A review of 100 iatrogenic nerve injuries: delays in referrals remain significant

Abstract

Introduction

This is a retrospective study of 100 consecutive patients with iatrogenic nerve injuries, as seen in a tertiary referral centre over a six-year period.

Materials and methods

Patients who presented with new-onset nerve palsy involving a motor or mixed motor/sensory nerve following an operation were studied.

Results

There were 44 male and 56 female patients with a mean age of 53 years (range 5–87 years). The median duration from the index procedures to referral was six months (range 0 days to 12 years). Approximately one third of referrals were made over 12 months since the index procedures. Twenty patients recovered spontaneously and were managed expectantly. Eighty patients underwent secondary interventions.

Discussion

There remains a significant delay in referring postoperative nerve palsy to a nerve specialist. The majority of these cases will warrant secondary reconstructive surgery and delay in treatment may have a negative effect on the ultimate outcomes.

Introduction

Nerve palsy following surgery is a distressing occurrence, both to the patient and the surgeon involved. Multiple case series have been published that highlight the importance of prompt recognition and prevention of such injuries.^1,2 However, the true incidence of iatrogenic nerve injuries remains elusive because the denominator is unknown and the lesions are often unrecognised or not detected in a timely manner.^1,3–5 In 1991, Birch et al reported 68 cases of intraoperative complete transection of a major nerve, with delays in diagnosis varying from seconds to 40 months.³ In nearly half of the cases, delays were more than six months from the index operations. Ten years later, Khan and Birch reported another 612 cases of iatrogenic injury to peripheral nerves, seen over a seven-year period.⁴ Of the total, 291 patients underwent secondary exploration. There was an average delay between diagnosis and intervention of 10 months, with the maximum delay of up to 40 months.

In a retrospective review of 722 surgically treated cases of peripheral nerve lesions over eight years, 126 (17.4%) were found to be iatrogenic in origin.⁶ According to the authors, only 35% of the patients with iatrogenic injuries were operated upon within a time frame of six months, as a consequence of delayed referral. This in turn was thought to be due to a failure in establishing a correct diagnosis promptly and the uncertainty regarding the potential benefit of a secondary surgery among the referring physicians. The authors advocated timely intervention and demonstrated that remedial surgery led to an improvement in 70% of their cases.

In 2017, Rasulic et al reported the outcome of 122 patients with iatrogenic nerve injuries treated over a 10-year period.⁵ Those who underwent secondary remedial surgery within six months of injury would appear to have a greater chance of achieving a satisfactory functional recovery than those who were operated after six months.

The British Orthopaedic Association guideline on the management of nerve injuries, published in 2012,⁷ stated that it must be accepted that nerve injury during operation may occur in spite of due care. Early recognition and prompt remedial action after the event are the keys to a good outcome. In spite of that, there appeared to be a prevailing tendency among surgeons to regard postoperative nerve palsy as a largely self-resolving phenomenon, leading to a delay in referral. We would therefore like to ascertain whether there had been a change in awareness and referral practice. The objective of our study was to report the current referral pattern and spectrum of iatrogenic nerve palsies as seen in a tertiary referral unit.

Materials and methods

We carried out a retrospective review of 100 consecutive patients who had been referred with new-onset postoperative nerve palsy to our unit. We included only nerve palsies involving a motor or mixed motor/sensory nerve. Cutaneous neuroma alone was excluded from this study. All patients were assessed by the senior author (CYN) and their data were entered into a prospectively maintained database.

In addition to basic demographics, we collected data on index procedures, delay between index procedure and date of referral, and secondary surgery performed. As there was a wide variation in referral practice and a number of outliers, the median (instead of mean) time to referral was reported as it was more representative of the dataset. Data analysis was performed using Microsoft Excel.

This study was approved by the institutional research and development department.

Results

Between 2013 and 2019, 100 cases of new-onset postoperative nerve palsies were managed by the senior author (CYN). There were 44 male and 56 female patients, with a mean age of 53 years (range 5–87 years). The specific nerves injured are summarised in Table 1. A total of 76 cases involved nerves in the upper limb or neck and 24 were in the lower limbs. A wide range of index procedures were associated with the nerve palsies (Table 2).

