Influence of arm position on ultrasound visibility of the axillary brachial plexus

Abstract

BACKGROUND 

Contemporary axillary brachial plexus block is performed by separate injections targeting radial, median, ulnar and musculocutaneous nerve. These nerves are arranged around the axillary artery, making ultrasound visualisation sometimes challenging. In particular, the radial nerve can be difficult to localise deep to the artery.

OBJECTIVES 

The primary aim of this study was to investigate which arm position optimises the visibility of the radial nerve. Secondary aims were the visibility and position of the other nerves during varying arm positions.

DESIGN 

A prospective observational study.

SETTING 

University teaching hospital, November 2012.

PARTICIPANTS 

Twenty volunteers, recruited by an advertisement on the Department's bulletin board. Inclusion criterion age more than 18 years. Exclusion criteria: refusal of ultrasound examination, restricted shoulder movement, local infection, BMI greater than 30 kg m^–2.

INTERVENTION 

One anaesthesiologist performed bilateral ultrasound examinations of the axillary brachial plexus on 20 volunteers. Each arm was placed in different positions [shoulder (S) 90° or 180° abduction, elbow (E) 0° or 90° extension] and scans were performed proximally in the axilla, and additionally 5 cm distally to this point [proximal (P) vs. distal (D)], resulting in eight different scans stored for off-line analysis performed by two blinded anaesthesiologists.

MAIN OUTCOME MEASURES 

For radial, median, ulnar and musculocutaneous nerve, visibility was assessed on a six-point visibility scale. Distances and angles of the nerves relative to the axillary artery and distances relative to the skin were measured.

RESULTS 

No significant differences between arm positions were found in the visibility score of radial (P = 0.359) and musculocutaneous nerves (P = 0.073). Visibility of the median nerve was improved in positions S90°/E0°/D and S180°/E0°/P (P = 0.02). The ulnar nerve was more visible in position S180°/E 0°/P and D (P = 0.007). The greatest distance between artery and radial nerve was 7.4 ± 4.7 mm at an angle of 120 ± 14° in position S180°/E 0°/D.

CONCLUSION 

The visibility of the radial nerve was not improved by varying positions of the arm. S180°/E0° provided the best overall visibility and accessibility of nerves.

TRIAL REGISTRATION 

https://www.toetsingonline.nl/to/ccmo_search.nsf/Searchform?OpenForm Identifier: NL42116.018.12.

Introduction

During the last decade, the use of high-definition ultrasound has renewed interest in peripheral regional anesthesia.¹ Surprisingly, although ultrasound is used to directly target nerves and plexus, extremities are still most often positioned as if performing landmark-oriented approaches. These positions were generally based on dissectional anatomical studies. For example, the brachial plexus in the axillary region is approached with the extremity positioned as described by Winnie with shoulder and elbow in 90°.² However, because of the mobility of the shoulder, the brachial plexus at the axillary level is particularly susceptible to rearrangement of its structures according to position.

Axillary brachial plexus block is one of the most commonly used methods of regional anesthesia.³ Separate blockade of the four main constituent nerves (radial, median, ulnar, musculocutaneous) significantly increases success rate.⁴ These nerves are arranged around the axillary artery within a neurovascular sheath. The position of the nerves inside the sheath is not fixed and allows a certain extent of movement. Furthermore, fibres are to a variable degree, exchanged between individual nerves.⁵ Anatomy in the axillary fossa is variable,⁶ which may render axillary brachial plexus block by single nerve blockade more difficult. Moreover, at the level of intersection of pectoralis major and biceps humeri muscles, wherein the axillary brachial plexus block is usually performed, only the radial, median and ulnar nerves are consistently found within the common neurovascular sheath. The musculocutaneous nerve usually separates more proximally and is usually found between the biceps and coracobrachialis muscles.^1,7 Due to its anatomic position deep to the axillary artery, the radial nerve is the most difficult to visualise when using ultrasound.⁸

Determining the optimal position of the arm for visualisation of the radial nerve during the performance of an ultrasound-guided axillary brachial plexus block might promote efficiency and safety. Thus, we investigated the influence of arm positioning on the sonoanatomy of the axilla and the visibility of the nerves most proximal in the axilla and 5 cm distally to this point. The primary objective of the study was to assess the ultrasound visibility of the radial nerve at two levels in four different arm positions. Secondary objectives were visibility, position and distance of all four nerves to the brachial artery and the skin.

Materials and methods

Following Ethical Committee approval by Medical Ethical Committee of Academic Medical Centre in Amsterdam on 05 November 2012 and registration in the national CCMO register (NL42116.018.12), this prospective observational study was conducted at the Department of Anaesthesiology of the Academic Medical Centre (AMC) Amsterdam in November 2012.

