Postoperative Neurosurgical Care

Postoperative cases account for a large number of neurological intensive care unit (neuro-ICU) admissions and in many centers there is a flexible relationship between neurosurgery recovery or postoperative wards and neuro-ICUs. Once the effects of anesthesia have receded, a neuro-ICU offers more intense surveillance for anticipated problems than is possible on the usual ward. Because few of the anticipated postoperative complications actually occur, a large number of patients produce only a few patient days. Knaus and colleagues (1) found that of all postoperative patients admitted to a neuro-ICU solely for “concentrated nursing care and intensive monitoring,” only 15% actually required and received active treatment. For example, of 82 of their electively admitted neurosurgical patients, only one received specific therapy for a complication and none had abnormalities detected by physiologic monitoring. This can be only roughly translated to experience in other centers, depending on how many patients are admitted for trauma and how many early aneurysmectomies are performed. They suggested that additional staffing on general wards is a more efficient way of caring for postoperative neurosurgical patients than routinely admitting patients to a neuro-ICU. Because there will undoubtedly be increasing scrutiny of the use of neuro-ICU beds for these “anticipatory” admissions and because the disparate problems associated with each type of operation must be clearly delineated to be able to detect them, it is useful to list the main postoperative complications (Table 10.1). Knowledge of many of these is embedded in neurosurgical operative practice because they derive from the details of the procedure; they are reviewed in neurosurgical and other textbooks (2,3), as well as in other sections of this book, particularly Chapter 16 on the neuro-ICU management of brain tumors. Some complications are generic, having to do with raised intracranial pressure (ICP), respiratory deterioration, and the general medical problems that arise during the postoperative period. Among the latter, the ones of most concern are the unpredictable difficulties that may occur after any operation, namely myocardial infarction (MI), pulmonary embolus (PE), and pneumonia. This chapter reviews the more specific postoperative neuro-ICU problems that are common to several neurosurgical operations. Less frequent ones, such as malignant hyperthermia, respiratory complications, or wound problems, are either too rare to include here, or are primarily surgical problems that are beyond the scope of the chapter but are covered well elsewhere (4).

ANESTHETIC EFFECTS IN NEUROSURGICAL PATIENTS

The main problem here is in distinguishing the residual effects of general anesthesia from the drowsiness and confusional state that are indicative of a postoperative intracranial hematoma or brain swelling. This distinction is particularly vexing because there has been a general impression in neurology that patients with cerebral diseases are excessively prone to the effects of anesthetics. However, our observation has been that there is little or
no enhancement or prolongation of the effect of anesthesia in patients who were fully awake preoperatively. Focal neurological deficits may worsen in the first hours after anesthesia as a result of the pharmacologic agents, but it is perhaps more surprising how little anesthesia affects focal syndromes. For example, patients who have had a stroke in the past generally show little deterioration after undergoing general anesthesia. The same cannot be said for demented patients who regularly have an increase in confusion in relation to general anesthesia (5). Nonetheless, the effects of anesthesia on the nervous system are complex, often involving a combination of depressant and excitatory influences, depending on the agent used.

TABLE 10.1. Neurosurgical procedures and their main postoperative complications

	Complication
Operation	Immediate	24-48 h	Management
Craniotomy/tumor (Chapter 16)	Cerebral edema Cerebral hemorrhage	Subgaleal cerebrospinal fluid leak	Avoid hypertension Slow steroid taper
	Subdural hemorrhage	Vasospasm	Fluid restriction
			Computed tomography and clinical follow-up
			Intracranial pressure monitoring when indicated
Aneurysmectomy (Chapter 15)	Stroke from vascular manipulation or clip	Vasospasm Cerebral edema	Blood pressure and aggressive fluid management (Chapter 3)
			Clinical monitoring
Arteriovenous malformation resection	Hemorrhage	Cerebral edema	Avoid excessive hypertension Fluid restriction if no vasospasm
Transsphenoidal hypophysectomy	Diabetes insipidus	Diabetes insipidus	Monitoring visual acuity
		Nasal cerebrospinal fluid leak	Preoperative endocrine evaluation and postoperative replacement
		Visual loss
Carotid endarterectomy	Hypotension	Hypotension, myocardial infarction, neck hematoma	(see text)
Tracheotomy	Hypocarbia	Wound drainage	Adjust ventilator
Gastrostomy	Bleeding		Wound care
	Airway occlusion
	Cuff leak
C-P angle tumor	Epidural hematoma	Hydrocephalus	Follow-up CT
		Apnea	Clinical monitoring
		Aspiration
C-P, cerebellopontine; CSF, cerebrospinal fluid; CT, computed tomography.

