Brenner and Rector's The Kidney, 8th ed.

CHAPTER 60. Intensive Care Nephrology

Matthew Dollins   Michael A. Kraus   Bruce A. Molitoris



Acute Respiratory Failure, 2037



Adult Respiratory Distress Syndrome, 2038



Clinical Features, 2038



Risk Factors, 2038



Pathophysiology, 2039



Treatment, 2039



Effects on Renal Function, 2041



Hypovolemic Shock, 2041



Pathogenesis, 2041



Oxygen Consumption and Delivery, 2042



Reperfusion, 2043



Clinical Manifestations, 2043



Diagnosis, 2043



Management, 2043



Fluid Resuscitation, 2044



Crystalloid versus Colloid for Resuscitation, 2044



Vasopressors, 2045



Management of Acidosis, 2045



Effects of Hypovolemic Shock on Renal Function, 2045



Sepsis, 2045



Definition, 2045



Source of Infection and Microbiology, 2046



Pathophysiology, 2046



Clinical Features, 2048



Management, 2049



Antibiotics, 2049



Hemodynamic Support, 2049



Treatment of the Coagulation Cascade, 2049



Immunomodulatory Therapy, 2050



Hemofiltration in Sepsis, 2050



Cardiogenic Shock, 2050



Pathophysiology, 2050



Clinical Features, 2051



Evaluation, 2051



Management, 2051



Fulminant Hepatic Failure, 2053



Definition, 2053



Causes, 2053



Clinical Features, 2054



Evaluation, 2054



Management, 2055

Approximately 5% to 25% of critically ill patients will develop kidney injury during the course of their illness, [1] [2] including between 9% and 40% of patients with sepsis,[3] 20% to 40% of patients with acute respiratory distress syndrome,[4] 33% of patients with cardiogenic shock,[5] and 55% of patients with fulminant hepatic failure.[6] The nephrologist is a critical component in the care of these patients, and an understanding of the underlying pathophysiology of respiratory failure, shock, and the management of mechanical ventilation is critical. The nephrologist should understand the literature behind the move to low tidal volume ventilation because this will affect the acid base status, and has implications for bicarbonate prescriptions during renal replacement therapy. The management of shock is also a rapidly developing field, and nephrologists, who are active in the care of many of these patients, should have an understanding of this area.


Acute respiratory failure can be defined as the inability of the respiratory system to meet the oxygenation, ventilation, or metabolic requirements of the patient.[7] This may occur in a previously healthy person with pneumonia or pulmonary embolism, or complicating chronic respiratory failure in the setting of pulmonary fibrosis or chronic obstructive pulmonary disease. Respiratory failure can be divided into two main types: hypoxemic respiratory failure, which is failure to maintain adequate oxygenation, and hypercapnic respiratory failure, which is inadequate ventilation with CO2 retention.

Respiratory failure is a common occurrence in the ICU, and many of these patients develop renal failure during the course of their illness. Because nephrologists are often asked to assist with the acid-base treatment of these patients, it is important that they have an understanding of mechanical ventilation and the newer treatment strategies for acute respiratory distress syndrome (ARDS).

When patients with respiratory failure are unable to maintain their oxygenation or ventilation by noninvasive means such as supplemental oxygen or continuous positive airway pressure (CPAP/BiPAP), they require intubation and mechanical ventilation, which improves gas exchange and decreases the work of breathing. Several modes of mechanical ventilation are now available. They can be classified into (1) volume cycled ventilation, in which a certain tidal volume is delivered by the ventilator [synchronized intermittent mandatory ventilation (SIMV) and continuous mandatory ventilation (CMV)]; (2) pressure cycled ventilation, in which volume is delivered until a preset maximum pressure is reached [pressure control ventilation (PCV)]; and (3) flow cycled ventilation, in which inspiration continues until a preset flow rate is reached [pressure support ventilation (PSV)]. SIMV allows the patient to breathe spontaneously between breaths assisted by the ventilator. The physician orders a set number of breaths, delivered every minute at a certain tidal volume, which is given in synchrony with inspiratory effort if the patient is able to generate inspiration. Any breaths beyond the set number the patients must generate themselves. CMV results in the ventilator delivering a breath every time the patient generates a negative inspiratory force, or at a set rate, whichever is the higher frequency. CMV minimizes the work of breathing done by the patient, and therefore should be used in the setting of myocardial ischemia or profound hypoxemia. One problem of CMV occurs in patients who are tachypneic or have obstructive lung disease. If there is inadequate time to exhale the full tidal volume, dynamic hyperinflation (breath stacking or auto-peep) may occur, which can result in increased intrathoracic pressure, decreased cardiac output, and possibly barotrauma. PCV differs from SIMV and CMV in that the physician sets an inspiratory pressure, not a tidal volume. During inspiration, a given pressure is imposed via the circuit, and the tidal volume delivered depends on how much flow can be delivered prior to the airway pressure equilibrating with the inspiratory pressure. The tidal volume can vary from breath to breath, and thus the minute volume is variable.

Continuous positive airway pressure (CPAP) is not a true form of mechanical ventilation, but provides a supply of fresh gas at a constant, specified pressure. It is most commonly used in weaning trials or in patients without respiratory failure who require an endotracheal tube to maintain an airway. PSV is a patient-triggered mode of ventilation in which a preset pressure is maintained throughout inspiration. When inspiratory flow falls below a certain level, inspiration is terminated. PSV is commonly used in patients who require minimal support, or to assist the spontaneous breaths during SIMV. Airway pressure release ventilation (APRV) is used in a spontaneously breathing patient who is using CPAP. At the end of each ventilator cycle, the lungs are allowed to briefly deflate to ambient pressure, then rapidly reinflated to the baseline (CPAP) pressure with the next breath. The perceived advantage is that lung expansion during exhalation is maintained with CPAP, but the brief interruption of this pressure at the end of exhalation allows for further carbon dioxide elimination, as well as enhanced venous return.

In addition to the mode of ventilation, the physician prescribes the oxygen concentration to be delivered, the level of positive end expiratory pressure (PEEP), the tidal volume, and the respiratory rate. When initially intubated, patients are typically placed on a high oxygen concentration and weaned down as quickly as possible. Multiple animal studies have supported the notion of oxygen toxicity, whereby higher oxygen concentrations lead to lung injury.[8] [9] [10] [11] Studies on healthy volunteers have shown that after 6 hours on 100% oxygen, there is a change in whole lung capacity and a decrease in the vital capacity,[12] and after 24 hours, there is a reduction in lung compliance likely secondary to increased interstitial edema.[13] Although there are no conclusive studies showing the effects of high levels of inspired oxygen on the lungs during acute illness, most clinicians try to reduce the inspired oxygen concentration to 50% or less as quickly as possible.

Positive end expiratory pressure provides a continuous airway pressure above atmospheric, preventing collapse of alveoli and small airways at end-expiration. By recruiting alveoli in this way, PEEP improves functional residual capacity and oxygenation. PEEP is most commonly set between 5 cm and 20 cm H2O and titrated until adequate oxygenation is achieved. The level of PEEP directly increases airway pressures, so high levels of PEEP can result in barotrauma. PEEP also increases intrathoracic pressure and can result in a decrease in cardiac output secondary to reduced filling volumes.

The tidal volume delivered during mechanical ventilation has recently undergone a dramatic change. Traditional tidal volumes were 10 to 15 ml/kg per breath, but as will be discussed in the ARDS section, recent studies support using lower tidal volumes in patients with lung injury. Patients who are felt to have acute lung injury or ARDS are now prescribed 4 to 8 ml/kg per breath with the goal of minimizing airway pressures. Mechanically ventilated patients without significant lung disease should be prescribed a maximal tidal volume of 10 ml/kg with the goal of preventing ventilator-induced lung injury from barotrauma.

The respiratory rate is set based on the patient's minute ventilation requirement. Patients who are septic or very metabolically active often require a high minute volume to adequately eliminate CO2, and with the lower tidal volumes used now, respiratory rates are often increased. However, care must be taken in the patient with asthma or obstructive lung disease because too high of a rate can lead to air trapping if these patients with prolonged expiratory phase requirements are not allowed to exhale the complete tidal volume prior to the next breath.


The adult respiratory distress syndrome was first described in 1967 when Ashbaugh and colleagues described 12 patients with acute respiratory distress, hypoxia refractory to oxygen therapy, decreased lung compliance, and diffuse infiltrates on chest tomography.[14] This clinical entity is now referred to as the acute respiratory distress syndrome because it is recognized to occur in children as well as adults affecting up to 150,000 patients per year in the United States.[15]

Adult respiratory distress syndrome generally has a poor prognosis, with recent studies reporting a 35% to 60% mortality rate. [16] [17] The lack of a uniform definition led to difficulty in designing studies and attempts at improving outcome, so in 1994 the American-European Consensus Conference developed a definition that is widely used today.[15] This conference established two categories, acute lung injury (ALI) and ARDS depending on the severity of hypoxemia. The acute onset of hypoxemic respiratory failure with bilateral infiltrates on chest tomography, and a pulmonary artery wedge pressure of less than 18, or no clinical evidence of left atrial hypertension characterizes both ALI and ARDS. ALI is present when the criteria mentioned previously are present with an arterial oxygen tension-fraction of inspired oxygen (paO2/FiO2) ratio of less than 300, and ARDS requires the criteria mentioned previously with a paO2/FiO2 ratio of less than 200. The syndromes represent two points in the same disease spectrum, and both appear to have similar outcomes.

Clinical Features

Acute lung injury and ARDS are usually diagnosed when a patient with a known risk factor develops acute dyspnea, hypoxemia, and tachypnea. Different stages are often apparent. The acute stage is characterized by the onset of acute respiratory failure, commonly associated with hypoxemia that is refractory to supplemental oxygen. Radiographic findings include bilateral infiltrates that may be indistinguishable from cardiogenic pulmonary edema,[18] and may be patchy or symmetric. Computed tomography shows the affected areas are primarily the dependent lung zones.[19] During this phase, patients previously requiring minimal oxygen often progress to requiring mechanical support as the work of breathing increases. Pathological findings include damage to the capillary endothelial and alveolar epithelial cells.[20] This disruption of the normal barrier results in increased permeability and filling of the alveoli with protein-rich fluid and inflammatory cells.[21] There is also direct damage to type II pneumocytes, which are responsible for surfactant production. Reduction in surfactant production leads to increased surface tension within the alveoli and results in atelectasis. Interstitial edema also results in the collapse of small airways. As these non-ventilated alveoli are perfused, severe, refractory hypoxemia develops, which accounts for the shunt physiology seen in this disorder. Mechanically ventilated patients with ARDS often have very high airway pressures, a result of fewer ventilated alveoli and reduced lung compliance from the influx of inflammatory cells. This often necessitates a high minute ventilation to maintain an acceptable pCO2.

Following the acute phase, many patients recover completely, yet some develop a fibrotic phase characterized by fibrosing alveolitis, persistent hypoxemia, and further worsening of pulmonary complications.[21] Right ventricular failure can develop due to destruction of the pulmonary capillary bed. Even following the fibrotic phase, improvement in hypoxemia and lung compliance can occur gradually, with many patients returning to normal pulmonary function over 6 to 12 months.[22]

Risk Factors

Acute lung injury and ARDS can develop in association with several clinical conditions ( Table 60-1 ), not all of which directly involve the pulmonary system. The most common condition associated with ARDS is sepsis, with up to 40% of septic patients developing ARDS. [20] [23] Other common risk factors include shock, the systemic inflammatory response syndrome (SIRS), pneumonia, multiple transfusions, near drowning, aspiration, trauma, pancreatitis, burns, coronary artery bypass grafting, and disseminated intravascular coagulation. [13] [14] [24] Multiple risk factors increase the risk for ARDS synergistically.[20]

TABLE 60-1   -- Risk Factors for Acute Respiratory Distress Syndrome



Pulmonary causes









Near drowning



Non pulmonary causes






Systemic inflammatory response syndrome









Multiple blood transfusions









Coronary artery bypass grafting



Disseminated intravascular coagulation





In ARDS, damage is evident to the pulmonary capillary endothelium, which results in an increased permeability and an influx of fluid into the alveoli. Alveolar epithelial cells, which contribute to the alveolar-capillary barrier, and are involved in alveolar fluid reabsorption as well as in the pathogenesis of fibrosis are also injured. Neutrophils play a prominent role in this process because they are found in the alveoli as well as the interstitium during ARDS. Pulmonary capillary endothelial cell activation results in recruitment, adhesion, and signaling of leukocytes They are recruited into the interstitium by cellular adhesion molecules such as selectins and beta-2 integrins.[25] The neutrophils are activated by complement, IL-1, -6, -8, and -10 as well as tumor necrosis factor-α, and these activated neutrophils can then secrete other inflammatory mediators, as well as highly reactive oxidant species, proteolytic enzymes, and metabolites of arachidonic acid that can directly injure alveolar and capillary endothelial cells, allowing for fluid to leak into the alveoli with resultant edema. [26] [27] Although neutrophils are a large component of the inflammatory response, they are not a requirement because neutropenic patients can also develop ARDS. Alveolar macrophages are also involved in ARDS, elaborating cytokines that contribute to the inflammatory process as well as clearing neutrophils from the alveoli and aiding in the resolution of ARDS.

Thromboxane A2 may interact with neutrophils to accentuate cell aggregation,[28] and lipoxygenase products are released in large quantities and may contribute to pulmonary vascular changes and permeability characteristics[21]resulting in “leaky capillaries”. Release of platelet activating factor leads to platelet aggregation in the microvasculature, which results in increased pulmonary vascular resistance and pulmonary hypertension. Infiltration of the interstitium with fibroblasts occurs during the late stage, resulting in the fibrosis seen in some cases.


The management of ARDS has historically been one of support, but the high mortality rates prompted significant research into its cause and propagation. The underlying or predisposing factor should always be addressed promptly, and because sepsis is the most common cause of ARDS, a search for an undiagnosed infection should be undertaken if no clear etiology is present.

The primary mechanism of support is mechanical ventilation. The ARDSNET trial[29] recently showed that a “lung protective mechanism” of mechanical ventilation could improve survival. The goal is to provide adequate oxygenation while avoiding further trauma to the lung that can worsen existing injury.

Traditionally, tidal volumes used during mechanical ventilation were in the range of 12 to 15 ml/min. Acute lung injury, similar to human ARDS, has been observed in animals mechanically ventilated with large tidal volumes.[30] It was reasoned that overdistention of the alveoli, leading to elevated airway pressure, was a primary element in this acute lung injury. Ventilation with high airway pressures has been shown to cause increased vascular permeability, acute inflammation, alveolar hemorrhage, and radiographic infiltrates. [31] [32] In persons with acute lung injury or ARDS, the large tidal volumes are shunted to the unaffected lung because they provide the least resistance, and overdistention results in damage to these previously unaffected segments.

The ARDSNET trial[29] was designed to assess if lower tidal volumes and hence lower airway pressures resulted in a clinical benefit in ARDS. This trial compared traditional ventilation treatment, which was an initial tidal volume of 12 ml/kg ideal body weight (IBW), to a lower tidal volume group that started at 6 ml/kg IBW (IBW=50+2.3 [height in inches – 60] for males, 45.5+2.3 [height in inches-60] for females). In each group, the tidal volume was decreased in increments of 1 ml/kg to maintain the plateau pressure (the airway pressure measured after a 0.5 second inspiratory pause) below 50 for the traditional ventilation group, and below 30 for the lower tidal volume group. The minimal tidal volume was 4 ml/kg. The level of PEEP and oxygen concentration was based on a sliding protocol ( Table 60-2 ). This study was stopped early after 861 patients were enrolled because of a mortality benefit seen in the lower tidal volume group. The mortality rate was 39.8% in the group treated with traditional tidal volumes and 31.0% in the group treated with lower tidal volumes (p=0.007). As expected, the group receiving lower tidal volumes had a slightly higher PaCO2 (43 versus 36 at day 3) and lower pH (7.38 versus 7.41 at day 3) than the traditional tidal volume group.

TABLE 60-2   -- Oxygen and Positive End Expiratory Pressure Titration in the ARDSNET Trial (Goal paO2 55–80 mm Hg)
































PEEP, positive end expiratory pressure.




Since the ARDSNET trial was published, two meta-analyses have suggested that volume-limited ventilation has a short-term survival benefit compared with conventional ventilation, [33] [34] giving further support to the use of low tidal volume ventilation in patients with ARDS or ALI.

Because ARDS is a process that results in decreased lung compliance, patients can generally tolerate higher respiratory rates without the risk of air trapping that is seen in obstructive diseases such as asthma or COPD. The ARDSNET trial used a maximum respiratory rate of 35 breaths per minute. Despite this high rate, the low tidal volumes used to maintain plateau pressures below 30 resulted in a minute ventilation too low to maintain acid-base balance in many patients, a result termed permissive hypercapnia. When the arterial pH fell below 7.30 with a respiratory rate of 35, an infusion of sodium bicarbonate was started.

Given that there have been no other studies in ARDS resulting in such an improvement in mortality, this mode of ventilation must be recommended for all patients with ARDS.

In addition to low-volume ventilation, other mechanical ventilation strategies have been evaluated in ARDS. High frequency oscillatory ventilation (HFOV) uses very low tidal volumes at 4 to 250 times the usual respiratory rate. This mode of ventilation has been studied more extensively in neonates with respiratory distress syndrome,[35] and studies in adults with ARDS have been less compelling. Small, uncontrolled studies have shown an improvement in oxygenation with HFOV, [36] [37] but no study has shown a statistically significant mortality benefit,[38] and there appears to be insufficient evidence to support its routine use in ARDS.

The “open lung” strategy of ventilation uses higher PEEP values to recruit collapsed alveoli, thus “opening up” non aerated regions of the lung. PEEP is kept above 15 cm H2O to prevent end expiratory collapse of alveoli, TV is kept below 6 ml/kg, and the peak airway pressure is kept below 40 cm H2O using pressure-controlled inverse ratio ventilation. One study comparing a higher PEEP strategy to the lower PEEP ARDSnet strategy did not show a difference in survival.[39]

Low tidal volume ventilation clearly has an impact on the nephrologist caring for the patient with ARDS with renal failure. Permissive hypercapnia may result in significant acidosis in a patient with renal failure who is unable to excrete the daily acid load. It also means that the nephrologist may have to use a higher bicarbonate bath during hemodialysis or continuous renal replacement therapy because increasing the minute volume to improve acid-base control is often not an option. In some patients with severe ARDS, large infusions of bicarbonate may not improve acidosis as carbon dioxide is produced, which the severely injured lungs may not be able to expel adequately. Tris-hydroxymethyl aminomethane (THAM) is a buffer that accepts one proton per molecule, generating bicarbonate but not carbon dioxide. It has been shown to control arterial pH without increasing carbon dioxide in the setting of refractory respiratory acidosis.[40] THAM is excreted by the kidneys, so it is not recommended in renal failure.

How best to manage volume status in a patient with ALI/ARDS is controversial. There is substantial data from animal experiments indicating fluid restriction can reduce pulmonary edema in the setting of increased pulmonary vascular permeability, such as in ALI/ARDS.[41] Human studies include an observational study where survival in ARDS was related to negative fluid balance,[42] a study in which patients with a 25% reduction in pulmonary capillary wedge pressure had a greater survival than other patients,[43] and a study where patients with less than 1 liter of fluid gain after 36 hours of recruitment had a better survival than other patients.[44] Yet there are other data suggesting patients with ALI/ARDS may do better with a strategy that increases oxygen delivery, usually requiring volume expansion. Fluid restriction can reduce cardiac output and tissue perfusion, leading to worsening of non-pulmonary organ dysfunction that is often seen in patients with ARDS. Several trials have assessed whether providing supranormal levels of oxygen delivery will improve outcome. [45] [46] [47] [48] [49] [50] Some feel that in systemic inflammatory conditions, such as sepsis or trauma, normal cardiac output and tissue oxygen delivery may be inadequate to prevent organ dysfunction. In postoperative treatment of trauma patients, there was a trend toward decreased mortality with supranormal oxygen delivery, [45] [46] [47] but there has been no benefit in patients with ALI/ARDS, [48] [49] and one study showed an increased mortality in patients who received supranormal levels of oxygen delivery.[50] Because there is no clear benefit to supranormal oxygen delivery, which requires volume expansion, and fluid restriction can lead to worsening of non-pulmonary organ dysfunction greatly increasing mortality, maintaining euvolemia (wedge 10-14, CVP 6-12) in patients with ARDS/ALI with use of fluids as guided by evidence of organ perfusion would be the most reasonable approach at this time.

The inflammatory nature of ARDS raises the possibility that glucocorticoids may be beneficial in managing this condition. High doses of glucocorticoids have not shown benefit when given to prevent ARDS in high-risk patients, [51] [52] [53] or when given early in the course of ARDS. [51] [54] However, the persistent inflammation and fibroproliferation seen in the late stage of ARDS may be improved by corticosteroids. One small study[55] (32 patients) evaluated prolonged methylprednisolone in patients with ARDS who did not improve after 7 days of respiratory failure, and found an improvement in severity scores and mortality in those treated. A randomized, controlled trial evaluating methylprednisolone in severe late-phase ARDS has been completed by ARDSnet. Although the manuscript has not been published yet, the results were presented at a meeting of the American Thoracic Society in May 2004, suggesting no improvement in mortality.

Prone positioning has been advocated as a means to ventilate the posterior lung regions that are more often atelectatic and flooded in ARDS. Once the patient is prone, these previously dependent lung regions open, as the anterior lung regions become dependent. Several personnel are required to safely move a patient into the prone position to ensure chest tubes, IVs, and the endotracheal tube are not dislodged. Patients are rotated every 12 to 18 hours, and studies have shown improved gas exchange and oxygenation in the prone position. [56] [57] Recent studies have shown no improvement in outcomes with prone positioning, [58] [59] although in a post-hoc analysis of one study,[60]mortality at study day 10 and at ICU discharge was lower in the prone positioning patients who were in the lowest quartile paO2/FiO2 (<88 mm Hg), lowest quartile of Apache II scores, or highest quartile of tidal volume (>12 ml/kg predicted body weight).

