Postoperative Coma and Delayed Awakening

Failure to arouse and delayed awakening are the most common early neurologic problems following general anesthesia. True prolonged postoperative coma is relatively uncommon, with estimates ranging from 0.005 to 0.08 percent following general surgery, but with higher rates reported after cardiac surgery. Advanced age, urgent surgery, preexisting brain disorders, perioperative hypotension, postoperative organ failure, and sepsis are independently associated with postoperative coma.

Delayed emergence from anest hesia is the most common cause of early failure to regain alertness after surgery; although this situation is benign, more serious alternative causes include stroke, anoxic-ischemic brain injury, and status epilepticus. Sepsis and multiorgan failure may produce severe encephalopathy and prolong unresponsiveness, sometimes for days after their resolution. The main causes of impairment of consciousness following surgery are found in Table 55-2 .

When a neurologist is asked to evaluate an unresponsive patient following surgery, the principal goal is to determine whether any form of severe acute neurologic disease is responsible. It is therefore important to obtain a detailed history of the patient’s preoperative status, review pertinent intraoperative information, and clarify the condition of the patient since surgery. In these situations, direct communication with the anesthesia and surgical teams may provide additional relevant information, such as the presence of clinical features suggesting seizure occurrence or a history of immediate awakening followed by unconsciousness, suggesting cerebral embolism or seizures. It should be determined whether the patient became hypoxemic, hypotensive, or had marked lability of blood pressure; profuse blood loss; poor hemostasis; or cardiac arrhythmias during or immediately after surgery. Intraoperative hypotension and fluctuations in hemodynamic parameters are common in patients with postoperative coma after cardiovascular surgery.

General clinical examination is important. A slow respiratory pattern and darkened skin color raises the possibility of CO ₂ narcosis. An irregular pulse may signify atrial fibrillation and its increased risk of postoperative stroke. Splinter hemorrhages in the conjunctiva or nail bed suggest fat embolism after long-bone fracture surgery. The neurologic examination should focus on the assessment of level of consciousness, brainstem reflexes, gaze (conjugate versus dysconjugate, presence of skew deviation), eye movements (roving, ocular dipping or bobbing, nystagmoid jerks), muscle tone, response to central (sternal or supraorbital) and peripheral (limb) pain, and presence of adventitious movements. The presence of lateralizing signs requires brain imaging. Brainstem findings or bilateral focal signs should raise suspicion of basilar thromboembolism, prompting vessel imaging. Subtle abnormal movements of the eyes, facial muscles, or fingers may be the only manifestations of underlying status epilepticus and necessitate urgent electroencephalography (EEG). Drug toxicity should be suspected in patients with rigidity, hyperreflexia, and tremors, with or without concurrent fever. Myoclonus may occur with any toxic or metabolic encephalopathy, but can also follow severe global brain anoxia. Asterixis may be a sign of hyperammonemia but also occurs in other metabolic encephalopathies and as a side effect of medications. In patients lacking motor responses after pharmacologic neuromuscular blockade, it must be determined—sometimes by electrophysiologic nerve-stimulation studies—whether blockade is continuing; when paralysis is persistent, administration of an antidote may require consideration.

Additional testing is directed by the history and physical findings. If reassuring and nonfocal, further investigations can be postponed. Brain imaging is indicated in patients with new focal deficits. Computerized tomography (CT) scan is reliable for the detection of acute hemorrhage and can document territorial ischemic infarctions and massive brain edema. Magnetic resonance imaging (MRI) may be necessary to diagnose an embolic shower and subtle manifestations of global anoxic-ischemic injury such as early cortical injury. In patients with brainstem signs, MRI of the brain can be combined with MR angiography to exclude vertebrobasilar occlusion. EEG should be reserved for those with suspected seizures, and the use of continuous monitoring then may increase the diagnostic yield. Lumbar puncture is rarely necessary in patients with postoperative encephalopathy unless central nervous system (CNS) infection is a possible cause. Serum studies should be used judiciously, including arterial blood gas to exclude CO ₂ retention, and serum glucose, sodium, creatinine and blood urea nitrogen, and ammonia to exclude various metabolic disturbances.