One of the most instructive studies on the effects of anesthesia was conducted by Rosenberg and colleagues (6). They recorded the neurological signs in patients without underlying neurological deficits who were awakening from halothane, enflurane, and nitrous oxide-narcotic (“balanced”) anesthesia. Not surprisingly, the pupillary response to light and the lash reflex were eliminated under general anesthesia. Four of 27 patients had depressed pupillary responses and three had a depressed lash reflex 40 minutes after the anesthetic. Both reflexes had returned to normal by the time the patients were fully awake (able to follow verbal commands) regardless of the anesthetic. Other reflexes responded differently to various anesthetics. Unsustained clonus was seen in virtually all patients after volatile inhalation anesthesia and in about half after nitrous oxide-narcotic anesthesia, most prominently during the period of unresponsiveness immediately after anesthesia. Sustained ankle
clonus was not seen after nitrous oxide-narcotic anesthesia, but occurred in half of patients after enflurane. Hyperactive quadriceps tendon reflexes occurred in 58% of patients after enflurane and in 37% after halothane but not after nitrous oxide-narcotic. Babinski signs occurred in half the patients anesthetized with Ethrane and in one fourth after halothane but not after nitrous oxide-narcotic anesthesia. The plantar response became flexor as patients awakened. Shivering, most common after enflurane anesthesia, was more frequent during periods of unresponsiveness but was not related to body temperature. A prospective study by McCulloch and Milne (7) found that transient neurological signs were more common in general surgical patients who had received enflurane-nitrous than after isoflurane-nitrous anesthesia. In their experience, quadriceps hyperreflexia and Babinski signs were most the most prevalent signs 5 to 20 minutes after anesthesia and always resolved within 1 hour in neurologically normal patients.

We have seen the following transient neurological signs in neurosurgical patients during the hours after anesthesia, all apparently related to the anesthetic rather than the procedure: Babinski signs (not previously present) up to 2 hours postoperatively; unilateral pupillary dilation in awake patients (possibly an Adie pupil phenomenon); eccentric pupil; mild worsening of previous hemiparesis for 1 to 2 hours postoperatively; worsened dysarthria; and, of course, asterixis. In contrast, rapidly increasing headache (although transient headache may occur after general anesthesia), especially with vomiting, progressive drowsiness, or evolving hemiparesis, new or worsening paresthesias, vertigo, facial paresis, or pupillary changes in a fully awake patient have almost invariably reflected neurological complications such as subdural or epidural hematoma or brain edema. It can be said in general that any progressive or fluctuating deterioration can be assumed to be from an operative complication rather than from anesthesia. Bradycardia likewise suggests a cerebral hemorrhage or brain swelling. Anticonvulsant toxicity often clouds the issue of postoperative deterioration. In our experience, the perioperative intravenous (i.v.) administration of phenytoin has caused postoperative hiccoughing and vomiting, sometimes with drowsiness or with slight agitation and mild confusion. Serum drug concentrations measured several hours later generally have been normal, but we have thought that transiently toxic levels caused the symptoms. There may be an additive effect of anticonvulsants with residual anesthetic, particularly in causing hiccoughs. Asterixis is also to be expected if anticonvulsant levels reach toxic levels and it should be kept in mind that an initial loading dose of phenytoin, even 1 g in a previously unexposed patient, may give rise to slight drowsiness and asterixis. Unilateral asterixis is indicative of a hemiparesis on the silent side or an anterior thalamic lesion on the affected side.

Another special problem arises in patients who have persistent weakness or ophthalmoplegia for hours after the administration of nondepolarizing neuromuscular blockers. Usually the remainder of the clinical state (alert, reactive pupils, flexor plantar response) make the benign nature of the weakness evident. However, ophthalmoplegia should under no circumstances be attributed to neuromuscular blocking agents or emergence from anesthesia without further consideration. We have seen four cases of basilar artery thrombosis that arose during the postoperative period, signs of which were attributed to anesthesia initially. In addition, rare cases of previously unrecognized myasthenia gravis may be uncovered by the appearance of prolonged postoperative ophthalmoplegia, weakness, or respiratory failure after the use of muscle relaxants. Problems with pseudocholinesterase deficiency are well known to practitioners and may cause similar problems after the use of succinylcholine.

In cases of prolonged unresponsiveness after anesthesia there may be concerns regarding an intraoperative ischemic or anoxic accident, or in the case of cardiac surgery,
multiple cerebral emboli. Somatosensory evoked potentials can distinguish an anatomic lesion in the hemispheres or global ischemia from an anesthetic effect because these responses are relatively unaltered by anesthesia (Chapter 8). The electroencephalogram (EEG) or its several computer processed derivatives generally do not distinguish among global cerebral ischemia, anesthesia, and a metabolic encephalopathy, although subtle differences in the processed signal may be helpful. The patient who is paralyzed and has ophthalmoplegia from neuromuscular blockers will have a normal EEG, of course.

Only gold members can continue reading. Log In or Register to continue