Surfactant is normally produced by the type II pneumocytes and allows patency of alveoli at lower airway pres-sures. In ARDS, surfactant production is decreased. Animal models of lung injury have shown a benefit from inhaled surfactant therapy, [61] [62] but a trial in 725 patients with ARDS showed no benefit from an artificial surfactant given by aerosol.[63] A more recent randomized controlled trial of 40 patients found no improvement in oxygenation or ventilator-free days with surfactant, but found a dose-dependent trend toward lower mortality at study day 28 (20% to 33% versus 38%).[64]

Nitric oxide (NO) is a vasodilator that when inhaled dilates pulmonary blood vessels perfusing aerated lung units, resulting in improved ventilation-perfusion mismatch without systemic vasodilatation. Inhaled NO has been found to improve oxygenation, [65] [66] but has not been found to improve mortality. [66] [67]

The inflammatory response seen in ALI/ARDS has led to many agents in addition to corticosteroids being evaluated as possible therapy. Prostaglandin E1,[68] ketoconazole[69] (an inhibitor of thromboxane and leukotriene synthesis), ibuprofen,[70] and procysteine/N-acetylcysteine[20] have all been evaluated and found to have no benefit. In addition, treatment of sepsis prior to or early in the development of ALI/ARDS with an anti-endotoxin monoclonal antibody, anti-TNF-α and anti-Interleukin-1 have not shown benefit.[41] Intravenous N-acetylcysteine has not shown a mortality benefit, [71] [72] but has been found in small studies to reduce the number of acute lung injury days. [71] [72] [73] Studies using IL-10 and recombinant human platelet activating factor are currently in the design stages.[74]

Extracorporeal carbon dioxide removal has been investigated and found to have no effect on mortality in one randomized controlled study.[75] Partial liquid ventilation with fluorocarbon liquids, which can dissolve 17 times more oxygen compared with water, has been evaluated with encouraging results, but more trials are needed prior to this therapy becoming widespread.[41]

Effects on Renal Function

Renal dysfunction is a common occurrence in patients with ARDS/ALI. In a retrospective study of 59 patients with ARDS, Valta and co-workers[4] found that 20% to 40% of patients with ARDS had renal dysfunction. Although many patients with ARDS are also septic or hemodynamically unstable, mechanical ventilation itself has been found to be a predictor of dialysis requirement.[76] Several studies have shown that mechanical ventilation can lead to reduced renal blood flow (RBF), decreased urine output, and sodium retention. [77] [78] [79] These changes in renal function are felt to be due to multiple factors. Hypercapnia has been shown to decrease RBF.[80] It acts by directly causing renal vasoconstriction and stimulates norepinephrine release. Hypercapnia also causes systemic vasodilation, which can result in decreased systemic vascular resistance and subsequently reduced RBF.[80] Positive pressure ventilation can also result in a decrease in cardiac output. Positive intrathoracic pressure from mechanical ventilation reduces venous return to the heart, resulting in decreased effective circulating volume, and increases pressure in the pulmonary vasculature, which results in elevated right ventricular afterload, factors that result in reduced cardiac output. The reduced airway compliance seen in ARDS leads to elevated intrathoracic pressures with even relatively small tidal volumes. In addition, the intrathoracic pressure increases linearly as the positive end-expiratory pressure (PEEP) is increased. Because many patients with ARDS require a high PEEP to maintain patency of alveoli and small airways for maintenance of oxygenation, this group of patients is particularly susceptible to the hemodynamic effects of mechanical ventilation.

Not all studies have shown a decrease in renal blood flow with positive pressure ventilation,[81] and it appears that volume status may play a role in the hemodynamic response to positive pressure ventilation. Those patients who are volume depleted are more susceptible to reduced cardiac output.

Factors other than hemodynamics are likely involved in the association of acute kidney injury (AKI) with mechanical ventilation and acute lung injury. Recent studies have shown that mechanical ventilation without changes in blood pressure or central venous pressure caused flattening of epithelial cells in the canine kidney.[82] Renal tubular apoptosis and biochemical markers of renal dysfunction were found in a rabbit model of mechanical ventilation as well.[83] Hormonal changes during mechanical ventilation have been evalu-ated. ADH levels are elevated in mechanically ventilated patients, but may act primarily as a vasoconstrictor and have minimal effect on water retention. [81] [84]A sympathetically mediated increase in plasma renin activity results in a decline in GFR by reducing renal blood flow, and stimulating sodium retention via aldosterone.[85] Atrial natriuretic peptide may also be reduced as a consequence of the decreased venous return and lower atrial pressures, resulting in reduced urine output.[86] Other factors such as nitric oxide and endothelin may also play a role, but their effect remains undetermined.

Adult respiratory distress syndrome is likely an early manifestation of a systemic inflammatory process that results in multiorgan dysfunction. Studies of bronchoalveolar fluid have shown increased TNF-α, IL-1β, and IL-6 concentrations during ARDS. [87] [88] It is possible that pulmonary cytokine production is increased during ventilator-induced lung injury, leading to elevations in systemic concentrations.[89] These cytokines have been shown to cause an ARDS-like condition in rats, but their role in human ARDS has not been fully unraveled.

TNF-α has been associated with the renal injury seen in ischemia-reperfusion models, as has IL-1, IL-2, IL-8, interferon-gamma, and granulocyte-macrophage colony-stimulating factor. [90] [91] Although the association between these cytokines, ARDS, and acute kidney injury has not been fully elucidated, it is likely ARDS represents an early stage in inflammation leading to multiorgan system failure including acute kidney injury. Whether primary cytokine production by the injured lung leads to further organ dysfunction has yet to be established.

Hypovolemic Shock

Hypovolemic shock can be defined as a reduction in effective circulating blood volume, which leads to an oxygen deficit in the tissues because oxygen supply is not able to meet oxygen demand. This imbalance in oxygen metabolism leads to reduced cellular metabolism, conversion to anaerobic metabolism, accumulation of CO2 and waste products (lactate), and, if prolonged, cellular death. Hypovolemic shock occurs most commonly from trauma and hemorrhage,[92] but can also be seen in the setting of volume depletion from vomiting, diarrhea, burns, uncontrolled diabetes mellitus, pancreatitis, or from addisonian crisis ( Table 60-3 ).

TABLE 60-3   -- Etiology of Hypovolemic Shock



Blood loss









Gastrointestinal bleeding



Massive hematuria






Aortic dissection/abdominal aortic aneurysm rupture






Splenic laceration/rupture



Hepatic laceration/rupture



Pelvis/long bone fractures



Ruptured ectopic pregnancy



Fluid losses



Diabetic ketoacidosis



Adrenal crisis












Lack of volume replacement






Comatose/found down





Loss of circulating volume is the primary stimulus for the manifestation of shock. Once 10% of circulating volume has been lost, compensatory mechanisms are activated to maintain cardiac output despite the decreased ventricular filling pressures and stroke volume ( Table 60-4 ).[93] Sympathetic discharge as well as adrenal catecholamine release leads to tachycardia, arterial vasoconstriction, and venoconstriction. [94] [95] As volume loss increases, the increase in heart rate is not able to overcome the loss of stroke volume, and cardiac output declines, which is initially detected as orthostatic hypotension and a fall in pulse pressure.[96] Once the loss of volume exceeds approximately 40%, or 20% to 25% if lost rapidly, hypotension and shock ensue.[97]

TABLE 60-4   -- Compensatory Response to Shock



Maximize intravascular volume



Redistribution of fluid to intravascular space



From interstitial compartments



From intracellular compartments



Renal adaptations



Increased aldosterone



Increased vasopressin



Maximize blood pressure



Increased sympathetic activity



Increased catecholamines



Increased angiotensin II production



Increased vasopressin



Maximize cardiac output



Sympathetic stimulation






Increased contractility



Maximize oxygen delivery



Metabolic acidosis



Increased RBC 2,3 DPG



Decreased tissue oxygen levels




During hypovolemic shock, peripheral vascular resistance is elevated as a result of several responses. These include catecholamine secretion by the adrenal glands, activation of the sympathetic nervous system, the vasoconstrictive effects of angiotensin II via activation of the renin-angiotensin-aldosterone system, and vasopressin released by the pituitary gland. [98] [99] [100] However, the rise in vascular resistance is not uniformly distributed throughout the organ systems. [95] [101] Although vasoconstriction increases vascular resistance, regional autoregulation to maintain blood flow can counteract this effect. Nitric oxide (NO) is produced by endothelial cells, which relaxes vascular smooth muscle cells,[102] and NO release may decrease responsiveness to endogenous and exogenous vasoconstrictors. [103] [104] Organs with reduced endothelium-dependent vasorelaxation have been found to have endothelial cell dysfunction, [105] [106] which may indicate that the NO-mediated relaxation provides protection against the response to shock. However, NO has been shown to inhibit mitochondrial respiration in vitro,[107] and one study[108] of 28 septic patients found an association between NO overproduction, antioxidant depletion, mitochondrial dysfunction, and decreased ATP concentration that related to organ failure. Carbon dioxide[109] and adenosine[110] may also play a role in regional autoregulation, having been shown to produce vasodilatation. The end result of these competing interactions is that blood flow is reduced to the kidneys, skin, intestines, and skeletal muscle, and increased to the heart and brain. [97] [111] [112]

Blood flow through capillaries is slowed during hypovolemic shock, with evidence that this is secondary to reduction in perfusion pressure.[113] Endothelial cell swelling may contribute,[114] and expression of endothelial adhesion molecules have been found to be up-regulated during hypovolemic shock on both the neutrophil and endothelium, [115] [116] [117] suggesting neutrophil aggregation may also contribute to sluggish capillary flow. Responses to increase the circulating volume include reabsorption of interstitial fluid into the vascular space, a result of the decline in capillary hydrostatic pressure greater than the decline in intersti-tial hydrostatic pressure.[118] The transport of protein from blood to interstitium is decreased,[119] cellular water is mobilized to the extracellular space,[120] and activation of the renin-angiotensin-aldosterone system as well as increased levels of antidiuretic hormone act to increase sodium and water reabsorption in the kidneys. Despite compensatory mechanisms to preserve the effective circulating volume, and maintain blood pressure, hypotension and shock will ensue if a large enough amount of fluid is lost. The reduction in perfusion to tissues results in an oxygen imbalance that is responsible for much of the organ failure seen in hypovolemic shock.

Oxygen Consumption and Delivery

Global oxygen delivery (Do2) is the total amount of oxygen delivered to the tissues per minute, and under resting conditions, it is more than adequate to meet the total oxygen requirements of the tissues.

Oxygen delivery is calculated by multiplying the cardiac output by the oxygen content in blood, the latter of which is dependent on the amount of dissolved oxygen (pO2), the oxyhemoglobin saturation (%HbO2), and the hemoglobin affinity for oxygen (typically expressed as 1.34):

Do2=cardiac output×(1.34)×(grams of Hb)×(%HbO2)+[(pO2 in mm Hg)×(0.003)]

Oxygen consumption (Vo2) can be measured directly from inspired and mixed expired oxygen concentrations and expired minute volume, or it may be derived from the cardiac output and arterial and venous oxygen contents:

Vo2=cardiac output×(arterial oxygen content - mixed venous oxygen content)

The amount of oxygen consumed (Vo2) as a fraction of oxygen delivery (Do2) is the oxygen extraction ratio (OER):


For a normal adult performing routine activities, Vo2 is approximately 250 ml/min with an OER of 25%, which can increase to 70% to 80% during maximal exercise.[121] In the setting of hypovolemic shock, a fall in hemoglobin or cardiac output can significantly reduce oxygen delivery to the tissues. It has been shown that below an oxygen delivery of approximately 8 ml/kg/min, oxygen uptake is maximal (near 100%),[122] and a decrease in oxygen delivery below this level results in cellular ischemia as oxygen tissue demand is not being met.

Global oxygen delivery in shock may be normal, despite evidence of cellular ischemia.[123] This is often due to the regional differences in blood flow seen in hypovolemic shock with some organs developing an oxygen debt despite normal global oxygen delivery.[121] Some authors have suggested supranormal levels of oxygen delivery may be able to overcome these regional differences and improve outcome, [124] [125] but this will be discussed later. Cellular hypoxia is manifested in several ways. Once oxygen demands exceed oxygen delivery, anaerobic metabolism ensues, with the production of lactate. Blood lactate level has traditionally been used as an indicator of tissue hypoxia, representing a balance between the production of lactate and consumption by the liver, as well as cardiac and skeletal muscle.[126] However, a single level may be unreliable, and serial levels may be more beneficial as an indicator of cellular hypoxia.[121]

Loss of function of cellular enzymes can occur during hypoxia, but there is a significant variation in sensitivity to hypoxia, with glucose oxidase being quite sensitive to hypoxia, whereas NADPH oxidase can function at 50%, and cytochrome 003 can function at 0.09% of the cellular oxygen required for glucose oxidase.[127]

Cellular ischemia in the gut may result in gastric ulcers, as well as disruption of the barrier function of the mucosa, which can result in translocation of bacteria from the bowel into the circulation. [128] [129] Hepatic ischemia can decrease the clearance of lactate,[130] drugs,[131] and centrilobular necrosis may result in elevated bilirubin and enzyme levels.[132] The spleen contracts during hypovolemic shock, releasing red blood cells into the circulation.[95]Myocardial ischemia can occur,[133] particularly in elderly patients who may have atherosclerotic coronary artery disease.


Although restoration of flow to an ischemic organ is critical to restore function, reperfusion itself may contribute to organ damage. Reactive oxygen species are formed once ischemic tissues are reperfused,[134] and these can cause direct cellular membrane damage by lipid peroxidation, as well as leukocyte activation and transmigration by stimulating leukocyte adhesion molecule expression.[135] The activated leukocytes contribute to cellular injury by releasing proteases, elastases, as well as cytokines that increases microvascular permeability, edema, and microthrombosis.[136] Ischemia-reperfusion also activates complement, which promotes leukocyte activation as well as altering vascular permeability, resulting in edema.[137] Data suggests that calcium influx into cells during reperfusion may contribute to injury by damaging cell organelles, inhibiting respiration and activating protease and prostaglandin synthesis. [138] [139] Reperfusion injury can manifest as myocardial stunning, reperfusion arrhythmias, breakdown of the gut mucosal barrier, acute kidney injury, hepatic failure, or multiorgan dysfunction syndrome.[140] [141] [142] [143] [144]

Clinical Manifestations

Early in the course of hypovolemia or blood loss, the patient may not be hypotensive, and attention should be paid to other signs of fluid loss. Tachycardia is common, and tachypnea can occur early in the course. Orthostatic hypotension is a reliable sign, whereas dry mucosal membranes and decreased skin turgor are less reliable, but indicative of hypovolemia.[145] If the patients are conscious, they may complain of thirst or diaphoresis. Once volume losses become profound, hypotension ensues, confusion may occur, and the patient may develop oliguria and peripheral cyanosis as a result of diminished perfusion. Hypovolemic shock due to trauma or bleeding is usually apparent, but internal bleeding or other causes listed in Table 60-3 may not be as obvious. The smell of acetone on the breath may point to uncontrolled diabetes mellitus, whereas adrenocortical insufficiency can result in brown discoloration of the mucous membranes.

Acidosis can occur, often from hypoperfusion of tissues resulting in lactate production. Disseminated intravascular coagulation can also occur during hypovolemic shock, resulting in microvascular thrombi formation, and may contribute to the multiple organ dysfunction often seen after traumatic or hypovolemic shock.[146]


The initial evaluation of the patient in shock should include a determination of the cause of shock. In most cases of hypovolemic shock, it is readily apparent that trauma or blood loss is the primary cause, but care must be taken not to overlook septic, cardiogenic, or anaphylactic shock. Initial resuscitation should begin during the evaluation. In the case of external blood loss, crossmatching blood should be done while fluids are infused for resuscitation. Gastrointestinal bleeding can be evaluated and potentially treated with upper or lower endoscopy once the patient is stabilized, as well as angiography. In the event of trauma, chest radiography should be performed to rule out tension pneumothorax or hemothorax. If abdominal trauma has occurred, peritoneal lavage can be performed to assess for hemorrhage, most commonly from splenic or hepatic lacerations.[147] If the patient is stabilized, computerized tomography or ultrasound can also assess for intra-abdominal hemorrhage as well as organ injury. Laboratory tests should include complete blood count, a chemistry panel including electrolytes, creatinine, glucose, and liver function tests; arterial blood gas, arterial lactate level, blood type and crossmatch, and urinalysis. In the event of trauma or bleeding, coagulation studies should include platelet count, prothrombin time, and partial thromboplastin time. If the cause of shock is not readily apparent, an electrocardiogram should be performed to rule out myocardial infarction.


Resuscitation of the patient in shock should begin immediately, and not delayed while diagnostic procedures are undertaken. Fluid resuscitation should begin once large bore intravenous catheters are placed. The primary goal in the management of hypovolemic shock is to return circulating volume to normal, and as a result, improve tissue perfusion, substrate delivery, and oxygen balance. Some authors have suggested that raising oxygen delivery and oxygen uptake to supranormal levels in the setting of trauma and hemorrhage may improve survival. [124] [125] Oxygen delivery can be maximized by increasing the cardiac output with either volume or dobutamine; by increasing the oxygen saturation above 90%; and by increasing the hemoglobin concentration. Care must be taken when transfusing, as a higher hematocrit can actually worsen oxygen balance by increasing viscosity and reducing capillary flow.[148]Although elderly patients with myocardial infarction may benefit from transfusion to a hematocrit of 30%,[149] large transfusions of blood have been associated with multiple organ dysfunction,[150] and a liberal transfusion policy to a hemoglobin of 10 to 12 has been associated with an increased mortality.[151] Measurement of oxygen delivery and consumption also requires pulmonary artery catheter placement, which may be an independent risk for mortality,[152] thus many physicians use improvement in blood pressure, metabolic acidosis, and serial lactate levels as markers that oxygen delivery and consumption are adequate. However, improvement of oxygen delivery in this setting may not improve cellular function. Tissue oxygen tension has been found to be increased in some studies of septic animals and patients with acidosis, [153] [154] indicating that dysoxia (inadequate utilization of oxygen), not hypoxia, may contribute to acidosis and organ failure.

Gastric tonometry has been proposed as a method to monitor a patient's perfusion status, and indirectly oxygen delivery. This is a device that is inserted nasally or orally and advance to the stomach, where it indirectly measures the pH of gastric mucosal cells. A low gastric mucosal pH (pHi) may indicate two things. First, it may be an early indicator of reduced global oxygen delivery as the splanchnic bed is prone to hypoperfusion and hypoxia due to redistribution of blood flow. Second, intestinal mucosal hypoxia may result in increased permeability with increased translocation of bacteria and endotoxin, resulting in multiple organ system dysfunction. Low pHi has been found to be a good indicator of poor outcome in the intensive care setting, [155] [156] [157] but no study has convincingly proven that therapy to improve pHi has any effect on outcome. [158] [159] [160] [161] Sublingual capnometry has been evaluated as a measure of tissue perfusion, but is not in widespread use.[162]

Further treatment depends on the etiology of shock. Traumatic shock will often require surgical exploration to treat the source of bleeding. Upper gastrointestinal bleeding due to ulcers can be treated medically with intravenous proton pump inhibitors, or endoscopically by electrocautery, laser coagulation, or injection therapy. Esophageal varices can be treated with infusion of somatostatin, or interventionally with injection sclerotherapy, or a sangsten-blakemore tube. Lower gastrointestinal bleeding can be treated with endoscopic therapies. Surgery is an option for recurrent bleeding. Diabetic ketoacidosis is treated with intravenous insulin, and adrenal crisis with intravenous hydrocortisone.

Fluid Resuscitation

Fluid resuscitation is the initial therapy in hypovolemic shock because this helps restore circulating volume and oxygen delivery. The types of fluids used are quite varied ( Table 60-5 ), and controversy exists as to which agent is the most efficacious. Both colloids (high molecular weight solutions) and crystalloids (electrolyte solutions) are used to manage shock.

TABLE 60-5   -- Fluid Used for Resuscitation


Sodium Chloride (0.9%)

Ringers Lactate

Sodium Chloride (3%)

Albumin (5%)

Hetastarch (6%)

Dextran 70 +NaCl


Sodium (mEq/L)








Chloride (mEq/L)








Potassium (mEq/L)








Osmolarity (mOsm/L)








Oncotic pressure (mm Hg)








Lactate (mEq/L)








Maximum dose (ml/kg/24h)



Limited by Serum Na+





Cost (liter)










Isotonic crystalloid solutions have traditionally been used as the primary fluid for volume expansion.[152] Normal saline (0.9%) and lactated ringers are both commonly used, although large volumes of lactated ringers should be avoided in the setting of renal failure because it can result in hyperkalemia, and probably should be avoided in hepatic failure because the damaged liver may not be able to convert lactate to bicarbonate. One advantage of isotonic crystalloids may be that they replace the interstitial fluid deficits seen after hypovolemic shock,[163] because 75% of the volume infused enters the interstitial space, whereas 25% remains intravascular.[164] However, the large volume of these fluids required leads to peripheral edema that may impair wound healing,[165] and has led to the study of hypertonic crystalloid and colloid solutions, which stay within the intravascular space to a greater degree, and thus require less total volume for a similar degree of resuscitation.

Hypertonic crystalloid solutions include 3%, 5%, and 7.5% sodium chloride, and are considered plasma expanders because they act to increase the circulatory volume via movement of intracellular and interstitial water into the intravascular space.[152] The primary disadvantage of these agents is the risk of hypernatremia, and the safety of these agents depends partially on how much water can be shifted from the intracellular to extracellular space without resulting in cellular damage.

Colloids are also plasma expanders because they are composed of macromolecules, and are retained in the intravascular space to a much greater extent than isotonic crystalloids. Albumin has a molecular weight of 69,000 daltons, and a half-life of 15 to 20 days. Albumin may serve as a free radical scavenger,[166] and the increased intravascular oncotic pressure may protect the lungs and other organs from edema.[167] Dextran is a colloid agent prepared from glucose polymers. Detran-40 has a molecular weight of 40,000 daltons, and dextran-70 has a molecular weight of 70,000 daltons. Dextran-70 has a longer intravascular retention time than dextran-40,[168] but both can cause histamine release from mast cells leading to anaphylactoid reactions.[169] Hydroxyethyl starch (hetastarch) is available in several different preparations (HES 200 or HES 450). HES is a natural starch of highly branched glucose polymers, similar in structure to glycogen, with a molecular weight of 200,000 or 450,000 daltons, depending on the preparation, and a plasma half-life of approximately 17 days. Its volume expansion properties are almost identical to albumin.[168] Pentastarch has a molecular weight of 260,000 daltons, but has a higher colloid oncotic pressure than HES or albumin, thus producing more intravascular volume expansion than these two agents.[170] There is recent evidence that the starches may be able to reduce capillary leak after ischemia or trauma, thereby decreasing edema formation. [171] [172] Both HES and pentastarch may increase the amylase level in blood. Gelatins are polypeptides from bovine raw material, have a lower molecular weight than the starches or dextrans, and are poorly retained in the intravascular space. Their duration of effect is approximately 2 hours; 3.5% urea-gelatin has a high concentration of potassium, which makes it unsuitable for patients with renal failure. The gelatins are not available in the United States at this time. Newer solutions consisting of hypertonic saline to which colloids have been added include NaCl 7.5% with dextran-70, NaCl 7.2% with dextran 60, and NaCl 7.5% with hetastarch.

