Critical Care of the Patient With Acute Stroke

Initial Assessment of Patients With Stroke

Initial clinical assessment of patients with severe stroke should focus on the following issues:

• 1.

  Vital functions (airway, pulmonary function, heart rate, blood pressure [BP])

• 2.

  Neurologic symptoms, severity of neurologic deficit based on validated stroke scales

• 3.

  Time of symptom onset, potential eligibility for specific treatment options

• 4.

  Blood sampling for electrolytes, full blood count, and coagulation studies

Many emergency measures depend on the type of stroke (ischemic versus hemorrhagic). Thus, appropriate neuroimaging studies must not be unnecessarily delayed. Additionally, measures should be avoided that have the potential to interfere with further treatment options (for example, insertion of a central venous line in a patient eligible for thrombolysis).

Ancillary Tests and In-House Transportation

Diagnostic studies and their indication in stroke patients are discussed in detail elsewhere in this book. As a general rule, the need for further diagnostics has to be weighted carefully against the potential risk that in-house transportations bear for the critically ill. These procedures usually require disconnection from the ventilator to a transportable ventilator or those suitable for the use in magnetic resonance imaging (MRI) scanners, which may not be tolerated easily by patients with severe pulmonary dysfunction. Moreover, monitoring is not as good as in the ICU, and the options for intervention during critical situations may be limited. Before transportation of a critically ill patient, one should always question the therapeutic consequences that are likely to be drawn from the results. If there are no consequences, one should resign from the procedure. Diagnostic procedures that do not require transportation of the patient should be preferred. Optimal neuromonitoring as will be described later in the chapter may help to reduce the need for imaging scans.

If a diagnostic or therapeutic procedure requiring transportation of the patient cannot be avoided, careful preparation is obligatory. The physician has to decide about the medications that may be discontinued and the adequate monitoring that has to be brought along; catheters need to be fixed properly, and extraventricular drains should be closed during transportation to avoid overdrainage. Emergency medication should be brought along, and unnecessary delays should be avoided. Deeper sedation frequently becomes necessary for transportation and to allow for acquisition of diagnostic studies of acceptable quality. However, since clinical evaluation of the patient may be markedly limited thereafter, short-acting sedatives should be preferred in these situations. All patients with severe stroke have to be accompanied by a physician or trained physician-extender during all diagnostic procedures.

Clinical Examinations

Neurointensive care patients should have a complete clinical examination at least three times per day and even more frequently (i.e., every 5 minutes to 1 hour) in the hyperacute phase of the disease or under particular circumstances. Level of arousal is the key focus of the exam; however, due to analgesics and sedation, the neurologic exam may frequently be restricted to evaluation of pupillary and brainstem reflexes, motor reaction upon painful stimuli, reflex status, and pathologic reflexes. Otherwise, care should be taken to assess a patient’s alertness, responsiveness to verbal stimulation, ability to track and regard, as well as comprehension in following simple commands. An acute worsening of mental status may be the first sign of neurologic deterioration. Other signs of raised intracranial pressure (ICP) and transtentorial herniation such as sequential loss of brainstem reflexes have to be recognized immediately. New motor deficits or pathologic reflexes may indicate enlargement of infarction. Changes in ventilator settings such as the need to switch from assisted ventilation to fully controlled mechanical ventilation under stable levels of sedation may indicate loss of brainstem function. Auscultation of heart and lung, palpation and auscultation of the abdomen, and careful inspection of the patient (edema, signs of dehydration, skin lesions, wounds) complete the clinical exam. Additionally, ventilator settings, blood gas analyses, laboratory parameters, temperature, urine excretion, and central venous pressure have to be reviewed regularly by the attending physicians.

