportant. In conclusion, a 2-injection technique using a nerve stimulator in which the radial and median nerves were injected offers a success rate in blocking the 3 nerves innervating the hand similar to that obtained with a 3-injection technique. The 3-injection technique should be considered when the musculocutaneous nerve distribution is involved in the surgical area.
References 1. Koscielniak-Nielsen ZJ, Stens-Pedersen HL, Knudsen Lippert F. Readiness for surgery after axillary block: Single or multiple injection techniques. Eur J Anaesthesiol 1997;14:164-171. 2. Koscielniak-Nielsen ZJ, Hesselbjerg L, Fejlberg V. Comparison of transarterial and multiple nerve stim-
Axillary Block: Double or Triple Injection
3.
4.
5.
6.
7.
ulation techniques for an initial axillary block by 45 ml of mepivacaine 1% with adrenaline. Acta Anaesthesiol Scand 1998;42:570-575. Koscielniak-Nielsen ZJ, Rotbøll Nielsen P, Loumann Nielsen S, Gardi T, Hermann C. Comparison of transarterial and multiple nerve stimulation techniques for axillary block using a high dose of mepivacaine with adrenaline. Acta Anesthesiol Scand 1999; 43:398-404. Koscielniak-Nielsen ZJ, Rotbøll Nielsen P, Risby Mortensen C. A comparison of coracoid and axillary approaches to the brachial plexus. Acta Anesthesiol Scand 2000;44:274-279. Sia S, Bartoli M, Lepri A, Marchini O, Ponzecchi P. Multiple-injection axillary brachial plexus block: A comparison of two methods of nerve localization– Nerve stimulation versus paresthesia. Anesth Analg 2000;91:647-651. Sia S, Bartoli M. Selective ulnar nerve localization is not essential for axillary brachial plexus block using a multiple nerve stimulation technique. Reg Anesth Pain Med 2001;26:12-16. Yamamoto K, Tsubokawa T, Shibata K, Kobayashi T.
8.
9.
10.
11.
12.
•
Sia et al.
503
Area of paresthesia as determinant of sensory block in axillary brachial plexus block. Reg Anesth 1995;20: 493-497. Baranowski AP, Pither CE. A comparison of three methods of axillary brachial plexus anaesthesia. Anaesthesia 1990;45:362-365. Lavoie J, Martin R, Tetrault JP, Cote` DJ, Colas MJ. Axillary plexus block using a peripheral nerve stimulator: Single or multiple injections. Can J Anaesth 1992;39:583-586. Bouaziz H, Narchi P, Mercier FJ, Labaille T, Zerrouk N, Girod J, Benhamou D. Comparison between conventional axillary block and a new approach at the midhumeral level. Anesth Analg 1997; 84:1058-1062. Inberg P, Annila I, Annila P. Double-injection method using peripheral nerve stimulator is superior to single injection in axillary plexus block. Reg Anesth Pain Med 1999;24:509-513. Coventry DM, Barker KF, Thomson M. Comparison of two neurostimulation techniques for axillary brachial plexus blockade. Br J Anaesth 2001;86:80-83.
