Extraplexus versus intraplexus ultrasound-guided interscalene brachial plexus block for ambulatory arthroscopic shoulder surgery: A randomized controlled trial

About the Authors:

Monica W. Harbell

Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Validation, Writing – original draft, Writing – review & editing

* E-mail: [email protected]

Affiliations Department of Anesthesia and Perioperative Care, University of California, San Francisco, California, United States of America, Department of Anesthesia and Perioperative Medicine, Mayo Clinic, Phoenix, Arizona, United States of America

ORCID logo https://orcid.org/0000-0003-4210-0942

Kerstin Kolodzie

Roles Formal analysis, Investigation, Methodology, Writing – review & editing

Affiliations Department of Anesthesia and Perioperative Care, University of California, San Francisco, California, United States of America, Department of Epidemiology and Biostatistics, University of California, San Francisco, California, United States of America

ORCID logo https://orcid.org/0000-0002-3395-6207

Matthias Behrends

Roles Formal analysis, Investigation, Methodology, Writing – review & editing

Affiliation: Department of Anesthesia and Perioperative Care, University of California, San Francisco, California, United States of America

C. Benjamin Ma

Roles Formal analysis, Investigation, Writing – review & editing

Affiliation: Department of Orthopedic Surgery, University of California, San Francisco, California, United States of America

Sakura Kinjo

Roles Data curation, Formal analysis, Investigation, Methodology, Writing – review & editing

Affiliation: Department of Anesthesia and Perioperative Care, University of California, San Francisco, California, United States of America

Edward Yap

Roles Data curation, Formal analysis, Investigation, Methodology, Writing – review & editing

Affiliations Department of Anesthesia and Perioperative Care, University of California, San Francisco, California, United States of America, Department of Anesthesia, Kaiser Permanente South San Francisco, South San Francisco, California, United States of America

Matthias R. Braehler

Roles Data curation, Formal analysis, Investigation, Methodology, Writing – review & editing

Affiliation: Department of Anesthesia and Perioperative Care, University of California, San Francisco, California, United States of America

ORCID logo https://orcid.org/0000-0001-8980-9524

Pedram Aleshi

Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – review & editing

Affiliation: Department of Anesthesia and Perioperative Care, University of California, San Francisco, California, United States of America

Introduction

Ultrasound-guided interscalene brachial plexus blocks provide effective analgesia after arthroscopic shoulder surgery and have been shown to reduce opioid consumption, decrease postoperative nausea and vomiting (PONV), improve patient satisfaction, and decrease recovery room length of stay [1–3]. There is interest in identifying the optimal location for local anesthetic administration during an interscalene block. The injection site chosen for any type of block can affect its quality, time to onset, and duration. For example, the injection of local anesthesia into the paraneural sheath during a popliteal sciatic nerve block can result in a faster onset and longer duration of the block [4]. Franco and colleagues identified a sheath surrounding the brachial plexus in a cadaveric study [5]. This has generated interest in comparing intraplexus injection (between C5 and C6 nerve roots within the sheath surrounding the brachial plexus) to extraplexus injections (outside of the brachial plexus sheath) for interscalene brachial plexus block. Previous studies that compared extraplexus to intraplexus injections have had conflicting results with one reporting no difference in block onset [6] and the other reporting a faster onset with the intraplexus approach, but increased paresthesia [7].

We performed a prospective, randomized study to investigate the efficacy and safety of an intraplexus injection between C5 and C6 nerve roots and an extraplexus injection anterior and posterior to the brachial plexus into the plane between the perineural sheath and scalene muscles. This study was registered at Clinicaltrials.gov, identifier: NCT01877330, submitted December 2012, but not publicly posted until June 2013.

Methods

After study approval from the Institutional Review Board at the University of California, San Francisco on January 16, 2013 (Study #12–10146), consent was informed and written informed consent was obtained from adult patients who were scheduled for arthroscopic shoulder surgery at an ambulatory surgery center between March 2013 and February 2016. All patients who were age ≥ 18 years, American Society of Anesthesiologists (ASA) Physical Status I to II classification, and scheduled for arthroscopic shoulder surgery by one of two surgeons were identified for possible inclusion in the study. Exclusion criteria included: inability to consent, non-English speaking, planned open shoulder surgery, any contraindication for regional anesthesia, such as allergy to local anesthetics, coagulopathy or severe thrombocytopenia, infection at puncture site, pre-existing neuropathy in operative limb, inability to abduct the shoulder to 90 degrees preoperatively, need for postoperative nerve function monitoring, pulmonary disease, low baseline oxygen saturation, dementia, patient refusal, and high preoperative opioid requirements.