Volunteers

Volunteers were recruited by placing an advertisement on the Department's bulletin board. Inclusion criterion was age more than 18 years. Exclusion criteria were refusal of ultrasound examination, restriction in shoulder movement, local infection and BMI greater than 30 kg m^–2. After obtaining written and informed consent from each volunteer, data such as sex, age, size and bodyweight as well as handedness were collected.

Ultrasound examinations

All examinations were performed by one anaesthetist (V.F.) experienced in regional anaesthesia, using one ultrasound machine (M-Turbo; Sonosite; Bothell, Washington, USA) with a linear multifrequency probe 13-6 MHz (HFL38X; Sonosite). After a short introduction and explanation of the procedure, volunteers were placed supine for a bilateral ultrasound examination of the axillary region. Depth and gain were optimised for each volunteer and ‘resolution mode’ was selected on the ultrasound machine. The probe was placed perpendicular to nerves, artery and humerus, in order to get a good short axis view of the nerves. To avoid shifting of the nerves during scanning, minimal probe pressure was exerted on the skin with only light compression of veins. Each arm was placed in four different positions: shoulder 90°/elbow 90° (=S90/E90); shoulder 90°/elbow 0° (=S90/E0); shoulder 180°/elbow 90° (=S180/E90); and shoulder 180°/elbow 0° (S180/E0). In these four positions, scans were performed at two levels: Proximal level (P): at the intersection between the lower border of the pectoralis major muscle and the biceps brachii muscle (marked as proximal). Distal level (D): Five centimetres distally from the first level (marked as distal). In all positions, the forearm was kept in a neutral position midway between pronation and supination.

Thus, eight different scans of each axilla were performed, results are summarised in Table 1.

Table 1: Eight positions for scanning the axillary plexus

During each scan, a 4-s long video clip was captured, saved and encrypted for subsequent blinded viewing and assessment.

Image assessments and measurements

After completion of all examinations, video clips were assessed independently by two blinded assessors (M.F.S, J.T.W.) experienced in regional anaesthesia. In each clip, radial, median, ulnar and musculocutaneous nerves were assessed on visibility using a six-point visibility scale:

1. 0, no nerve identified,
2. 1, nerve identified with a high probability,
3. 2, nerve identified, but most of it not visible,
4. 3, nerve identified, more than 50% of its borders can be precisely distinguished from surrounding structures,
5. 4, nerve completely visible, but fascicles poorly defined,
6. 5, nerve completely visible and multiple fascicles identifiable.

Any discrepancy in visibility scores was discussed afterwards and clips were reviewed in order to find a consensus for the score.

Distances from each nerve to the skin and to the artery and angle to the artery were measured. The shortest distances from nerves to skin and to the axillary artery were measured in millimetres. The centres of the axillary artery and of each nerve were reference points for angle measurement (degrees). In cases wherein the nerve was not visible on the clip, distances and angles were not recorded. All data obtained were entered in a computer spreadsheet (SPSS, Chicago, Illinois, USA) for statistical analysis. Mean measurements of distances and angles from the nerves to the artery were geometrically visualised.

Statistical analysis

Power analysis had revealed that to detect a clinical meaningful increase of visibility of the radial nerve of 30%, assuming a standard deviation of 20% with a power of 80 and an alpha of P value less than 0.001 (compensated for six comparisons), a group size of n = 18 would be required. Assuming a 10% dropout, we included 20 volunteers (nQuery Advisor 7.0, Janet D. Elashoff, Republic of Ireland). We suggested that a 30% increase in visibility would be clinically meaningful, and did some preliminary experiments to estimate the standard deviation to be expected.

Volunteer data are expressed as mean ± SD or median and range, where appropriate. Visibility scores of the radial, median, ulnar and musculocutaneous nerve, distances to the axillary artery, distances to the skin and angle with respect to artery are represented as mean ± SD in each of the eight scan positions. One-way repeated measures analysis of variance was used to compare visibility scores in different scan positions and validated by Mauchly's sphericity test to reduce the likelihood of type I errors. Therefore, visibility scores were taken at interval level. Posthoc Bonferroni analysis was used to account for the multiple comparisons where appropriate. A value of P less than 0.05 was considered to be statistically significant. Statistics were calculated with use of SPSS 20.0 for Windows (SPSS, Chicago, Illinois, USA).

Results

Participant flow

Twenty volunteers were recruited in November 2012. All volunteers signed written informed consent without any dropout. None of the volunteers experienced any harm or discomfort during the examinations. Data from the volunteers are presented in Table 2.

