Analysis of Upper and Lower Extremity Peripheral Nerve Injuries in a Population of Patients with Multiple Injuries

Abstract

Background 

The purpose of this study was to determine the prevalence, cause, severity, and patterns of associated injuries of limb peripheral nerve injuries sustained by patients with multiple injuries seen at a regional Level 1 trauma center.

Methods 

Patients sustaining injuries to the radial, median, ulnar, sciatic, femoral, peroneal, or tibial nerves were identified using a prospectively collected computerized database, maintained by Sunnybrook Health Science Centre, and a detailed chart review was undertaken.

Results 

From a trauma population of 5,777 patients treated between January 1, 1986, and November 30, 1996, 162 patients were identified as having an injury to at least one of the peripheral nerves of interest, yielding a prevalence of 2.8%. These 162 patients sustained a total of 200 peripheral nerve injuries, 121 of which were in the upper extremity. The mean patient age was 34.6 years (SEM +/- 1.1 year), and 83% of patients were male. The mean injury severity score was 23.1 (+/- 0.90), and the mean length of hospital stay was 28 days (+/- 1.8).

Conclusions 

Motor vehicles crashes predominated (46%) as the cause of injury. The most frequently injured nerve was the radial nerve (58 injuries), and in the lower limb, the peroneal nerve was most commonly injured (39 injuries). Diagnosis of a peripheral nerve injury was made within 4 days of admission to Sunnybrook Health Science Centre in 78% of the cases. Surgery was required to treat 54% of patients. Head injuries were the most common associated injury, occurring in 60% of patients. Other common associated injuries included fractures and dislocations. The present report aims to aid in identification and treatment of peripheral nerve injuries.

Key Words: Demographics, Epidemiology, Nerve injury, Peripheral nerve, Prevalence, Trauma.

Peripheral nerve injuries (PNI), not uncommon in trauma, are increasingly being recognized with improved trauma services. To date, the main focus of PNI reports in the literature has been the pathophysiology, management, and the operative results. ^[1-3] The medical literature is deficient in comprehensive studies regarding the epidemiology of such serious nerve injuries. This paper will serve to identify the demographics, cause, incidence, severity, associated injuries, and outcome of both upper and lower limb PNI in a large trauma population.

Early identification of PNI is important, because untreated PNI may be sources of serious disabilities in later life, given the highly productive age group of the trauma population. ^[4-6] For example, the injury admissions data from Ontario Trauma Registry for 1994 stated that 34% of home-discharged trauma patients required support services and 16% needed rehabilitation centers. Although it is not clear from this information whether PNI contributed to any of these disabilities, it is certainly possible that part of the disability could be arising from the PNI, sustained in trauma to the extremities, especially in orthopedic and vascular trauma. ^[7,8]

PATIENTS AND METHODS

Patient Population

Sunnybrook Health Science Centre (SHSC) is designated as the principal Level 1 trauma center in the province of Ontario. SHSC treats approximately 650 patients with multiple injuries each year. These patients are assessed by a trauma team consisting of the trauma leader (a general trauma surgeon, an emergency room physician or anesthetist, with comprehensive training in trauma management), anesthetist, surgical residents, nurses, and respiratory technologists. From the time of admission, standardized trauma assessment forms, developed at SHSC, are used to document all of the injuries to patients. ^[9,10] Specially trained trauma data clerks subsequently enter this prospectively collected information into computerized databases. Further information about hospital stay and follow up visits is also compiled and added to the database. Hence, a comprehensive overview of each patient's injuries and treatment course at SHSC is available.

Inclusion Criteria

For the purposes of this study, a PNI was defined as an injury to a major nerve in the upper (i.e., radial, median, or ulnar nerve) or lower (i.e., sciatic, femoral, peroneal, or tibial nerve) limb. Patients with exclusively nerve root, plexus (brachial or lumbar), digital, or any other minor nerve injury were excluded. Patients with brachial plexus injuries have been previously reported. ^[11]

By using the database, it was found that from January 1, 1986, to November 30, 1996, SHSC attended to 5,777 trauma patients, 295 (5.1%) of which were coded as having been involved in a trauma resulting in any form of a PNI. Of the 295 patients identified, 113 were excluded because the peripheral nerve trauma was found to involve plexus or nerves not of interest to the present analysis. Therefore, there was a total population of 182 trauma patients who sustained a major limb PNI.