Pulmonary Function and Mechanical Ventilation

Maintenance of adequate oxygenation is essential in patients with acute stroke as hypoxia could be deleterious for the ischemic penumbra. Avoidance of hypo- and hypercapnia is of equal importance, as the first may cause cerebral vasoconstriction and secondary ischemia and the latter to vasodilation in the cerebral arterioles supplying healthy brain tissue, possibly raising ICP and stealing blood from the lesion site, where cerebral vessels are already maximally dilated under resting conditions. The need to secure the airway and mechanically ventilate a patient with severe stroke may be one of the main reasons, if not the main one, to transfer a patient to the NCCU. Several publications in the 1990s suggested that cerebrovascular patients requiring mechanical ventilation have a very bad prognosis and questioned its usefulness while other studies indicated that a considerable part of these patients, even those on long-term ventilation, can have a good outcome. ^, Of note, mechanical ventilation in these studies was almost invariably just an indicator of illness severity, and neither described in its details nor seen as a potent management tool. More recent studies have explored the utility of endotracheal intubation in the setting of mechanical thrombectomy, ^, as will be discussed later, as well as early tracheostomy in ventilated stroke patients. Today, with evidence from several studies that treatment on specialized NCCUs improves outcome, with more treatment options for cerebrovascular patients, as well as advances in ventilation techniques, the only justifications to withhold these life-saving procedures from neurocritical stroke patients are obvious futility or the patient’s explicit will. Since the further course and eventual outcome are often unclear in the acute phase of the disease, the physician in charge should initiate airway management and ventilation without delay when indicated. Although standard, basic principles of airway and ventilation management from the general ICU can mostly be applied to the NCCU patient as well, the NCCU stroke patient merits some more specific considerations.

There are distinct neurologic disease–related causes of respiratory failure. All sorts of severe cerebrovascular compromise of the central nervous system, such as supra- and infratentorial acute ischemic (AIS) or hemorrhagic stroke (HS), subarachnoid hemorrhage (SAH, covered elsewhere in this book), or cerebral venous and sinus thrombosis (CVST) and their sequelae, such as hydrocephalus and/or increased ICP, can cause respiratory failure. The complex connections between the central respiratory centers, that is, the cortex and the autonomic centers in pons and medulla, as well as their connections to the phrenic nerve and the upper motor neurons, can be affected on every level. This does not necessarily (only) result in loss of respiratory drive or respiratory rhythm but may also cause loss of protective airway reflexes and airway patency, and thus impair ventilation. Major reasons for intubation and ventilation in stroke patients are decline in level of consciousness (LOC), loss in glossopharyngeal muscle tone, or loss of protective reflexes and dysphagia with risk of aspiration ( Box 56.1 ). Specific patterns of pathologic breathing (e.g., Cheyne-Stokes, Cluster, Biot) have beensuggested for topographic diagnosis of lesion levels. However, that correlation seems to be less consistent in clinical reality.

During the course of the disease, however, respiratory function can be further compromised by the development of various pathologic conditions, including hypoventilation-associated atelectasis or pneumonia due to immobilization, decline in LOC, epileptic seizures, critical illness polyneuropathy, and stroke-related immunosuppression. About 10% of unselected patients with stroke have to undergo mechanical ventilation. Differences in the rate of mechanical ventilation have been related to different causes of stroke; Mayer and associates reported that 5% of patients with AIS, 26% of patients with intracerebral hemorrhage (ICH), and 47% of patients with SAH underwent mechanical ventilation, whereas Gujjar and colleagues found rates of 6% for AIS and 30% for ICH. In 54 out of 218 patients with AIS who needed mechanical ventilation, Berrouschot et al. found that 90% had to be ventilated for neurologic deterioration and 10% for cardiopulmonary compromise. Clinical signs of impending respiratory failure are tachypnea exceeding 35 breaths/min, dyspnea with use of accessory muscles, sweating, and paradoxical breathing. Transcutaneous pulse oximetry for assessing arterial saturation of oxygen (SpO ₂ ) constitutes the minimal requirement for monitoring of pulmonary function. Enhanced intermittent arterial blood sampling for blood gas analysis is recommended in suspected respiratory compromise. Intubation should be undertaken as soon as clinical signs of pulmonary dysfunction or aspiration risk are present or SpO ₂ drops below 90% despite O ₂ application, PaO ₂ values drop below 60 mm Hg, PaCO ₂ values exceed 60 mm Hg, or both ( Box 56.2 ). A prospective observational study in patients with large hemispheric infarction (LHI) suggested Glasgow Coma Scale (GCS) less than 10 or respiratory failure as indications for intubation with a history of hypertension and infarct size greater than two-thirds middle cerebral artery (MCA) territory as additional predictors of need for mechanical ventilation. The principles of noninvasive airway support by head positioning, airway clearing, and application of nasal prongs or an oronasal mask for O ₂ insufflation, and possibly the insertion of oro/nasopharyngeal airways, apply to almost all stroke patients with respiratory deterioration in the beginning and might be followed by intubation later. For patients with less severe stroke these noninvasive measures might suffice and be followed by noninvasive ventilation (i.e., BiPAP). It has to be confirmed regularly, however, whether those patients are still awake, cooperative, and have not lost their airway protective reflexes or developed other indications for intubation. Otherwise rapid initiation of the latter is warranted.