Patients were randomly assigned to receive either an intraplexus or an extraplexus injection by computer-generated simple randomization. The allocation sequence was determined prior to patient enrollment and was concealed to all investigators and anesthesiologists involved in the trial. The random group assignment was stored in consecutively numbered sealed opaque envelopes which were only opened by the anesthesiologist performing the block after informed consent was signed. While the anesthesiologist performing the block was not blinded to group allocation once enrolled and randomized, the operating room team, post-anesthesia care team and research staff recording the sensory and motor data were blinded to group assignment.

Procedures

The interscalene brachial plexus blocks were performed preoperatively with the patient in the supine position with the head of bed elevated 30 degrees and the patient’s head slightly turned away from the operative side. For the block placement, patients received midazolam 0-2mg and/or fentanyl 0–100 mcg intravenously, according to the anesthesiologist’s discretion. Using sterile technique, a 6-13-MHz linear probe (LOGIQe, GE Healthcare, Wauwatosa, WI, USA) transducer was used to identify the C5-C6-C7 nerve roots of the brachial plexus. 2–3 mL of lidocaine 2% was used to anesthetize the skin using a 27-G needle. A 5 cm, 22-gauge insulated needle (SonoPlex Stim cannula, Pajunk, Geisingen, Germany) was used to perform a real-time ultrasound-guided interscalene nerve block using a posterior-to-anterior, in-plane needle insertion approach. A total of 25mL of ropivacaine 0.5% was injected in either the intraplexus or extraplexus location (S1 File). In the intraplexus group, the needle was advanced in a posterior-to-anterior direction through the middle scalene muscle, through the brachial plexus sheath in-between the C5-C6 nerve roots. Less than 1 mL of the local anesthetic was injected to confirm that the needle was positioned in the intraplexus location. The remainder of the local anesthetic was injected in divided doses. If the injection did not result in the expansion of the brachial plexus sheath, then it was repositioned so that the tip was within the brachial plexus sheath. In the extraplexus group, the needle was advanced above the C5 nerve root and anterior to the brachial plexus. One mL of local anesthetic was given to confirm the extraplexus location anterior to the brachial plexus sheath. A total of 12mL of ropivacaine 0.5% was administered in anterior to the brachial plexus. The needle was then withdrawn to the posterior surface of the brachial plexus and an additional 12mL of ropivacaine 0.5% was given. If any swelling of the brachial plexus sheath was seen, the needle was repositioned to the extraplexus location. All patients received less than 3 mg/kg of ropivacaine. Patients were instructed to notify the study team if they experienced any paresthesia during block placement. Paresthesia was defined as an abnormal sensation either during or after block placement that felt like “tingling, pins and needles, or electricity down the arm.” Anesthesia providers were instructed to stop the block procedure if a paresthesia occurred, to redirect the needle, and to not resume local anesthetic administration until the paresthesia resolved. The block performance time, number of needle passes, and paresthesias were recorded by the study team.

After block placement, sensory and motor examinations were performed at one minute intervals, until there was decrease in sensory exam or until the patient left the preoperative area for the operating room. Patients were assessed for time to complete loss of shoulder abduction by having the patient sit upright and attempt to abduct the arm to 90 degrees. Sensation was assessed using pinprick and comparing to the contralateral side in the sensory distributions of the brachial plexus (axillary, supraclavicular, musculocutaneous, median, ulnar and radial). The time when the sensation first started to feel less sharp compared to the contralateral, non-blocked side was recorded to the nearest minute. Patients were also assessed for signs of Horner’s syndrome, hoarseness and dyspnea post-block administration.

All patients then underwent a standardized general anesthetic with a propofol induction and laryngeal mask airway placement. Anesthesia was maintained using a combination of volatile anesthetics (sevoflurane or desflurane with oxygen/air mixture) and propofol infusion (25–100 mcg/kg/min). Patients received dexamethasone 4 mg IV after induction and ondansetron 4 mg IV prior to emergence as PONV prophylaxis. Fentanyl was administered intravenously during the case per the anesthesiologist’s discretion. Sensory exam was reassessed in the recovery room using pinprick. Pain scores were assessed using the Numerical Rating Scale (NRS) from 0–10, where zero is no pain and ten is the worst imaginable pain. Highest NRS pain score in recovery room, perioperative opioid and antiemetic use were recorded, as well as recovery room length of stay and incidence of PONV. Patients were counselled to take oral analgesics on a schedule for the first 24 hours post-operatively.

