Three layers of tissue closely invest peripheral nerves: the epineurium, perineurium, and endoneurium. Although needle placement or injection within any of these layers is undesirable, recent evidence suggests that it is injection within the perineurium that is most likely to cause significant injury [5]. The epineurium is an external enveloping layer surrounding the fascicles and connective tissue within the nerve. The perineurium is a multilayered epithelial sheath that surrounds individual or groups of fascicles. Needle placement within the fascicle can cause injury directly or through pressure-related ischemia after injection [5].

Short bevel needles are purportedly less likely to injure nerves than long bevel needles [6]; however, there is little epidemiological evidence to support this assertion. What is more certain is that the injury is less severe if the needle is inserted with the bevel “parallel” to the line of the axons, rather than at right angles when fascicles will be transected rather than split longitudinally [7]. Neurotmesis, complete disruption of axon and myelin sheath, is far more likely to cause permanent injury than neuropraxis due to compression or stretch because the myelin sheath is preserved.

The use of ultrasound guidance should, theoretically, reduce the risk of direct needle trauma to the nerve. However, because of the relatively low incidence of complications from peripheral nerve blocks, studies have not been able to establish a benefit of ultrasound guidance compared to nerve stimulation alone in terms of nerve injury [8, 9].

The peripheral nerves are reasonably well protected from chemical injury, but solutions containing preservatives and their accidental contamination are best avoided. There is no evidence to suggest that local anesthetics, in clinically used concentrations, have any more adverse effect on peripheral nerves than they do on the neuraxis. However, occasional laboratory studies, such as the observation that local anesthetics have toxic effects on cell cultures [10], do raise questions. Such “toxicity” is related to both concentration and duration of exposure, but the implications of these findings to man are unclear given the large numbers of patients who receive the drugs annually. However, it would seem prudent to use the lowest concentration of drug possible to achieve the desired effect, especially when an infusion technique is used [11].

Local anesthetic adjuvants (other than epinephrine) are used to prolong the duration of single injection techniques. Multiple agents have been studied including clonidine, dexmedetomidine, midazolam, neostigmine, and dexamethasone. Electron micrography of isolated rat nerves demonstrates histologic evidence of damage associated with some additives when used alone or in combination [12]. Further, most have limited prolonging effects at the expense of unwanted side effects such as excessive sedation, hypotension, and bradycardia with clonidine [13]. Dexamethasone alone (667 μg/mL) combined with ropivacaine (0.25 %) appears to have minimal issues with neurotoxicity in rats [12], and minimal side effects with the greatest effect on prolonging block duration in human trials [14]. Combining multiple perineural adjuvants enhances toxicity in rat neuronal cell cultures [12]. Other important factors to consider include the preparation of the adjuvant. Specifically, multidose preparations often contain preservatives that can be neurotoxic. While sodium bisulfite appears to be fairly neutral, benzyl alcohol is a potent neurotoxin. Furthermore, use of perineural adjuvants is “off-label” and practitioners should carefully weigh the risks and benefits of their use. Finally, specifically regarding dexamethasone and dexmedetomidine, there is emerging evidence suggesting that intravenous administration of these drugs may also prolong analgesic duration, without the safety concerns of perineural injections [15, 16].

Most local anesthetic drugs have a vasoconstrictor effect at low concentrations, but there are no data to suggest that this contributes to injury, even when a solution containing clinical concentrations of epinephrine is used. Many upper and lower limb blocks will be performed in patients whose surgery will be performed under tourniquet and they can have a much more profound effect especially if poorly applied, so that there is mechanical distortion or excessive pressure. Compression due to hematoma or abscess is possible, these lesions having the same risk factors as after central blocks, but they tend to spread more readily through the peripheral tissues, so high pressures are not generated.

It has been argued, in the context of nerve entrapment syndromes [17], that patients with a pre-existing neurological problem are more likely to suffer injury if a second, more distant insult occurs: the “double crush” phenomenon. However, the relevance of this to the risk of a patient with peripheral nerve disease (e.g., diabetic neuropathy) suffering injury from a peripheral block is unclear. Peripheral nerve injury requires disruption of the perineurium and, in practice, this is very difficult to accomplish; recent experience with ultrasound indicates that nerves are difficult to impale, tending to move away from approaching needles. Even if a nerve is pierced, it is difficult to maintain the needle within the nerve, most of the solution leaking out of the epineurium after a small volume has been injected [18].