Orotracheal intubation, although the approach of first choice, almost always involves an episode of hypotension or at least variance in BP. This can be detrimental in cerebrovascular disease in case of impaired cerebral autoregulation and subsequently decreased cerebral perfusion pressure (CPP). Hypotension during pharmacologic induction for intubation is more common in patients with more severe underlying disease, a baseline mean arterial pressure (MAP) less than 70 mm Hg, age greater than 50 years, and with use of propofol or increasing doses of fentanyl as induction drugs. Therefore, the less vascular-active etomidate might be more appropriate for induction in cerebrovascular patients; it can be accompanied by fasciculations that should not be taken for seizures. Ketamine as an alternative induction agent was suggested to be associated with increases in ICP, but this was not confirmed in several more recent studies. Contrary to other sedatives, ketamine does not have depressing circulatory effects; instead, it has activating ones. It can thus cause tachycardia and hypertension and should not be used in patients already in the upper ranges of these parameters. The concept of rapid sequence induction (RSI) commonly used in non-fasting patients involves the rapidly acting muscular blocker succinylcholine for excellent intubation conditions (i.e., glottis wide open). The depolarizing succinylcholine has been reported to induce small, but at times (e.g., in traumatic brain injury [TBI]) relevant ICP increases and can cause rhabdomyolysis and hyperkalemia in patients with seizures, or post-immobilization. As the non-depolarizing muscular blocker rocuronium has been found very comparable to succinylcholine in a Cochrane analysis of almost 50 good-quality studies on the subject, rocuronium may be preferred for RSI in neurocritically ill stroke patients; or muscular blockers may be avoided. Any drugs used for intubation should be short-acting. Continuous infusion of sedatives and analgesics is warranted for the duration of mechanical ventilation. Further details on sedation and analgesia are given in the next section of this chapter.

Optimal ventilation management in stroke patients has not been clarified. Ventilation modes and parameter settings should be chosen according to general ICU principles. The following table ( Box 56.3 ) contains suggestions for initial ventilator settings that have to be adapted to the clinical state and blood gas analysis results in the further course.

Oxygenation is the main goal of mechanical ventilation to provide this essential brain nutrient. However, it is important to aim for tissue oxygenation, and not for arbitrary levels of oxygen in the blood. Toxic levels of oxygen in patients that are not hypoxemic should be avoided. There is ample evidence that hyperoxia is associated with tissue damage resulting from free oxygen radical formation, lipid peroxidation, and other mechanisms. Additionally, hyperoxia may disrupt brain perfusion by an incompletely understood process called hyperoxia-related cerebral vasoconstriction that may theoretically lead to secondary ischemia. ^, It thus seems reasonable to aim for normoxemia but not more. Improving oxygenation is not limited to increasing FiO ₂ or the degree of ventilation; it can also be achieved by reducing cerebral oxygen demand by reducing work of breathing, treating infections and fever, controlling agitation, delirium, and shivering, using anticonvulsants in seizures, and/or employing certain sedatives that reduce the cerebral metabolic rate of oxygen (CMRO ₂ ). Normocarbia may be an even more important aim to strive for in the NCCU patient, as PaCO ₂ plays such a prominent role in influencing cerebral blood flow (CBF) via pH changes in both directions as long as cerebral autoregulation is intact (which is often not the case in brain-lesioned patients, but this is difficult to determine). Both hypercarbia, with subsequent fall in pH, cerebral vasodilation, increase in CBF, and rise in ICP, and hypocarbia, with subsequent rise in pH, cerebral vasoconstriction, and decrease in CBF, increase the risk of secondary ischemia. PaCO ₂ fluctuations may be detrimental, depending on the extent and duration of the derangement, the specific neurologic disease, and the stage of injury (acute vs subacute). The concept of lung-protective ventilation with permissive hypercarbia as part of treating acute respiratory distress syndrome (ARDS) may be problematic in brain-injured patients, and neuromonitoring should be installed if this is performed. Hyperventilation may be used transiently to induce hypocarbia and high pH to manage raised ICP rapidly. However, if applied chronically or prophylactically, it may be associated with higher morbidity and mortality in TBI patients, ^, or with cerebral ischemia in SAH and ICH patients. The application of increased positive end-expiratory pressure (PEEP) might theoretically cause raised intrathoracic pressure with subsequently reduced venous return from the brain and thus increased ICP. Alternatively, or additionally, reduced cardiac output and reduced MAP (see above, in itself potentially detrimental via reduced CPP) can lead to raised ICP indirectly via reductions in CBF and brain oxygenation. However, the individual patient’s ICP reaction to raised PEEP seems to vary greatly, possibly according to their lung and ventricular compliance—those with normal or poor pulmonary compliance do not demonstrate relevant ICP crises (TBI and SAH ). In patients with NCCU-dependent stroke, raised PEEP did not transduce into significant ICP changes (but did result in reductions in MAP and thus CPP). Also, PEEP can be absolutely necessary to achieve adequate oxygenation, which is the primary prerequisite for brain integrity and should not be subordinated to potential changes in ICP. Furthermore, no increase in mortality has been linked to the use of PEEP in brain-injured patients so far. In essence, higher PEEP should not be withheld from NCCU stroke patients that are in need of improved oxygenation. However, neuromonitoring should be installed in these patients to detect changes in ICP and thus CPP and be able to take measures to achieve a reasonable compromise, such as by raising MAP.