Chart Review

After the PNI patients were identified, a detailed chart review was undertaken to verify the database information and to collect additional information not included in the database. In the course of this review, 14 patients were found to be misdiagnosed with a PNI. During their initial clinical presentation, these patients were presumed to have a PNI by the examining physician. However, detailed review of their clinical findings and subsequent diagnostic testing revealed that these patients did not have a PNI. Six other charts were not attainable for review. Thus, 20 charts were excluded from the study, leaving a population of 162 trauma patients with a verified PNI.

The following information was collected and entered into a spreadsheet for each patient/nerve injured: age, sex, specific nerve injured, side of injury, cause of injury, Injury Severity Score ^[12] (ISS), length of hospital stay, associated injuries, the level and grade of nerve injury, time from admission to diagnosis of PNI, electrodiagnostic testing, treatment (both surgical and nonsurgical) received, and eventual functional outcome.

Classification Criteria

Mechanism or cause of injury was characterized using the preexisting criterion of the SHSC database. Causes of injury were classified under one of the following categories: motor vehicle crash (MVC), motorcycle crash, recreational MVC, pedestrian, gunshot wound, fall, industrial accident, stab wound, bicycle accident, and assault. Injuries resulting from any other causes were grouped together.

Level of nerve injury was defined using the following criteria: upper extremity injuries were classified into the levels of the arm (between shoulder and elbow), elbow, forearm, wrist, and palm. Injuries in the lower extremity was classed according to pelvis/buttock, upper leg, knee, lower leg, and ankle levels.

On the basis of available clinical information in the chart, the grade of nerve injury was assigned using the Sunderland Grading System. ^[13] Grade 1 was defined as neuropraxia, in which the patient recovered from the nerve injury within days to weeks; grade 2 designated axonal severance, in which the patient had demonstrated clinical or electrophysiologic evidence of regeneration at a rate of approximately 1 inch per month; a grade 5 involved a complete laceration of a nerve. PNI that failed to improve but did not involve complete laceration were arbitrarily classed as grade 3 and 4, which were combined. Note that no distinction between grade 3 and 4 injuries were made, because the histologic analysis required for such a distinction was not available.

Clinical outcome of the nerve injury was determined using the following classification system, ranging from 0 to 4: a score of 0 designated no function, 1 indicated poor function, 2 indicated average function, 3 good function, and a 4 indicated return to normal function. Scoring of outcome was based on the last available assessment in the patients' charts at the time of chart review. It was also noted whether the patient demonstrated signs of improvement or recovery from the nerve injury.

The incidence of major associated injuries was also collected. Major injuries to the entire body as well as the occurrence of injuries to the bones anatomically adjacent to the peripheral nerves were noted. Associated injuries of particular interest included lacerations, fractures, dislocations, vascular injuries, and head injuries.

Statistical Analysis

All data collected were assessed using commercially available PC software (Statistica, version 3.5. Statsoft, Tulsa, Okla). Descriptive, parametric, and nonparametric statistics were generated where appropriate. An alpha level of p < 0.05 was used for all analysis.

RESULTS

Overview

The 162 patients identified had a total of 200 nerve injuries. The prevalence of patients with limb PNI in the 5,777 trauma patients treated at SHSC was therefore 2.8%. The mean age of the population was 34.6 years, with a range of 13 to 76 (Figure 1). The majority (59%) of the population was between the ages of 18 and 35. The ratio of males to females was 5:1, with 83% of the patients being male. Table 1 further summarizes the demographic and descriptive statistics of this population. The mean ISS was found to be 23.1 (SEM +/- 0.90). Patients had a range of ISS scores from 2 to 57. The majority of patients (71%) had ISS scores between 10 and 29. In the population with peripheral nerve injuries, only one patient died, 10 days after admission after an MVC.

Figure 1: Age and gender distribution of a population of patients with multiple injuries including peripheral nerve injuries (n = 162).