Table 2: Volunteer characteristics

Visibility scores

For analysis of visibility scores, 320 video clips were captured from 40 axillary regions of 20 volunteers in eight different scan positions. Reviewers agreed primarily on the scores in 92% of the cases; in the rest of the cases, an agreement could easily be found on reviewing the videos. Mean ± SD visibility score in eight different scan positions are shown in Fig. 1 . We failed to identify the radial nerve in 10% of the clips (visibility score = 0) in scan position S180°/E90°/D, in 12.5% of cases in scan position S180°/E0°/P and in 22.5% in scan positions S90°/E90°/P, S90°/E0°/P and S90°/E0°/D. No significant differences in visibility score of the radial (P = 0.359) and musculocutaneous nerve (P = 0.073) were found among the eight scan positions, whereas significant differences were found in visibility of the median (P = 0.02) and ulnar nerve (P = 0.007). Posthoc testing demonstrated significantly improved visibility of median nerve in scan positions S90°/E0°/D and S180°/E0°/P compared with the ‘classical’ position of S90°/E90°/P (Fig. 1 b). Visibility of the ulnar nerve was significantly better in positions S180° than S90° (except S180°/E90°/D) (Fig. 1 c).

Fig. 1: Visibility scores of the radial nerve (a), median nerve (b), ulnar nerve (c) and musculocutaneus nerve (d) in eight different positions, represented in corresponding colour (see Table 1). Visibility score is 0 to 5, where score 0 represents ‘nerve not visible’ and score 5 represents ‘nerve completely visible’. ^* P < 0.05, ^** P < 0.005.

Fig. 1: (Continued). Visibility scores of the radial nerve (a), median nerve (b), ulnar nerve (c) and musculocutaneus nerve (d) in eight different positions, represented in corresponding colour (see Table 1). Visibility score is 0 to 5, where score 0 represents ‘nerve not visible’ and score 5 represents ‘nerve completely visible’. ^* P < 0.05, ^** P < 0.005.

Orientation of the nerves to the artery and the skin in different positions

Mean positions of the nerves in relation to the artery in the eight different scan positions with the artery as reference points are geometrically demonstrated in a transverse view in Fig. 2 .

Fig. 2: Geometric presentation of the distances (mm) and angles (degrees) of the radial nerve (a), median nerve (b), ulnar nerve (c) and musculocutaneus nerve (d) in relation to the brachial artery in eight different positions, represented in corresponding colour (Table 1). Lines represent SD of mean distance and curves represent SD of mean angle to the brachial artery. Brachial artery is the reference point on the X-axis and Y-axis. Note: These circles are just representations of the centres of the nerves, but much smaller than the actual nerves. Thus, in the original ultrasound pictures, the nerves appear much closer than in this schematic drawing.

Fig. 2: (Continued). Geometric presentation of the distances (mm) and angles (degrees) of the radial nerve (a), median nerve (b), ulnar nerve (c) and musculocutaneus nerve (d) in relation to the brachial artery in eight different positions, represented in corresponding colour (Table 1). Lines represent SD of mean distance and curves represent SD of mean angle to the brachial artery. Brachial artery is the reference point on the X-axis and Y-axis. Note: These circles are just representations of the centres of the nerves, but much smaller than the actual nerves. Thus, in the original ultrasound pictures, the nerves appear much closer than in this schematic drawing.

Radial nerve

Significantly greater distances between the radial nerve and the artery were found in shoulder positions of 180° with accompanied angles of 108°–124°, whereas the smallest distance was found in position shoulder 90°, elbow 90° or 0°, distal (P < 0.001). Smallest distance to the skin of 6.9 mm (2.6) was found in a position shoulder 180°, elbow 0°, proximal and greatest distance was 13.2 mm (3.5), found in a position shoulder 90°, elbow 90°, distal (P < 0.001).

Median nerve

No significant differences in distance from the median nerve to the artery were found, ranging from 9.9 ± 5.9 to 15.0 ± 12.9 mm in different scan positions. However, significant differences were found in distances to the skin (P < 0.001) with the smallest distance of 3.0 ± 1.4 mm in position shoulder 90°, elbow 0°, proximal and a greatest distance of 5.4 ± 2.1 in position shoulder 90°, elbow 90°, distal.

Ulnar nerve

No significant differences in distance from the ulnar nerve to the artery were found at different scan positions, ranging from 2.7 ± 4.0 to 4.3 ± 3.9 mm, whereas distance to the skin was greatest in position shoulder 90°; elbow 0°; distal (5.3 ± 1.8 mm) and smallest in position shoulder 180°; elbow 90°; proximal (P < 0.001).

