Perioperative nerve injury is a potential complication of regional anesthesia. In 1914, Neuhof (1) described neurologic deficits after brachial plexus anesthesia that could not be attributed to preoperative findings. More recently, Cheney et al. (2), by using the ASA Closed Claims database, found that the use of regional anesthesia was more frequently associated with claims involving nerve damage than general anesthesia. It is interesting that despite extensive investigation, the actual mechanism of injury was rarely identified within these patients. Perhaps multiple variables, including patient, surgical, and anesthetic risk factors, may all play a role in perioperative nerve injury. Anesthetic risks include needle- or catheter-induced mechanical trauma, local anesthetic toxicity, or neural ischemia during regional blockade. Unfortunately, despite contributions from patient or surgical factors, regional techniques are often implicated as the sole contributor to new or worsening perioperative nerve injuries (3).
Patients with preexisting peripheral neuropathies are theoretically at even greater risk for perioperative nerve damage as a result of underlying mechanical, ischemic, or metabolic derangements. The performance of a regional technique within this patient population remains controversial because of the potential risk of progressive nerve injury. However, no prior investigation has specifically examined these concerns. This study evaluated the safety of axillary blockade compared with general anesthesia in patients with a preexisting ulnar neuropathy undergoing surgery to the affected nerve (i.e., ulnar nerve transposition). Anesthetic, surgical, and patient risk factors were also evaluated as potential contributors to perioperative outcomes and the development of new or worsening neurologic symptoms.
After IRB approval, the medical records of all patients who underwent transposition of the ulnar nerve at the Mayo Clinic from 1985 to 1999 were retrospectively reviewed. Demographic data including age, race, sex, weight, and height were collected for each patient. A documented medical history of diabetes mellitus, hypertension, peripheral vascular disease, obesity, or excessive alcohol use or abuse was recorded. The presence of a preexisting neurologic disease (proximal neuropathy, mononeuropathy, or distal sensorimotor peripheral neuropathy) in addition to each patient’s ulnar neuropathy was also examined.
The etiology of each patient’s ulnar neuropathy was categorized by use of surgeon documentation as either traumatic, occupation related, multifactorial, or unknown. The duration of symptoms before surgery and the clinical characteristics of the neuropathy (pain, paresthesias, numbness, motor weakness, muscle atrophy) and their distribution were recorded. The overall severity of peripheral nerve injury was graded with the previously described McGowan Scale (Grade I–III) (4). Grade I injuries are those lesions involving mild sensory loss and no detectable motor weakness; Grade II injuries include lesions with moderate sensory loss and mild motor weakness; and Grade III injuries are those with severe sensory disturbances and paralysis of one or more of the ulnar intrinsic muscles. The results of electrophysiologic studies performed before each patient’s surgical intervention were also documented. Specifically, the sensory and motor components of both ulnar nerve conduction velocity and action potential amplitude were recorded. The surgical data collected included the attending surgeon, precise surgical procedure, and duration from incision to skin closure. Tourniquet use, inflation pressure, and the duration of tourniquet inflation were also documented.
Each anesthetic technique was categorized as either general anesthesia, axillary blockade, or local infiltration. The technique of each axillary block was further categorized as transarterial technique only, paresthesia technique only, nerve stimulation only, combined technique (paresthesia or nerve stimulation combined with transarterial), or sheath field block. The distribution of all paresthesias and motor responses elicited was recorded. The local anesthetic solution, addition of vasoconstrictors, and total volume used were also documented.
The duration of hospitalization and postoperative surgical outcome were assessed at each patient’s initial postoperative examination and at their 2- to 3-wk and 6-wk surgical follow-ups. Patient outcomes for each preoperative symptom (pain, paresthesias, numbness, and motor weakness) were categorized as complete resolution, partial resolution, no change, or worsening of the symptom. The onset of new neurologic symptoms was also recorded at each of the time intervals above.
Postoperative complications related specifically to the axillary block (hematoma, infection, and new neuropathy) were identified and their duration recorded. In patients who developed a new neuropathy, the clinical characteristics and duration of each symptom were reviewed.
Patient demographics and procedural characteristics were compared between patients receiving general anesthesia versus axillary blockade by using the χ2 test for classification variables and the ranked sum test for continuous variables. Fisher’s exact test was used to compare the frequency of successful outcomes at the time of the initial postoperative examination and at subsequent follow-ups for patients receiving general anesthesia versus axillary blockade. In patients who were lost to follow-up or in whom there was incomplete follow-up, we carried forward information from the patient’s last documented physical examination. The percentage of patients with incomplete follow-up was summarized and compared between the two groups. In all cases, two-tailed P values ≤0.05 were considered statistically significant.
There were 360 patients who underwent ulnar nerve transposition during the study period. Patient demographics and medical history were not significantly different between the two groups with the exception of age; patients undergoing axillary block were significantly older (Table 1). The severity of ulnar neuropathy as assessed by the McGowan scale was similar between groups. Preoperative electromyography also demonstrated comparable levels of neural dysfunction. Action potential conduction velocities (motor and sensory) and motor amplitudes were similar between groups. However, patients receiving axillary blockade had significantly lower sensory action potential amplitudes (specific numeric values are not reported for electromyographic findings because of institution-specific “range of normals”). There was no difference between groups with respect to surgical time, use of a tourniquet during surgery, or total tourniquet time.