Table 1: Descriptive statistics of peripheral nerve injuries in a trauma population of 5,777 patients

Nerve injuries more frequently (60.5%) occurred in the upper extremity than in the lower extremity. The most frequently injured nerve was the radial nerve, with a predominance of 58 injuries. The most commonly injured nerve in the lower extremity was the peroneal nerve, with a total of 39 injuries. Of 162 patients, there were 10 patients (6.2%) who had both an upper and lower extremity nerve injury. An approximately equal number of injuries were sustained to both right and left sides (91 vs. 109, respectively). No notable differences were observed for side predominance for each individual nerve. Incidence of PNI was determined for each year of the study period and was as follows: four injuries in 1986, seven injuries in 1987, eight injuries in 1988, 12 injuries in 1989, 13 injuries in 1990, 14 injuries in 1991, 17 injuries in 1992, 26 injuries in 1993, 31 injuries in 1994, 15 injuries in 1995, and 15 injuries in 1996.

Of the 200 nerve injuries, 87 received surgical treatment directed toward a specific nerve injury. Electrodiagnostic tests were used for 80 (49%) of the 162 patients with PNI within SHSC. The mean length of hospital stay was 28.1 days, with a range of 0 to 128 days. Although the mean time required from admission to diagnosis was 6.7 days, 78% of PNI were diagnosed within 4 days of admission to SHSC. As a general trend, upper extremity nerve injuries were diagnosed earlier than lower extremity injuries (Figure 2). There was a considerable delay in diagnosis of femoral nerve injuries (mean, 51.8 days), with time from admission to diagnosis of injury differing significantly from the other nerve injuries (analysis of variance, F sub (6,193) = 6.44, p < 0.001). However, within the present population, there were only four femoral nerve injuries.

Figure 2: Time from admission to diagnosis of individual nerve injuries. Mean +/- SEM in days, n = 200. Whereas upper extremity nerve injuries were generally diagnosed earlier than lower extremity injuries, the delay in diagnosis of femoral nerve injuries was significantly different from the other nerve injuries (p < 0.001).

Cause of Injury

The majority (52%) of trauma cases seen at SHSC are secondary to MVC. This finding was also seen in our PNI population (Figure 3), as 46% of patients with nerve injuries were involved in an MVC. Although MVCs were responsible for 46% of PNI, only 2.5% of all MVC resulted in trauma to a peripheral nerve (Figure 4). Conversely, industrial accidents were the source of only 6.1% of PNI, but 6.0% of industrial accidents resulted in an injury to a peripheral nerve. Other relatively high-risk causative factors included gunshot wounds, motorcycle crashes, recreational motor vehicles such as snowmobiles, and stab wounds (Figure 4).

Figure 3: Causes of peripheral nerve injuries, expressed as a percent. Note, n = 163, as one patient's injury was the result of two separate causes. GSW, gunshot wound; MVC, motor vehicle crash.

Figure 4: Percent of peripheral nerve injury resulting from specific causes within a population of patients with multiple injuries (n = 5,777). MVC, motor vehicle crash; GSW, gunshot wound.

Anatomic Level of Nerve Injury

Upper extremity PNI most frequently occurred at the upper arm level (44%). Sixty-two percent of the 58 radial nerve injuries occurred in the arm, 6.9% were at the elbow level, 19% in the forearm, and 12% in the wrist. Median nerve injuries most commonly were at the forearm level (10 of 25, or 40%), 36% were in the arm, 8% were in the elbow, 12% in the wrist, and 4% were in the palm. Ulnar nerve injuries occurred with equal frequency (13 of the 38 injured ulnar nerves) at the forearm and elbow levels, 21% were in the arm, and the remaining 11% in the wrist.

Of the 28 sciatic nerve injuries, 61% of the injuries were at the pelvic level, the remainder were in the upper leg. Only one tibial nerve injury occurred at the ankle level, the remaining seven injuries were in the lower leg. Of the 39 peroneal nerve injuries, 56% of the injuries were at the level of the lower leg, 33% were at the knee level, and 10% were at the ankle level. Femoral nerve injuries occurred with equal frequency at the pelvic and upper leg levels (n = 2 at each level).