Crystalloid versus Colloid for Resuscitation

There has been much debate as to which type of fluid is best for resuscitation of shock. Colloids offer the theoretical advantage of expanding the intravascular space with less volume, and have been shown to increase blood pressure more rapidly than crystalloids.[173] One liter of dextran-70 increases intravascular volume by 800 ml, one liter of HES by 750 ml, one liter of 5% albumin by 500 ml, and one liter of saline by 180 ml.[174] Yet in the setting of sepsis, where there is significant capillary leak, both albumin and normal saline were found to increase interstitial volume to the same extent.[175] Small studies have also found a lower incidence of pulmonary edema during resuscitation with colloids compared with crystalloids,[176] potentially less reperfusion injury to the myocardium after colloid resuscitation,[177] and better blood flow to the myocardium.[178] Yet there is also evidence that colloids can inhibit the coagulation system, [179] [180] and cause anaphylactoid reactions.[181] Albumin has been shown in one study to have negative inotropic effects,[182] and another study found impaired salt and water excretion when albumin was used for resuscitation from shock.[183] Hetastarch may increase the risk of acute kidney injury when given for resuscitation of sepsis,[184] which may be due to inadequate free water replacement in the setting of a potent volume expander.

Meta-analyses of fluid administration and mortality have not supported a benefit for colloids over crystalloids. [185] [186] [187] [188] [189] [190] In trauma patients, Wade and colleagues[185] found no difference in survival between those receiving hypertonic saline with dextran 60 or isotonic saline, whereas Choi and colleagues[186] found crystalloids were associated with a significantly lower risk of death [relative risk (RR) 0.39[186]]. One meta-analysis[187] did find that hypertonic saline with dextran did improve survival compared to crystalloids in the setting of head injury. Several meta-analyses [188] [189] [190] have shown a trend toward increased mortality in heterogenous groups of critically ill patients resuscitated with colloids. Although the Cochrane Injuries Group Albumin Reviewers[191] found that the risk of death was significantly increased in critically ill patients who received albumin (RR 1.68[191]), a trial of 6997 ICU patients requiring fluid resuscitation compared albumin with normal saline, and found no difference in outcomes for the two groups.[192]

Volume resuscitation in critically ill patients is a matter of debate at this time, with many practitioners favoring crystalloids whereas others favor colloids. Although the available data does not strongly favor one therapy over the other, patients with profound volume deficits treated with crystalloids may benefit from the addition of colloid solutions to hasten restoration of circulating volume. A search is underway for red blood cell substitutes that can rapidly expand blood volume, as well as carry and deliver oxygen to tissues. Diaspirin cross-linked hemoglobin showed an increase in mortality when this agent was used in the management of severe traumatic shock,[193] and no agent is currently in widespread use.


The use of vasopressors in hypovolemic shock should be reserved for the setting in which adequate fluids are not yet available, or for the patient in whom adequate fluid infusion has not improved hypotension.[147] In this setting, a pulmonary artery catheter can help guide therapy because persistent shock can be caused by either peripheral vasodilatation or myocardial dysfunction. A wedge pressure of 12 mm Hg to 16 mm Hg is indicative of adequate volume expansion. Animal studies have shown that vasopressin can reverse shock unresponsive to fluids and catecholamines,[194] and can improve survival after cardiac arrest in hypovolemic shock.[195] Although vasopressin has been shown in one small study[196] to improve blood pressure in septic shock, others have suggested it can cause a reduction in cardiac output,[197] and only case series of improvement in hypovolemic shock are available. [198] [199]

Management of Acidosis

Lactic acidosis due to tissue hypoperfusion is common in hypovolemic shock. Improvement of the effective circulating volume and restoring tissue oxygen balance will diminish the production of lactate, allowing for improvement of acidosis. Yet for cases of intractable shock, metabolic acidosis may persist despite aggressive therapy.

Acidosis has been shown to decrease cardiac contractility in animal models,[200] and reduces cardiac contractility response to catecholamines.[201] However, the effect of acidosis on cardiac function in the clinical setting is less well documented. Decreased cardiac contractility in the setting of lactic acidosis may be partially due to hypoxemia, hypoperfusion, or sepsis, and establishing direct effects of the low pH are difficult.[202] In fact, many patients treated with permissive hypercapnia-low tidal volume ventilation develop acidosis that is well tolerated with minimal change in the cardiac output.[203]

Management of acidosis with sodium bicarbonate has not been shown to be beneficial. Animal models of lactic acidosis fail to show improvement in hemodynamics from sodium bicarbonate compared with normal saline, [203] [204] [205] and human studies have shown no improvement in hemodynamics or catecholamine responsiveness. [206] [207] Furthermore, bicarbonate infusion has been theorized to cause worsening intracellular acidosis because the carbon dioxide, produced when bicarbonate reacts with acids, can diffuse rapidly across the cell membrane, whereas bicarbonate cannot. Studies of intracellular pH changes have been mixed, with some showing an increase,[208] decrease,[209] [210] [211] or no change [201] [212] [213] in pH. Management of acidosis with bicarbonate has also been shown to increase hemoglobin affinity for oxygen in healthy volunteers, resulting in reduced oxygen delivery.[214]

Because there is no documented benefit, and the potential for adverse effects appears real, management of lactic acidosis should not include administration of sodium bicarbonate unless further compelling evidence becomes available.

Effects of Hypovolemic Shock on Renal Function

Acute kidney injury is a common finding in a patient with shock. Diminished perfusion to the kidneys with resultant ischemia is a primary cause. Early in hypovolemia, renal perfusion can be maintained by intrarenal production of NO and prostaglandins that have vasodilatory actions. [215] [216] However, once hypovolemia becomes severe and shock ensues, these mechanisms are not enough to prevent ischemia. Other factors seem to play a role as well. Disseminated intravascular coagulation can occur during traumatic or hypovolemic shock, and the resultant microvascular thrombi can cause renal ischemia.[146]

Hypovolemic shock has been shown to increase tumor necrosis factor-α and interleukin-1 release, [217] [218] and can activate the complement cascade.[219] These substances may contribute to acute kidney injury, and their effects of renal function are discussed more fully in the section on sepsis.


It is estimated that sepsis accounts for up to 10% of admissions to the ICU,[220] and there are 400,000 to 500,000 episodes of sepsis each year in the United States[221] resulting in greater than 100,000 deaths per year.[222] Sepsis and septic shock are common causes of acute kidney injury, and the nephrologist is frequently involved in the care of this disease. Despite improvements in our ability to monitor and treat patients in the intensive care unit, the mortality rate for sepsis has actually increased.[221] A complete understanding of the pathophysiology and newer therapeutic approaches for sepsis is critical for any clinician involved in patient care.


The American College of Chest Physicians/Society of Critical Care Medicine consensus conference in 1991 led to a uniform definition of systemic inflammatory response syndrome (SIRS), sepsis, severe sepsis, and septic shock[223] ( Table 60-6 and Fig. 60-1 ). SIRS describes the common systemic response to a wide variety of clinical insults ( Table 60-7 ). It is characterized by two or more of the following: (1) temperature more than 38°C or less than 36°C; (2) heart rate more than 90 beats/min; (3) respiratory rate >20 breaths/min; (4) white blood cell count more than 12,000 cells/mm3, less than 4,000 cells/mm3, or more than 10% immature neutrophils. Sepsis is present when SIRS is diagnosed in the setting of a confirmed infection. Severe sepsis is defined as sepsis plus either organ dysfunction or evidence of hypoperfusion or hypotension. Septic shock is a subset of severe sepsis and is present when sepsis-induced hypotension persists despite fluid resuscitation, and is accompanied by hypoperfusion abnormalities or organ dysfunction. In a study of 2527 patients who met SIRS criteria in a single ICU, the mortality rate was found to increase as patients fulfilled more criteria and advanced along the spectrum. The mortality of patients with two SIRS criteria was 7%, three SIRS criteria 10%, four SIRS criteria 17%, sepsis 16%, severe sepsis 20%, and septic shock 46%.[224]

TABLE 60-6   -- Definition of SIRS, Sepsis, Severe Sepsis, Septic Shock



SIRS: Presence of two or more of the following



Temperature >38°C or <36°C



Heart rate >90 beats/min



Respiratory rate >20 breaths/min



White blood count >20,000/mm3, <4000/mm3, or >10% immature neutrophils



Sepsis: SIRS in the presence of documented infection



Severe sepsis: Sepsis with hypotension, hypoperfusion, or organ dysfunction

Septic shock: Sepsis with hypotension despite volume resuscitation and evidence of organ dysfunction or hypoperfusion


SIRS, systemic inflammatory response syndrome.




FIGURE 60-1  The spectrum of systemic inflammatory response syndrome (SIRS), sepsis, septic shock, and multiorgan dysfunction syndrome (MODS).


TABLE 60-7   -- Common Causes of Systemic Inflammatory Response Syndrome




















Autoimmune diseases



Source of Infection and Microbiology

In the 1960's and 1970's, gram-negative organisms were the most common causes of septic shock,[225] but gram-positive organisms have now increased in prevalence. Gram-negative organisms are now estimated to be responsible for 25% of all cases of sepsis, with gram-positive organisms responsible for 25%, mixed gram positive and gram negative 20%, fungal 3%, anaerobic organisms 2%, and 25% of organisms unknown.[226] The most common gram-negative organisms are Escherichia coli (25%), Klebsiella (20%), and Pseudomonas aeruginosa (15%). The most common gram-positive organisms are Staphylococcus aureus (35%), Enterococcus (20%), and Coagulase negative staphylococcus (15%).[226] The most common primary sites of infection in sepsis are the respiratory tract (50%), intra-abdominal/pelvis (20%), urinary tract (10%), skin (5%), and intravascular catheters (5%).[226]

There has been a rise in the incidence of sepsis and septic shock over the past several decades.[226] Factors that are potentially responsible include the increasing number of immunocompromised people, from acquired immunodeficiency syndrome or cytotoxic and immunosuppressant therapy, and the increase in interventional procedures. Other risk factors for sepsis include malnutrition, alcoholism, malignancy, diabetes mellitus, advanced age, and chronic renal failure.[221]


Although much has yet to be learned about the pathophysiology of sepsis, scientific advances have shed light on many of the factors that lead to the complex cascade that can result in septic shock and death. It has been hypothesized that the manifestations of sepsis result from excessive inflammatory response to bacterial organisms.[227] Gram-negative bacteria contain lipopolysaccharide (LPS) as a cell wall component, which can activate macrophages, as well as the complement cascade.[228] Gram-positive bacteria produce exotoxins, which can activate T cells and macrophages,[227] and release cell membrane components that can activate the inflammatory process.[229] Both pro-inflammatory and anti-inflammatory components are released in response to bacterial invasion, and these two systems are usually tightly controlled to destroy the infection while preventing damage to the host.[226] It is theorized now that sepsis is a result of an imbalance in these two processes, with the pro-inflammatory component overexpressed, [226] [230] [231] [232] however it has been shown that neutrophils in critically ill patients demonstrate functional abnormalities, including reduced migration, superoxide production, and bacterial killing, all factors that may impair host defense. [233] [234] Whether this neutrophil dysfunction leads to worsening of the sepsis syndrome in unknown.

Pro-inflammatory cytokines released in response to infectious stimuli include tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), interleukin-6 (IL-6), and interleukin-8 (IL-8). Anti-inflammatory cytokines include interleukin-10, interleukin-13, and transforming growth factor-β (TGF-β). TNF-α and IL-1 have wide-ranging effects, including activation of macrophages, lymphocytes and neutrophils, increasing expression of adhesion molecules, and increasing production of other pro-inflammatory cytokines.[226] Anti-inflammatory cytokines reduce the production of IL-1 and TNF-α, and inhibit antigen presentation to T- and B-lymphocytes. Animal models have demonstrated that cytokines are a key component of sepsis, with infusions of TNF-α and IL-1 producing a state similar to septic shock, [235] [236] [237] and administration of antibodies to these cytokines resulting in attenuation of the shock-like state. [237] [238] [239] [240]

Other mediators of sepsis include metabolites of the arachidonic cascade such as prostaglandin E2 (PGE2), prostaglandin I2 (PGI2), and thromboxane A2. PGE2 causes vasodilatation seen in septic shock,[226] whereas thromboxane A2 causes platelet and leukocyte aggregation and vasoconstriction.[241] Platelet activating factor (PAF) is produced by many cells in response to inflammatory stimuli,[242] amplifies many cytokines released in sepsis, and stimulates leukocyte activation and adherence to endothelial cells.[226]

Recently, interest has developed in the role nuclear factor-kB (NF-kB) plays in multiple disease processes. NF-kB is a transcription factor located in the cytoplasm of most cell types.[243] Stimulation of the cells by cytokines or byproducts of bacterial and viral infection leads to translocation of NF-kB from the cytoplasm to the nucleus, where it regulates transcription of target genes.[244] It appears that the genes affected by NF-kB activate and modulate cytokines, chemokines, and receptors involved in diseases such as sepsis, SIRS, ARDS, and multiorgan dysfunction.[245] Research is ongoing to see the extent NF-kB is involved in the processes mentioned previously, and if its activation can be regulated.

The coagulation system also plays a role in the manifestation of sepsis. Levels of protein C are decreased, [236] [246] and its conversion to activated protein C, which inhibits thrombosis ( Fig. 60-2 ), is down-regulated during sepsis.[247] Antithrombin III, which is an inhibitor of thrombin and factor X, has been found to be dramatically reduced in septic shock.[248] Tissue factor pathway inhibitor, which inhibits the highly thrombogenic compound tissue factor, has also been found to be reduced in the setting of sepsis.[246] These factors contribute to the widespread microvascular thrombosis that occurs during sepsis, a result of which is reduction in perfusion to various tissues,[249] which may lead to the multiorgan dysfunction syndrome (MODS) seen in many patients with sepsis.

FIGURE 60-2  Proposed actions of activated protein C in modulating the systemic inflammatory, procoagulant, and fibrinolytic host responses to infection. The inflammatory and procoagulant host responses to infection are intricately linked. Infectious agents and inflammatory cytokines such as tumor necrosis factor alpha (TNF-α) and interleukin-1 activate coagulation by stimulating the release of tissue factor from monocytes and the endothelium. The presentation of tissue factor leads to the formation of thrombin and fibrin clot. Plasminogen-activator inhibitor 1 (PAI-1) is a potent inhibitor of tissue plasminogen activator, the endogenous pathway for lysing a fibrin clot.  (From Bernard GR, Vincent JL, Laterre PF, et al: Efficacy and safety of recombinant human Activated Protein C for severe sepsis. N Engl J Med 344(10):699–709, 2001.)



Toll-like receptors (TLRs) are proteins that recognize specific patterns on pathogens. Ten human TLRs have been identified, and when stimulated, they initiate a cascade of signaling events leading to the production of multiple cytokines and effector molecules, including TNF-α, IL-1, and IL-6, and appear to be the primary transducers of the inflammatory response to invading microorganisms.[250] Yet TLR signaling is necessary for host defense because animals with a TLR 4 mutation have higher mortality rates when injected with live gram-negative bacteria.[251] TLRs are present in cells of the innate immune system, as well as solid organs. The kidney expresses most of the TLRs, but TLR 2 and 4 have been most extensively studied because they have a potential role in mediating gram-positive and gram-negative signaling, respectively. A possible role for local renal TLRs in mediating renal failure is being evaluated. TLR4 knockout mice have been found to be protected against renal failure in an LPS model of sepsis, and when kidneys from wild-type animals were transplanted into a TLR4 knockout strain, significant kidney injury developed.[252] It is thus possible that local TLRs directly mediate the kidney injury seen in sepsis.

Clinical Features

Sepsis is a systemic process, and is defined by clinical manifestations. Common clinical manifestations include changes in body temperature (fever or hypothermia), tachycardia, tachypnea, and leukocytosis or leukopenia. Many of the signs and symptoms of sepsis are induced by the inflammatory cytokines. Fever, for example, can be caused by TNF-α and IL-1,[221] and the failure to develop fever has been associated with increased mortality.[253]Hypoglycemia, hyperglycemia, hypokalemia, hyponatremia, hypocalcemia, hypomagnesemia, and hypophosphatemia can also be seen. Tachycardia is a common, but nonspecific manifestation. Patients with severe sepsis and septic shock have hypotension, due in part to nitric oxide release,[254] as well as due to decreased effective circulating volume. The intravascular volume depletion is related to several factors, including decreased systemic vascular resistance (SVR),[255] increased microvascular permeability, and increased insensible losses. Once volume resuscitated, most patients with septic shock have evidence of hyperdynamic cardiovascular function with a normal or elevated cardiac output and decreased SVR.[255] Despite these findings from pulmonary artery catheterization, the heart may not be as hyperdynamic because it should given the clinical setting. Studies have shown that sepsis induces a depression in myocardial function, [256] [257] characterized by elevated left ventricular end-diastolic volume and decreased left ventricular systolic work index. This myocardial depression is believed to be caused by a myocardial depressant substance, which has not been fully identified. Leading candidates are TNF-α and IL-1,[258] but nitric oxide has also been shown to have a negative effect on systolic function.[259]

Tachypnea and hypoxemia are common in sepsis, and the acute respiratory distress syndrome (ARDS) has been reported to occur in up to 40% of patients with sepsis.[260] Many view ARDS as an initial manifestation of multiorgan dysfunction syndrome, and believe it represents diffuse endothelial injury resulting from the exaggerated inflammatory response. [260] [261] [262] [263]

Adrenal insufficiency is a common finding in septic shock, with a reported incidence of 25% to 40%. [264] [265] [266] Some authors now feel the threshold for diagnosing adrenal insufficiency should be a cortisol level below 25 mg/ml to 30 mg/ml, instead of the usual 18 mg/ml to 20 mg/ml, and low dose (1–2 mg) ACTH stimulation should be used for diagnosis because it represents physiologic stress levels of ACTH, in contrast to the standard ACTH stimulation test, which uses doses that are 100-200 fold higher than maximal stress levels of ACTH.[267] It is also advocated that if a fluid-resuscitated patient is hypotensive and requires pressors, a baseline cortisol less than 25 mg/ml should be considered diagnostic of adrenal insufficiency.[268]

Disseminated intravascular coagulation (DIC) is often seen in sepsis, and is characterized by enhanced activation of coagulation, with intravascular fibrin formation and deposition. The resulting microvascular thrombi can reduce blood flow to portions of organs, contributing to the onset of MODS. A reduction in circulating coagulation factors and platelets is often seen because they are consumed in the production of microthrombi, and this can lead to bleeding episodes.[269] Laboratory studies in DIC typically show thrombocytopenia, with an elevation of the prothrombin time, activated partial thromboplastin time, as well as D-dimer.

Central nervous system alterations are frequently found in patients with sepsis,[270] and septic encephalopathy is the most common form of encephalopathy in intensive care units.[271] Impaired mitochondrial function and oxygen extraction by the brain, increased permeability of the blood-brain barrier, and disruption of astrocyte end-feet, all caused by inflammatory mediators, contribute to the diffuse neuronal injury seen in septic encephalopathy.[272]Confusion, disorientation, lethargy, agitation, obtundation, and coma are the common clinical manifestations.

Critical illness polyneuropathy is a common occurrence in the setting of sepsis, and is often first recognized when the patient cannot be weaned from ventilatory support.[270] This illness is caused by axonal degeneration, is characterized by hyporeflexia, weakness distally greater than proximally, and normal or slightly elevated creatine kinase levels,[226] and may take up to 6 months for recovery.[270]

Renal dysfunction is found in 9% to 40% of patients with sepsis,[273] and the mortality rate in these patients is greater than 50%.[274] The clinical manifestations vary from acute tubular necrosis to bilateral cortical necrosis. Hypotension is commonly seen in sepsis, and renal hypoperfusion likely plays a major role in the incidence of acute kidney injury (AKI), as does the administration of nephrotoxic agents to manage sepsis. However, studies evaluating renal blood flow have had conflicting results, with some showing reduction in global renal blood flow, [275] [276] whereas others showed no reduction in renal blood flow or GFR. [277] [278] It is apparent that there are other factors involved in AKI induction. TNF-α released from mesangial cells causes leukocyte accumulation in the glomerulus, as well as apoptotic death of glomerular endothelial cells.[279] IL-1 can induce vasoconstriction, neutrophil aggregation, and further cytokine release.[280] PAF levels, which are increased in sepsis and correlate with the severity of AKI,[281] increase both afferent and efferent arteriolar resistance, producing a decline in the glomerular filtration rate (GFR).[282] Endothelin-1 is secreted in response to septic mediators, including TNF-α, [283] [284] and has been found to cause renal vasoconstriction,[285] as well as inhibition of sodium and water reabsorption by the collecting duct. [286] [287] Thromboxane A2 decreases GFR and renal blood flow, and preferentially vasoconstricts the afferent arteriole.[288] Leukotrienes are released in endotoxemia[289] and also reduce GFR and renal blood flow.[288] Other mediators implicated in septic AKI include the renin-angiotensin system, atrial natriuretic factor, IL-1, adenosine, and catecholamines.[223] The role of DIC and diminished levels of activated protein C in the generation of microvascular thrombi has been discussed, and the diminished renal perfusion caused by the thrombi likely contributes to septic AKI. No specific pharmacologic therapy for septic AKI has proven beneficial.

As improvements in care of the critically ill patient have been developed, death from the initial disease process in sepsis has become less common, patients have lived longer, and the development of the multiorgan dysfunction syndrome has become more common. MODS is now the most common cause of death among patients with sepsis.[290] The exact pathophysiologic mechanism leading to MODS has not been fully defined, but mitochondrial dysfunction, microvascular thrombi, hypoperfusion, ischemia/reperfusion injury, circulating inflammatory factors, diffuse endothelial cell injury, bacterial-toxin translocation, and increased tissue nitric oxide are all potential contributors. [274] [291] [292]


The management of sepsis is primarily based on eradication of the infection and supporting the patient's hemodynamics and other organ systems. Activated protein C has been approved for use in sepsis, but other immunomodulatory therapies are still being evaluated.