Axillary Brachial Plexus Block Using Peripheral Nerve Stimulator: A Comparison Between Double- and Triple-Injection Techniques Salvatore Sia, M.D., Antonella Lepri, M.D., and Paolo Ponzecchi, M.D. Background and Objectives: The multiple-injection technique for axillary block, in which the main 4 nerves of the plexus are located by a nerve stimulator and separately injected, has been shown to produce a high success rate. However, this technique may prove to be more difficult and time-consuming than other methods. Therefore, a simplified technique, with a reduced number of injections, might be desirable. A comparison between 2- and 3-injection techniques was made in the present double-blind study. Methods: One hundred patients were randomly allocated to 2 groups. In group 3N, the radial, median, and musculocutaneous nerves were located by a nerve stimulator and injections made. In group 2N, the radial and median nerves were located and injections made. Forty milliliters of local anesthetic was used. Results: A greater success rate for anesthetizing the musculocutaneous nerve was found in group 3N (98% v 80%; P ⬍ .005). No differences between the groups were found in the success rate for blocking the radial, median, and ulnar nerves. The rate of complete block (all the sensory areas distal to the elbow) was 90% in group 3N and 76% in group 2N. The time to perform the block was shorter in group 2N (5 ⫾ 1 v 6 ⫾ 1 minutes; P ⬍ .001). Conclusions: The 2-injection technique offers a success rate in blocking the 3 nerves innervating the hand similar to that obtained with the 3-injection technique. The latter approach should be considered when the musculocutaneous nerve distribution is involved in the surgical area. Reg Anesth Pain Med 2001;26:499-503. Key Words:
Anesthetic techniques, Regional, Brachial plexus, Axillary approach, Nerve stimulation.
T
he multiple-nerve stimulation technique for axillary block, in which the 4 distal nerves of the plexus are identified by a nerve stimulator and separate injections for each performed, has been shown to produce a high success rate and a rapid onset of block.1-5 However, because this technique requires more time to perform than other axillary approaches, a simplification of the technique might be desirable. We previously found that a 3-injection method in which the ulnar nerve was not purposely
See Editorial page 495
From the Department of Anesthesiology, Centro Traumatologico Ortopedico, Azienda Ospedaliera Careggi, Firenze, Italy. Accepted for publication April 19, 2001. Reprint requests: Salvatore Sia, M.D., Via Santelli, 41, 50134 Firenze, Italy. E-mail: [email protected] © 2001 by the American Society of Regional Anesthesia and Pain Medicine. 1098-7339/01/2606-0101$35.00/0 doi:10.1053/rapm.2001.25896
located provided a spread and a latency of sensory block equal to that obtained with a 4-injection technique.6 As fewer injections may be required to block the nerves in this region, this study was designed to compare the efficacy of the 2– and 3–stimulator-assisted injection techniques. Therefore, we compared the onset time and success rate of a 3-injection technique in which the radial, median, and musculocutaneous nerves were identified to a 2-injection technique in which the radial and the median nerves were identified.
Methods After written informed consent and institutional approval, 100 American Society of Anesthesiologists (ASA) physical status I and II patients, scheduled for elective surgery of the hand, wrist, or forearm under axillary brachial plexus block, were included in the study. All patients received fentanyl 1 g/kg and midazolam 20 g/kg intravenously (IV) 5 minutes before the block. The patients were randomly assigned to 1 of the following 2 groups:
Regional Anesthesia and Pain Medicine, Vol 26, No 6 (November–December), 2001: pp 499 –503
499
500
Regional Anesthesia and Pain Medicine Vol. 26 No. 6 November–December 2001
group 2N (n ⫽ 50) and group 3N (n ⫽ 50). The patients were placed supine, with the arm abducted 90° and bent at the elbow with the forearm supinated. The pulse of the axillary artery was palpated at the level of the major pectoral muscle crossing the axilla. A mixture of equal parts of 0.5% bupivacaine and 2% lidocaine was used. The subcutaneous tissue overlaying the artery was infiltrated with 4 mL of the solution before the block to anesthetize the medial cutaneous nerves of the arm and forearm. A 22-gauge 50-mm long, short-bevelled insulated needle (Stimuplex; Braun, Melsungen, Germany) was connected to the negative lead of the nerve stimulator (Stimuplex; Braun). The stimulating current was set to 0.5 mA and the stimulus frequency to 2 Hz. In group 2N, the needle was inserted superior (lateral) to the artery to identify the median nerve, and inferior (medial) to identify the