A history of delayed arousal after a previous surgery makes a diagnosis of delayed emergence from anesthesia more likely. Older age and renal and hepatic insufficiency may compromise drug metabolism and clearance, increasing the risk of delayed arousal. Review of the anesthesia course may guide the use of naloxone (for reversal of opioid effect), flumazenil (to reverse benzodiazepines), and physostigmine (to reverse neuromuscular blockade) in appropriate cases.

The prognosis of patients with postoperative coma varies markedly depending on etiology. Although the prognosis is favorable in patients with reversible toxic or metabolic causes, the mortality may exceed 80 percent in patients with multifocal infarctions revealed by brain MRI. In survivors of prolonged postoperative coma, long-term functional outcome is frequently poor.

Postoperative Delirium

Postoperative delirium is common, with a reported incidence as high as 40 to 60 percent of patients. Delirium can occur immediately following anesthesia or after some interval from a seemingly normal recovery from the anesthetic. The latter is more commonly persistent and often multifactorial in etiology. The risk of delirium is increased in the elderly and in patients with previous cognitive impairment, a history of substance use (including alcohol), and those receiving polypharmacy before surgery and narcotics or benzodiazepines perioperatively. A previous episode of postoperative delirium predicts recurrence. Perioperative hypoxemia and hemodynamic instability, profuse intraoperative blood loss, and postoperative organ failure may also increase the risk of postoperative delirium.

There is limited evidence that the type and route of anesthesia influence the risk of postoperative delirium, with higher rates following general anesthesia. Deeper intraoperative levels of sedation with propofol increase the risk of postoperative delirium in elderly patients undergoing surgery for hip fracture. Direct alteration of brain function by the anesthetic agent may be responsible.

Evaluation of patients with postoperative delirium begins by considering risk factors for its development (see Table 54-2 in Chapter 54 ). Time should be spent reviewing the medications that the patient was taking prior to surgery, and those administered during and after the operation. Toxic and metabolic factors are the most frequent causes. Exclusion of acute structural brain injury, seizures, and meningoencephalitis (an exceptional cause of perioperative coma) is the primary responsibility of the consulting neurologist. Primary neurologic diagnoses to be considered include ischemic infarctions (often bilateral or in the brainstem), intracranial hemorrhage (intraparenchymal, subdural, or rarely subarachnoid), nonconvulsive status epilepticus, and posterior reversible encephalopathy syndrome.

Management of the delirium depends on its cause. Discontinuation of sedatives and narcotics is often the only intervention required. When agitation is present and patients may harm themselves or others, oral quetiapine, olanzapine, or, in more severe cases, intravenous haloperidol may be necessary. Dexmedetodimine, a central α ₂ -adrenergic agonist, is particularly useful for additional sedation or when agitation is accompanied by sympathetic hyperactivity. Benzodiazepines may worsen delirium and therefore should be reserved for specific indications, such as alcohol withdrawal.

In most cases, postoperative delirium is transient and reversible. Complications sometimes stemming from it include aspiration, self-removal of the endotracheal tube or indwelling catheters, and delays in mobilization that, in turn, result in an increased risk of thromboembolism, infections, and deconditioning. In addition, the occurrence of postoperative delirium is related to a greater risk of postoperative cognitive dysfunction, particularly during the first few months following surgery.

Cognitive Dysfunction

Postoperative cognitive dysfunction occurs more often in older patients and in those with postoperative delirium. Its incidence is greatest after major cardiovascular surgery, ranging from 30 to 80 percent during the first few weeks after surgery and 10 to 60 percent after 3 to 6 months. Additional risk factors include alcohol abuse, lower educational level, and previous stroke. Patients affected by this complication, even when transient, may have worse long-term cognitive outcomes, greater disability (particularly in the elderly), and a higher mortality. Perhaps surprisingly, there is no definitive evidence that the choice of anesthetic regimen (such as general or regional anesthesia) or the depth of anesthesia alters the risk of postoperative cognitive dysfunction, which has been related to a dysregulated neuroinflammatory response. Maintenance of physiologic homeostasis during and after surgery may minimize the risk of this complication.