Table 1 .

Specific nerves injured

Nerve	Injuries (n)
Radial	18
Spinal accessory	12
Sciatic	12
Axillary	10
Ulnar	10
Median	6
Tibial	6
Posterior interosseous	6
Musculocutaneous	5
Common peroneal	4
Medial cord	3
Long thoracic	2
Pan-infraclavicular brachial plexus	2
Femoral	2
Lateral cord	1
C5 root	1
Total	100

Table 2 .

Types of index procedures

Type of procedure	Procedures (n)
Arthroplasty	38
Plating	18
Soft tissue excision biopsy	14
Wiring	6
Tenodesis	4
Nailing	4
Removal of metalwork	3
Entrapment release	3
Arthroscopy	3
Fusion	1
Manipulation	1
Ligament Repair	1
Ligament Reconstruction	1
Tendon Reconstruction	1
Injection	1
Thoracotomy	1
Total	100

The median delay from the index procedure to the date of referral was 6 months (interquartile range 2–14 months). The delay ranged from 0 days to 12 years. Each end of the spectrum was represented by an intraoperative referral of an ulnar nerve transection during elbow surgery and two spinal accessory nerve (SAN) palsies, which were referred after approximately 12 years of delay. Overall, 68 patients were referred within one year of their index procedures and 32 referrals were made after one year’s delay. Among the various nerves injured, SAN had the longest average delay. Even after the two (12-year) outliers were excluded, the average delay in referral with a SAN palsy was approximately two years after their index operations.

Twenty patients reported spontaneous improvement by the time they were reviewed or showed signs of spontaneous recovery while under active monitoring, and were thus managed nonoperatively. These included the following nerves: posterior interosseous (3), radial (3), sciatic (3), ulnar (2), femoral (2), spinal accessory (1), C5 root (1), lateral cord (1), pan-infraclavicular brachial plexus (1), median (1), recurrent motor branch of median nerve (1) and tibial (1). Eighty patients proceeded with secondary surgery, as listed in Table 3.

Table 3 .

Secondary remedial operations performed

Type of surgery	Surgeries (n)
Exploration and neurolysis ± revision of index	56
Nerve grafting	8
Nerve transfer	6
Tendon transfer	3
Distal release	3
Direct muscular neurotisation	2
Combination of procedures	2
Total	80

Discussion

Iatrogenic nerve injuries result in avoidable morbidity to the patients, additional healthcare costs and potential medicolegal disputes. Delay in recognition and treatment have been highlighted as the main reasons for litigation.⁸ In a national survey of a no-fault compensation scheme in New Zealand in 2009, iatrogenic nerve injury was the fourth most common cause of claims.⁹ Of the 313 iatrogenic nerve injuries, orthopaedics accounted for one third of the cases, which was the highest number among the various surgical specialties.⁹

Our study demonstrates that approximately one third of referrals were made one year after the index operations. This observation echoes the findings of Kretschmer et al in 2001.⁶ There are a multitude of explanations that could account for such delays. Poor understanding of nerve injuries, failure to recognise early clinical signs, inadequate clinical examination and over dependence on neurophysiological investigations have been proposed as some of the reasons.⁸ The observed procrastination among some surgeons might reflect an inherent assumption that the postoperative nerve palsy is a form of neurapraxia, which would invariably recover spontaneously. If the nerve has been visualised at the time of operation and deemed intact, a period of watchful observation is reasonable, as long as there are clear signs of nerve recovery. However, if the mechanism of nerve damage is not clear and the potential for a more severe injury remains, early referral is mandated.