radial nerve. Twenty milliliters of the solution was injected near each nerve. In group 3N, 10 mL of the solution was injected near the musculocutaneous nerve, 10 mL near the median nerve, and 20 mL near the radial nerve. All blocks were performed or supervised by the first author and assessed by an investigator unaware of group assignment. Pain associated with the injections was assessed after completion of the block using the visual analog scale (VAS). The time to perform primary block was defined as the time between the initial insertion of the needle to infiltrate the subcutaneous tissue overlying the artery and the removal of the insulated needle. The sensory onset time of the primary block was assessed in the areas supplied by 6 nerves (musculocutaneous, radial, median, ulnar, and medial cutaneous nerves of the arm and of the forearm) at 10, 20, and 30 minutes after the end of the initial block. Sensory loss was assessed via a 22-gauge needle and defined as analgesia (loss of pinprick) or anesthesia (loss of touch). The primary block was defined as complete when analgesia or anesthesia (surgical analgesia) was observed at 30 minutes in all the sensory areas below the elbow. After 30 minutes, in case of incomplete block, the unblocked nerve(s) implicated in the surgical site were blocked at the midhumeral, elbow, or wrist level and the sensory assessment continued 40 and 50 minutes after the primary block. Patients were declared ready for surgery when they had a complete block or, in case of incomplete block, surgical analgesia in all areas necessary for surgery. Latency time was considered as the time between the end of the primary block and the time when the patient was pronounced ready for surgery. Total anesthetic time was calculated by adding the times for block performance and latency. Primary block effectiveness was calculated as the
percentage number of patients in each group in which a complete block was obtained at 30 minutes. Primary block was also assessed at 30 minutes for each nerve separately. Secondary block effectiveness was calculated as percentage of patients in each group who had surgical analgesia after supplementary blocks. Motor block was assessed at 30 minutes and defined as complete (no movements against gravity), satisfactory (minor movements of the digits possible), or absent. Midazolam, in 1-mg increments, was administered IV to patients who requested sedation during surgery. Fentanyl, in 50-g increments, was administered IV in case of tourniquet pain. The incidence of adverse effects was noted. The incidence of acute nerve injury was evaluated at 48 hours. Neurological sequelae were recorded during the surgical follow-up visits at 10 and 30 days. The estimation of the necessary sample size was based on the results of our previous study,6 in which a 90% primary block effectiveness and a 97% rate of musculocutaneous nerve block were obtained using a 3-injection technique. A pilot study using a 2-injection technique was performed on 20 patients. An 80% primary block effectiveness was recorded and the musculocutaneous nerve was unblocked in all the cases of incomplete block. Assuming a primary block effectiveness of 90% against 80%, with an ␣ error of 5% and a  error of 20%, 195 patients would be required in each group. However, because we expected a difference between the groups only for the rate of the musculocutaneous nerve block and not for the rate of block of the other 3 nerves innervating the hand, we decided to choose as study outcome the frequency of blocking of the musculocutaneous nerve instead of the primary block effectiveness. Therefore, assuming a success rate of 97% against 80%, 47 patients in each group would be required to show significant difference at 80% power. Obviously, because only 50 patients in each group were enrolled in the study, the risk of  error for the primary block effectiveness cannot be excluded. Parametric variables were described as mean ⫾ SD and compared with Student’s t-test. Qualitative variables were described as number (percentage) in each category and analyzed with Fisher’s exact test. VAS score was described as median and ranges and analyzed with Mann-Whitney U-test. P ⬍ .05 was considered significant.
Results Demographics and type and duration of surgery were not significantly different between the groups (Table 1). The block characteristics are shown in
Axillary Block: Double or Triple Injection Table 1. Patient Characteristics and Type and Duration of Surgery
Age (yr) Sex (M/F) Weight (kg) Surgical site (forearm/wrist/hand) Duration of surgery (min) Tourniquet time (min)
Group 2N (n ⫽ 50)
Group 3N (n ⫽ 50)
43 ⫾ 10 29/21 65 ⫾ 12 8/10/32 66 ⫾ 22 52 ⫾ 19
40 ⫾ 14 26/24 62 ⫾ 11 12/10/28 72 ⫾ 26 60 ⫾ 20
NOTE. 2N, 2 injections; 3N, 3 injections. Data are expressed as mean ⫾ SD or n.