A general anesthetic was performed in 260 (72%) patients. The remaining 100 (28%) patients received an axillary blockade, including 64 patients in whom an ulnar paresthesia or motor response was elicited at the time of block placement (Table 2). Anesthetic technique did not affect neurologic outcome immediately after surgery or at 2 or 6 wk after surgery (Table 3). Overall, 6 (6%) patients within the Axillary group and 15 (6%) patients within the General Anesthesia group experienced new or worsening neurologic symptoms during the study period.
All six patients in the Axillary Block group who reported new or worsening neurologic symptoms after surgery received bupivacaine (0.375% or larger) in combination with either an ulnar paresthesia or motor response (Table 4). By using logistic regression, bupivacaine was identified as an independent risk factor for worsening of ulnar nerve function compared with other local anesthetics (6 of 34 vs 0 of 65;P = 0.001), whereas elicitation of an ulnar paresthesia or motor response did not demonstrate an association with neurologic injury (6 of 64 vs 0 of 35;P = 0.087).
Patients undergoing axillary blockade were significantly more likely to be discharged from the hospital on their operative day when compared with patients undergoing general anesthesia (51% vs 31%;P = 0.001). After surgery, there were no reported cases of anesthetic-related (nonulnar) neurologic complications, axillary infection, or hematoma formation.
Patient, surgical, and anesthetic risk factors have all been identified as potential contributors to perioperative nerve injury. Surgical risk factors include direct intraoperative trauma, vascular compromise, infection, hematoma formation, tourniquet ischemia, or improperly applied casts. Horlocker et al. (3) examined the etiology of perioperative nerve injury in 607 patients undergoing 1614 blocks for upper-extremity surgery. Surgical variables were believed to be the etiology in 55 (88.7%) of 62 neurologic complications identified. Direct surgical trauma or stretch occurred in 40 (73%) cases, inflammation or infection in 6 (11%) cases, hematoma or vascular compromise in 4 (7%) cases, cast irritation in 3 (5%) cases, and tourniquet ischemia in 2 (4%) cases. It is interesting that all complications involving motor deficits had a surgical cause. Fourteen patients (25%) required subsequent surgical intervention to restore nerve function.
Anesthetic risk factors such as needle- or catheter-induced mechanical trauma, local anesthetic neurotoxicity, or ischemic injury secondary to vasoconstrictors or neural edema may also contribute to perioperative nerve injury. Auroy et al. (5) prospectively evaluated the incidence and characteristics of serious complications related to regional anesthesia. A total of 103,730 regional anesthetics, including 21,278 peripheral nerve blocks, were performed over a five-month period. Neurologic complications related to regional anesthetic technique occurred in 34 patients. In all cases of neurologic injury after peripheral nerve blockade, needle placement was associated with either a paresthesia during needle insertion or pain on injection of local anesthetic. Furthermore, the postoperative radiculopathy had the same topography as the associated paresthesia. The authors concluded that needle trauma and local anesthetic toxicity were the causes of most neurologic complications.
Finally, patient risk factors most often associated with perioperative ulnar injury include male sex, increasing age, extremes of body habitus, and preexisting diabetes mellitus (6). However, patients with preexisting neuropathies may be at increased risk of perioperative nerve injury. The application of one or more patient, surgical, or anesthetic risk factors on a dysfunctional, but clinically asymptomatic, nerve may result in the onset of new symptoms (7–10). This “double crush” phenomenon was originally described by Upton and McComas (7) in 1973 after noting that 81 (70%) of 115 patients with an electrophysiologically proven upper-extremity entrapment neuropathy (carpal tunnel) also had evidence of an ipsilateral cervical root lesion. They hypothesized that axons compressed or injured at one site may be particularly susceptible to damage at a more distal location. Osterman (8) emphasized not only that two low-grade compressions along a nerve trunk are worse than a single site, but also that the damage of the dual compression far exceeds the expected additive damage caused by each isolated compression (Fig. 1).
Patients undergoing ulnar nerve transposition would potentially be at increased risk for perioperative worsening of ulnar nerve function because these patients have a preexisting neuropathy and are undergoing surgery to the affected nerve. Additional injury could occur during the performance of a regional anesthetic (double crush). It is interesting that this investigation does not seem to support these concerns. We found no evidence to suggest that the use of axillary blockade increased the risk of new or worsening neurologic symptoms in patients undergoing ulnar nerve transposition. This suggests that axillary blockade may be performed in these patients with the same degree of safety and confidence as general anesthesia. However, our results should cautiously be extrapolated to patients whose neuropathy is the result of underlying toxic (chemotherapy) or metabolic (diabetes mellitus) disease processes or in those with coexisting neurologic disorders (multiple sclerosis, radiculopathy, or polio). Patients with generalized neuropathic processes may represent more risk compared with our patients who presented with a mononeuropathy.