Patients were contacted by telephone on the day following surgery (at least 24 hours after block placement) to assess block duration, highest pain score at rest and with movement in the 24 hours after block placement, opioid consumption, post-discharge nausea/vomiting and patient satisfaction score. Pain scores were also recorded one week postoperatively.

Data collection and measurements

Study data were collected and managed using REDCap electronic data capture tools, which is a secure, web-based application designed to support data capture for research studies [8].

Efficacy outcomes.

The primary outcome of this study was the time to onset of motor blockade, which was defined as the time to complete loss of shoulder abduction. Secondary efficacy outcomes included time to onset of sensory blockade, block performance time, number of needle passes, block duration, block failure, perioperative opioid consumption, pain scores, recovery room length of stay, incidence of PONV and patient satisfaction. These outcomes are further defined in Table 1.

[Figure omitted. See PDF.]

Table 1. Outcome definitions.

https://doi.org/10.1371/journal.pone.0246792.t001

Total opioid consumption was calculated as total oral morphine equivalents, using a previously published equation [9,10] which is included in S2 File.

Safety outcomes.

The secondary outcomes relating to safety of the block included the incidence of paresthesia, Horner’s syndrome, dyspnea, hoarseness, as described in Table 1, as well as nerve injury, local anesthetic toxicity and infection at site of block injection. If patients reported any residual numbness or weakness during the postoperative day 1 phone call, then they were contacted daily by the study team until it resolved.

Sample size calculation

The total sample size for a two-sample comparison of survivor functions for the primary outcome variable of time to onset of motor block was calculated to be 104 patients per group (208 patients in total). In accordance with Freedman’s method, the number of events was calculated as 191 with a two-sided α level of 5%, power of 80% and a hazard ratio of 0.66. The total sample size was set at 208 patients to account for patients administratively censored. The proportional hazard assumption was assessed using the log-log plot.

Statistical analysis

Statistical analysis was performed using STATA 13.1 (StataCorp, College Station, TX). Data was analyzed by intention-to-treat. Data were reported as mean (SD) for continuous variables, as median (interquartile ranges) for non-normally distributed continuous variables and as count (percentage) for categorical variables. Time-to-event data was compared using the log-rank test. Categorical data was analyzed using Chi squared analysis. Comparison of means was performed using the Mann-Whitney U test.

Results

A total of 208 ASA I-II adults were enrolled in this study. Due to simple randomization, there were 105 patients in the extraplexus group and 103 patients in the intraplexus group (Fig 1). After enrollment and allocation, 3 patients in the extraplexus group had a small amount of local anesthetic deposited in the intraplexus space, while 8 in the intraplexus group had a small amount of local anesthetic deposited in the extraplexus space. These patients were analyzed in an intention-to-treat manner. There were two patients (1 in extraplexus group, 1 in intraplexus group) that were missing post-operative sensory exam data and three patients who were lost to telephone follow-up on postoperative day 1; this missing data was excluded from analysis in the post-operative sensory exam analysis and postoperative day 1 analysis, respectively.

[Figure omitted. See PDF.]

Fig 1. CONSORT diagram of patient recruitment.

https://doi.org/10.1371/journal.pone.0246792.g001

The proportional hazard assumption was confirmed for the primary outcome, time to loss of shoulder abduction.

There were no differences in baseline characteristics between the two groups (Table 2). Surgical technique did not vary between surgeons. A similar percentage of patients in both groups converted to open shoulder surgery, which was open biceps tenodesis (Table 2). The patient, operating room team and post-anesthesia care team were blinded to group assignment throughout the study. However, in approximately 10% of the enrolled patients, the outcome assessor performing the sensory and motor assessments was aware of the study group assignment.

[Figure omitted. See PDF.]

Table 2. Patient demographics and baseline characteristics.

https://doi.org/10.1371/journal.pone.0246792.t002

Efficacy outcomes

Loss of shoulder abduction (motor block) was faster in the intraplexus group (log-rank p-value <0.0005; Fig 2). The median value of time to onset of motor blockade was 4 minutes [interquartile range: 2–6] in the intraplexus group compared to 6 minutes in the extraplexus group [4–10], p<0.0005). The onset of sensory blockade was faster in the intraplexus group for the axillary nerve (p = 0.0013), musculocutaneous nerve (p = 0.0001), median nerve (p = 0.0128) and radial nerve (p = 0.0046) distributions (Fig 3). No difference was noted in the onset of sensory blockade for supraclavicular and ulnar nerve distributions between the two groups.