Table 2. The time to perform the block was significantly shorter in group 2N than in group 3N (5 ⫾ 1 v 6 ⫾ 1; P ⬍ .001). The primary block effectiveness was 90% in group 3N and 76% in group 2N (P ⫽ .054). A greater success rate for anesthetizing the musculocutaneous nerve was seen in group 3N (98% v 80%; P ⬍ .005). No differences were found for the rate of block of the 3 nerves innervating the hand (90% in both groups) (Table 3). Six group 2N and 4 group 3N patients underwent supplementary blocks. None of the patients experienced pain from the site of surgery; however, 2 patients in group 2N and 3 in group 3N reported tourniquet pain and fentanyl was administered to these 5 patients. Intraoperative sedation was requested by 6 patients in group 2N and 4 in group 3N. Immediate symptoms of intravascular injection were observed in 6 patients, 3 in each group; none of these patients required treatment. Unintentional elicitation of paresthesia was observed in 5 patients in group 2N and in 7 in group 3N. Neurologic complications were not observed.
Discussion The 4-stimulation technique, in which each distal nerve of the plexus is identified in the axilla and Table 2. Characteristics of the Block
Performance time (min) Onset time (min) Latency time (min) Total anesthetic time (min) Primary block effectiveness n (%) Secondary block effectiveness n (%) Motor block (complete/satisfactory) Venous puncture (n) VAS at block performance
Group 2N (n ⫽ 50)
Group 3N (n ⫽ 50)
5⫾1 16 ⫾ 8 21 ⫾ 10 26 ⫾ 11 38 (76) 50 (100) 35/15 5 1 (0-5)
6 ⫾ 1* 17 ⫾ 7 19 ⫾ 9 25 ⫾ 10 45 (90) 50 (100) 38/12 4 1 (0-5)
NOTE. 2N, 2 injections; 3N, 3 injections. Data are expressed as mean ⫾ SD, n (%), or median (range). Primary block effectiveness: percentage of patients in which a complete block was obtained at 30 minutes. Secondary block effectiveness: percentage of patients who had surgical analgesia after supplementary block. *P ⬍ .001.
•
Sia et al.
501
Table 3. Spread of Analgesia 30 Minutes After the Primary Block
All nerves All nerves but musculocutaneous Radial nerve Musculocutaneous nerve Median nerve Ulnar nerve Medial cutaneous arm nerve Medial cutaneous forearm nerve
Group 2N (n ⫽ 50)
Group 3N (n ⫽ 50)
38 (76) 45 (90) 46 (92) 40 (80) 48 (96) 49 (98) 40 (100) 40 (100)
45 (90) 45 (90) 47 (94) 49 (98)* 47 (94) 49 (98) 40 (100) 40 (100)
NOTE. Numbers of patients with percentages in parentheses. 2N, 2 injections; 3N, 3 injections. *P ⬍ .005.