Paraparesis

Epidural hematoma is probably the most feared complication of spinal and epidural anesthesia ( Table 55-3 ). Although it is rare, patients at higher risk include those with advanced age, alcoholism, anatomic abnormalities of the vertebral column leading to difficulties with needle placement, and, most importantly, coagulopathy due to a bleeding diathesis or anticoagulant effect.

Symptoms of epidural hematoma typically present within the first few hours after the surgery. Attributing them to anesthetic effects may delay the diagnosis and decrease the chances of recovery after surgical evacuation. The neurologist should contact the anesthesiologist to clarify the anesthetic approach used, the effects of the drugs administered, and the expectation of the duration of their effect. In the acute phase, it may be difficult to differentiate between a compressive epidural hematoma and spinal cord ischemia. Preservation of posterior column sensation is expected with an anterior spinal artery infarction. Spinal CT scan allows visualization of the hematoma in most cases, but its sensitivity is suboptimal when hardware is present. Therefore, MRI of the spine should be obtained urgently if possible. Urgent surgical evacuation of the hematoma may result in excellent functional recovery when undertaken promptly.

The American Society of Regional Anesthesia and Pain Medicine has published recommendations for the management of patients receiving antithrombotics who require regional anesthesia. Warfarin should be discontinued 5 days before planned regional anesthesia, and anticoagulation with warfarin should be reversed until the international normalized ratio (INR) is 1.5 or less before urgent interventions. Patients at high risk of thromboembolism should receive bridging with unfractionated heparin (to be stopped 4 hours before surgery) or low-molecular-weight heparin (reduced to half of the usual daily dose and stopped 24 hours before surgery). After surgery, warfarin can be resumed on the first day and heparin products restarted after 24 hours following minor surgeries or 48 to 72 hours after major surgeries. Although continuation of aspirin is acceptable if necessary, clopidogrel should be discontinued 5 to 10 days before the surgery and can be restarted 24 hours after it. These recommendations do not provide guidance for patients receiving any of the newer anticoagulants (e.g., dabigatran, rivaroxaban, apixaban). Depending on the estimated bleeding risk, it is prudent to discontinue dabigatran 3 to 5 days before the procedure. The factor Xa inhibitors rivaroxaban and apixaban should be discontinued 1 to 3 days before the intervention depending on the estimated bleeding risk.

Epidural abcess can also cause postoperative paraparesis, but its presentation is usually delayed (at least for days after surgery), and cases exclusively related to neuroaxial blockade are rare ( Table 55-4 ).

Neuropathy

The incidence of neuropathy has been estimated to be less than 0.04 percent after neuroaxial blockade and less than 3 percent following peripheral blockade. Neurologic symptoms and deficits are typically transient. Although prospective studies with blinded assessment are lacking, persistent postoperative neuropathy after regional anesthesia is quite infrequent.

Neuropathies can occur due to various mechanisms, some of which are not strictly related to anesthesia. Mechanical stress from stretch or compression of nerves as a result of malpositioning during surgery is a common cause; neuropathies are more likely to occur after prolonged surgeries. Other risk factors include male gender, diabetes mellitus, smoking, hypertension, vascular disease, obesity or extremely thin body habitus, and preexisting clinical or subclinical nerve dysfunction. Some of these associations suggest that ischemic and metabolic insults play a role in pathogenesis.

The most frequently affected nerves in the upper extremities are the ulnar nerve and brachial plexus and in the lower extremities the sciatic, femoral, and fibular (peroneal) nerves and the lumbosacral plexus. The predisposition of these particular nerves probably relates to their susceptibility to stretch and compression. The most common deficits that result are pure sensory and sensorimotor disturbances. Positioning strategies to reduce the risk of each of these peripheral neuropathies have been outlined. Avoidance of arm abduction beyond 90 degrees in supine patients diminishes stretch of the brachial plexus. Minimizing pressure over the ulnar groove, avoidance of excessive flexion of the elbow ( Fig. 55-1 ), and use of protective elbow padding are recommended to reduce ulnar nerve injuries. Restricting stretch of the hamstring muscle and the degree of hip flexion when the patient is in the lithotomy position lessens the risk of sciatic neuropathy. Perioperative fibular (peroneal) neuropathies can be averted through careful avoidance of prolonged pressure over the fibular head and the use of protective padding. Limb position and external compression play a lesser role in the pathogenesis of perioperative femoral neuropathies, which are attributed most often to damage from inappropriate retractor placement ( Fig. 55-2 ).