Identifying the source of sepsis should be one of the primary goals while treatment is being initiated. Antibiotic choice often depends on the suspected site of infection. Initial antibiotic therapy usually requires multiple antibiotics to cover the likely pathogens,[226] and if culture results identify a source, coverage can be narrowed. Double antibiotic coverage is indicated in the management of Pseudomonas aeruginosa infection, febrile-neutropenic patients, and severe intra-abdominal infections.[293] If no organism is isolated, initial broad spectrum antibiotics can be continued so long as the patient is improving. Immediate institution of antibiotic therapy is critical because there is a 10% to 15% higher mortality in patients not treated promptly.[293] A full discussion of antibiotic selection is beyond the scope of this chapter.

Hemodynamic Support

Intravascular volume depletion, peripheral vasodilatation, and increased microvascular permeability all contribute to hypotension seen in severe sepsis and septic shock,[226] and aggressive volume resuscitation should be the primary initial therapy.[294] The fluid requirements for resuscitation are very large, and often underestimated by the clinician. Up to 10 liters of crystalloid are often required in the first 24 hours.[294] Boluses of fluid should be given until blood pressure, heart rate, or evidence of end-organ perfusion such as urine output have improved. Early therapy is crucial, and a recent study showed that early, goal-directed therapy using central venous pressure, mean arterial pressure, hematocrit, and central venous oxygen saturation as end points improved mortality.[295] Crystalloids and colloids are both used for resuscitation, and there is no evidence-based advantage of one over another.

Despite adequate fluid resuscitation, many patients remain hypotensive,[226] and these patients require vasopressor agents. Dopamine and norepinephrine are both considered first choice vasopressors in the Surviving Sepsis Campaign guidelines.[296] However, chronotropic sensitivity to dopamine is increased in sepsis; thus tachycardia and arrhythmias may limit its use.[226] Norepinephrine is as effective in raising blood pressure as dopamine, but has less cardiac effects; it does not raise cardiac output as much as dopamine, and causes less tachycardia.[297] Phenylephrine has purely alpha effects, and fewer risks of tachyarrhythmias, but experience in septic shock is limited.[297]Epinephrine can be used for refractory hypotension, but has been shown to cause a rise in serum lactate levels.[297] Vasopressin is produced in the posterior pituitary gland and has both vasoconstrictor and antidiuretic properties. Vasopressin levels rise 20- to 200-fold in shock states,[298] but patients with septic shock have been found to have significantly lower vasopressin levels than patients with cardiogenic shock.[299] Vasopressin use in septic shock has been shown to spare norepinephrine use as well as maintain mean arterial pressure and cardiac index.[300] Although norepinephrine causes profound constriction of glomerular afferent arterioles, vasopressin has been shown to constrict the glomerular efferent arteriole, thus increasing the glomerular filtration rate.[301] Vasopressin can cause coronary artery vasoconstriction leading to MI, and because it does not have positive inotropic actions, the increased afterload during vasopressin use may decrease cardiac output.[301] Dobutamine has been used in sepsis to improve oxygen delivery, but can potentiate hypotension due to B2-mediated vasodilatation. Dobutamine is recommended for patients with a low cardiac index (<2.5 L/min/m2) after volume resuscitation,[290] but if profound hypotension is present (systolic blood pressure <80 mm Hg), should be used in conjunction with an agent with more peripheral vasoconstrictor effects such as norepinephrine or phenylephrine.

Several studies have examined whether resuscitation of septic patients to pre-determined end points of global oxygen delivery improves outcome. Earlier studies showed no benefit, [40] [58] [302] but enrolled patients up to 72 hours after admission. A more recent study showed early therapy to a central venous oxygen saturation level (Scvo2) of 70% improved mortality.[295] This was accomplished by sequentially giving fluids to a target CVP of 8 mm Hg to 12 mm Hg, followed by red blood cell transfusion to a hematocrit of 30%, followed by dobutamine infusion to a maximum of 20 mg/kg/min until the Scvo2 reached 70%. The Surviving Sepsis Campaign guidelines now recommend targeting a CVP of 8 mm Hg to 12 mm Hg, MAP>65 mm Hg, urine output>0.5 ml/kg/hr, and Scvo2>70%.[296]

Treatment of the Coagulation Cascade

Disseminated intravascular coagulation is a common finding, and has been discussed earlier. Management of the underlying condition will accelerate resolution of DIC. No specific therapy is recommended for DIC unless severe or life-threatening hemorrhage occurs, at which time replacement with platelets, fresh frozen plasma, and possibly cryoprecipitate is indicated.[246]

The finding that protein C levels are reduced in sepsis and are associated with an increased risk of death[87] led to the study of activated protein C (APC) in the management of sepsis. The PROWESS trial[303] was a randomized, controlled trial that evaluated APC in 1690 septic patients. Patients were eligible if they had known or suspected infection, three or more signs of SIRS, and at least one organ dysfunction. Patients were treated with placebo or a continuous infusion of APC for 96 hours. Patients treated with APC had a 19.4% reduction in the relative risk of death, and an absolute reduction in the risk of death of 6.1%. A post hoc analysis of this study by the Food and Drug administration,[304] however, found that the benefit of APC seemed to be restricted to patients with more severe illness (an Acute Physiology and Chronic Health Evaluation [APACHE II] score of 25 or more). A study is currently underway to assess the efficacy of APC in patients with an APACHE II score of less than 25. PROWESS was the first randomized controlled trial to show a survival benefit of a therapeutic intervention in sepsis. Because APC acts on the coagulation system, there is an increased risk of bleeding associated with its use, and the incidence of serious bleeding in the treated group in PROWESS was higher than in the controls (3.5% versus 2.0%) despite fairly stringent criteria to exclude those at risk for bleeding. Activated protein C should be considered in all patients at high risk for death from sepsis (APACHE II score > 25, sepsis-induced multiple organ failure, septic shock, or sepsis-induced ARDS). A recent study[305] showed no benefit from APC in septic patients at low risk of death (APACHE II score < 25 or single organ dysfunction). Patients with conditions that were exclusion criteria in the PROWESS trial, including pregnancy, breast feeding, chronic renal failure requiring dialysis, acute pancreatitis without a known source of infection, cirrhosis, and HIV infection with CD4+ count less than 50/mm3 should be evaluated carefully. Care should also be taken in the patient at risk for bleeding because this is the main side effect. Antithrombin III has been the subject of many clinical studies in sepsis, [306] [307] [308] [309] and has shown an improvement in DIC. However, no clinical study has shown a survival advantage with AT III. These studies were all of small size, and a large multicenter study would better address the role of ATIII. Recombinant tissue factor pathway inhibitor has been shown to be beneficial in animal models of septic shock, [310] [311] but has not shown survival benefit in human studies.[312] Site inactivated factor VIIa, which competitively inhibits factor VIIa from binding to tissue factor and initiating coagulation, has been shown to prevent sepsis-induced acute lung injury and acute kidney injury in baboons,[313] but has not been evaluated in humans.

Immunomodulatory Therapy

Corticosteroids have long been the subject of studies in sepsis, the rationale being that minimization of the inflammatory cascade could improve outcome. Short-term therapy with glucocorticoids has not improved outcome in sepsis,[314] and a meta-analysis of 10 studies showed no beneficial effect.[315] Yet two recent small studies (40 and 41 patients) did show an improvement in outcome, [316] [317] and a randomized, placebo-controlled trial of hydrocortisone 50 mg IV every 6 hours plus fludrocortisone 50 mg daily via a nasogastric tube for 7 days in patients with septic shock showed a significantly lower 28-day mortality rate in the treated group (55%) compared with patients who received placebo (61%).[318] Low-dose corticosteroids (200–300 mg/day in 3 to 4 divided doses for 7 days) is now recommended for patients with septic shock who, despite adequate fluid resuscitation, require vasopressor therapy.[296] A 250 mg ACTH stimulation test can be used to determine responders (>9 mg/dl increase in cortisol 30 to 60 minutes after ACTH administration), and corticosteroid therapy can be discontinued in these patients. Fludrocortisone (50 mg orally once daily) can be added to the regimen.

Antibodies to TNF have been studied in sepsis, and although most studies have found no benefit, [60] [319] [320] a recent study showed a risk-adjusted relative reduction in mortality of 14.3% for septic patients with an interleukin-6 level of more than 1000 pg/ml when treated with anti-TNF antibody.[321]

Studies have been done to evaluate the benefit of anti-endotoxin,[322] PAF antagonists,[323] bradykinin antagonists,[324] prostaglandin antagonists, [325] [326] IL-1 receptor antagonists,[327] nonselective nitric oxide synthase inhibitor,[328] n-acetyl cysteine,[329] granulocyte colony-stimulating factor,[330] and IVIG[331] in sepsis, and none have been shown to be beneficial. A recent study assessing C1-inhibitor in 40 patients found an improvement in serum creatinine at day three and four, but no survival benefit.[332]

Hemofiltration in Sepsis

The role of cytokines in sepsis and septic shock has led to the theory that removing them by hemofiltration may improve outcomes. Many studies have evaluated the effect of hemofiltration on cytokine levels and have shown clearance of cytokines, including TNF, [333] [334] IL-1, [333] [335] [336] IL-6, [231] [337] and IL-8,[337] by hemofiltration. Although a few studies have shown a reduction in the amount of cytokines in the plasma with hemofiltration,[338] [339] the preponderance of studies have shown no reduction in plasma cytokine levels. [340] [341] [342] [343] [344] [345] The high production rate and rapid endogenous clearance of many cytokines[346] likely results in the amount being removed by hemofiltration to be too minor to change circulating levels. It also appears that a large percentage of the clearance of cytokines occurs as a result of adsorption to the dialysis membrane,[347] which becomes saturated after a short time, limiting the clearance.

In animal models, hemofiltration has improved survival in some studies, [348] [349] but these studies initiated hemofiltration before or shortly after the septic insult, which is generally not possible in clinical practice. Studies using a true infection model have not shown an effect on survival. [350] [351] Prospective human studies to evaluate the benefit of hemofiltration in sepsis have generally been small. Reduction in the hyperdynamic response, including improved systemic vascular resistance,[352] improvement in APACHE II scores,[353] improvement in vasopressor requirement,[354] and beneficial hemostatic changes[355] have been found. One prospective uncontrolled study found that short-term (4 hour) high-volume hemofiltration improved septic shock in 11 of 20 patients treated,[356] but no randomized controlled trial at this time has shown an improvement in survival when hemofiltration is used for sepsis. Recent studies have shown that adsorption coupled with hemofiltration[357] and high-volume ultrafiltration with frequent membrane changes[347] may improve cytokine clearance, but currently there is no evidence to support the routine use of hemofiltration for sepsis.


Cardiogenic shock is a state of decreased cardiac output in the setting of adequate intravascular volume, resulting in inadequate tissue perfusion. The diagnosis can be made clinically by the findings of poor tissue perfusion such as oliguria or cool extremities, along with the hemodynamic criteria of sustained hypotension (systolic blood pressure <90 mm Hg), reduced cardiac index (<2.2 L/min/m2), and congestion (pulmonary capillary wedge pressure > 18).[358]

Cardiogenic shock occurs in 4.2% to 7.2% of myocardial infarctions [359] [360] and is the most common cause of death among patients suffering from myocardial infarction.[361] The first series, in 1967, evaluating the outcome of cardiogenic shock found a mortality of 81%,[362] but a multicenter trial in 2000 found a mortality rate of 60%.[363] Also, the in-hospital mortality rate from ischemic cardiogenic shock in the National Registry of Myocardial Infarction (NRMI) decreased from 60.3% in 1995 to 47.9% in 2004, presumably due to more aggressive revascularization.[364]

The most common cause of cardiogenic shock is massive myocardial infarction,[365] but can also be caused by a smaller infarction in a patient with reduced left ventricular function, acute mitral regurgitation (from papillary muscle rupture), rupture of the interventricular septum, myocarditis, end-stage cardiomyopathy, valvular heart disease, tamponade, or hypertrophic cardiomyopathy. In the SHOCK (should we emergently revascularize occluded arteries for shock) registry and trial[363] of 1422 patients with cardiogenic shock, 79% of patients had left ventricular failure as the cause of shock, whereas 6.9% had severe mitral regurgitation, 3.9% had ventricular septal rupture, 2.8% had isolated right ventricular shock, and 1.4% had tamponade. Shock may be found at presentation in patients with acute myocardial infarction, but can be delayed. In the SHOCK registry, mean time to development of shock was 7 hours after infarction.[363]


In patients with myocardial infarction or ischemia, cardiogenic shock may occur once 40% of the myocardium is lost.[365] The resultant clinical sequelae can potentiate the myocardial damage. Hypotension and tachycardia can increase ischemia in this setting. Coronary blood flow is dependent on the duration of diastole, which is significantly shortened in the patient with tachycardia, resulting in reduced perfusion of the already ischemic myocardium.[365]The elevated wall stress resulting from left ventricular dilatation and pump failure increases myocardial oxygen requirements, which also worsens ischemia.[365] The cellular hypoxia seen in ischemia leads to a reduction in ATP levels, and eventual myocyte swelling.[366] Apoptosis of myocytes occurs after myocardial infarction,[367] and may contribute to the state of reduced cardiac output.

Clinical Features

Hypotension is universal in cardiogenic shock. Tachycardia is often present, but in the SHOCK registry, the mean heart rate was 96.[368] Arrhythmias may be present, and jugular venous distention, pulmonary rales, and a third heart sound are usually found. Signs of hypoperfusion may include confusion, mottling of the skin, and oliguria. In a study of 118 patients with cardiogenic shock, acute kidney injury occurred in 33% of patients within 24 hours, and increased mortality from 53% to 87%.[9] Multiple organ failure develops in many patients, primarily due to ischemia from decreased cardiac output. However, systemic inflammation may play a role in many patients. High plasma levels of IL-6 have been associated with multiple organ failure in this population,[369] 18% of patients in the SHOCK registry showed signs of severe systemic inflammation, leading to a diagnosis of suspected sepsis,[370] and a minority of patients develop distributive shock, possibly secondary to occult sepsis or mesenteric ischemia.[371] Lactic acidosis may also occur from hypoperfusion.[372]


The primary goal in the evaluation of cardiogenic shock is to determine the primary cause. As stated previously, myocardial infarction (MI) with reduced left ventricular systolic function is the most common cause,[362] and is often readily apparent on physical exam. However, other causes of shock, such as sepsis, hypovolemia, and pulmonary embolism, must be considered. An electrocardiogram (ECG) should be performed on arrival, and if an inferior MI is suspected, a right-sided ECG should be performed to evaluate for right-sided involvement. Routine blood tests including cardiac enzymes should be performed, a Foley catheter should be placed to monitor urine output, and a chest radiograph should be obtained.

Echocardiography is a valuable tool to confirm the diagnosis of cardiogenic shock, and can evaluate for potential mechanical causes that require surgical intervention, such as mitral regurgitation, papillary muscle rupture, tamponade, or left ventricular free wall rupture.

Although brain natriuretic peptide (BNP) has been found to be effective in the diagnosis of congestive heart failure (CHF), it is not typically used in the diagnosis of cardiogenic shock. BNP is a 32-aa polypeptide found in the cardiac ventricles. BNP release from the ventricles, and therefore serum levels, are indirectly proportional to ventricular volume expansion and pressure overload.[373] A BNP level greater than 100 pg/ml has a sensitivity of 82% for CHF, increasing to 99% for NYHA class IV CHF.[368] BNP levels have been found to correlate closely with New York Heart Association (NYHA) classification, with mean levels in NYHA class IV to be greater than 900 pg/ml.[374] BNP levels have also been found to decline as pulmonary capillary wedge pressure decreases,[375] making it a possible monitoring tool in the management of severe CHF.


Airway management and maintenance of adequate oxygenation should be the first concern during resuscitation. Bilevel positive airway pressure (BiPAP) or continuous positive airway pressure (CPAP) can improve oxygenation and prevent intubation in severe congestive heart failure,[376] although use in cardiogenic shock has not been well studied. Intubation and mechanical ventilation may be required if supplemental oxygen or noninvasive ventilation cannot maintain adequate oxygenation with minimal work of breathing.

Patients treated with beta-blockers, ace-inhibitors, and nitrates for myocardial infarction that subsequently develop shock should have these agents discontinued because they may worsen the clinical state. Patients with mechanical causes of shock should be evaluated for surgical repair.

A minority of patients with cardiogenic shock may develop hypotension without evidence of pulmonary edema.[377] In these patients, a fluid challenge of 100 cc to 250 cc of normal saline should be given. In the patient who does not respond to fluids, or has pulmonary congestion, inotropic agents should be administered.

Dobutamine is primarily a β1 agonist, but is a weak β2 and a stimulator, and can improve myocardial contractility and cardiac output. Dobutamine is the drug of first choice when the systolic blood pressure (SBP) is greater than 80 mm Hg, but it can induce hypotension as a result of the β2 effect, so should either not be used when blood pressure is less than 80 mm Hg, or used in conjunction with another vasopressor. Dobutamine may worsen tachycardia, and can cause arrythmias,[378] and has not been shown to improve outcomes in patients with cardiogenic shock.

Dopamine should be used when the SBP is less than 80 mm Hg. At low doses (<5 mg/kg/min) β1 effects predominate, and as the dose is increased, a effects become more prevalent. Ischemia of the periphery, tachycardia, and arrhythmias can occur.[378] Norepinephrine is a pure a agonist and can be used when there is an inadequate response to dopamine. Vasopressin has been shown in a retrospective study to increase mean arterial pressure without affecting pulmonary capillary wedge pressure or cardiac index.[379]

Milrinone and amrinone are phosphodiesterase inhibitors (PDE) that increase cyclic amp levels in the myocardium. Increased cAMP increases intracellular calcium in the myocyte, leading to increased contractility. These agents increase inotropicity and cardiac output, without increasing myocardial oxygen consumption.[380] They do not induce direct tachycardia, but do cause a peripheral vasodilatation, which can lead to hypotension and reflex tachycardia, and they may cause arrhythmias. Amrinone, as well as dopamine and dobutamine, may also improve myocardial mitochondrial function during shock.[381] Although several small studies have shown improved hemodynamics with PDE inhibitors, a recent meta-analysis on the effectiveness of catecholamines and PDE inhibitors concluded that such treatment provides little evidence for improved symptoms and may not be safe.[382]

If hemodynamics have stabilized after the initiation of pressors, management of pulmonary edema with diuretics may be initiated. Direct vasodilator therapy can then be considered to decrease preload and afterload, which can improve ischemia. Sodium nitroprusside and nitroglycerin both have short half-lives, and can be carefully titrated, observing for worsening of the hemodynamic state.

Recombinant human brain natriuretic peptide (BNP) has been found to increase cardiac output, decrease pulmonary capillary wedge pressure and systemic vascular resistance, and improve natriuresis and diuresis in decompensated heart failure. [383] [384] It is FDA approved for management of NYHA class IV CHF, but is controversial in patients with CKD. A study in 2005 found a decline in GFR during use of BNP for acute decompensations of heart failure,[385] whereas another study found no effect on renal function.[386] BNP has not been evaluated in cardiogenic shock.

Several drugs are being evaluated for management of decompensated heart failure and cardiogenic shock. Toborinone is a PDE inhibitor that does not induce tachycardia or increase myocardial oxygen consumption in stable congestive heart failure patients. Levosimendan acts by enhancing myofilaments sensitivity to calcium. It has positive inotropic effects, but does not increase intracellular calcium, thus reducing risk of arrhythmia or increased oxygen consumption. Endothelial receptor antagonism is being evaluated as vasodilator therapy.

The SHOCK II randomized clinical trial will evaluate nitric oxide inhibition in cardiogenic shock. NO has a biphasic effect on myocardial function. At low levels, it promotes coronary and myocardial relaxation, but at high levels, NO results in adverse effects and is associated with decreased contractility and reduced effect of β-adrenergic stimulation. Two small studies of NO blockade in cardiogenic shock have been published, one showing improvement in MAP and wedge pressure,[387] and one showing improvement in 4-month survival compared to placebo.[388]

Intra-aortic balloon pumping (IABP) can improve diastolic blood pressure,[389] improve coronary perfusion,[390] and increase cardiac output.[389] In the Global Utilization of Streptokinase and TPA for Occluded Coronary Arteries (GUSTO-1) trial, patients who had early IABP placement had a trend toward improved survival,[391] and the SHOCK trial found a lower in-hospital mortality in patients who received IABP.[392] A recent study evaluated 23,180 patients with cardiogenic shock, 7268 of whom had IABP. IABP was associated with a significant reduction in mortality among patients who received thrombolytic therapy (67% versus 49%), but was not of benefit in patients treated with primary angioplasty (45% versus 47%).[393] IABP has been shown to increase clot lysis when used in conjunction with thrombolytics in animal models,[394] likely due to increased coronary blood flow, and this may explain the improved mortality when IABP and thrombolytics are used together.

Intra-aortic balloon pumping has been reported to have a complication rate between 2.6% and 15%, with a mortality rate of 0.05% to 0.4%. [395] [396] Most of the complications were vascular, with major bleeding occurring in 4.6%, and limb ischemia occurring in 3.3% of patients.[396] A study of 71 patients found a rate of bacteremia and sepsis to be 15% and 12%, respectively.[397] Acute kidney injury was not reported as a complication in these studies, likely because the associated shock made attribution of acute kidney injury to IABP difficult.