injected with a quarter of the total local anesthetic mixture, provides a fast onset and a high success rate.1-5 However, this technique might prove to be more time-consuming than other approaches. The nerves are clustered closely together; therefore, selective block of individual nerves might be difficult to achieve because local anesthetic solution may permeate to adjacent nerves after earlier injections. Moreover, the withdrawal and redirection of the stimulating needle to elicit 4 different muscular twitches could increase patient discomfort. Therefore, a simplified technique with a reduced number of injections might be desirable. In a previous study, we found that a 3-injection method in which the ulnar nerve was not purposely located and 20 mL of solution was injected near the radial nerve provides a spread and a latency of sensory block equal to that obtained with a 4-injection technique.6 To further reduce the number of injections, we decided to study the effectiveness of a 2-injection technique. The 2 nerves located in group 2N were not chosen neither randomly (e.g., the 2 nerves located first) nor on the basis of the surgical site. We planned to inject with 20 mL of solution both the radial and median nerves. The radial nerve was chosen on the basis of our previous results.6 The injection near the median nerve was made to promote spread to the musculocutaneous nerve, which is distant from the remaining brachial plexus at this level. The results of Yamamoto et al.7 supported this choice. These investigators found that, when using a single-injection technique, the most reliable block of the musculocutaneous nerve was achieved when anesthetic was injected to the area supplied by the median nerve compared with the ulnar or radial nerves. The possibility that a different combination of blocked nerves or a different distribution of the solution might be equally or even more effective cannot be excluded. An axillary block is considered successful by most physicians when analgesia is present in all areas
502
Regional Anesthesia and Pain Medicine Vol. 26 No. 6 November–December 2001
subject to surgical interventions. This definition is sufficient clinically, but may imply a falsely high success rate and makes comparison difficult.2 Therefore, to standardize the criteria of success, we considered successful (complete block) only the primary blocks including all nerves to the forearm. To further standardize, a time of 30 minutes was arbitrarily chosen for the evaluation of the primary block, even if a delay might have extended the area of analgesia. In the present study, there was a higher but not statistically significantly improved primary block effectiveness in group 3N (90% v 76%; P ⫽ .054). This result was related to a greater success rate for anesthetizing the musculocutaneous nerve (97% v 80%; P ⬍ .005), while no differences between the groups in success rate for anesthetizing the 3 nerves innervating the hand were found. In fact, when the musculocutaneous nerve is not considered in the statistical analysis, the rate of successful block is 90% in both groups. The success rate for anesthetizing the median, ulnar, and radial nerves obtained in both groups was similar to that reported in other studies using a 4-nerve stimulation approach.1-5 The success rate for anesthetizing the musculocutaneous nerve in studies using a 4-injection approach1-5 is similar to that recorded in group 3N. Few studies have reported on the effectiveness of a different number of injections and on the importance of a separate location of the musculocutaneous nerve when using a multiple-injection technique with a nerve stimulator. The results of these studies are conflicting. Baranowski and Pither,8 using a technique in which 1 to 3 nerves were located, found that the more nerves detected, the higher the success rate of the block. They also found that the musculocutaneous nerve was blocked in 67% of the cases without a specific stimulation of the nerve, while the block of the nerve was unsuccessful in the 4 cases in which the nerve was identified and injected. Lavoie et al.9 found that stimulation of the musculocutaneous nerve and one other nerve innervating the surgical site gives the same success rate (no supplemental block needed) as stimulation of all 4 major nerves of the plexus. However, these results were not confirmed by Bouaziz et al.,10 who found that a complete block (all 4 nerves) was obtained in only 54% of patients in which the musculocutaneous plus another nerve were located and injected at the axilla, compared with 88% of patients in which all 4 nerves of the plexus were located at the midhumeral level. The results obtained by Bouaziz et al.10 in the 2-injection group (musculocutaneous plus another nerve) are similar to those obtained with a single-injection technique1,8,11 and might suggest, in our opinion, that
at least 2 injections near the brachial plexus sheath are needed to obtain a satisfactory rate of complete block. Inberg et al.,11 using a 2-injection technique in which the median plus radial or ulnar nerves (depending on the site of surgery) were located and injected, obtained a primary success rate (no need for supplementary block) of 92% at 40 minutes and an 88% success rate for the musculocutaneous nerve block. They stated that the incidence of incomplete block of the musculocutaneous nerve could be minimized with a 2-injection technique. Unlike these investigators, we believe that a selective location and injection of the musculocutaneous nerve permits a higher success rate for the block of this nerve. A recent study by Coventry et al.12 compared 2 neurostimulation techniques using 30 mL of local anesthetic solution. In one group, the musculocutaneous and median nerves were identified and injected, while in the other group the musculocutaneous, median, and radial nerves were used. Complete sensory block (all the nerves to the forearm) occurred in 53% and in 97% of patients in the 2-injection and 3-injection groups, respectively (P ⬍ .001). The results of this study are in agreement with ours. A single injection near the brachial plexus, even if associated with a selective block of the musculocutaneous nerve, does not produce a reliable brachial plexus block. Specific ulnar nerve localization and injection is not essential for the block of the plexus. A 3-injection technique in which the musculocutaneous, median, and radial nerves are located and injected offers a high success rate for block of all the nerves innervating the forearm and the hand. The time to perform the block was significantly shorter in group 2N than in group 3N (5 ⫾ 1 v 6 ⫾ 1 minutes; P ⬍ .001). However, a difference of 1 minute in the mean performance time could be considered clinically unimportant. In conclusion, a 2-injection technique using a nerve stimulator in which the radial and median nerves were injected offers a success rate in blocking the 3 nerves innervating the hand similar to that obtained with a 3-injection technique. The 3-injection technique should be considered when the musculocutaneous nerve distribution is involved in the surgical area.