Regional techniques during outpatient surgical procedures such as ulnar nerve transposition provide several advantages, including decreased nausea and vomiting, prolonged analgesia, and a more rapid recovery when compared with inhaled techniques (11–15). In our study, patients undergoing axillary block were less likely to be admitted to the hospital compared with those who received a general anesthetic. Thus, regional anesthesia may improve the perioperative management without affecting neurologic outcome among this patient population. Despite these advantages, clinicians must still consider the safety of injecting local anesthetic after elicitation of a paresthesia or motor response in the distribution of an already compromised nerve. All six patients in the Axillary Block group with new or worsening neurologic symptoms received bupivacaine in combination with either an ulnar paresthesia or motor response. Unfortunately, because of the small number of patients experiencing neurologic complications, it is difficult to perform meaningful statistical analysis of potential risk factors within this group. However, complication rates were significantly higher in patients receiving bupivacaine versus other local anesthetics. It is unclear whether these findings are the result of local anesthetic toxicity, poor arm positioning during prolonged postoperative sensory anesthesia, mechanical trauma, or a combination of variables. Additional data are needed to determine the optimal local anesthetic on the basis of the duration of analgesia, decreased incidence of hospital admission, and risk of neurologic complications.
There are several limitations caused by the retrospective nature of this study, particularly with respect to patient randomization. However, despite the inability to randomize patients, the two groups within the study are remarkably similar. Known patient risk factors such as male sex, advanced age, extremes of body habitus, and diabetes mellitus were represented equally between groups. It is important to note that the severity of preoperative ulnar nerve dysfunction, as evaluated by both McGowan injury scores and electromyography, was comparable between groups. Differences in preoperative sensory action potential amplitudes were not clinically significant. Surgical variables, including the attending surgeon and intraoperative technique, tourniquet use, casting preferences, and postoperative care, were also similar. The accuracy of data collection is also demonstrated by our neurologic outcomes, which are comparable to those previously reported (16).
In summary, although regional anesthetic techniques may theoretically place patients at risk of mechanical trauma, local anesthetic toxicity, or neural ischemia, these factors do not seem to worsen neurologic outcome in patients with preexisting ulnar neuropathy undergoing ulnar nerve transposition. Therefore, axillary blockade may be considered a suitable anesthetic technique for this procedure and should not be regarded as a contraindication within this patient population.
The authors would like to thank Paul Decker for his assistance with data analysis.
1. Neuhof H. Supraclavicular anesthetization of the brachial plexus. JAMA 1914; 62: 1629–32.
2. Cheney FW, Domino KB, Caplan RA, Posner KL. Nerve injury associated with anesthesia. Anesthesiology 1999; 90: 1062–9.
3. Horlocker TT, Kufner RP, Bishop AT, et al. The risk of persistent paresthesia is not increased with repeated axillary block. Anesth Analg 1999; 88: 382–7.
4. McGowan AJ. The results of transposition of the ulnar nerve in traumatic ulnar neuritis. J Bone Joint Surg Br 1950; 32: 293–301.
5. Auroy Y, Narchi P, Messiah A, et al. Serious complications related to regional anesthesia. Anesthesiology 1997; 87: 479–86.
6. Warner MA, Warner ME, Martin JT. Ulnar neuropathy. Anesthesiology 1994; 81: 1332–40.
7. Upton AR, McComas AJ. The double crush in nerve-entrapment syndromes. Lancet 1973; 2: 359–62.
8. Osterman AL. The double crush syndrome. Orthop Clin North Am 1988; 19: 147–55.
9. Yu J, Bendler EM, Mentari A. Neurological disorders associated with carpal tunnel syndrome. Electromyogr Clin Neurophysiol 1979; 19: 27–32.
10. Hurst LC, Weissberg D, Carroll RE. The relationship of the double crush syndrome to carpal tunnel syndrome (an analysis of 1000 cases of carpal tunnel syndrome). J Hand Surg [Br] 1985; 10: 202–4.
11. Davis WJ, Lennon RL, Wedel DJ. Brachial plexus anesthesia for outpatient surgical procedures on an upper extremity. Mayo Clin Proc 1991; 66: 470–3.
12. Bridenbaugh LD, Soderstrom RM. Lumbar epidural block anesthesia for outpatient laparoscopy. J Reprod Med 1979; 23: 85–6.
13. Jankowski CJ, Horlocker TT, Rock MG, Stuart MJ. Femoral 3-in-1 nerve block decreases recovery room time and charges and time to hospital discharge after outpatient knee arthroscopy. Reg Anesth 1998; 23 (Suppl): 60.
14. Patel NJ, Flashburg MH, Paskin S, Grossman R. A regional anesthetic technique compared to general anesthesia for outpatient knee arthroscopy. Anesth Analg 1986; 65: 185–7.
15. Brown AR, Weiss R, Greenberg C, et al. Interscalene block for shoulder arthroscopy: comparison with general anesthesia. Arthroscopy 1993; 9: 295–300.
© 2001 International Anesthesia Research Society
16. Froimson AI, Anouchi YS, Seitz WH Jr, Winsberg DD. Ulnar nerve decompression with medial epicondylectomy for neuropathy at the elbow. Clin Orthop 1991; 265: 200–6.