[Figure omitted. See PDF.]

Fig 2. Kaplan-Meier curve of time to onset of motor blockade.

Extraplexus (solid line); intraplexus (dashed line).

https://doi.org/10.1371/journal.pone.0246792.g002

[Figure omitted. See PDF.]

Fig 3. Kaplan-Meier curve of time to onset of sensory blockade.

Extraplexus (solid line); intraplexus (dashed line).

https://doi.org/10.1371/journal.pone.0246792.g003

The intraplexus approach required fewer needle passes than the extraplexus approach (2 vs. 3, p<0.0005); however, there was no difference in block performance time between the two groups (Table 3). There was no difference in perioperative opioid consumption, block duration, highest NRS pain score in the recovery room, recovery room length of stay between groups, nor highest NRS pain score at 1 week postoperatively (Tables 3 and 4). There were no failed blocks as confirmed by sensory testing in the recovery room and no patients required a rescue block. Overall, patients were highly satisfied with their care with no difference between the two groups.

[Figure omitted. See PDF.]

Table 3. Block characteristics and efficacy outcomes.

https://doi.org/10.1371/journal.pone.0246792.t003

[Figure omitted. See PDF.]

Table 4. Pain scores and opioid consumption data.

https://doi.org/10.1371/journal.pone.0246792.t004

Safety outcomes

The intraplexus approach resulted in a higher incidence of paresthesia during block placement (35.9% vs. 14.5%, p = 0.0004). The longest duration of paresthesia postoperatively was 7 days, which occurred in one patient that was in the intraplexus group. Both groups exhibited equal incidence of Horner’s syndrome, dyspnea, hoarseness, and PONV (Table 5). No complications associated with nerve blocks (nerve injury, local anesthesia toxicity, nor infection) were observed in the study.

[Figure omitted. See PDF.]

Table 5. Safety outcomes.

https://doi.org/10.1371/journal.pone.0246792.t005

Discussion

Although an intraplexus injection of local anesthetic for interscalene block can result in the faster onset of motor and sensory brachial plexus blockade, the difference is not clinically significant. Further in this study, we found no difference in block duration, success rate, or analgesic benefit between intraplexus and extraplexus approaches to interscalene brachial plexus injection. The higher incidence of paresthesia in the intraplexus group compared to the extraplexus group is particularly concerning. Although the rate of post-block neuralgia associated with transient paresthesia is not well defined, there are multiple cases of chronic debilitating pain associated with paresthesia techniques, which most likely reflect an intraneuronal injection of local anesthetic [11]. Fortunately, in our study, any paresthesia associated with the block was transient, with the majority resolving during block placement with adjustment of the needle and there were no long-term neurologic deficits. Notably, the incidence of paresthesia in our study was much lower than the 96.7% incidence seen with intraplexus interscalene injection by Maga and colleagues [7]. Given how similar the block profile is between the two approaches, in terms of onset, duration, and analgesic benefit, we believe the results of this study call into question the benefits and safety of intraplexus injection for interscalene block.

The results of this study are fairly consistent with other studies that have also examined the efficacy of intraplexus and extraplexus injection for interscalene block. As in this study, Spence and colleagues also concluded that there is little advantage to intraplexus injection [6]. Unfortunately, the study conducted by Spence et al was designed as a noninferiority study and the anesthetic protocol was changed mid-enrollment [6]. Ultimately, their study was terminated early due to a difficulty enrolling patients after a clinical practice change unrelated to the safety or efficacy of interscalene blocks. Further, their extraplexus injection only consisted of injection posterior to the brachial plexus in the interscalene groove whereas in our study the injection was both anterior and posterior to the brachial plexus. Of note, there is a potential risk of phrenic nerve injury with an extraplexus approach that traverses above C5 to the space anterior to the brachial plexus and in close proximity to the anterior scalene muscle. However, in our study, there was no difference in the incidence of dyspnea in either group.