Musculocutaneous nerve

Distance to the artery was greatest in position shoulder 90°, elbow 90°, distal (13.9 ± 5.4 mm) and smallest in position shoulder 180°, elbow 0°, proximal (9.0 ± 4.4 mm, P < 0.001). Distance to the skin was greatest in position shoulder 180°, elbow 90°, distal (12.6 ± 4.2 mm) and smallest in position shoulder 90°, elbow 90°, proximal (9.8 ± 3.4 mm) and in position shoulder 90°, elbow 0°, proximal (9.8 ± 3.2 mm, P < 0.001).

Discussion

Ultrasonographic visibility of the radial nerve was not influenced by changing the position of the shoulder, elbow or the scan level at the axilla. Consequently, the primary aim of this study – to determine the optimal arm and scan position for improved visibility of the radial nerve – was not reached by any of the eight positions.

However, within individuals, analysis demonstrated a significantly improved visibility of the median nerve when scanning in position shoulder 180°; elbow 0°; proximal or the position shoulder 90°; elbow 0°; distal. The arm position providing optimal overall visibility is shoulder 180°; elbow 0°; proximal.

Ustuner et al.⁹ recently demonstrated a considerable variability of the anatomy of the axillary brachial plexus in 60 patients undergoing plexus block. These results resemble our results, although the parametric representation of the position we give in Fig. 2 underestimates the actual variation. This is because we not only wanted to describe the variability of the axillary brachial plexus, but wanted to investigate the influence of arm position on nerve visibility and location. Thus, we also saw a wide variation in distances and angles from the radial nerve to the artery (Fig. 2 a), although the varying positions of the arm and scan levels did not influence the visibility of the nerve. This may be due to the fact that the radial nerve is often obscured deep to the axillary artery by dorsal enhancement and lateral shadowing artefacts. Whether this improved visibility is due to anatomical rearrangement of the plexus or might be caused by veins or other tissues being compressed in varying positions cannot be delineated from the current results. In the position shoulder 90°; elbow 0 °; proximal, the radial nerve is most often vertically located under the artery, making invisibility of the radial nerve very likely due to acoustic enhancement of the artery (Fig. 2 a). The distance from the radial nerve to the artery was greatest when the shoulder was abducted 180° and the scan performed distal in the axilla, theoretically reducing the risk of a deterioration in visibility as a result of artefacts induced by the artery.¹⁰ However, this scan position did not improve the visibility of the radial nerve. Aside from the influence of the radial artery, there are several other possible causes of poor to moderate visibility of the radial nerve. The radial nerve does not travel parallel to the artery and the skin, but after passing superficial to the latissimus dorsi tendon and teres major muscles in the axilla, it runs diagonally in the fascial plane of the long and median head of the triceps, and spirals obliquely across the posterior surface of the humerus.¹¹ Thus, scanning of the brachial plexus and artery in the short axis at the level of the axillary artery is not perpendicular to the axis of the radial nerve, resulting in poor reflection and visualisation. Moreover, muscular branches to the heads of the triceps arise from the radial nerve at the level of the axilla and proximal humerus in highly variable numbers and levels, resulting in individual differences in visibility scores.¹² In a similar study of the sciatic nerve undertaken in the popliteal fossa, visualisation of the division of the sciatic nerve into the tibial nerve and common peroneal nerve was difficult because of differences in angulation, direction, depth and internal architecture of nerve tissue.¹³

Our results are partly in agreement with findings obtained by Wong et al.,¹⁴ who found it impossible to visualise the radial nerve in two of 48 patients. In contrast to our current study, the latter examination was not performed in standardised views, and radial nerve identification was verified by nerve stimulation.¹⁴ The examiners identified nerves purely by watching films of short ultrasound sweeps. Although the examiners have a long expertise in ultrasound guided regional anaesthesia, only scanning of a nerve over a longer distance or possibly using nerve stimulation can identify a nerve with a higher degree of certainty. The examiners therefore were cautious in identifying nerves and could not identify 10 to 22% of the cases, whereas Wong et al.¹⁴ were able to identify 95% of the nerves correctly by ultrasound. Furthermore, even when the observers identified a nerve, the visibility scoring system quantifies the level of uncertainty. Thus, the investigation resembled the clinical situation, when the anaesthesiologist identifies a nerve on the basis of the ultrasound picture.