Temporary sensory or motor impairment may occur in nearly 3% of patients after a peripheral nerve block, but most symptoms resolve within days or weeks of surgery [1]. Permanent injury is more rare; Auroy and colleagues reported this in an average of only 2.4 instances per 10,000 blocks [19]. There is quite marked variation in the incidence of both temporary and permanent syndromes, with popliteal block having the greatest incidence of permanent injury (31.5 per 10,000 patients). Risk factors can be related to patient, surgical, and anesthetic factors. Patient factors include pre-existing neurological disorder, diabetes mellitus, extremes of obesity, male gender, and extremes of age. Surgical factors include direct surgical trauma, compressive dressings, tourniquet pressure and/or time, compression by hematoma or abscess, and poor patient positioning [20].

Symptoms suggestive of postoperative neuropathy (persisting sensory or motor symptoms beyond 48 h after last dose of local anesthetic) should prompt a full history to identify any predisposing risk factor or causative element in the anesthetic or surgical technique, plus a detailed neurological examination to define the problem precisely. As noted above, most injuries have an excellent prognosis, but the symptoms of even a minor deficit can be distressing, so considerable patient reassurance may be required. More severe deficits, or those which fail to resolve within 2–3 weeks, should be referred to a neurologist or neurosurgeon for further investigation and management. However, it is important that the anesthetist is fully involved in this process because surgical colleagues (and patients) can, quite correctly, become irritated if the anesthetist fails to follow up such problems. Conversely, the anesthetist (or department) taking this seriously will build a relationship between colleagues that will, in turn, facilitate the practice of regional anesthesia and future referral if necessary.

Having made that point it is important that the evaluation of any postoperative neurological deficit includes a search for factors unrelated to anesthesia technique. The incidence of nerve injury arising during surgery is several orders of magnitude greater than during regional anesthesia, so this must be considered before assigning responsibility to any one technique or practitioner. For example, the risk of nerve injury during total hip arthroplasty is as high as 1–2 % [21], and similar estimates are quoted for other orthopedic procedures. Surgery may predispose to nerve injury through direct trauma or stretching, tourniquet pressure, compressive dressings, hematoma or abscess formation, and improper patient positioning during surgery. However, some of these are considered to be the joint responsibility of surgeon and anesthetist .

As with central blocks, thorough preoperative assessment and careful attention to the detail of block technique are essential in the prevention of neurological injury after peripheral blocks. It also seems advisable to use the lowest concentration of local anesthetic possible, particularly when infusions are used for postoperative analgesia.

More specifically, evidence suggests that a major factor predisposing to peripheral nerve injury is intraneural (or more accurately intrafascicular) needle placement or injection. Traditional teaching states that both of these produce severe pain and should therefore be easily detectable, so that needle position can be corrected if it occurs. However, recent studies suggest that neither intraneural needle placement nor injection is always painful, with ultrasound suggesting that they have been performed in an unrecognized manner for many years. Robards and colleagues [22] found that when typical final currents (0.2–0.4 mA) were achieved for popliteal nerve block the needle tip was intraneural in 100 % of cases. Similarly, Macaire and colleagues [23], using nerve stimulation for median nerve block at the wrist, identified that the needle tip was intraneural on several occasions, with this placement explaining the faster block onset seen in their nerve stimulation group .

Such observations may have led to the suggestion that intentional intraneural injection might optimize success without invariably leading to injury [24]. However, a paucity of evidence remains, even with from animal models, to support deliberate intraneural injection [25]. Further, it is difficult to avoid the conclusion that the high incidence of neuropathy after popliteal block [1] is directly related to the ease with which intraneural needle placement occurs during this technique [22]. Until there is much evidence to the contrary, intraneural needle placement and solution injection are practices to be avoided. Recent evidence suggests that several factors, apart from a gentle technique and inserting the needle with the bevel parallel to the nerve, may reduce the likelihood of severe injury, all of them aimed at preventing accidental intraneural injection (Table 11.3).

Many of the warning symptoms of direct nerve contact outlined above require a conscious, or only mildly sedated, patient to report them, and this would suggest that blocks should not be performed after administration of heavy sedation or anesthesia. However, this is a matter of some controversy .