Changing the I:E ratio to 1:1 or even higher to improve oxygenation has also been thought to reduce venous return from the brain and raise ICP. Studies in ventilated patients with ischemic stroke, ICH, and TBI have not confirmed this hypothesis. ^, Pre-existing pulmonary disorders such as chronic obstructive pulmonary disease (COPD), pneumonia, or the development of ARDS may require more invasive ventilator settings. Lung-protective strategies comprise the application of low tidal volume, low inspiratory pressure, and sufficiently high PEEP levels and should be applied in stroke patients as well, although data suggest that this is neglected when compared to other ventilated NCCU patients.

Theoretically, higher PEEP levels result in higher intrathoracic pressure and reduced venous return and thereby may promote an increased ICP. Moreover, the PEEP could affect CPP by lowering MAP. In a recent study in SAH patients, stepwise elevation of PEEP to 20 cm H ₂ O (14.7 mm Hg) resulted in an increase in central venous pressure and a significant decrease of MAP and regional CBF. However, reduction of CBF depended on MAP changes as a result of disturbed cerebrovascular autoregulation, and normalization of MAP restored regional CBF to baseline values. Likewise, PEEP levels up to 12 mm Hg did not increase ICP in patients with acute stroke. In summary, PEEP application seems to be safe, provided that MAP is maintained. However, monitoring of MAP, ICP, and CPP is desirable.

Weaning from the ventilator should not be delayed in NCCU stroke patients, even if they are still comatose. The best method of weaning—continuous versus discontinuous—is unclear (as in general ICU patients). Discontinuous weaning methods, however, involve successive spontaneous breathing trials and thus wake-up trials. These have been found to be associated with a release of stress hormones and rises in ICP in brain-injured patients, particularly those with a higher ICP from the outset. In a small randomized pilot study, treating ventilated patients with severe stroke by a gradual weaning method employing the adaptive support ventilation mode had a shorter duration of ventilation.

Extubation can only be considered in patients that are on minimal ventilator settings and are cardiocirculatorily stable. A central problem, however, is that classical extubation criteria involve having an awake and cooperative patient, something rarely encountered in the NCCU, where patients might present with aphasia, anarthria, apraxia, agitation, delirium, or reduced LOC, depending on their brain lesion. Classical extubation criteria (see above) have failed to predict extubation failure in NCCU patients, which occurs far more often than in non-neurologic ICU patients, at a rate of 15%–35% in cerebrovascular patients. However, to delay extubation in NCCU patients for not meeting classical extubation criteria, especially the one regarding consciousness, leads to complications, such as more ventilator-associated pneumonia (VAP) and prolonged ICU length of stay (LOS), while earlier or later extubated patents do not seem to differ with regard to the re-intubation rate. Coma should not be the only reason to withhold weaning or extubation from these patients. Rather, particular attention should be paid to presence of dysphagia, which is very frequent in the NCCU population. Endoscopic swallowing tests that do not necessarily demand much cooperation from the patient have been successfully applied in stroke patients and help to guide the extubation decision.