Maga and colleagues found a significant difference in onset of complete motor and sensory blockade between intraplexus and extraplexus injection for interscalene nerve block with 70% of the intraplexus approach blocked at 10 minutes compared to only 37% in the extraplexus group [7]. A notable difference between Maga’s study and our study is that Maga’s extraplexus approach involved out-of-plane intramuscular injections in the scalene muscles adjacent to the brachial plexus and in our study, the extraplexus injection was performed in-plane and not intramuscular, but rather in the space between the scalene muscles and outside of the brachial plexus sheath. This difference in injection approach would likely affect the distribution of the local anesthetic and potentially affect efficacy and safety profiles. Injecting directly into the muscle to perform interscalene brachial plexus blocks is particularly concerning given the growing evidence of the local anesthetic-induced myotoxicity [12].

There is growing evidence that an intraplexus injection for interscalene block can increase the potential for neurologic injury. In a cadaveric study by Szerb and colleagues, 11.5% of intraplexus injections were actually subepineural, while there were no subepineural injections in the periplexus group [13]. Another concern with intraplexus injection is if providers inject between what may appear as C6 and C7, they are likely to have an intraneural injection, as C6 can frequently have an intraroot splitting in the interscalene groove [14]. Unfortunately, the use of ultrasound does not necessarily protect against intraneural injection as in an observational study of 257 patients receiving ultrasound-guided interscalene and supraclavicular nerve blocks, 17% of patients had an unintentional intraneural injection [15]. Furthermore, in a cadaveric study by Orebaugh et al., there was 50% rate of unintentional injection into the subepineurium of the brachial plexus during ultrasound-guided interscalene nerve block [16]. Permanent neurologic injury has occurred with interscalene block even with the use of ultrasound guidance [17].

This study has some limitations. Although patients and providers administering medications intraoperatively and postoperatively were always blinded to the treatment groups, the staff assessing sensory and motor blockade were not blinded to group assignment in approximately 10% of the patients. While not ideal, we still consider the risk of bias as low [18], as the patients’ ability to abduct the shoulder is unlikely to be influenced by the lack of blinding of the outcome assessor. Shoulder abduction is a rather objective outcome, especially when blinding of the patients themselves is intact. In addition, the proportion of cases with broken blinding of the outcome assessor was small (20 of 208 patients). Another limitation was the relatively high volume of local anesthetic used in this study, which was commonly used at the time that this study was designed. One could argue that not as much difference was seen between the groups due to the large volume of local anesthetic use, however, we were able to see a difference in motor onset in this study which strengthens the results of this study. Further, larger amounts of local anesthetics were used in the studies by Spence et al and Maga et al [6,7]. An additional limitation of this study is the reliance on patient report for the duration of the block, which can be a subjective metric. In this study, patient-reported dyspnea was used as a surrogate of hemidiaphragmatic paralysis; however, this is not as precise as diaphragmatic function measured using ultrasound or spirometry and represents another limitation of this study.

In summary, the intraplexus approach to the interscalene brachial plexus block results in a faster motor onset than the extraplexus approach. It also results in a faster sensory block, requires fewer needle passes, and provides equally effective analgesia as the extraplexus approach. However, the time difference in block onset between groups is small and not a clinically meaningful difference. Given that the intraplexus group was associated with increased incidence of transient paresthesia and there were few differences in block quality between intraplexus and extraplexus approaches, an extraplexus approach to interscalene brachial plexus block may be a more appropriate choice.

Supporting information

S1 File. Ultrasound images of extraplexus versus intraplexus injection.

https://doi.org/10.1371/journal.pone.0246792.s001

(DOCX)

S2 File. Oral morphine equivalence conversion equation.

https://doi.org/10.1371/journal.pone.0246792.s002

(DOCX)

S3 File. Anonymized data set.

https://doi.org/10.1371/journal.pone.0246792.s003

(XLS)

S4 File. CONSORT checklist.

https://doi.org/10.1371/journal.pone.0246792.s004

(DOC)

S5 File. Interscalene study protocol.

https://doi.org/10.1371/journal.pone.0246792.s005

(DOC)

Acknowledgments

The authors would like to thank Dr. Jeffrey Ghassemi for his help in patient recruitment and data collection.