Typical anatomic site of ulnar nerve compression during surgery.

Illustration of the common mechanism of compression of the femoral nerve through pressure applied with the surgical separator.

Cauda equina syndrome is a rare complication of spinal anesthesia, attributed speculatively to maldistribution of the local anesthetic into the subarachnoid space. Most reported cases have occurred after continuous spinal anesthesia, and the predominant manifestations are related to the S1 nerve roots, presumably because of their more posterior location.

Full recovery from these neuropathies is often seen after days to weeks, but some patients suffer persistent disability. There is insufficient information to guide prognostication, except that electrodiagnostic evidence of demyelination (conduction block) rather than axonal loss implies a better prognosis than otherwise.

Transient Neurologic Symptoms

A syndrome known to anesthesiologists as transient neurologic symptoms is characterized by early postoperative pain in the buttocks and legs after spinal anesthesia. The pain, which may be severe, typically develops after an initial asymptomatic period of a few hours and is frequently bilateral. The syndrome resolves spontaneously within 2 to 5 days. The risk of this complication is higher with lidocaine and mepivacaine than with other local anesthetics and is not influenced by the dose or concentration of the drug. The main differential diagnoses include cauda equina syndrome, aseptic meningitis, adhesive arachnoiditis, and epidural hematoma or abscess. Imaging reveals no structural abnormality that can explain the symptoms, and the etiology remains uncertain.

Visual Loss

Posterior ischemic optic neuropathy is a rare complication of complex spine surgeries that require prolonged prone positioning, associated with intraoperative blood loss and hemodynamic instability. The performance of staged surgeries, a more liberal intraoperative transfusion strategy, avoidance of intraoperative hypotension, and maintenance of the head at or above the level of the heart have each been recommended, but the etiology of the condition is unclear and the value of these strategies is unproven. Although the visual impairment can be transient, it is often severe and bilateral and may be permanent in some cases. Cardiac surgeries also may be complicated in rare instances by visual loss, typically due to anterior ischemic optic neuropathy or retinal embolism. Anesthesia technique does not appear to be associated with the risk of perioperative visual loss.

Hearing Loss

Ear blockage may occur after general anesthesia with nitrous oxide in patients susceptible to eustachian tube obstruction. Presumably mucosal edema facilitates the development of the condition. Prognosis is favorable, with resolution over hours to a few days.

Transient hypoacusis, typically subtle, may follow spinal anesthesia. The proposed mechanism is a decrease in cerebrospinal fluid volume reducing perilymphatic pressure and leading to endolymphatic hydrops. The hypoacusis is frequently, but not always, associated with headache. Use of smaller-diameter spinal needles may reduce the risk.

Cranial Nerve Palsies

Abducens nerve palsy has been described after spinal anesthesia, and it can be unilateral or, rarely, bilateral. Typically accompanied by positional headache, it occurs in cases with substantial cerebrospinal fluid loss. The risk is related to the diameter of the spinal needle.

Pupillary changes may follow neuroaxial blocks. Horner syndrome sometimes occurs after epidural anesthesia, typically after labor; the condition lasts only for the duration of the anesthetic block. It can also occur transiently after brachial plexus blockade. Anisocoria found after surgery should not be attributed to anesthetic blockade except when the patient is fully lucid and the examination clearly reveals an isolated Horner syndrome.

Bilateral pupillary dilatation in the immediate perioperative period may be caused by “total” spinal anesthesia, a condition provoked by the unintentional injection of the anesthetic drug into the subarachnoid space. Marked mydriasis and unconsciousness are common, and the problem is typically recognized in the operating room.

Exceptional cases of unilateral and transient trigeminal and facial nerve dysfunction have also been reported after epidural anesthesia for women in labor.