The outcome of cardiogenic shock in the setting of myocardial infarction is directly related to the patency of the involved coronary arteries.[398] Therefore, interventions to open occluded arteries are crucial. Thrombolytics have been shown to be able to reduce the incidence of shock when given for acute myocardial infarction,[362] but once shock is established, there is conflicting data. The GISSI trial[399] found a similar in-hospital mortality between patients with cardiogenic shock treated with streptokinase (69.9%), and controls (70.1%). However, the SHOCK trial registry of 1190 patients found a lower in-hospital mortality among patients treated with thrombolytics (54%) compared to those who did not receive thrombolytic therapy (64%),[392] and a study of 23,180 patients in the National Registry of Myocardial Infarction 2 found a mortality rate of 59% for patients receiving only thrombolytic therapy compared with 77% in the group who received no reperfusion therapy.[395]

Despite the apparent benefit of thrombolytics, recent evidence has favored an even more aggressive approach to the management of cardiogenic shock. The SHOCK trial[400] randomized patients to early revascularization or initial medical stabilization, with 63% of the latter group receiving thrombolytics. The early revascularization group had a similar 30-day mortality compared with the medical therapy group (47% versus 56%, p = 0.11), but 6-month mortality[400] (50% versus 63%, p < 0.03), and 1-year mortality[401] (53% versus 66%, p < 0.03) were improved. The survival benefit was limited to patients younger than 75 years old, however other studies have reported different outcomes. In the Shock registry,[402] there was an improved mortality rate in elderly patients who received early revascularization compared with those who had delayed or no revascularization, and a multicenter study of 310 patients older than 75 years old undergoing percutaneous coronary intervention (PCI) reported a hospital mortality rate of 46%.[403] The choice of revascularization in the SHOCK trial was individualized, but generally PCI (64% of procedures) was used for one or two vessel disease, and surgery for left main stenosis or three vessel disease. The mortality in the medical treatment group is lower than in many other studies, and may be related to aggressive use of thrombolytics (63% of patients) and IABP (86% of patients). Other, small studies suggest that administration of platelet glycoprotein IIb/IIIa inhibitor to coronary stenting may improve outcomes further. [404] [405]

Ventricular assist devices have been used in peri-infarction cardiogenic shock, acute myocarditis, and post-cardiotomy shock to bridge patients to either recovery of adequate myocardial function or transplantation.[406]

As stated, acute kidney injury develops in one third of patients in cardiogenic shock, often necessitating a continuous modality of renal replacement therapy (CRRT), given the hemodynamic instability. CRRT in the critically ill will be discussed later.

Ultrafiltration by CRRT has not been evaluated in cardiogenic shock, but it has been proposed as a treatment for severe refractory heart failure, even in patients without renal failure.

Chronic heart failure results in neuroendocrine activation that is initially a response to the decreased effective circulating volume, but eventually serves to worsen heart failure as volume overload progresses. [407] [408] The renin-angiotensin-aldosterone system (RAAS) is activated, resulting in renal sodium and water retention as a result of increased aldosterone. The sympathetic nervous system is activated, and along with angiotensin II, results in peripheral vasoconstriction, which increases wall stress on the left ventricle, worsening heart failure. This increased wall stress due to volume and pressure overload results in ventricular remodeling, which potentiates heart failure,[404]resulting in a further activation of the RAAS. Arginine vasopressin release results in both vasoconstriction and free water retention. In patients who are diuretic resistant, there is no pharmacological means to remove fluid and potentially shut down the cycle of neuroendocrine activation and worsening heart failure. Ultrafiltration, either by continuous or intermittent methods, has been proposed to accomplish this. [405] [409]

A study of 32 patients[410] found that circulating renin, aldosterone, and norepinephrine levels were highest in patients with the most advanced heart failure (NYHA class III to IV and urine output <1000 ml/24 hrs), and treating with ultrafiltration led to significant reductions in these values along with a nearly 500% increase in diuresis (379 ml/24 hrs to 2195 ml/24 hrs), and a doubling of natriuresis (30 mEq/L to 63 mEq/L). However, those patients with urine output of greater than 1000 ml per day were found to have a reduction in diuresis and natriuresis in this study. In addition to improving the cycle of neuroendocrine activation, ultrafiltration has been proposed to improve refractory heart failure by removing the “myocardial depressant factor”, which has been found in ultrafiltrate.[411]

Most studies have been retrospective or case reports. Studies that have been performed have shown that ultrafiltration can improve the hemodynamics in patients with acute heart failure,[412] and have shown that patients can respond to ultrafiltration with improved diuresis, reduced heart failure symptoms, and improved sensitivity to diuretics, [413] [414] [415] [416] but there is conflicting data that hemofiltration may offer no benefit. [417] [418] Unfortunately, all of the studies in this area are small.

Although hemofiltration may improve congestive heart failure in some patients, removal of intravascular volume may not be tolerated by all, and can clearly lead to permanent renal dysfunction. The choice of patients to undergo hemofiltration for refractory heart failure should be made very carefully, and the risks of permanent renal failure should be discussed with the patient in advance.


Fulminant hepatic failure is an acute, and frequently fatal, process that results in severe metabolic abnormalities, neurological complications, and often multiorgan failure. Treatment in a critical care setting is required, and has helped improve the survival of many of these patients over the past several decades,[419] although mortality still remains high. Liver dialysis has been investigated as a possible treatment for encephalopathic patients, and if it becomes widespread, may fall under the domain of the nephrologist.


Acute liver failure from hepatocyte dysfunction results in decreased protein levels, diminished synthesis of clotting factors, manifested by prolongation of the prothrombin time and a decrease in the level of factor V, and often cerebral edema. Fulminant hepatic failure (FHF) is defined as severe acute liver failure, in a patient with no preexisting liver disease, with encephalopathy developing within 2 weeks of the first manifestation of liver disease, which is usually jaundice.[420] Liver failure that is complicated by the onset of encephalopathy between 3 and 12 weeks after the onset of jaundice has been termed subfulminant hepatic failure (SFHF).[420] Because the rate of onset of this disease process is an indicator of prognosis, with the patients having the most rapid onset of encephalopathy also having the best chance of recovery, a newer definition has been proposed to classify fulminant and subfulminant hepatic failure.[421] Hyperacute, acute, and subacute liver failure are defined by the amount of time between the onset of jaundice and the development of encephalopathy (0 to 7 days, 8 to 28 days, and 29 days to 12 weeks, respectively). The survival of hyperacute liver failure has been reported to be 36%, whereas acute has a survival of 7% and subacute 14%.[412] The most common cause of hyperacute liver failure is acetaminophen overdose, although hepatitis A and B can also result in this condition. Acute liver failure is predominantly caused by viral hepatitis and drug reactions,[422] whereas subacute liver failure is most often caused by a hepatitis in which no viral cause can be found.[423]

Patients with chronic liver disease can also develop decompensation with complications similar to acute hepatic failure, except they typically do not develop brain edema.


There are many causes for FHF, and there are regional variations in the etiology. Acute viral hepatitis is the leading cause of FHF worldwide, although the vast majority of cases of viral hepatitis do not manifest as FHF.[424] Hepatitis A virus (HAV) is the most common cause of hepatitis worldwide, but accounts for less than 1% of all cases of FHF.[424] Of all the viral causes of FHF, HAV has the best prognosis with greater than 60% of patients surviving without a transplant.[425] HAV does not lead to chronic hepatitis, and is diagnosed by the presence of HAV anti-IgM antibodies.

Hepatitis B virus (HBV) currently accounts for approximately 15% of cases of FHF in the United States,[422] and 23% in Europe.[424] Coinfection with hepatitis D virus (acquiring HBV and HDV simultaneously) is found in up to 30% of cases of FHF with hepatitis B.[426] Superinfection with HDV can occur in a patient known to have HBV, and can lead to FHF. Hepatitis C virus (HCV) coinfection with HBV may also precipitate FHF.[427] The diagnosis of acute HBV infection is based on the presence of anti-IgM antibodies to HbcAg because HbsAg is often not detected in the acute setting.

Hepatitis C virus has traditionally not been felt to be a cause of FHF, but one recent study identified HCV RNA in 19% of patients with FHF.[428] A cause and effect between HCV and FHF has not been fully established at this time. Hepatitis E virus (HEV) is known to cause FHF, particularly in the third trimester of pregnancy, and is endemic in Asia and Africa. To date no causes of FHF in the United States have been attributed to HEV,[427] but should be considered in patients who have traveled to, or emigrated from endemic areas. Herpes simplex virus and cytomegalovirus have been reported to cause FHF, but usually in the setting of immunosuppression or pregnancy. Adenovirus,[429] human herpes virus 6,[430] Epstein-Barr virus, and influenza virus type B have also been reported to cause FHF.[431] In patients with FHF that is suspected to be viral in origin, 10% to 20% of all cases cannot be identified, and has been termed non-A, non-B, non-C FHF. This disease usually has a subacute presentation, and does not seem to recur after liver transplantation.[422]

Drug toxicity is the second most common cause of FHF, either by direct hepatotoxic effect or by an idiosyncratic reaction. Acetaminophen is the most common drug to cause FHF in the United States and United Kingdom. Acetaminophen is partially converted to the toxic metabolite N-acetyl-p-benzoquinone-imine (NADQI), which is neutralized by reacting with glutathione. If the amount of acetaminophen ingested acutely or chronically overwhelms the ability of glutathione to inactivate NADQI, hepatotoxicity occurs. Persons with depleted glutathione stores from alcoholism or malnutrition are more susceptible to the hepatotoxicity, as are persons taking p450 enzyme inducing drugs. Single doses of as low as 10 g can lead to FHF, but doses within the therapeutic range taken chronically by alcoholics can also lead to FHF. Halothane is known to be hepatotoxic, but is rarely used now in clinical practice. The newer halogenated agents enflurane, methoxyflurane, and isoflurane have a much lower risk of hepatotoxicity. Halothane hepatotoxicity is idiosyncratic, and there is cross-reactivity between the halogenated agents. Many other drugs have been associated with ALF and FHF, including NSAIDS, isoniazid, phenytoin, valproic acid, sulfonamides, propylthiouracil, amiodarone, and ectasy (methyldioxyamphetamine). A thorough history is critical for diagnosing drug related FHF, which can be difficult in a profoundly encephalopathic patient.

Ingestion of the mushroom Amanita phalloides is a rare but well-recognized cause of FHF. Hepatic, pancreatic, renal, and cerebral damage can occur as a result of the toxin a-amantin.[432] Profuse watery diarrhea and abdominal pain develop within 24 hours of ingestion and last 1 to 6 days prior to the development of liver failure. As little as three mushrooms may lead to FHF.

Wilson disease presents as FHF in 10% of cases. This disease may be associated with hemolytic anemia secondary to massive copper release from the liver, and Kayser-Fleischer rings on slit-lamp examination. Although no set of laboratory values can distinguish Wilson disease from other causes of FHF,[433] a low ceruloplasmin level and a low serum alkaline phosphatase are suggestive. Rapid recognition is key because this condition is 100% fatal without a liver transplantation.[421] Other conditions that can present as fulminant hepatic failure are listed in Table 60-8 .

TABLE 60-8   -- Additional Causes of Fulminant Hepatic Failure

Autoimmune hepatitis

Metastatic tumor

Acute fatty liver of pregnancy

Budd-Chiari syndrome

HELLP syndrome

Portal vein thrombosis

Reye syndrome

Right heart failure

Malignant hyperthermia

Acute rejection of liver transplantation




Clinical Features

Fulminant hepatic failure presents with a variety of symptoms. Nausea and vomiting may be the first indicators, followed by jaundice. Encephalopathy may develop rapidly. Several metabolic disorders result from the loss of hepatocyte function. Hypoglycemia results from impaired gluconeogenesis, high insulin levels, and the inability to utilize stored glycogen. Metabolic acidosis is a consequence of poor tissue perfusion and inability to clear lactate. Hypokalemia and hyponatremia also occur frequently. By definition, fulminant hepatic failure requires that encephalopathy be present, the etiology of which is felt to be multifactorial.[434] Ammonia, nitric oxide, manganese, and inhibition of the Na+K+ATPase pump in neuronal cells may all play a role in acute hepatic encephalopathy.[435] Increased production or diminished clearance of “endogenous” benzodiazepines may contribute to hepatic encephalopathy, [436] [437] and enhanced gamma-aminobutyric acid (GABA)ergic inhibitory neurotransmission has also been postulated to have an effect.[438] Hepatic encephalopathy is graded on a scale of 1 to 4 as listed in Table 60-9 . Strong consideration should be given to intubation for airway protection as encephalopathy progresses through stage 3.

TABLE 60-9   -- Stages of Hepatic Encephalopathy

Stage 1: Euphoria, anxiety, disruption of sleep, shortened attention span, mild confusion, slight asterixis

Stage 2: Slurred speech, lethargy, inappropriate behavior, asterixis, hypoactive reflexes, loss of continence

Stage 3: Marked confusion, incoherent speech, hyperactive reflexes, somnolent but “arousable”

Stage 4: Coma, unresponsive to pain, lacking asterixis




Cerebral edema has been found in 40% of patients with FHF and advanced encephalopathy.[439] In patients dying with FHF, uncal or cerebellar herniation (or both) was found to be the cause of death in 80%.[440] The rapid increase in water content of the brain results from increased permeability of the blood-brain barrier.[441] The edematous brain is confined by the cranium, leading to increased intracranial pressure and decreased cerebral perfusion.[442]Cerebral edema is manifested clinically by abnormal pupillary reflexes, systemic hypertension, and bradycardia. Decerebrate posturing and brain stem respiratory patterns are classic, but late findings. Invasive monitoring by a subdural transducer does have a risk of bleeding and infection, [439] [443] but in the sedated and ventilated patient may be the only mechanism to assess intracranial pressure (ICP). Some authors have recommended ICP monitoring for all patients with grade 3 and 4 encephalopathy, [444] [445] and ICP monitoring has been shown to improve the outcome of liver transplantation by excluding those patients with low cerebral perfusion pressure (MAP– ICP) who likely have permanent neurological damage. [446] [447] However, there is currently no consensus on which patients should receive ICP monitoring.[437] Most patients who recover from FHF with associated cerebral edema have full recovery of neurologic function, but permanent brain damage can occur.[448]

Patients with ALF are often found to have profound circulatory changes. A reduction in the systemic vascular resistance, which manifests as hypotension, is common, and has been attributed to high levels of circulating endotoxin and tumor necrosis factor.[446] There is also microthrombi formation in small vessels, resulting in diminished perfusion of metabolically active tissues.[437] These two factors combine to reduce oxygen delivery and extraction, resulting in anaerobic metabolism, and contributing to the lactic acidosis and multiorgan failure seen in many patients. Coagulopathy is a common finding in FHF. Half of all patients develop thrombocytopenia,[449] caused by consumption, reduced bone marrow production, or, in a patient with pre-existing liver disease, hypersplenism. Reduced production of factors I, II, V, VII, IX, and X lead to an increase in the prothrombin time (PT). A reduction in Factor V is seen rapidly because it has a short half-life. The level of factor V is often used as a marker of disease progression, and is considered an independent prognostic factor.[419] The synthesis of coagulation factors II, IX, and X may also be reduced, prolonging the partial thromboplastin time (PTT). Disseminated intravascular coagulation may be seen resulting from a combination of factors, including the release from necrotic hepatocytes of thromboplastic material, platelet activation from circulating bacterial endotoxins not cleared by the liver, and expression of tissue factor on activated endothelial cells stimulating the extrinsic coagulation cascade.[450] Although hemorrhage is uncommon, the gastrointestinal tract is the most common site of bleeding, and intracranial hemorrhage may rarely occur spontaneously.[451]

Bacterial infections may occur in up to 80% of patients with ALF,[452] caused in part by diminished opsonic activity,[453] complement deficiency, and impaired neutrophil function.[454] Fungal infections are also common, the predominant organisms being candida albicans or glabrata and aspergillus. There may be an absence of clinical signs of infection in FHF, so a high index of suspicion must be maintained, particularly when patients have a sudden deterioration. Prophylactic antibiotics have not been shown to be beneficial,[455] but are commonly used.

Renal dysfunction is common, present in up to 55% of all patients with FHF.[448] Direct toxicity can be a result of acetaminophen overdose, radiocontrast, or antibiotic use. The circulatory changes seen in FHF predispose patients to renal dysfunction because they have reduced renal blood flow.[437] Hepatorenal syndrome is a well recognized occurrence and is discussed in Chapter 27 . Many patients with FHF develop severe renal dysfunction requiring renal replacement therapy (RRT), and studies comparing intermittent renal replacement therapy with continuous modes found a significant increase in ICP with intermittent RRT,[456] but no increase with continuous renal replacement therapy (CRRT). [457] [458] There was also no increase in ICP with the CRRT, but a significant increase with intermittent methods. [457] [458] CRRT should be considered first-line therapy, even in hemodynamically stable patients with FHF.


Initial laboratory tests should include chemistry profiles, coagulation studies, complete blood count, toxicology screen, viral serologies, ceruloplasmin (in patients younger than 40 years), creatinine kinase, and urinalysis. Transaminases can be strikingly high, but the levels do not predict outcome. Increases in Bilirubin, prothrombin, and a reduction in factor V do have prognostic value and should be observed closely.

Signs of cardiac or renal failure should prompt consideration of Swan-Ganz catheter placement because the intravascular volume status can be difficult to determine otherwise in FHF. Sepsis must be evaluated for, including a search for fungal infections.

Intravenous H2 blockers are considered routine to prevent gastric bleeding in the setting of coagulopathy. Liver biopsy is needed for diagnosis in a minority of cases because the etiology is usually evident. Transjugular biopsy has become the favored method in the setting of coagulopathy because it has less risk of bleeding than the percutaneous approach.[459]


There are few causes of FHF for which specific therapy is useful. Acetaminophen toxicity should be treated with N-acetyl cysteine (NAC), which enhances the availability of glutathione, and administration up to 36 hours after an overdose of acetaminophen may improve outcome.[460] Therefore NAC should be given anytime acetaminophen overdose is suspected, even if levels are undetectable. The oral dose is 140 mg/kg initially, followed by 70 mg/kg every four hours for 17 doses. When given intravenously, the dose is 150 mg/kg bolus followed by 70 mg/kg IV every four hours for 12 doses. Acyclovir should be used for herpes simplex infection and lamivudine has been proven to be of some benefit for hepatitis B infection.[461] Acute fatty liver of pregnancy and the HELLP (hemolysis, elevated liver enzymes, and low platelets) syndrome require immediate delivery of the fetus. Amanita poisoning should be treated with high-dose penicillin (300,000 to 1 million U/kg/day),[462] which has an antagonistic effect on the mushroom toxin, amatoxin, and sylibin (20 to 48 mg/kg/day),[463] which blocks the hepatocellular uptake of amatoxin. The remainder of therapy for FHF is supportive. Fluid resuscitation is often required in the acute setting because there is hypotension from decreased systemic vascular resistance, as well as increased endothelial permeability, which leads to redistribution. Albumin and fresh frozen plasma have traditionally been considered first-line agents for fluid resuscitation [421] [464] because these patients often have renal sodium retention with ascites formation, and saline tends to lead to third spacing. FFP continues to be used as a first-line agent for volume resuscitation, although there are no studies to support its use over normal saline. Once volume resuscitation is complete, dextrose with 0.45% normal saline should be used for maintenance fluids, with careful monitoring of serum electrolytes.

Hypotension that persists after fluid resuscitation is adequate, as evidenced by a wedge pressure of 12 to 14, requires vasopressors. Norepinephrine is most commonly used due to its preferential effect on peripheral alpha receptors.

In the seriously ill patient, intravenous NAC has been used to improve cardiac output and oxygen extraction, particularly in the setting of pressor use, which may increase tissue hypoxia due to vasoconstriction. Studies of NAC for this purpose have not been consistent, with some showing a benefit,[465] and others showing no improvement in oxygen extraction or markers of tissue hypoxia.[466] Epoprostenol, a prostacyclin that has microcirculatory vasodilatation properties, has been studied in patients receiving vasopressors and found to improve oxygen delivery and the oxygen extraction ratio,[467] but in a study of patients with renal failure, intravenous prostacyclin was found to decrease blood pressure and significantly reduce the cerebral perfusion pressure.[468]

The management of hepatic encephalopathy is directed at limiting the production of ammonia and avoiding benzodiazepines. Most ammonia is derived from intestinal bacteria, and lactulose can help reduce the absorption of ammonia, although it has not been shown to impact survival in patients with advanced encephalopathy.[469] Neomycin has been used, but has not been shown to be useful in hepatic encephalopathy.[470] It also has the potential for ototoxicity and nephrotoxicity, and therefore should be avoided. Flagyl is used to reduce intestinal bacteria and decrease ammonia production. Flumazenil, a benzodiazepine antagonist has been shown to offer short-term improvement in encephalopathy.[471]

Cerebral edema is a common cause of death in patients with FHF and should be treated aggressively. When ICP monitoring is performed, the cerebral perfusion pressure (CPP) should be maintained above 50. An ICP greater than 30 and a CPP less than 40 for 2 hours has been associated with permanent neurological damage. [442] [469] Raising the head of the bed to 20 to 30 degrees may improve ICP. Fever can increase ICP and should be aggressively treated. Mannitol is commonly used in doses of 0.5 to 1.0 g/kg, and is effective in reducing the ICP in patients with normal renal function. A diuresis of twice the volume of mannitol given should be expected in one hour. This dose can be repeated but the serum osmolality must be monitored and mannitol stopped when the osmolality reaches 320 mOsm/kg. Mannitol can lead to intravascular volume depletion and acute kidney injury, and its use in patients with ongoing renal dysfunction will likely lead to acute kidney injury. In the setting of renal failure, mannitol can be coupled with hemofiltration to maintain the osmotic effect. Twice the volume of mannitol administered should be removed by ultrafiltration to ensure this effect.[437] Hyperventilation to a Pco2 to 30 is routinely used to transiently reduce the intracranial pressure, but one study showed this had no benefit,[472] and a Pco2 below 24 is associated with cerebral vasoconstriction.[473] Agitation should be kept to a minimum. If the previously mentioned measures are inadequate to keep the CPP above 50, pentobarbital may be used to induce coma. Hypothermia has been evaluated in one small study, but is not considered standard of care.[474]

Most patients with FHF develop coagulopathy, but spontaneous hemorrhage is uncommon.[41] Parenteral vitamin K should be given for 3 days if coagulopathy develops, and fresh frozen plasma should only be given for bleeding or in advance of invasive procedures. Other therapies including insulin and glucagon,[475] corticosteroids,[476] and exchange transfusion[477] have not been shown to be beneficial, whereas plasmapheresis has had mixed results, and is not widely used. [478] [479]

Patients with acute liver failure are placed at the top of the liver transplant list, but the shortage of donor organs means that many will die waiting for transplantation. The King's College criteria ( Table 60-10 ) are utilized to help decide when a patient should be listed for a transplant. Contraindications include uncontrolled intracranial hypertension, sepsis, adult respiratory distress syndrome, and dependence on pressors.[424] Transplantation of human hepatocytes into the splenic bed has been tried with some success.[480] Living related transplants have been used in Japan for FHF with a reported 1-year survival of 90% in a series of 14 patients,[481] and has been used successfully in the United States.[482]

TABLE 60-10   -- King's College Criteria for Liver Transplantation

Acetaminophen overdose patient

Arterial pH < 7.3 with any degree of encephalopathy

INR >6.5

Creatinine > 3.4 mg/dl in presence of grade III or IV encephalopathy

Non-acetaminophen related liver failure

INR > 6.5 with encephalopathy or any three of the following variables:

Age <10 or >40

Etiology: idiopathic nonviral hepatitis, idiosyncratic drug reaction

Jaundice >7 days before onset of encephalopathy

Serum bilirubin >17.5 mg/dl

INR >3.5

Modified from Shakil AO, Kramer D, Mazariegos GV, et al: Acute liver failure: Clinical features, outcome analysis, and applicability of prognostic criteria. Liver Transpl 6:163, 2000.