Citation: Harbell MW, Kolodzie K, Behrends M, Ma CB, Kinjo S, Yap E, et al. (2021) Extraplexus versus intraplexus ultrasound-guided interscalene brachial plexus block for ambulatory arthroscopic shoulder surgery: A randomized controlled trial. PLoS ONE 16(2): e0246792. https://doi.org/10.1371/journal.pone.0246792

References

1. Hadzic A, Williams BA, Karaca PE, Hobeika P, Unis G, Dermksian J, et al. For outpatient rotator cuff surgery, nerve block anesthesia provides superior same-day recovery over general anesthesia. Anesthesiology 2005;102:1001–7. pmid:15851888

2. Warrender WJ, Syed UAM, Hammoud S, Emper W, Ciccotti MG, Abboud JA, et al. Pain Management after Outpatient Shoulder Arthroscopy: A Systematic Review of Randomized Controlled Trials. Am J Sports Med 2017;45:1676–86. pmid:27729319

3. Fontana C, Di Donato A, Giacomo G Di, Costantini A, Vita A De, Lancia F, et al. Postoperative analgesia for arthroscopic shoulder surgery: A prospective randomized controlled study of intraarticular, subacromial injection, interscalenic brachial plexus block and intraarticular plus subacromial injection efficacy. Eur J Anaesthesiol 2009;26:689–93. pmid:19593887

4. Missair A, Weisman RS, Suarez MR, Yang R, Gebhard RE. A 3-dimensional ultrasound study of local anesthetic spread during lateral popliteal nerve block: what is the ideal end point for needle tip position? Reg Anesth Pain Med 2012;37:627–32. pmid:23080350

5. Franco CD, Rahman A, Voronov G, Kerns JM, Beck RJ, Buckenmaier CC. Gross anatomy of the brachial plexus sheath in human cadavers. Reg Anesth Pain Med 2008;33:64–9. pmid:18155059

6. Spence BC, Beach ML, Gallagher JD, Sites BD. Ultrasound-guided interscalene blocks: understanding where to inject the local anaesthetic. Anaesthesia 2011;66:509–14. pmid:21568985

7. Maga J, Missair A, Visan A, Kaplan L, Gutierrez JF, Jain AR, et al. Comparison of Outside Versus Inside Brachial Plexus Sheath Injection for Ultrasound-Guided Interscalene Nerve Blocks. J Ultrasound Med 2016;35:279–85. pmid:26715658

8. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—A metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42:377–81. pmid:18929686

9. Harbell MW, Cohen JM, Kolodzie K, Behrends M, Braehler MR, Kinjo S, et al. Combined preoperative femoral and sciatic nerve blockade improves analgesia after anterior cruciate ligament reconstruction: A randomized controlled clinical trial. J Clin Anesth 2016;33:68–74. pmid:27555136

10. McPherson M. Demystifying Opioid Conversion Calculations: A Guide for Effective Dosing. Bethesda, MD: ASHP; 2009.

11. Kaufman BR, Nystrom E, Nath S, Foucher G, Nystrom A. Debilitating chronic pain syndromes after presumed intraneural injections. Pain 2000;85:283–6. pmid:10692629

12. Hussain N, McCartney CJL, Neal JM, Chippor J, Banfield L, Abdallah FW. Local anaesthetic-induced myotoxicity in regional anaesthesia: a systematic review and empirical analysis. Br J Anaesth 2018;121:822–41. pmid:30236244

13. Szerb JJ, Greenberg JL, Kwofie MK, Baldridge WH, Sandeski RE, Zhou J, et al. Histological confirmation of needle tip position during ultrasound-guided interscalene block: a randomized comparison between the intraplexus and the periplexus approach. Can J Anaesth 2015;62:1295–302. pmid:26335906

14. Franco CD, Williams JM. Ultrasound-Guided Interscalene Block: Reevaluation of the "Stoplight"; Sign and Clinical Implications. Reg Anesth Pain Med 2016;41:452–9. pmid:27203394

15. Liu SS, YaDeau JT, Shaw PM, Wilfred S, Shetty T, Gordon M. Incidence of unintentional intraneural injection and postoperative neurological complications with ultrasound-guided interscalene and supraclavicular nerve blocks. Anaesthesia 2011;66:168–74. pmid:21320084

16. Orebaugh SL, McFadden K, Skorupan H, Bigeleisen PE. Subepineurial Injection in Ultrasound-Guided Interscalene Needle Tip Placement. Reg Anesth Pain Med 2010;35:450–4. pmid:20814286

17. Cohen JM, Gray AT. Functional deficits after intraneural injection during interscalene block. Reg Anesth Pain Med 2010;35:397–9. pmid:20607900

18. Higgins JPT, Savovic J, Page MJ, Elbers RG SJ. Assessing risk of bias in a randomized trial. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ WV, editor. Cochrane Handb. Syst. Rev. Interv. 6.1, Cochrane; 2020.