In our series, it was evident that some clinicians had difficulty in distinguishing a non-degenerative nerve lesion from a degenerative nerve lesion. The former refers to a neurapraxic injury resulting in a conduction block.^10,11 In contrast, the latter refers to a lesion where there is Wallerian degeneration of the distal nerve segment.¹² Clinically, the presence of neuropathic pain in the cutaneous distribution of a nerve is a strong indicator of nerve division and has attracted consensus for immediate operative exploration and decompression.^3,13 Similarly, loss of sudomotor and vasomotor functions in the cutaneous distribution of the nerve concerned should be regarded as red flags of nerve division.^3,14

While neurophysiology may provide further insights regarding a specific nerve lesion, requesting neurophysiology must not delay referral or treatment.⁷ If the nerve conduction studies demonstrate evidence of absent nerve conduction or electromyography shows signs of muscle denervation (fibrillations, positive sharp waves), the suspected nerve lesion is no longer neurapraxia. It is at least axonotmesis, if not neurotmesis. There is, however, no reliable clinical means of differentiating the two from a single examination. Serial assessments may shed light on the distinction. If there is an advancing Tinel sign along the known course of the nerve within a predicted time frame, the lesion is likely to be a recovering axonotmesis. In this circumstance, observation may be continued. If there is a static Tinel with repeated examinations, this could imply a hold-up of nerve regeneration or neurotmesis. In these situations, surgical exploration of the nerve is recommended.

Surgical exploration allows the nature and extent of a nerve injury to be established. Some may have reservation regarding exploration of nerve palsies, such as following arthroplasty, because of concern for secondary infection and the assumption that these palsies are often the result of stretch injury.^15–17 However, even in those with nerve in continuity, it has been shown that neurolysis may improve neuropathic pain.¹⁶ It may also address any structural hold-up of neural regeneration, thus facilitating recovery. In cases with nerve disruptions, secondary neural reconstruction can be performed. Apart from that, exploration provides an opportunity to manage any associated secondary issues such as joint contracture, non-union and hardware failure. Revision of the primary procedure can then be performed in conjunction with nerve exploration, if appropriate.

Published series have shown that common surgical procedures causing nerve injuries include osteosynthesis, arthroplasty, arthrodesis, lymph node biopsy in the neck and entrapment release.^6,8,18–21 One also has to consider inappropriate patient positioning and pressure area management as possible causes. In our series, arthroplasty and plating accounted for more than half the cases. Of note, seemingly benign procedures such as removal of metal work and manipulation of joint, which are often delegated to less-experienced surgeons, could lead to disastrous nerve injury if due care is not observed.

Consistent with other published series, the majority of our cases involved nerves of the upper limb.^18,22–27 This may reflect a higher propensity for nerve injuries in the upper compared with the lower limbs. We postulate that this may be due to the relative proximity of the nerves to bones and joints. In addition, the superficial nature and numerous tether points along the course of some of these nerves make them vulnerable to injury. This highlights the importance of having greater anatomical awareness of nerves at risk and their variants, in the prevention of such injuries.

There are limitations to our work. This study is subject to the usual confounders of a retrospective review. We recognise that a tertiary unit perspective would be skewed towards the worse end of the spectrum and there could be many transient postoperative nerve palsies that are being managed adequately by the primary surgeons without the need for onward referral. Owing to the lack of denominator, we are unable to establish the true incidence of iatrogenic nerve injuries. There is a preponderance of orthopaedic procedures in the series, due to the nature of our practice.

Because of the heterogeneous nerve injuries, presentations, treatments and follow-up periods, uniform reporting of validated comparable outcomes is not possible. The outcome of a nerve lesion could be influenced by a number of patient factors (age, comorbidity), injury factors (specific nerve damaged, depth of lesion, distance from target muscle) and surgical factors (specific reconstruction, delay to treatment). To quantify the potential effect of delay on the ultimate outcome, future studies on the individual nerves that would allow control of the other factors are being planned.

In conclusion, our review would suggest that delay in referring an iatrogenic nerve injury to a nerve specialist remains a significant issue. This is despite professional guidelines and large series of cases published over the last few decades. When faced with a nerve palsy following an operation, the primary surgeon could make a huge impact on a patient’s care by diligently recognising the nerve injury and initiating referral in a timely manner.