References 1. Koscielniak-Nielsen ZJ, Stens-Pedersen HL, Knudsen Lippert F. Readiness for surgery after axillary block: Single or multiple injection techniques. Eur J Anaesthesiol 1997;14:164-171. 2. Koscielniak-Nielsen ZJ, Hesselbjerg L, Fejlberg V. Comparison of transarterial and multiple nerve stim-
Axillary Block: Double or Triple Injection
3.
4.
5.
6.
7.
ulation techniques for an initial axillary block by 45 ml of mepivacaine 1% with adrenaline. Acta Anaesthesiol Scand 1998;42:570-575. Koscielniak-Nielsen ZJ, Rotbøll Nielsen P, Loumann Nielsen S, Gardi T, Hermann C. Comparison of transarterial and multiple nerve stimulation techniques for axillary block using a high dose of mepivacaine with adrenaline. Acta Anesthesiol Scand 1999; 43:398-404. Koscielniak-Nielsen ZJ, Rotbøll Nielsen P, Risby Mortensen C. A comparison of coracoid and axillary approaches to the brachial plexus. Acta Anesthesiol Scand 2000;44:274-279. Sia S, Bartoli M, Lepri A, Marchini O, Ponzecchi P. Multiple-injection axillary brachial plexus block: A comparison of two methods of nerve localization– Nerve stimulation versus paresthesia. Anesth Analg 2000;91:647-651. Sia S, Bartoli M. Selective ulnar nerve localization is not essential for axillary brachial plexus block using a multiple nerve stimulation technique. Reg Anesth Pain Med 2001;26:12-16. Yamamoto K, Tsubokawa T, Shibata K, Kobayashi T.
8.
9.
10.
11.
12.
•
Sia et al.
503
Area of paresthesia as determinant of sensory block in axillary brachial plexus block. Reg Anesth 1995;20: 493-497. Baranowski AP, Pither CE. A comparison of three methods of axillary brachial plexus anaesthesia. Anaesthesia 1990;45:362-365. Lavoie J, Martin R, Tetrault JP, Cote` DJ, Colas MJ. Axillary plexus block using a peripheral nerve stimulator: Single or multiple injections. Can J Anaesth 1992;39:583-586. Bouaziz H, Narchi P, Mercier FJ, Labaille T, Zerrouk N, Girod J, Benhamou D. Comparison between conventional axillary block and a new approach at the midhumeral level. Anesth Analg 1997; 84:1058-1062. Inberg P, Annila I, Annila P. Double-injection method using peripheral nerve stimulator is superior to single injection in axillary plexus block. Reg Anesth Pain Med 1999;24:509-513. Coventry DM, Barker KF, Thomson M. Comparison of two neurostimulation techniques for axillary brachial plexus blockade. Br J Anaesth 2001;86:80-83.