The percentage of between 10 and 22.5% of the video-clips wherein the radial nerve was not identifiable seems high; however, for methodological reasons, the scans were very standardised and very little movement was allowed to identify the nerve. Thus, it is not surprising that the rate of identification is lower than when free scanning is possible. One may argue that nerve identification just by looking at short clips is fallible. Thus, transcutaneous identification might be an option in volunteers, but this has been shown in earlier studies to be almost of no value in locating the nerves where they are superficially located.¹⁵ Furthermore, even needle stimulation cannot be seen as a gold standard because the relationship of ‘distance of the nerve – stimulation threshold’ is highly variable.¹⁶

Although the visibility score of the radial nerve does not change significantly between the different arm positions, it is most superficial to skin and most distant to the artery at position shoulder 180°; elbow 0°; proximal. An additional benefit in this same position is the high visibility score of the median and ulnar nerve.

In addition, the location of the musculocutaneus nerve varied widely in differing arm positions, without affecting nerve visibility. Therefore, although this position (shoulder 180°; elbow 0°; proximal) does not increase the visibility of the radial nerve significantly, it seems to be the most advantageous for the integral visibility and location of all nerves.

Movement of the shoulder to 180° extends the coracobrachial muscle, and straightening the elbow extends the biceps muscles whilst shifting and decreasing the cross-section of the muscle layers. This is one of the reasons why in this arm position the nerves are most superficial to the skin and in the case of the median and ulnar nerve most visible.

Limitations

Before using this new position of the shoulder as a standard positioning in clinical practice, a few things must be considered. The scans taken were short and the probe was only marginally moved in order to have a high degree of standardisation. In clinical practice, the probe is variably moved according to the individual anatomy. Nevertheless, we could identify a position that results in the most superficial nerve depth and leads to the highest visibility of two of the nerves. This position may actually be uncomfortable or even impossible in some patients, especially elderly or those with systemic diseases such as rheumatoid arthritis. However, many patients undergoing operation of the forearm, wrist or hand have isolated disease without restriction of shoulder mobility. Furthermore, our volunteers were rather lean, young and healthy in comparison to a patient population undergoing operation of the forearm or hand. In conclusion, the population of our volunteers might not be representative for the average patient population and therefore our results cannot be generalised to all patient populations. Nevertheless, in clinical daily practice, it may be worth attempting to position the patient arm to shoulder 180 °; elbow 0°, proximal to optimise the visibility of the median and ulnar nerve.

Furthermore, ultrasound is a dynamic technique and structures should never be identified by purely looking at frozen pictures, but following the path of structures. However, in order to standardise the technique of achieving films, we had to limit the sweeps to one of two locations and with very little movement. As the aim of the study was to investigate the influence of positioning of the arm on visibility and position of the nerves in the axilla, a compromise between standardising the technique for scientific reasons and using ultrasound as a dynamic technique was accepted.

We used a six-point scale for the nerve visibility score, whilst a four-step scale was used in the study of Wong et al.¹⁴ and in a study about the visualisation of the sciatic nerve.¹⁷ However, this is not a standardised or validated scale for ultrasonic nerve visibility. When planning the study, we did not find any other scoring system in the literature. Validation of a scoring system would be restrained by incongruity of examiner's proficiency, the extent of the sweeps, taken pictures, the settings and quality of the ultrasound machine, the anatomical structure visualised and the volunteer or patient population. Therefore, even a subjective scoring system will always be sensitive to many influences. However, the examiners judged most videos equally and could easily agree on those videos with an initial discrepancy. Other studies have been performed to determine a recommended patient position on the basis of nerve to skin distances measured with ultrasound for an infraclavicular block.¹⁸ In a study by Bigeleisen and Wilson¹⁹, success rate, performance and onset time were measured in patients undergoing a supraclavicular block. We did not determine these variables, as our study was performed in volunteers to explore a recommended position for the axillary brachial plexus block in order to increase visibility of the often ‘problematic’ radial nerve. Although we could not identify a position that enhanced the visibility of the radial nerve, we did identify a position that increased the visibility of median and ulnar nerve, and exposed most nerves to a more superficial position more distant from the artery. It remains to be determined whether selecting for an arm position with good overall visibility will also translate into a clinically appreciable benefit, such as shorter time to block, or a higher block success rate.

In conclusion, an arm position with the shoulder in 180° abduction, elbow 0° and a proximal scan level tended to improve most aspects of the brachial plexus visibility on ultrasound. However, the ultrasonographic visibility of the radial nerve could not significantly be improved in any of the investigated positions.

Acknowledgements relating to this article

Assistance with the study: none.

Financial support and sponsorship: VF was sponsored by an ESA fellowship (Trainee exchange programme) during the time of the study. Further financing was done only by institutional sources.

Conflict of interest: none.

Presentation: preliminary data from this research were presented at 32nd annual congress of European Society of Regional Anaesthesia, Glasgow, UK, 04 to 07 September 2013.