INR, international normalized ratio.





Although the toxins that lead to hepatic encephalopathy as well as the other manifestations of FHF have not been identified, extracorporeal detoxification has been examined as a way to remove these toxins, and potentially improve outcome. The Liver Dialysis unit utilizes hemodiabsorption (hemodialysis with a thick suspension of pulverized sorbents replacing the dialysate solution in the dialyzer) to remove potential toxins from the blood. The small particle size of the charcoal provides 300,000 M2 of surface area, much more area for absorption than typical charcoal columns that only have a few thousand M2.[483] Blood is pumped at a rate of 200 ml/min to 250 ml/min, and toxins pass directly across the cellulose membrane where they bind to the small particles of charcoal or cation exchangers in the sorbent solution. Treatment of 6 hours daily for 5 days was studied in a trial of patients with either FHF or acute on chronic liver disease.[484] Fifty-six patients were enrolled, 31 treated with Liver Dialysis, 25 as controls. In the acute on chronic liver failure group, 72% of those treated with Liver Dialysis had recovery of liver function or improvement to transplantation, whereas only 36% of the control group had these outcomes (p = .036), but patients with FHF showed no benefit from treatment with Liver Dialysis. It is unclear why the difference was so striking between the acute on chronic liver failure group and the FHF group, but it has been postulated that the toxins in FHF are more protein bound, such as the toxins of sepsis, and are not removed as effectively.[480] Both groups did show an improvement in neurologic outcomes after treatment, with over 50% of treated patients showing improvement in encephalopathy.

Another sorbent system that has been developed for management of hepatic failure is the molecular adsorbent recirculating system (MARS). In this system, a non-albumin permeable polysulfone dialysis membrane is used, with blood perfused on one side and 20% albumin on the dialysate side. Albumin bound substances such as bilirubin, bile acids, tryptophan, and fatty acids are cleared across the membrane to the dialysate albumin,[485] which is regenerated by passing through a charcoal column, followed by an anion exchange resin column and finally dialyzed against a bicarbonate buffered dialysate to remove the small toxins such as ammonium and aromatic amino acids. Results of trials are similar to Liver Dialysis, with improvement in outcomes in acute on chronic liver failure, but not FHF, and improved encephalopathy scores in both groups.[462] MARS was also found in one small study of 13 patients to improve survival in hepatorenal syndrome,[486] with none of 5 patients treated with hemodiafiltration alone alive at 7 days, whereas 3 of 8 patients also treated with MARS were alive at 7 days and 2 of 8 alive at 30 days. MARS is available in Europe for patients with liver failure, but is currently available for investigational use only in the United States.

The high mortality rate in FHF patients awaiting a liver transplant has made clear the need for a temporary support system that can act as a “bridge” until the failing liver regenerates, or until liver transplantation is available. A bioartificial liver (BAL) is a device in which hepatocytes are inoculated into one side of a semi-permeable membrane. Blood is passed into a plasma separator, the plasma is warmed, oxygenated, and then passed through the device that houses the hepatocytes.[487] The membrane acts to prevent antibodies from entering the cell compartment from the plasma, but allows hepatocytes to extract oxygen, nutrients, and toxins from the plasma, and allows metabolites to pass from the hepatocytes into the plasma. The current BAL devices undergoing clinical trials are using either porcine hepatocytes or immortalized human hepatocytes. Three BAL devices using porcine hepatocytes have been evaluated in phase I trial and found to be safe, but only one, the HeatAssist, has been evaluated in a larger study.[488] In this study of 171 patients, including 147 with FHF or SFHF, no survival advantage was found in patients treated with BAL compared to controls. But when the subgroup of FHF/SFHF was controlled for transplantation, there was a significant improvement in relative risk for mortality in those treated with the BAL. One BAL device utilizing immortalized human hepatocytes had similar positive results in early evaluation,[489] but in a randomized, controlled trial of 24 patients demonstrated no survival benefit.[490] Further trials are needed to assess the efficacy of these devices.

Renal Replacement Therapy in the Intensive Care Unit

Acute kidney injury in the ICU is a grave situation. ICU patients suffer multiple organ failure concomitant with or prior to the onset of AKI. These patients are more likely to be septic, volume overloaded, and profoundly acidotic. They often require blood pressure support and mechanical ventilation and they are twice as catabolic as patients with acute kidney injury outside the ICU setting. Secondary to these co-morbidities the ICU AKI patient is likely to require renal replacement therapy. Those ICU patients who develop AKI and require renal replacement therapy have significantly higher mortalities. Clermont and colleagues compared mortality in ICU patients without AKI, those with AKI not requiring RRT, those with ESRD on RRT, and those with AKI requiring RRT. The respective mortalities in these populations were 5%, 23%, 11%, and 57%.[491] An increase in mortality clearly reflected a worse overall medical co-morbid condition, but also suggested an effect on worsening mortality of renal failure. [491] [492] Whether dose, timing, or type of renal replacement therapy can have an effect on their outcome has just recently become an area of intense research to the nephrology and ICU communities.[493]

Historically, the decision to initiate renal replacement therapy was based on severe life-threatening complications of renal failure. Little attention had been placed on the nuances of the ICU patient with acute kidney injury. Hence renal replacement therapy in the ICU had been administered with similar goals and techniques applied in the outpatient dialysis setting. Although a large body of literature exists for the outpatient ESRD population, it should not be assumed to apply in the ICU setting. There are marked differences in these populations: the principle of “steady-state” kinetics are not applicable in AKI. The PCR of patients in the ICU is typically twice that of other populations.[494] [495] Furthermore, severe volume overload not only influences ultrafiltration requirements but also changes solute distribution volumes, and the use of vasopressors may decrease the efficiency of renal replacement therapy. Finally, adequate therapy including the timing of RRT initiation and dose has not been defined for AKI.

Decisions regarding how to deliver RRT in the ICU should be made with every effort to improve the patient's mortality and morbidity. When approaching the patient who requires renal replacement therapy the following questions should be asked: What access is best? When is it best to initiate RRT? Which RRT modality should be utilized? What degree of solute clearance is required?

Indications for Renal Replacement Therapy

Indications for initiation of RRT in the ICU should be expanded over those for initiation of RRT for ESRD. The presence of AKI requiring renal replacement therapy is associated with a significant rise in ICU mortality. Mehta appropriately separated indications for RRT in the ICU into “renal replacement” therapy and “renal support” therapy.[496] The indications for renal replacement are similar to those typically used for initiation of dialysis in ESRD patients, albeit expanded in the care of the ICU patient. The mnemonic AEIOU is useful when considering replacement therapy in the ICU.

Acidosis has historically been an indication for RRT when associated with azotemia. RRT is initiated as the serum bicarbonate declines and intravenous supplementation is inadequate or unacceptable due to sodium overload. However, with continuous renal replacement therapies lactic acidosis can be controlled successfully and acidosis secondary to permissive hypercapnia can be modified.

Electrolyte abnormalities are an indication for RRT in the ICU. Although hyperkalemia is the most common indication for intervention, all electrolyte abnormalities can be modified with RRT. Sodium can usually be modified with changes in free water intake/delivery but in rare instances RRT may be necessary. Calcium, phosphate, and uric acid abnormalities, as seen in tumor lysis syndrome and hypercalcemia from any cause, may require RRT to correct.

Intoxications frequently require RRT—Lithium, Theophylline, Ethylene Glycol, Methanol, Aspirin, Phenobarbital, and Cytoxan overdoses may all be managed with RRT.

Volume overload is often and easily managed with RRT, but is usually reserved for patients with oligoanuria unresponsive to diuretics. These patients are initiated on a form of RRT when they have pulmonary edema, severe hypertension, or significant edema.

Uremia is also an indication for RRT. Unfortunately, no clear definition of uremia in the ICU patient exists. Certainly mental status changes and pericarditis are easily defined complications of renal failure, they occur late in the course, and likely should not be markers for initiation of therapy.

Renal support indications as defined by Mehta[496] represent a change in therapy from ameliorating the conditions directly resulting from lack of intrinsic renal function to one that supports the patient and the effects of the complications from other organ failure. The goal of therapy becomes increasing survival time to allow for recovery of multiple organ systems including the recovery of renal function. Potential indications for renal support are extensions of renal replacement and novel approaches for care in the ICU. Volume overload without oligoanuria or even significant azotemia is an example of renal support. Continuous renal replacement therapies can be used in the patient with total body overload, with less than adequate urine output, despite a response to diuretics. Renal support can allow for administration of total parenteral nutrition, fluid removal in congestive heart failure, and total fluid management in the patient with multiorgan failure. A patient in the ICU may require inputs of greater than 3 liters per day if nutrition is maintained and antibiotics or blood products are required ( Table 60-11 ). Continuous therapies allow for continuous fluid removal in excess of input despite hypotension or pressor requirements. Postoperative mortality rises with the percentage of body weight increase in the ICU[497] and CRRT allows for this fluid removal postoperatively, potentially reducing morbidity and mortality. In addition fluid removal with continuous therapies has been shown to maintain urine output and GFR when compared to intermittent therapies by Manns and colleagues.[498] Finally, McDonald and co-workers have demonstrated better nutrition in the ICU patient on CRRT versus intermittent therapy.[499]

TABLE 60-11   -- Daily Fluid Requirements for a Typical Patient in the Intensive Care Unit

Fluid Administration in the Intensive Care Unit


Volume (l/day)

Medications (antibiotics/pressors)


Blood products


Alimentation (TPN, enteral feeds)


Total obligate input

3–7 liters/day


TPN, total parenteral nutrition.




Knowing when to initiate RRT in the ICU is a complex decision. Traditionally, waiting for a life-threatening indication for renal replacement has dictated timing. However, as we consider renal support requirements, and the extremely high mortality in these patients, earlier intervention seems appropriate. Gettings and associates retrospectively reviewed survival in trauma patients receiving CRRT in the ICU. Both groups received equivalent clearances and had similar characteristics. However, the group initiated on CRRT earlier, with a BUN less than 60 mg/dl (mean 42) had a survival of 39% compared to only 20% when CRRT was initiated after the BUN exceeded 60 mg/dl (mean 96).[500]

In the absence of controlled trials, definitive guidelines are not available. If the patient in the ICU is approached with consideration of renal support as well as renal replacement as proposed by Mehta, the patient should be initiated on extracorporeal therapy when clinical prowess suggests an improvement can be obtained over the next 24 hours that outweighs the risk. Similarly, the decision to withhold RRT in the ICU should be based on the estimation that a lack of intervention with an extracorporeal therapy will not be detrimental to the patient. In other words, if return of renal function is likely or conservative management with furosemide or other therapies are likely to succeed without harm to the patient, it is reasonable to observe the patient without extracorporeal therapy.

Vascular Access

Vascular access in the ICU is an important and often overlooked aspect of extracorporeal therapy. Poor access can lead to significant recirculation and inadequate flows, resulting in less efficient therapy and delivery of less than prescribed Kt/V with intermittent hemodialysis. Poor access flow and high recirculation are also detrimental in continuous therapies. Increases in hematocrit in the system, resulting from recirculation, rises leads to clotting of the extracorporeal circuit. Factors determining catheter function include location of catheter placement and catheter design. Other factors of unstudied significance primarily include patient characteristics.

Most present day non-cuffed dialysis catheters are manufactured using polyurethane. This is fairly firm for easy insertion but relaxes when remaining at body temperature. These catheters are placed via the Seldinger technique. Some catheters are made of silicone; these are thicker walled and more flexible. Placement of these catheters requires a peel away sheath or stiffening stylet. Non-cuffed catheters range in length from 15 cm to 24 cm. The 15 cm catheters are designed for placement via the right internal jugular vein. Although in most men and larger women 19 cm to 20 cm may be required to reach the superior vena cava/right atrial junction. Catheters 20 cm and longer are designed for the left internal jugular and femoral approaches.

Cuffed catheters are designed for a tunneled placement. In general these are placed when expected use of the catheter exceeds two to three weeks.[501] These catheters are silicone, hence more flexible than acute catheters, and longer. Placement is performed with a modified Seldinger technique using a peel away sheath, and is more technically challenging. Cuffed hemodialysis catheters vary in length from 50 cm to 90 cm and are designed for internal jugular, femoral, trans-hepatic, or translumbar placement.

Catheter tip design varies by manufacturer. Designs include step tip (with or without sideholes), single lumen with a septum (varying geometric designs), split tip, or two separate single lumen catheters. Studies comparing design variations, in patients, have not been performed with the acute catheters. Presently all seem to have adequate initial flow and function when placed in the internal jugular vein. Cuffed dialysis catheters have been studied in patients. Achieved blood flow using either the split tip catheter, large step tip catheters (14.5 french), and twin catheters was similar to blood pump setting as measured with ultrasound dilution—100% of desired flow at blood pump 300 cc/min and 93% to 95% of flow at 400 cc/min. A 15.5 french step tip tunneled dialysis catheter had lower flows—97% of desired flow at 300 cc/min and only 82% at blood pump speed of 400 cc/min. [502] [503] [504]

Recirculation has been measured by catheter design and catheter placement. Recirculation in non-tunneled dialysis catheters varies by location. Studies have demonstrated significantly higher recirculation values in femoral vein catheters than central-subclavian or internal jugular veins. Most catheter recirculation is less than 5% in the subclavian or internal jugular veins. [504] [505] However, in the femoral vein, catheter recirculation may exceed 50% with a mean of 20% to 30% at blood pump speeds of 300 cc/min. The amount of recirculation increases with blood pump speed and decreases when catheter length exceeds 19 cm.[501] Catheter design may also play a role in recirculation.

When catheter recirculation has been studied in cuffed tunneled dialysis catheters, the recirculation in the split tip catheter design was superior. At blood flows of 400 cc/min the split tip design had a mean recirculation of 1.3% to 4.9% compared to 5.2% with 14.5 french step tip catheters, 5.7% to 7.2% with 15.5 french split tip catheters, and 10.9% with twin catheters. [502] [503] [504]

In the acute dialysis unit flow and recirculation can be measured with ultrasound dilution at varying pump speeds to maximize clearance with IHD. Also recirculation can be monitored with CRRT on a daily basis and catheters changed or repositioned if recirculation rises. This theoretically could reduce or eliminate the degree of anticoagulation required and improve the circuit life of the CRRT system. However, the value of this approach to date has not been evaluated.

Despite improved flow and function of the subclavian placement versus femoral vein, the subclavian access should be avoided whenever possible. Subclavian catheter insertion is associated with an unacceptable rate of central vein thrombosis and stenosis.[506] This damage to the vessels leads to loss of future arteriovenous fistulas and graft sites, and frequently in patients with acute kidney injury, it is difficult to determine who might need chronic renal replacement therapy either at discharge or in the future. Despite the well-documented problems with subclavian catheters and clear guidelines,[501] a surprisingly high number of subclavian catheters are still used. According to the 2002 DOPPS study, 18% of European acute catheters and 46% of acute catheters in the United States are subclavian.[507] Dialysis catheters in the ICU should be placed in the internal jugular or the femoral position, the right internal jugular is usually preferred, and femoral catheter length should exceed 19 cm.

Care of the Catheter

Catheter care is designed to prevent malfunction and infection. Malfunction is usually related to thrombus or fibrin sheath formation. Prevention with any particular type of locking solution is not proven. Most institutions continue to use heparin or 4% citrate. When catheter dysfunction is present, line reversal may be successful and has acceptable recirculation characteristics as demonstrated by Twardowski.[508] Alternatively, locking agents are used based on local preferences but no controlled data are available to demonstrate effectiveness.

Infection is prevented by use of excellent local care and observation of DOQI guideline number 15 ( Table 60-12 ).[501] In addition, Oliver and colleagues[509] have demonstrated improved infection rates in acute catheters with the use of local antibiotic ointment on a dry gauze at the exit site. Finally, the risk of bacteremia is associated with the development of an exit site infection and duration of catheter use. For femoral catheters, the risk of infection rises dramatically after 1 week and the risk of IJ catheter associated bacteremia rises after 3 weeks.[510] It is reasonable to try to limit catheter duration to within these periods. Once catheter exit site infection is recognized, the catheter should be removed as risk of bacteremia rises within days—2% at 24 hours and 13% at 48 hours.[510] Catheter-associated bacteremia is treated with catheter removal and intravenous antibiotics.[501]

TABLE 60-12   -- Recommendations for the Care of the Temporary Hemodialysis Access

Catheter Care Item


Catheter insertion site

Femoral—Single use, severe CHF, use in bed-bound; use in ICU when other sites too risky.


Internal jugular—Preferred site when accessible (lowest recirculation and lowest risk of stenosis).


Subclavian—Avoid due to risk of stenosis.

Catheter placement

Femoral—Use >19 cm catheter. Consider monitoring recirculation to optimize efficiency.


Internal jugular and subclavian—right, use 15–20 cm catheters. Left approach needs 20 cm–24 cm catheters. Place tip at SVC/RA junction.

Exit site care

Dressing care and catheter manipulations performed by trained staff.


Examine catheter site each dialysis.


Dressing—Dry gauze with skin disinfection, use povidone iodine ointment or mupirocin ointment with dressing changes.


Use sterile technique at all times; patient and nurse wear mask. Nurse wears gloves.


Adapted from Rolando N, Gimson A, Wade J, et al: Prospective controlled trial of selective parenteral and enteral antimicrobial regimen in fulminant liver failure. Hepatology 17:196–201, 1993.

Duration of use

Femoral—Less than 1 week


Internal jugular—Less than 3 weeks


From Lee WM: Management of acute liver failure. Semin Liver Dis 16:369–378, 1996.


CHF, congestive heart failure; SVC/RA, superior vena cava/right atrial.




Modalities of Renal Replacement Therapy

Of the therapies of RRT available in the ICU, a superior modality has not been definitively demonstrated. CRRT offers recognized advantages in continuous fluid and solute control but has never been convincingly shown to improve mortality. Intermittent hemodialysis (IHD) can be performed and recently attention has shifted to increasing the frequency above traditional thrice weekly. Slow low efficient daily dialysis (SLEDD) increases volume control and solute clearance when compared to IHD and approaches CRRT. Peritoneal dialysis (PD) is also used in the ICU but requires an intact peritoneal cavity. A decision regarding which modality is selected depends on local preferences, cost, and availability of therapies. However, ideally the therapy should be tailored to the needs of the ICU patient. Regardless of the RRT utilized, the goals should always be to improve the patient's fluid, electrolyte, and acid base balances, allowing for the greatest chance of renal and patient recovery.

Continuous Renal Replacement Therapies

With the advent of safe placement of double lumen venous catheters, the CRRT options in general have developed into continuous venovenous modalities. Continuous arterial venous therapies, as initially described by Kramer in the 1970's, have the advantage of blood flow and filtration determined by blood pressure. This gives the theoretical advantage of fewer hypotensive episodes. Arterial venous circuitry also requires a simple, low volume extracorporeal system with low resistance, thus obviating the need for expensive and complex pump-assisted devices. However, this therapy cannot obtain the higher volumes of ultrafiltration required for consistent metabolic control.[511] Arterial venous circuitry also requires a large bore femoral artery catheter with a risk of vascular complication estimated to up to 10%.[512] For these reasons arterial venous forms of CRRT are not widely used.

Pump-assisted circuits provide multiple options for fluid, electrolyte, and metabolic control (Figs. 60-3 and 60-4 [3] [4]). In slow continuous ultrafiltration (SCUF) the extracorporeal system is simplified to include a blood system hooked inline with a high efficiency or high flux membrane. As blood passes through the membrane, plasma water and solutes pass through the membrane to allow formation of an ultrafiltrate, which is discarded. No replacement fluids or dialysate fluids are required. Although SCUF is a purely convective modality, the ultrafiltrate volume is limited to inputs plus desired losses and hence not enough volume to control azotemia or significant metabolic disorders. This therapy is reserved for patients with residual renal function, but high volumes of input and/or significant fluid overload with a “relative” oliguria.

FIGURE 60-3  Continuous renal replacement circuitry. Circuits will also include a pre-pump port for anticoagulant administration, a post filter air detector and pressure monitoring for access, filter, and filtrate. A, Circuitry for slow continuous ultrafiltration. B, Pre-filter replacement for continuous venovenous hemofiltration. Note, in some systems the replacement may be post blood pump but pre dialysis filter. C, Post filter replacement continuous venovenous hemofiltration. D, Schematic of continuous venovenous hemodialysis. E, Circuit of continuous venovenous hemodiafiltration. Replacement fluid could be added pre or post pump and pre filter, or replacement fluid could be added post filter.


FIGURE 60-4  Illustration of pre and post blood composition as blood traverses a hemodialysis filter versus blood composition pre and post filter using hemofiltration. A, Note the significant clearance of small molecules by diffusion in hemodialysis. Comparatively, in hemofiltration, there is little change in solute concentration, except for a rise in hematocrit. The clearance in hemofiltration is solely convective. B, In hemofiltration a change in blood solute concentration is achieved by adding replacement fluid.