Grade of Nerve Injury and Functional Outcome

The frequency distribution of grade of nerve injury, using a modification of the Sunderland grading system (as outlined in the Patients and Methods section) is presented in Figure 5. Note, grade of nerve injury information could not be ascertained in four of the 200 injured nerves. There were no significant differences in the frequency of injury grade in the overall, nor in the individual nerve injury population (chi squared = 15.3, p = 0.08).

Figure 5: Frequency of nerve injuries as graded using a modification of the Sunderland grading system (see Patients and Methods section) (n = 196; grades of four injuries were unknown).

Functional outcome could be determined with good accuracy for 125 of the 200 injured nerves. Of these, 40% returned to normal function, 16% had good function, 10% had average recovery, 22% had poor recovery, and 12% had no function in the injured nerve territory. Final functional outcome was assessed from 5 to 1,786 days after injury (mean, 300 days; median, 133 days).

There was no relationship found between grade of nerve injury and cause of injury. However, severity of nerve injury was found to strongly correlate with functional outcome (rho = -0.744, p < 0.001). Of the cases in which both nerve injury grade and outcome were both known (n = 125), 97% of grade 1 nerve injuries attained a normal functional outcome and 3% had good function; grade 2 nerve injuries had mostly good to normal outcome; grade 3 and 4 had more variable outcome; and 83% of grade 5 nerve injuries had no or poor functional recovery. No grade 5 injury attained a normal functional outcome.

Associated Injuries

Sixty percent of patients in the study had an associated head injury, and in 7.4% of patients, the head injury was severe enough to result in a coma. Thoracic spine injuries occurred in 9.3%, lumbar spine injuries in 9.9%, fractured rib(s) in 39%, internal thoracic injuries in 34%, vascular injuries in 34%, and 9% of patient with peripheral nerve injuries suffered tendon injuries. Injuries occurring in musculoskeletal structures anatomically adjacent to injured nerve are listed in Table 2. Humeral fractures accompanied 72% of radial nerve injuries, whereas radius/ulna injuries occurred with 45% of median and ulnar nerve injuries. Pelvic fractures occurred with 79% of sciatic nerve injuries, and an injury to the tibia/fibula was present in 56% of peroneal nerve injuries.

Table 2: Number of associated injuries occurring anatomically adjacent to injured peripheral nerves

In the population of trauma patients with humeral fractures (444 cases), the incidence of upper extremity nerve injuries was as follows: radial nerve, 9.5%; median nerve, 1.4%; and ulnar nerve, 3.8%. Within the population of patients with radius/ulna fractures (703 cases), 1.3% median nerve and 2.4% ulnar nerve injuries were diagnosed. Femur fractures (818 cases) were accompanied by sciatic nerve injuries in 1.1% of cases. Within the population of patients with pelvic fractures (1,283 cases), the incidence of sciatic nerve injuries was 1.7% and 0.16% for femoral nerve injuries. Tibia and fibula fractures (996 cases) were accompanied by peroneal nerve injuries 2.2% of the time and tibial nerve injuries 0.50% of the time. Hip dislocations (168 cases) occurred with sciatic nerve injuries in 7.1% of cases.

DISCUSSION

The overall incidence of PNI was 2.8% from our population of patients with multiple injuries over the past 11 years. There have been only rare publications on PNI that provide some comparison to the inferences of the present study. ^[14,15] For example, a large series on neurotrauma in 21,973 patients from New South Wales treated in 1977 reported a similar PNI incidence of 2%. ^[15] The dominance of male sex (83% of patients were male) and age distribution is consistent with the existing literature on trauma. ^[11,15-18]

The annual incidence of PNI in the present study started to increase in the beginning of the decade with a steep rise in 1993 and 1994. The initial increase may have reflected more thorough data collection from the use of standardized trauma assessment forms, introduced in 1989 at SHSC. ^[9,10] There is no obvious explanation for the marked rise in 1993 and 1994; however, there was an increase in total trauma admissions at SHSC during 1993 and 1994 and 1994 and 1995 (approximately 20%). This increase may be one contributing factor, although PNI incidence decreased in subsequent years, despite overall trauma admissions staying more or less constant.