While 10%–15% of ICU patients receive a tracheostomy during their stay, this rate is about 35% in NCCU patients. This may reflect the fact that neurologic ICU patients often are not compromised with regard to their pulmonary function, but rather to their capacity to protect the airway and handle secretions. The question of the optimal time to tracheostomize patients with severe stroke has hardly been studied. A retrospective study suggested that among ICU patients, the neurologic/neurosurgical ones were those fastest to be weaned from the ventilator. Two retrospective studies investigated predictors of tracheostomy need in ICH and found ganglionic location, hematoma volume, hydrocephalus, midline-shift, GCS, and presence of COPD as predictors. ^, The optimal time point for tracheostomy was retrospectively investigated in cerebrovascular patients, ^, the studies suggesting that duration of ventilation and ICU-LOS were reduced in patients receiving earlier tracheostomy. This could not be confirmed in the only prospective randomized trial on early tracheostomy (up to day 3 vs. days 7–14 from intubation) in ventilated NCCU stroke patients. However, that pilot study SETPOINT showed that early tracheostomy is safe, feasible, and reduces sedative demand. Whether benefits in functional outcome can be achieved by early tracheostomy in ventilated stroke patients is currently being investigated in the multicenter, randomized trial SETPOINT2.

Until the potential benefits of early tracheostomy in NCCU patients are clarified, it is probably reasonable to proceed to tracheostomy as part of a weaning protocol if extubation trials failed or were deemed not feasible, i.e., the patient appears to require mechanical ventilation for more than 1 week.

Sedation and Analgesia

All patients treated in the ICU are exposed to various stress factors, including anxiety, unfamiliar auditory and visual stimuli, awareness of severe illness, and sleep disturbances, and stroke patients are no exception. Medical conditions such as pain, respiratory insufficiency, cardiovascular impairment, and sepsis constitute further stress factors. Furthermore, mechanical ventilation via an orotracheal tube, invasive therapeutic, or diagnostic procedures can be equally stressful and are often not feasible without proper analgesia and sedation. Inadequate sedation and analgesia can cause agitation and combativeness, which results in greater metabolic rate and oxygen consumption, potentially supporting secondary brain damage. General goals of analgesia and sedation are freedom from pain, anxiety, and agitation. In stroke patients requiring NCCU treatment, some more specific goals have to be added: these comprise preservation of CPP, reduction of ICP, compensation of disturbed autoregulation, avoidance of seizures, the option of neurologic examination, and the prevention of prolonged coma and delirium, the latter of which brain-injured NCCU patients are particularly prone to. Risks of sedation and analgesia are circulatory compromise, immunosuppression, gastrointestinal tract (GIT)-disorders, deep vein thrombosis (DVT), and pulmonary embolism, critical illness neuropathy and myopathy, prolonged coma, and delirium. Therefore, the daily goal must be to reduce or even discontinue sedation, if the cerebral situation allows this. It is important not to use sedatives to treat conditions that have to be treated more specifically in other ways, such as pain, seizures, fever, infections, or psychologic disorders. The quite controversial issue of how to sedate the patient during stroke thrombectomy is covered further below in a special section.

Sedation in the general ICU has undergone considerable changes during the last two decades. These include the overall reduction of sedative dosing, and with that, sedation depth, the emphasis on analgesia, the employment of protocols and sedation scores, and the implementation of “drug holidays” and daily “wake-up trials.” These measures have hardly been studied in NCCU patients, but still deserve to be considered. A patient-directed, protocol-based sedation has led to benefits such as reduced ICU-LOS, reduced duration of ventilation, and even reduced mortality in general ICU patients. The implementation of analgesia-based sedation protocols (only using sedatives as add-on if necessary) as part of this individualized form of treatment has been studied in two trials with 215 NCCU patients and 162 NCCU patients, both of which included many patients who had suffered SAH. These studies revealed reduced sedative need and increased pain-free days as well as equal feasibility and better neurologic judgability, respectively. Modern analgesia and sedation involve scores that have, again, hardly been validated in NCCU patients. The previously widespread RAMSAY score has shown weaknesses in several more recent studies and has largely been substituted by the well-validated Richmond Agitation Sedation Score (RASS) and the Sedation-Agitation Scale (SAS). RASS and SAS were found useful at least in a few small studies in NCCU patients. ^, Analgesia has been assessed by the nociception coma scale (NCS) to differentiate between minimal vegetative state and minimal cognitive state in comatose post-NCCU patients. Although overall supported by a very weak base of evidence, we think that RASS, SAS, and NCS are feasible and probably useful to guide sedation and analgesia in NCCU stroke patients as well. Finally, daily wake-up trials are a matter of controversy in this setting. Studies on this question in NCCU patients have reported increases in ICP and the release of stress hormones and other safety concerns, such as regional cerebral hypoxia. Also, a recent study in 413 general ICU patients has not confirmed the benefits that were suggested by previous studies. In essence, we recommend that a wake-up trial be carried out in sedated stroke patients after the acute management is completed and if the brain lesion allows this, but to refrain from further such trials if there is physiologic derangement and rather choose a gradual reduction of sedation over time or early tracheostomy.