Continuous renal replacement therapy modalities designed for fluid and metabolic control require higher volumes of ultrafiltrate and hence require replacement fluid, dialysate fluid, or a combination of both. Continuous venovenous hemofiltration (CVVH) is an extracorporeal circuit with a double lumen venous catheter hooked to an extracorporeal system with a blood pump, high efficiency or high flux dialysis membrane, and replacement fluid. As in SCUF, pure convection produces the ultrafiltrate; however, much greater volumes are generated. Volume status and metabolic improvements are maintained by the addition of replacement fluid in the circuit. Replacement fluid can be added pre or post filter. Pre filter replacement carries a benefit of less hemoconcentration within the dialysis membrane but decreases clearances up to 15%.[513] Post filter replacement maintains efficiency of the circuit but may be associated with an increase in thrombosis of the extracorporeal circuit.[514] Continuous venovenous hemodialysis (CVVHD) is an extracorporeal circuit with a double lumen venous catheter hooked to an extracorporeal system with a blood pump, high efficiency or high flux dialysis membrane, and dialysate fluid. As in intermittent hemodialysis, the dialysate runs countercurrent to the blood pathway. As in other forms of CRRT, low efficiency is maintained by limiting dialysate volume to 1 to 3 liters per hour. While considered a diffusive therapy, certainly convection occurs due to high permeability of the membrane and back filtration. The degree of back filtration and convective clearance is likely to vary by membrane and has not been studied to date. However, Brunet and co-workers has shown significant middle molecule clearance with a PAN dialyzer (Gambro M100) and a dialysate rate of 2 liters per hour. Beta2 microglobulin clearance was 80% of that achieved with 2 liters per hour of CVVH therapy.[513]

Continuous venovenous hemodiafiltration (CVVHDF) consists of an extracorporeal circuit with a double lumen venous catheter hooked to an extracorporeal system with a blood pump, high efficiency or high flux dialysis membrane, and both replacement and dialysate fluid. In general, this combination is used to increase clearance; others utilize CVVHDF to simplify citrate delivery for anticoagulation. [515] [516] Clearly, both diffusive and convective forces determine solute clearances.

Intermittent Therapies

Conventional hemodialysis is typically delivered in the ICU setting three to four times a week with a standard Kt/V prescribed similar to that of ESRD.[517] Recently Schiffl and colleagues demonstrated improved mortality and renal recovery when the frequency was increased to daily.[518] Multiple studies have demonstrated an improvement in survival with use of biocompatible membranes.[519] Conventional hemodialysis is diffusional with ultrafiltration added as desired or tolerated to control fluid overload. Blood flow with intermittent therapies in the ICU ranges between 200 cc/min to 500 cc/min. It may be varied to reduce recirculation or improve hemodynamic tolerance of therapy. Dialysate composition and flow are similar to ESRD therapy with dialysate flows of 500 to 800 cc/min. Duration of therapy is typically 3 to 5 hours. SLEDD has been developed to be a hybrid therapy between the continuous and intermittent therapy. Blood flows and dialysate flows are slowed to 200 to 300 cc/min and 100 cc/min respectively, and the therapy is prolonged to 8 to 12 hours daily. [511] [520] This slower form of dialysis theoretically allows for more hemodynamic stability with increased clearances when compared to conventional hemodialysis. SLEDD also allows for time off the extracorporeal circuit, allowing for time for the patient to travel to diagnostic studies. Because SLEDD utilizes standard hemodialysis machinery, dialysate is generated on line and the cost is potentially lower than that of CRRT, particularly if ICU personnel can be trained to monitor the dialysis session obviating the need for dialysis nurses to be bedside for the 8 to 12 hours of therapy.

Peritoneal dialysis (PD) can also be utilized in the ICU. An intact peritoneum is required and a temporary or permanent PD catheter is placed. Standard peritoneal dialysis solutions can be used. Clearance is convective and diffusive. Increasing dextrose concentration in the dialysate and increasing frequency of exchanges adjusts the quantity of ultrafiltration. Transport kinetics are likely quite variable between patients and have not been studied in the setting of AKI. Successful peritoneal dialysis has been described in the ICU setting with adequate results,[521] but recently PD has been demonstrated to be inferior to CRRT in AKI due to sepsis and malaria.[522] Peritoneal dialysis offers the advantages of lack of anticoagulation and vascular problems, hemodynamic stability, and are performed with relatively inexpensive and simple systems. The disadvantages include the need for an intact peritoneum, hyperglycemia, potential respiratory embarrassment due to increased abdominal pressure, risk of peritonitis, and less clearance than can be obtained with SLEDD or CRRT.

Prescription and Comparisons of Therapeutic Options

Unlike ESRD, there are no established standards of therapy in AKI. Absence of steady state, variable excesses of body water, variable protein catabolic rates, variable frequency of therapies, and potential lack of correlation of urea as a marker for toxins in AKI hamper quantification of delivered therapy. An adequate dose of dialysis, the lowest clearance that maximizes survival and decreases morbidity, has not been established.

Liao and colleagues modeled the dose capabilities of renal replacement therapies in acute kidney injury.[523] This allows a comparison of removal of urea, inulin, and Beta2 microglobulin between commonly used prescriptions of these extracorporeal therapies. Liao and co-workers modeled clearances of IHD, SLEDD, and CVVH to compare clearances. Based on a 70 kg male with 10 kg of volume excess and an initial BUN of 90 mg/dl, comparisons were modeled with CVVH of 3 liters per hour with pre-filter dilution, daily (6/week) IHD of 4 hours duration with a blood flow of 350 cc/min and dialysate flow of 600 cc/min, and SLEDD performed 7 days per week for 12 hours with a blood flow of 300 cc/min and dialysate flow of 100 cc/min. Using these assumptions the equivalent renal clearance (EKR) of urea, inulin, and Beta2 microglobulin were estimated with each therapy. Relative to SLEDD and daily IHD effective urea clearance was 8% and 60% higher in CVVH (EKR of urea in cc/min with CVVH is 33.7, SLEDD is 31.3 and IHD 21.1). Differences in clearance of middle molecules was even more pronounced in CVVH (EKR of inulin in cc/min was 11.8 with CVVH, 5.4 with IHD, and 3.0 with SLEDD. EKR of Beta2 microglobulin was 18.2 in CVVH, 7.0 in IHD, and 4.2 in SLEDD). These superior middle molecule clearances achievable with CVVH are due to the combination of convective clearance and continuous operation. This prescriptive model must be clinically validated.

Even if the adequate dose of dialysis was known, substantial barriers remain in obtaining this dose. With intermittent hemodialysis, Evanson has shown that delivered dose of dialysis was significantly below prescribed (Kt/V prescribed 1.25+/-0.47 and delivered 1.04+/-0.49).[517] Schiffl and co-workers also noted this in their comparison of daily versus thrice weekly dialysis.[518] This difference between prescribed and delivered dose of dialysis can be attributed to difficulty in obtaining and maintaining blood flows secondary to hemodynamic instability, inability to achieve adequate ultrafiltration, shortened treatment times due to hemodynamic instability or studies, and catheter malfunction. Recently Venkataraman and colleagues have found similar differences between delivered and prescribed delivery in CRRT. In 110 patients the delivered CRRT dose was 68% of the prescribed dose.[524] As opposed to IHD, no treatments were interrupted due to hemodynamic instability, but the authors attributed the low delivery of therapy to recurrent clotting of the CRRT system. Because SLEDD is low flow and of shorter duration, prescribed therapy may be easier to obtain. Significant evidence is lacking, but one recent abstract evaluated delivered versus prescribed Kt/V in 9 patients and found no significant difference between prescribed and delivered dose.[525]

In addition, factors that effect therapy delivery are also machine related. Newer generation of CRRT machines now can increase rates of hemofiltration, dialysis, and hemodiafiltration. The amount and type of clearance delivered in any of these modalities is also influenced by blood flow rate. Increasing blood flow leads to decreased filtration fraction in post-dilution CVVH. Filtration Fraction is equivalent to replacement rate (Qr) divided by plasma flow rate (Qp):

Filtration Fraction=Qr/Qp

Increasing the blood flow in pre-dilution CVVH decreases the dilution fraction. Dilution fraction is equivalent to the flow of blood water (Qbw) divided by the sum of Qbw and replacement rate.

Dilution Fraction=Qbw/(Qbw+Qr)

Thus, increasing blood flow allows for increased clearances achievable in pre- or post-filter CVVH therapies.[526]

Recently an increasing body of evidence suggests that increasing the intensity and therapy dose of RRT in AKI influences outcome.[527] Dosing of intermittent dialysis has rendered some conflicting reports. An early study (1980's) of 34 ICU patients by Gillum and co-workers compared daily intensive therapy versus non-intensive therapy. In this small number of patients there was no difference in survival.[528]

Paganini and colleagues[529] reviewed the outcomes of 844 patients with AKI requiring RRT in the ICU between 1988 and 1994. He described no difference in survival in the most critically ill or those with low severity of illness scores based on dose of dialysis. However in the majority of the patients, those with an intermediate severity of illness score, there was a significant improvement in survival with higher delivery of intermittent dialysis (Kt/V>1.0).

Schiffl and colleagues[518] prospectively compared alternate day to daily dialysis in the ICU. They described a decrease in mortality from 46% to 28% in the patients who received daily dialysis when compared to the alternate day group. They also described a reduction in sepsis and increased rate of renal function recovery in the daily group.

Similarly studies in CRRT also suggest a benefit to increasing the delivery of therapy. Storck and co-workers[530] compared survival in a non-randomized prospective study of CAVH versus CVVH. Survival was significantly higher in the CVVH group and correlated with increasing volumes of ultrafiltration delivered (7.5 liters per day in the CAVH group and 15.5 liters per day in the CVVH group).

More recently, Ronco and colleagues[531] reported the results of a randomized study comparing the dose of therapy in CVVH. Using a lactate-based post filter CVVH system, the differences in outcome in 420 ICU AKI patients randomized to receive replacement fluid rates of 20 cc/kg/hr, 35 cc/kg/hr, and 45 cc/kg/hr. Delivered ultrafiltration rates were 31, 56, and 68 liters per day. Survival was found to be significantly higher in the 35 cc/kg/hr and 45 cc/kg/hr groups (57% and 58%) when compared to the group of patients receiving 20 cc/kg/hr (survival of 41%, p < 0.001). Although not primary endpoints, this study noted prolonged time to death with the higher therapies, BUN at time of initiation was significantly lower in survivors versus non-survivors, and there was a tendency toward improved survival in the septic patients with 45 cc/kg/hr (although the number of septic patients was too small to be significant).

Comparisons of modalities of RRT on AKI are few. Numerous historical control studies have suggested that CRRT may be superior to IHD. However, the data comparing CRRT to IHD is limited. Swartz did compare IHD to CRRT survival in 349 patients in the ICU at the University of Michigan. The study was not randomized and clearly comorbidities were higher in the group receiving CRRT. After adjusting for comorbidities, no difference in survival was seen.[532] Mehta and colleagues performed a prospective randomized multicenter trial comparing CRRT to IHD in 166 patients. Survival was actually higher in the group receiving IHD (58.5% survival in IHD versus 40.5% in CRRT). This study excluded hemodynamically unstable patients and unfortunately the randomization was flawed because the patients in the CRRT group had significantly higher APACHE III scores and a greater percentage of liver failure.[533] Kellum and colleagues have tried to reconcile the differences in an excellent meta-analysis of studies comparing CRRT to IHD. Overall there was no difference in therapy, but when adjusting for studies with relative risk of death and sufficient quality relative risk of death was substantially lower in CRRT. However, the authors concluded that given the quality of studies, there was insufficient evidence to make strong conclusions.[534]

Although studies in the late 1970's and early 1980's demonstrated efficacy of PD when compared to intermittent hemodialysis, these were not randomized nor do they represent current methods of extracorporeal therapy.[521] Phu and associates studied 70 adult patients with AKI from malaria (48 patients) or sepsis (22 patients) in Vietnam. The patients receiving PD had a mortality rate of 47% compared to 15% in the CVVH (p<0.005). Peritoneal dialysis was performed with 2-liter exchanges and only 30-minute dwell times—thus limiting clearance significantly. The patients receiving CVVH were prescribed clearances of only 25 liters per day. The study suggests that peritoneal dialysis is inferior to CVVH in the management of AKI associated with malaria or sepsis.[522]

Acute kidney injury in the ICU requiring RRT is associated with significant mortality and morbidity. Over time etiology of death has changed from that due to complications of uremia to that of complications of co-morbidities. RRT must be prescribed based on the individual patient needs, supporting the patient and allowing for time and other therapies to improve patient recovery. For the time being, local preferences and expertise in general determine the delivery of RRT. In the future, ability to achieve goals of toxin clearance, fluid stabilization, control of electrolyte, and acid base derangements may dictate type, frequency, and duration of extracorporeal therapies.


1. Hojs R, Ekart R, Sinkovic A, et al: Rhabdomyolysis and acute renal failure in intensive care unit.  Ren Fail  1999; 21:675-684.

2. de Mendonca A, Vincent JL, Suter PM, et al: Acute renal failure in the ICU: Risk factors and outcome evaluated by the SOFA score.  Intensive Care Med  2000; 26:915-921.

3. Thijs A, Thijs LG: Pathogenesis of renal failure in sepsis.  Kidney Int Suppl  1998; 66:S34-S37.

4. Valta P, Uusaro A, Nunes S, et al: Acute respiratory distress syndrome: Frequency, clinical course, and costs of care.  Crit Care Med  1999; 27:2367-2374.

5. Koreny M, Karth GD, Geppert A, et al: Prognosis of patients who develop acute renal failure during the first 24 hours of cardiogenic shock after myocardial infarction.  Am J Med  2002; 112:115-119.

6. Ring-Larsen H, Palazzo U: Renal failure in fulminant hepatic failure and terminal cirrhosis: A comparison between incidence, types, and prognosis.  Gut  1981; 22:585-591.

7. Greene KE, Peters JI: Pathophysiology of acute respiratory failure.  Clin Chest Med  1994; 15:1-12.

8. de los Santos R, Coalson JJ, Holcomb JR, et al: Hyperoxia exposure in mechanically ventilated primates with and without previous lung injury.  Exp Lung Res  1985; 9:255-275.

9. Bryan CL, Jenkinson SG: Oxygen toxicity.  Clin Chest Med  1988; 9:141-152.

10. Hayatdavoudi G, O'Neil JJ, Barry BE, et al: Pulmonary injury in rats following continuous exposure to 60% O2 for 7 days.  J Appl Physiol  1981; 51:1220-1231.

11. Kistler D, Caldwell P, Weibel E: Development of fine structural damage to alveolar and capillary lining cells in oxygen poisoned rat lungs.  Am Rev Respir Dis  1988; 137:A78.

12. Comroe JJ, Dripps R, Dumke P, et al: Oxygen toxicity-the effects of inhalation of high concentration of oxygen for 24 hours on normal men at sea level and at simulated altitude of 18,000 feet.  JAMA  1945; 128:710.

13. Pierce A: Oxygen toxicity.  Basic Respir Dis  1979; 1:1.

14. Ashbaugh DG, Bigelow DB, Petty TL, et al: Acute respiratory distress in adults.  Lancet  1967; 2:319-323.

15. Bernard GR, Artigas A, Brigham KL, et al: The American-European Consensus Conference on ARDS. Definitions, mechanisms, relevant outcomes, and clinical trial coordination.  Am J Respir Crit Care Med  1994; 149:818-824.

16. Brun-Buisson C, Minelli C, Bertolini G, et al: Epidemiology and outcome of acute lung injury in European intensive care units. Results from the ALIVE study.  Intensive Care Med  2004; 30:51-61.

17. Vincent JL, Sakr Y, Ranieri VM: Epidemiology and outcome of acute respiratory failure in intensive care unit patients.  Crit Care Med  2003; 31(Suppl 4):S296-S299.

18. Aberle DR, Wiener-Kronish JP, Webb WR, et al: Hydrostatic versus increased permeability pulmonary edema: Diagnosis based on radiographic criteria in critically ill patients.  Radiology  1988; 168:73-79.

19. Gattinoni L, Bombino M, Pelosi P, et al: Lung structure and function in different stages of severe adult respiratory distress syndrome.  JAMA  1994; 271:1772-1779.

20. Ware LB, Matthay MA: The acute respiratory distress syndrome.  N Engl J Med  2000; 342:1334-1349.

21. Fulkerson WJ, MacIntyre N, Stamler J, et al: Pathogenesis and treatment of the adult respiratory distress syndrome.  Arch Intern Med  1996; 156:29-38.

22. McHugh LG, Milberg JA, Whitcomb ME, et al: Recovery of function in survivors of the acute respiratory distress syndrome.  Am J Respir Crit Care Med  1994; 150:90-94.

23. Hudson LD, Milberg JA, Anardi D, et al: Clinical risks for development of the acute respiratory distress syndrome.  Am J Respir Crit Care Med  1995; 151:293-301.

24. Pepe PE, Potkin RT, Reus DH, et al: Clinical predictors of the adult respiratory distress syndrome.  Am J Surg  1982; 144:124-130.

25. Weinacker AB, Vaszar LT: Acute respiratory distress syndrome: Physiology and new management strategies.  Annu Rev Med  2001; 52:221-237.

26. Tate RM, Repine JE: Neutrophils and the adult respiratory distress syndrome.  Am Rev Respir Dis  1983; 128:552-559.

27. Henderson Jr WR: Eicosanoids and lung inflammation.  Am Rev Respir Dis  1987; 135:1176-1185.

28. Bernard GR, Reines HD, Halushka PV, et al: Prostacyclin and thromboxane A2 formation is increased in human sepsis syndrome. Effects of cyclooxygenase inhibition.  Am Rev Respir Dis  1991; 144:1095-1101.

29. The Acute Respiratory Distress Syndrome Network: Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and acute respiratory distress syndrome.  N Engl J Med  2000; 342:1301.

30. Dreyfuss D, Saumon G: Ventilator-induced lung injury: lessons from experimental studies.  Am J Respir Crit Care Med  1998; 157:294-323.

31. Parker JC, Hernandez LA, Longenecker GL, et al: Lung edema caused by high peak inspiratory pressures in dogs. Role of increased microvascular filtration pressure and permeability.  Am Rev Respir Dis  1990; 142:321-328.

32. Webb HH, Tierney DF: Experimental pulmonary edema due to intermittent positive pressure ventilation with high inflation pressures. Protection by positive end-expiratory pressure.  Am Rev Respir Dis  1974; 110:556-565.

33. Moran JL, Bersten AD, Solomon PJ: Meta-analysis of controlled trials of ventilator therapy in acute lung injury and acute respiratory distress syndrome: An alternative perspective.  Intensive Care Med  2005; 31:227-235.

34. Petrucci N, Iacovelli W: Ventilation with smaller tidal volumes: A quantitative systematic review of randomized controlled trials.  Anesth Analg  2004; 99:193-200.

35. Courtney SE, Durand DJ, Asselin JM, et al: Pro/con clinical debate: High-frequency oscillatory ventilation is better than conventional ventilation for premature infants.  Crit Care  2003; 7:423-426.

36. David M, Weiler N, Heinrichs W, et al: High-frequency oscillatory ventilation in adult acute respiratory distress syndrome.  Intensive Care Med  2003; 29:1656-1665.

37. Mehta S, Lapinsky SE, Hallett DC, et al: Prospective trial of high-frequency oscillation in adults with acute respiratory distress syndrome.  Crit Care Med  2001; 29:1360-1369.

38. Wunsch H, Mapstone J, Takala J: High-frequency ventilation versus conventional ventilation for the treatment of acute lung injury and acute respiratory distress syndrome: A systematic review and cochrane analysis.  Anesth Analg  2005; 100:1765-1772.

39. Brower RG, Lanken PN, MacIntyre N, et al: Higher versus lower positive end-expiratory pressures in patients with the acute respiratory distress syndrome.  N Engl J Med  2004; 351:327-336.

40. Kallet RH, Jasmer RM, Luce JM, et al: The treatment of acidosis in acute lung injury with tris-hydroxymethyl aminomethane (THAM).  Am J Respir Crit Care Med  2000; 161:1149-1153.

41. Brower RG, Ware LB, Berthiaume Y, et al: Treatment of ARDS.  Chest  2001; 120:1347-1367.

42. Simmons RS, Berdine GG, Seidenfeld JJ, et al: Fluid balance and the adult respiratory distress syndrome.  Am Rev Respir Dis  1987; 135:924-929.

43. Humphrey H, Hall J, Sznajder I, et al: Improved survival in ARDS patients associated with a reduction in pulmonary capillary wedge pressure.  Chest  1990; 97:1176-1180.

44. Eisenberg PR, Hansbrough JR, Anderson D, et al: A prospective study of lung water measurements during patient management in an intensive care unit.  Am Rev Respir Dis  1987; 136:662-668.

45. Shoemaker WC, Appel PL, Kram HB, et al: Prospective trial of supranormal values of survivors as therapeutic goals in high-risk surgical patients.  Chest  1988; 94:1176-1186.

46. Fleming A, Bishop M, Shoemaker W, et al: Prospective trial of supranormal values as goals of resuscitation in severe trauma.  Arch Surg  1992; 127:1175-1179.discussion 1179-1181

47. Boyd O, Grounds RM, Bennett ED: A randomized clinical trial of the effect of deliberate perioperative increase of oxygen delivery on mortality in high-risk surgical patients.  JAMA  1993; 270:22.2699-2707

48. Yu M, Levy MM, Smith P, et al: Effect of maximizing oxygen delivery on morbidity and mortality rates in critically ill patients: A prospective, randomized, controlled study.  Crit Care Med  1993; 21:830-838.

49. Tuchschmidt J, Fried J, Astiz M, et al: Elevation of cardiac output and oxygen delivery improves outcome in septic shock.  Chest  1992; 102:216-220.

50. Hayes MA, Timmins AC, Yau EH, et al: Elevation of systemic oxygen delivery in the treatment of critically ill patients.  N Engl J Med  1994; 330:1717-1722.

51. Brower R, Ware LB, Berthiaume Y: Treatment of acute respiratory distress syndrome.  Chest  2001; 120:1347.

52. Luce JM, Montgomery AB, Marks JD, et al: Ineffectiveness of high-dose methylprednisolone in preventing parenchymal lung injury and improving mortality in patients with septic shock.  Am Rev Respir Dis  1988; 138:62-68.

53. Bone RC, Fisher Jr CJ, Clemmer TP, et al: Early methylprednisolone treatment for septic syndrome and the adult respiratory distress syndrome.  Chest  1987; 92:1032-1036.

54. Bernard GR, Luce JM, Sprung CL, et al: High-dose corticosteroids in patients with the adult respiratory distress syndrome.  N Engl J Med  1987; 317:1565-1570.

55. Meduri GU, Headley AS, Golden E, et al: Effect of prolonged methylprednisolone therapy in unresolving acute respiratory distress syndrome: A randomized controlled trial.  JAMA  1998; 280:159-165.