The majority of the patient (52%) treated by the SHSC trauma center had been involved in MVCs, a Figure thatis in agreement with other trauma center studies. ^[16] Although MVCs were responsible for 46% of the PNI, only 2.5% of patients treated for injuries after an MVC were found to have sustained a PNI. Industrial accidents, gunshot wounds, motorcycle and recreational motor vehicle crashes, and stab wounds represented much higher risk activities. Many of these injuries resulted from a sharp mechanism of damage to the nerve. Upper extremity PNI, analyzed by McAllister et al., ^[5] reported a sharp object most frequently producing the lesion, either by domestic or industrial accidents. However, this report was based on a patient population referred to a hand surgery unit, with most cases seen after diagnosis of injury. In a report of neurovascular trauma of upper limb, penetrating injuries dominated. ^[19] The cause of PNI from developing countries is also predominantly a penetrating injury; 98% of 74 patients over a 5-year retrospective study reported from Pakistan had penetrating injury, with 83.8% injured by gunshot wounds, whereas less than 1% had been injured in an MVC. ^[18]

The literature is consistent in reporting the higher incidence of PNI in the upper compared with the lower extremity. ^[18] Additionally, in the upper extremity, the majority of PNI occur at the level of the upper arm. ^[6,20] This area is where the radial nerve was most commonly injured in our study. The exceptions reported in the literature of ulnar nerve injury predominating at elbow ^[6] and the median nerve injury occurring most often at the wrist ^[5] may reflect the populations studied. Stone and Keenan investigated patients with brain injuries, and included the pressure neuropathy of the ulnar nerve, which is often associated with head injuries. ^[6] McAllister et al. studied patients with sharp injuries to the upper extremities predominantly at wrist, at a referral center for hand surgery. ^[5] It is interesting here to observe that lower extremity peripheral nerves are much more commonly affected in childhood trauma and almost all had an underlying bony injury. ^[21]

PNI are frequently associated with traumatic brain injury. The incidence depends on the population studied. In our patients, from a Level 1 trauma center, 60% of patients with PNI also had some form of head injury. Moreover, the incidence of PNI in our overall trauma population (2.8%) was almost identical to the incidence of PNI in the patients who sustained a head injury. Our data also did not reveal a higher incidence in the patients with the more severe head injuries. However, a prolonged depressed level of consciousness and inability to perform a full peripheral neurologic assessment may have resulted in some cases of PNI being missed. In studies of patients admitted with severe traumatic brain injury to rehabilitation units, the reported frequency of peripheral nerve dysfunction varies from 7%, ^[21] 10%, ^[22] 11%, ^[4] to a high of 34%. ^[6]

Evaluation of the type of PNI with associated head injuries revealed that many could be explained on the basis of initial trauma, for example, humeral fracture with radial nerve injury, posterior acetabular displacement with sciatic nerve injury, and pelvic fracture with femoral nerve injury. ^[23] Some of the delayed neuropathies were attributed to poor bed positioning and pressure, especially ulnar and peroneal. ^[4,6,22] Both initial trauma and a delayed posttraumatic mechanism contribute to peripheral neuropathy. Garland and Bailey ^[4] reported an 11% incidence of initially missed musculoskeletal injury and a similar incidence of delayed diagnosis of peripheral neuropathy in their retrospective study of 91 head injured adults. In their report, the majority of the PNI had associated local skeletal injury, or resulted from pressure by hematomas and bed positioning.

In the present study, 78% of PNI were diagnosed within 4 days. Delay in diagnosis of PNI are often attributed to the severity of associated head injuries, because the dysfunction is not brought to the attention of the physician by the patient with a head injury or all the neurologic deficits are assumed to be due to central nervous system injuries. ^[6] However, in a recent study, no statistical association between the severity of head injury and the delay in diagnosis of brachial plexus injury was identified. ^[11] For the more distal PNI, an ischemic or badly injured limb may also make clinical assessment difficult. In the present study, upper extremity nerve injuries tended to be diagnosed much earlier than lower extremity ones. This finding may have occurred because of the bedridden nature of the patient with multiple injuries, with the lower limb neurologic deficit not clearly manifested until the patient was mobilized and ambulated. Alternatively, the onset of PNI may truly have occurred in a delayed fashion, secondary to the underlying leg musculoskeletal injury (compression from hematoma, compartment syndrome) or from a complication of the injuries management (traction or surgery). The delay in diagnosis of femoral nerve injuries is well recognized, ^[24] often being attributed to the need for emergency management of an adjacent life-threatening (vascular) injury, with the initial femoral nerve injury being overlooked. Moreover, the femoral palsy often does not declare itself until the patient attempts to ambulate.