Fluid and Electrolyte Balance

Fluid and electrolyte disturbances are common findings in ICU patients. They may be due to (1) sympathetic responses to ischemic or hemorrhagic neuronal injury, (2) unbalanced fluid and electrolyte substitution, (3) unbalanced nutritional regimen, or (4) administration of diuretics and other drugs (particularly osmotherapeutics). Sympathetic nervous system stimulation reduces renal blood flow, thus activating the renin-angiotensin system and increasing the secretion of aldosterone, an effect that in turn causes sodium retention and kaliuresis. Antidiuretic hormone (ADH) secretion may also be affected by central nervous system lesions, resulting in sodium and water retention and decreased urine output (syndrome of inadequate antidiuretic hormone secretion [SIADH]) or in diabetes insipidus. Finally, release of brain-derived natriuretic peptide (BNP) has been associated with cerebral salt waste syndrome (CSWS, see below).

Fluid disturbances can be assessed by (preferably a combination of) (1) clinical observation, (2) evaluation of fluid intake and output, (3) measurement of CVP via a central venous line or pulse contour continuous cardiac output analysis (PiCCO), and (4) measurements of serum osmolarity, urine osmolarity, and serum sodium concentration. Sodium, the main electrolyte of the extracellular fluid, accounts for more than 90% of its osmolarity. There is a close relationship between sodium and water shifts. Sodium concentrations and the hydration state of the patient provide the required information for diagnosis and the treatment of fluid imbalances. Isotonic volume depletion is the most common abnormality encountered. The treatment of choice is enteral or IV administration of isotonic fluids. Careful fluid balancing and monitoring of the CVP are warranted to allow determination of the amount of fluids needed. In patients with concomitant left ventricular failure, chest radiograph, echocardiography, PiCCO analysis should be used to avoid potentially deleterious fluid overload.

For hypernatremic or hyponatremic states, the therapeutic regimen depends on the hydration state of the patient. Hyponatremia is the most common electrolyte disturbance in the NCCU. Rapidly developing and severe hyponatremia less than 110 mmol/L may lead to generalized seizures, coma, and brain edema. The differential diagnosis includes the SIADH (release) versus CSWS. The correct and early diagnosis is important, because the recommended fluid management is different in the two conditions. While SIADH is caused by excessive secretion of ADH leading to water retention, hypervolemia, and secondary diuresis and natriuresis, CSWS is associated with the release of brain natriuretic factor, resulting in diuresis, natriuresis, and concomitant hypovolemia. Therefore, treatment of SIADH is based on water restriction while in contrast treatment of CSWS requires fluid and sodium administration. Differential diagnosis between both syndromes requires correct estimation of hydration state, measurement of urine volume, serum and urine osmolality and the ratio of both, and urine sodium concentration. Aggressive fluid restriction may not be feasible or safe in all NCCU patients, such as those with SAH where euvolemia is paramount. Furthermore, it still remains controversial whether both syndromes are separate entities or rather different manifestations of a common pathophysiologic origin. Some investigators entirely doubt the concept of CSWS. ^,

In the setting of a central diabetes insipidus, SC or IM administration of 5–10 units of aqueous vasopressin repeated 2 or 3 times daily as needed or 2–4 μg in 2 divided doses of its analogue, desmopressin (DDAVP), effectively reduces water diuresis. Further details concerning fluid management are provided in textbooks of general intensive care.

Blood Pressure Control

When and how to manage BP in stroke is covered elsewhere in this book and in the parts on specific types of NCCU stroke below.