The skeletal injuries were found to have a very close relationship to the PNI in our study, which is well supported in the literature. ^[25,26] Fractures and other musculoskeletal injuries are common and may damage the adjacent nerves either by direct trauma or stretch with hemorrhage in the nerve. ^[27] Adduction injury of the knee and dislocation of proximal tibial fibular shaft very often injures the peroneal nerve, as is the radial nerve injured frequently in cases of humeral fractures. In children, 52% of type III supracondylar humerus fractures injured the median nerve when displaced posterolaterally, whereas radial nerve involvement was 28% (always with posteromedial displacement). ^[28] Similarly, axillary nerve injury has a very close association with shoulder dislocation and blunt injury to the shoulder. ^[29,30] Due to the associated movement at the shoulder, cases of scapular fracture also injure the axillary nerve along with the brachial plexus. ^[25] Our observations regarding brachial plexus injury and shoulder trauma in the SHSC trauma population was reported earlier. ^[11] Femoral nerve injury is known to occur secondary to an iliopsoas hematoma from high energy sports and also with pelvic fractures. ^[24,31] The sciatic nerve injury is very often associated with dislocated hip and is, unlike the femoral neuropathy, diagnosed early. ^[32] It is important to recognize the PNI with musculoskeletal trauma at the earliest possible point, because approximately 50% of neuropathies are potentially preventable and early diagnosis yields optimal results with surgical or nonsurgical treatment.

Similar to the close association between the injuries to the subclavian artery and brachial plexus, close proximity of median and ulnar nerves to the brachial artery makes them more vulnerable to concomitant injury. As many as 50% of the brachial artery injuries had associated PNI, often with an underlying skeletal fracture. ^[19,33] It is crucial to recognize the PNI, because the long-term disabilities are caused by nerve injury, whereas vascular repair itself is often successful in revascularizing the limb. ^[19] At exploration for vascular repair, the nerve injury may not be recognized unless looked for. In penetrating injuries, acute repair of the divided nerve produces the best long-term results, which approach close to 50% good outcomes. ^[33] On the other hand, bluntly injured nerves and transected, but heavily contused, nerves should not be repaired primarily. ^[34]

The Abbreviated Injury Scale score and peripheral nerve function recovery had a statistically significant relationship, and the Abbreviated Injury Scale was the most valid prognostic classification in assessment of war inflicted PNI. ^[35] Most of our patients (71%) had an ISS between 10 and 29. There was no relationship found between the grade of nerve injury and the cause of PNI. The outcome, however, strongly correlated with the severity of the nerve injury. Both ISS and Abbreviated Injury Scale at higher scores reflect the overall injury severity and the nerve injuries are also likely to be severe (e.g., crush, laceration). The majority of these are vehicular crashes. Permanent disability (2 years after injury) has strong association with increasing ISS scores. ^[36] In a group of 49,500 patients analyzed by van der Sluis et al., ^[17] 97% of residual disabilities caused by neuropathy and myositis ossificans were direct consequences of the original injuries, whereas 3% were complications during hospitalization. In their study, many thoracic and abdominal injuries were accompanied by injuries that caused disabilities in other body parts. Disablement (Glasgow Outcome Scale (GOS) scores 3 and 4) was especially related to injuries of the nervous system (spinal cord injuries and peripheral neuropathies), in the majority of patients with thoracoabdominal trauma, indicating the vulnerability of the nervous system. ^[17,36] However, even moderately disabled patients (GOS 4) continued to show recovery until the end of first year of follow-up and up to two thirds of patients with multiple injuries eventually returned to work. ^[16,37] The recovery pattern of PNI in our series also shows a similar trend. These observations support a thorough initial examination and evaluation of a patient with multiple injuries for early diagnosis and management of PNI.

Acknowledgments

The authors thank Cyndy Rogers for assistance with data collection and Dr. Michael Schwartz for critical review of the manuscript.