56. Mure M, Martling CR, Lindahl SG: Dramatic effect on oxygenation in patients with severe acute lung insufficiency treated in the prone position.  Crit Care Med  1997; 25:1539-1544.

57. Fridrich P, Krafft P, Hochleuthner H, et al: The effects of long-term prone positioning in patients with trauma-induced adult respiratory distress syndrome.  Anesth Analg  1996; 83:1206-1211.

58. Gattinoni L, Tognoni G, Brazzi L, et al: Ventilation in the prone position. The Prone-Supine Study Collaborative Group.  Lancet  1997; 350:815.

59. Gattinoni L, Tognoni G, Pesenti A, et al: Effect of prone positioning on the survival of patients with acute respiratory failure.  N Engl J Med  2001; 345:568-573.

60. Abraham E, Anzueto A, Gutierrez G, et al: Double-blind randomised controlled trial of monoclonal antibody to human tumour necrosis factor in treatment of septic shock. NORASEPT II Study Group.  Lancet  1998; 351:929-933.

61. Loewen GM, Holm BA, Milanowski L, et al: Alveolar hyperoxic injury in rabbits receiving exogenous surfactant.  J Appl Physiol  1989; 66:1087-1092.

62. van Daal GJ, So KL, Gommers D, et al: Intratracheal surfactant administration restores gas exchange in experimental adult respiratory distress syndrome associated with viral pneumonia.  Anesth Analg  1991; 72:589-595.

63. Anzueto A, Baughman RP, Guntupalli KK, et al: Aerosolized surfactant in adults with sepsis-induced acute respiratory distress syndrome. Exosurf Acute Respi-ratory Distress Syndrome Sepsis Study Group.  N Engl J Med  1996; 334:1417-1421.

64. Spragg RG, Lewis JF, Wurst W, et al: Treatment of acute respiratory distress syndrome with recombinant surfactant protein C surfactant.  Am J Respir Crit Care Med  2003; 167:1562-1566.

65. Rossaint R, Falke KJ, Lopez F, et al: Inhaled nitric oxide for the adult respiratory distress syndrome.  N Engl J Med  1993; 328:399-405.

66. Taylor RW, Zimmerman JL, Dellinger RP, et al: Low-dose inhaled nitric oxide in patients with acute lung injury: A randomized controlled trial.  JAMA  2004; 291:1603-1609.

67. Sokol J, Jacobs SE, Bohn D: Inhaled nitric oxide for acute hypoxic respiratory failure in children and adults: A meta-analysis.  Anesth Analg  2003; 97:989-998.

68. Bone RC, Slotman G, Maunder R, et al: Randomized double-blind, multicenter study of prostaglandin E1 in patients with the adult respiratory distress syndrome. Prostaglandin E1 Study Group.  Chest  1989; 96:114-119.

69. The ARDS Network Authors : Ketoconazole for early treatment of acute lung injury and acute respiratory distress syndrome.  JAMA  2000; 283:1995-2002.

70. Mathay M: Severe sepsis: A new treatment with both anticoagulant and anti-inflammatory properties.  N Engl J Med  2001; 344:759-762.

71. Bernard GR, Wheeler AP, Arons MM, et al: A trial of antioxidants N-acetylcysteine and procysteine in ARDS. The Antioxidant in ARDS Study Group.  Chest  1997; 112:164-172.

72. Domenighetti G, Suter PM, Schaller MD, et al: Treatment with N-acetylcysteine during acute respiratory distress syndrome: A randomized, double-blind, placebo-controlled clinical study.  J Crit Care  1997; 12:177-182.

73. Jepsen S, Herlevsen P, Knudsen P, et al: Antioxidant treatment with N-acetylcysteine during adult respiratory distress syndrome: A prospective, randomized, placebo-controlled study.  Crit Care Med  1992; 20:918-923.

74. Hite RD, Morris PE: Acute respiratory distress syndrome: Pharmacological treatment options in development.  Drugs  2001; 61:897-907.

75. Morris AH, Wallace CJ, Menlove RL, et al: Randomized clinical trial of pressure-controlled inverse ratio ventilation and extracorporeal CO2 removal for adult respiratory distress syndrome.  Am J Respir Crit Care Med  1994; 149:295-305.

76. Chertow GM, Lazarus JM, Paganini EP, et al: Predictors of mortality and the provision of dialysis in patients with acute tubular necrosis. The Auriculin Anaritide Acute Renal Failure Study Group.  J Am Soc Nephrol  1998; 9:692-698.

77. Annat G, Viale JP, Bui Xuan B, et al: Effect of PEEP ventilation on renal function, plasma renin, aldosterone, neurophysins and urinary ADH, and prostaglandins.  Anesthesiology  1983; 58:136-141.

78. Jarnberg PO, de Villota ED, Eklund J, et al: Effects of positive end-expiratory pressure on renal function.  Acta Anaesthesiol Scand  1978; 22:508-514.

79. Hemmer M, Suter PM: Treatment of cardiac and renal effects of PEEP with dopamine in patients with acute respiratory failure.  Anesthesiology  1979; 50:399-403.

80. Kuiper JW, Groeneveld AB, Slutsky AS, et al: Mechanical ventilation and acute renal failure.  Crit Care Med  2005; 33:1408-1415.

81. Andrivet P, Adnot S, Sanker S, et al: Hormonal interactions and renal function during mechanical ventilation and ANF infusion in humans.  J Appl Physiol  1991; 70:287-292.

82. Chien C, Jakobe A, McVerry B, et al: Mechanical ventilation associated lung injury in canines causes renal histologic changes consistent with early acute tubular necrosis.  J Am Soc Nephrol  2003; 14:354A.

83. Imai Y, Parodo J, Kajikawa O, et al: Injurious mechanical ventilation and end-organ epithelial cell apoptosis and organ dysfunction in an experimental model of acute respiratory distress syndrome.  JAMA  2003; 289:2104-2112.

84. Hemmer M, Viquerat CE, Suter PM, et al: Urinary antidiuretic hormone excretion during mechanical ventilation and weaning in man.  Anesthesiology  1980; 52:395-400.

85. Marquez J, Guntupalli K, Sladen A, et al: Renal function and renin secretion during high frequency jet ventilation at varying levels of airway pressure.  Crit Care Med  1983; 11:930-932.

86. Pannu N, Mehta RL: Mechanical ventilation and renal function: an area for concern?.  Am J Kidney Dis  2002; 39:616-624.

87. Bellingan GJ: The pulmonary physician in critical care * 6: The pathogenesis of ALI/ARDS.  Thorax  2002; 57:540-546.

88. Martin TR: Lung cytokines and ARDS: Roger S. Mitchell Lecture.  Chest  1999; 116(Suppl 1):2S-8S.

89. Ranieri VM, Suter PM, Tortorella C, et al: Effect of mechanical ventilation on inflammatory mediators in patients with acute respiratory distress syndrome: A randomized controlled trial.  JAMA  1999; 282:54-61.

90. Goes N, Urmson J, Ramassar V, et al: Ischemic acute tubular necrosis induces an extensive local cytokine response. Evidence for induction of interferon-gamma, transforming growth factor-beta 1, granulocyte-macrophage colony-stimulating factor, interleukin-2, and interleukin-10.  Transplantation  1995; 59:565-572.

91. Chiao H, Kohda Y, McLeroy P, et al: Alpha-melanocyte-stimulating hormone protects against renal injury after ischemia in mice and rats.  J Clin Invest  1997; 99:1165-1172.

92. Heckbert SR, Vedder NB, Hoffman W, et al: Outcome after hemorrhagic shock in trauma patients.  J Trauma  1998; 45:545-549.

93. Hosomi H, Sagawa K: Effect of pentobarbital anesthesia on hypotension after 10% hemorrhage in the dog.  Am J Physiol  1979; 236:H607-H612.

94. Schwartz S, Frantz RA, Shoemaker WC: Sequential hemodynamic and oxygen transport responses in hypovolemia, anemia, and hypoxia.  Am J Physiol  1981; 241:H864-H871.

95. Koyama S, Sawano F, Matsuda Y, et al: Spatial and temporal differing control of sympathetic activities during hemorrhage.  Am J Physiol  1992; 262:R579-R585.

96. Wong DH, O'Connor D, Tremper KK, et al: Changes in cardiac output after acute blood loss and position change in man.  Crit Care Med  1989; 17:979-983.

97. Weil M, Rackow E: Hemmorhagic shock.   In: Schwartz G, ed. Principles and Practices of Emergency Medicine,  4th ed. Baltimore: Williams & Wilkins; 1999:36-45.

98. Forrest P: Vasopressin and shock.  Anaesth Intensive Care  2001; 29:463-472.

99. Schadt JC, Ludbrook J: Hemodynamic and neurohumoral responses to acute hypovolemia in conscious mammals.  Am J Physiol  1991; 260:2.Pt 2 H305-318

100. Bond RF, Johnson 3rd G: Vascular adrenergic interactions during hemorrhagic shock.  Fed Proc  1985; 44:281-289.

101. Edouard AR, Degremont AC, Duranteau J, et al: Heterogeneous regional vascular responses to simulated transient hypovolemia in man.  Intensive Care Med  1994; 20:6.414-420

102. Alexander RW, Dzau VJ: Vascular biology: The past 50 years.  Circulation  2000; 102(Suppl 4):112-116.

103. Szabo C, Billiar TR: Novel roles of nitric oxide in hemorrhagic shock.  Shock  1999; 12:1-9.

104. Pieber D, Horina G, Sandner-Kiesling A, et al: Pressor and mesenteric arterial hyporesponsiveness to angiotensin II is an early event in haemorrhagic hypotension in anaesthetised rats.  Cardiovasc Res  1999; 44:166-175.

105. Szabo C, Farago M, Horvath I, et al: Hemorrhagic hypotension impairs endothelium-dependent relaxations in the renal artery of the cat.  Circ Shock  1992; 36:238-241.

106. Szabo C, Csaki C, Benyo Z, et al: Role of the L-arginine-nitric oxide pathway in the changes in cerebrovascular reactivity following hemorrhagic hypotension and retransfusion.  Circ Shock  1992; 37:307-316.

107. Clementi E, Brown GC, Feelisch M, et al: Persistent inhibition of cell respiration by nitric oxide: Crucial role of S-nitrosylation of mitochondrial complex I and protective action of glutathione.  Proc Natl Acad Sci U S A  1998; 95:7631-7636.

108. Brealey D, Brand M, Hargreaves I, et al: Association between mitochondrial dysfunction and severity and outcome of septic shock.  Lancet  2002; 360:219-223.

109. Stepp DW, Kroll K, Feigl EO: K+ATP channels and adenosine are not necessary for coronary autoregulation.  Am J Physiol  1997; 273:H1299-H1308.

110. Marshall JM: Adenosine and muscle vasodilatation in acute systemic hypoxia.  Acta Physiol Scand  2000; 168:560-573.

111. Fan FC, Chen RY, Schuessler GB, et al: Effects of hematocrit variations on regional hemodynamics and oxygen transport in the dog.  Am J Physiol  1980; 238:H522-H545.

112. Maitra SR, Pan W, Geller ER, et al: Alterations in renal gluconeogenesis and blood flow during hemorrhagic shock.  Circ Shock  1993; 41:67-70.

113. Hansell P, Borgstrom P, Arfors KE: Pressure-related capillary leukostasis following ischemia-reperfusion and hemorrhagic shock.  Am J Physiol  1993; 265:H381-H388.

114. Botha AJ, Moore FA, Moore EE, et al: Early neutrophil sequestration after injury: A pathogenic mechanism for multiple organ failure.  J Trauma  1995; 39:411-417.

115. Maekawa K, Futami S, Nishida M, et al: Effects of trauma and sepsis on soluble L-selectin and cell surface expression of L-selectin and CD11b.  J Trauma  1998; 44:460-468.

116. Boyd AJ, Rubin BB, Walker PM, et al: A CD18 monoclonal antibody reduces multiple organ injury in a model of ruptured abdominal aortic aneurysm.  Am J Physiol  1999; 277:H172-H182.

117. Mazzoni MC, Intaglietta M, Cragoe Jr EJ, et al: Amiloride-sensitive Na+ pathways in capillary endothelial cell swelling during hemorrhagic shock.  J Appl Physiol  1992; 73:1467-1473.

118. Prist R, Rocha-e-Silva M, Scalabrini A, et al: A quantitative analysis of transcapillary refill in severe hemorrhagic hypotension in dogs.  Shock  1994; 1:188-195.

119. Tucker VL, Bravo E, Weber CJ, et al: Blood-to-tissue albumin transport in rats subjected to acute hemorrhage and resuscitation.  Shock  1995; 3:189-195.

120. Bock JC, Barker BC, Clinton AG, et al: Post-traumatic changes in, and effect of colloid osmotic pressure on the distribution of body water.  Ann Surg  1989; 210:395-403.discussion 403-395

121. Leach RM, Treacher DF: The pulmonary physician in critical care * 2: oxygen delivery and consumption in the critically ill.  Thorax  2002; 57:170-177.

122. Ronco JJ, Fenwick JC, Tweeddale MG, et al: Identification of the critical oxygen delivery for anaerobic metabolism in critically ill septic and nonseptic humans.  JAMA  1993; 270:1724-1730.

123. Curtis SE, Cain SM: Regional and systemic oxygen delivery/uptake relations and lactate flux in hyperdynamic, endotoxin-treated dogs.  Am Rev Respir Dis  1992; 145:348-354.

124. Shoemaker WC, Appel PL, Kram HB: Measurement of tissue perfusion by oxygen transport patterns in experimental shock and in high-risk surgical patients.  Intensive Care Med  1990; 16(Suppl 2):S135-S144.

125. Girbes A, Groenveld A: Circulatory optimization in patients with or at risk for shock.  Clin Intensive Care  2000; 11:77-81.

126. Vincent JL, Roman A, Debacker D, et al: Oxygen uptake/supply dependency.  Am Rev Respir Dis  1990; 142:2-7.

127. Robin ED: Of men and mitochondria: coping with hypoxic dysoxia. The 1980 J. Burns Amberson Lecture.  Am Rev Respir Dis  1980; 122:517-531.

128. Xu D, Lu Q, Deitch EA: Calcium and phospholipase A2 appear to be involved in the pathogenesis of hemorrhagic shock-induced mucosal injury and bacterial translocation.  Crit Care Med  1995; 23:125-131.

129. Arden WA, Yacko MA, Jay M, et al: Scintigraphic evaluation of bacterial translocation during hemorrhagic shock.  J Surg Res  1993; 54:102-106.

130. Tashkin DP, Goldstein PJ, Simmons DH: Hepatic lactate uptake during decreased liver perfusion and hyposemia.  Am J Physiol  1972; 223:968-974.

131. Chandel B, Shapiro MJ, Kurtz M, et al: MEGX (monoethylglycinexylidide): A novel in vivo test to measure early hepatic dysfunction after hypovolemic shock.  Shock  1995; 3:51-53.discussion 54-55

132. Fuchs S, Bogomolski-Yahalom V, Paltiel O, et al: Ischemic hepatitis: Clinical and laboratory observations of 34 patients.  J Clin Gastroenterol  1998; 26:183-186.

133. Miyazaki K, Hori S, Inoue S, et al: Characterization of energy metabolism and blood flow distribution in myocardial ischemia in hemorrhagic shock.  Am J Physiol  1997; 273:H600-H607.

134. Collard CD, Gelman S: Pathophysiology, clinical manifestations, and prevention of ischemia-reperfusion injury.  Anesthesiology  2001; 94:1133-1138.

135. Toyokuni S: Reactive oxygen species-induced molecular damage and its application in pathology.  Pathol Int  1999; 49:91-102.

136. Panes J, Perry M, Granger DN: Leukocyte-endothelial cell adhesion: Avenues for therapeutic intervention.  Br J Pharmacol  1999; 126:537-550.

137. Collard CD, Lekowski R, Jordan JE, et al: Complement activation following oxidative stress.  Mol Immunol  1999; 36:941-948.

138. Gunter TE, Gunter KK, Sheu SS, et al: Mitochondrial calcium transport: Physiological and pathological relevance.  Am J Physiol  1994; 267:C313-C339.

139. Murphy AN: Calcium mediated mitochondrial dysfunction and the protective effects of Bcl-2.  Ann N Y Acad Sci  1999; 893:19-32.

140. Carden DL, Granger DN: Pathophysiology of ischaemia-reperfusion injury.  J Pathol  2000; 190:255-266.

141. Yamazaki S, Fujibayashi Y, Rajagopalan RE, et al: Effects of staged versus sudden reperfusion after acute coronary occlusion in the dog.  J Am Coll Cardiol  1986; 7:564-572.

142. Weight SC, Bell PR, Nicholson ML: Renal ischaemia—reperfusion injury.  Br J Surg  1996; 83:162-170.

143. Delva E, Camus Y, Nordlinger B, et al: Vascular occlusions for liver resections. Operative management and tolerance to hepatic ischemia: 142 cases.  Ann Surg  1989; 209:211-218.

144. Maxwell SR, Lip GY: Reperfusion injury: A review of the pathophysiology, clinical manifestations and therapeutic options.  Int J Cardiol  1997; 58:95-117.

145. McGee S, Abernethy 3rd WB, Simel DL: The rational clinical examination. Is this patient hypovolemic?.  JAMA  1999; 281:1022-1029.

146. Nast-Kolb D, Waydhas C, Gippner-Steppert C, et al: Indicators of the posttraumatic inflammatory response correlate with organ failure in patients with multiple injuries.  J Trauma  1997; 42:446-454.discussion 454-445

147. Canizaro PC, Pessa ME: Management of massive hemorrhage associated with abdominal trauma.  Surg Clin North Am  1990; 70:621-634.

148. Voerman HJ, Groeneveld AB: Blood viscosity and circulatory shock.  Intensive Care Med  1989; 15:72-78.

149. Wu W, Rathore S, Wang Y, et al: Blood transufsions in elderly patients with acute myocardial infarctions.  N Engl J Med  2002; 345(17):1230-1236.

150. Sauaia A, Moore FA, Moore EE, et al: Early predictors of post injury multiple organ failure.  Arch Surg  1994; 129:39-45.

151. Hebert PC, Wells G, Blajchman MA, et al: A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion Requirements in Critical Care Investigators, Canadian Critical Care Trials Group.  N Engl J Med  1999; 340:409-417.

152. Gould SA, Sehgal LR, Sehgal HL, et al: Hypovolemic shock.  Crit Care Clin  1993; 9:239-259.

153. Rosser DM, Stidwill RP, Jacobson D, et al: Oxygen tension in the bladder epithelium rises in both high and low cardiac output endotoxemic sepsis.  J Appl Physiol  1995; 79:1878-1882.

154. Boekstegers P, Weidenhofer S, Pilz G, et al: Peripheral oxygen availability within skeletal muscle in sepsis and septic shock: Comparison to limited infection and cardiogenic shock.  Infection  1991; 19:317-323.

155. Doglio GR, Pusajo JF, Egurrola MA, et al: Gastric mucosal pH as a prognostic index of mortality in critically ill patients.  Crit Care Med  1991; 19:1037-1040.

156. Friedman G, Berlot G, Kahn RJ, et al: Combined measurements of blood lactate concentrations and gastric intramucosal pH in patients with severe sepsis.  Crit Care Med  1995; 23:1184-1193.

157. Marik PE: Gastric intramucosal pH. A better predictor of multiorgan dysfunction syndrome and death than oxygen-derived variables in patients with sepsis.  Chest  1993; 104:225-229.

158. Gutierrez G, Palizas F, Doglio G, et al: Gastric intramucosal pH as a therapeutic index of tissue oxygenation in critically ill patients.  Lancet  1992; 339:195-199.

159. Ivatury RR, Simon RJ, Islam S, et al: A prospective randomized study of end points of resuscitation after major trauma: Global oxygen transport indices versus organ-specific gastric mucosal pH.  J Am Coll Surg  1996; 183:145-154.

160. Pargger H, Hampl KF, Christen P, et al: Gastric intramucosal pH-guided therapy in patients after elective repair of infrarenal abdominal aneurysms: Is it beneficial?.  Intensive Care Med  1998; 24:769-776.

161. Gomersall CD, Joynt GM, Freebairn RC, et al: Resuscitation of critically ill patients based on the results of gastric tonometry: A prospective, randomized, controlled trial.  Crit Care Med  2000; 28:607-614.

162. Maciel AT, Creteur J, Vincent JL: Tissue capnometry: does the answer lie under the tongue?.  Intensive Care Med  2004; 30:2157-2165.

163. Shires GT, Braun FT, Canizaro PC, et al: Distributional changes in extracellular fluid during acute hemorrhagic shock.  Surg Forum  1960; 11:115-150.

164. Waikar SS, Chertow GM: Crystalloids versus colloids for resuscitation in shock.  Curr Opin Nephrol Hypertens  2000; 9:501-504.

165. Hauser CJ, Shoemaker WC, Turpin I, et al: Oxygen transport responses to colloids and crystalloids in critically ill surgical patients.  Surg Gynecol Obstet  1980; 150:811-816.

166. Holt ME, Ryall ME, Campbell AK: Albumin inhibits human polymorphonuclear leucocyte luminol-dependent chemiluminescence: Evidence for oxygen radical scavenging.  Br J Exp Pathol  1984; 65:231-241.

167. Demling RH: Effect of plasma and interstitial protein content on tissue edema formation.  Curr Stud Hematol Blood Transfus  1986; 53:36-52.

168. Imm A, Carlson RW: Fluid resuscitation in circulatory shock.  Crit Care Clin  1993; 9:313-333.

169. Thompson WL: Rational use of albumin and plasma substitutes.  Johns Hopkins Med J  1975; 136:220-225.

170. Kohler H, Zschiedrich H, Clasen R, et al: [The effects of 500 ml 10% hydroxyethyl starch 200/0.5 and 10% dextran 40 on blood volume, colloid osmotic pressure and renal function in human volunteers (author's transl)].  Anaesthesist  1982; 31:60-67.

171. Myers GA, Conhaim RL, Rosenfeld DJ, et al: Effects of pentafraction and hetastarch plasma expansion on lung and soft tissue transvascular fluid filtration.  Surgery  1995; 117:340-349.

172. Allison KP, Gosling P, Jones S, et al: Randomized trial of hydroxyethyl starch versus gelatine for trauma resuscitation.  J Trauma  1999; 47:1114-1121.

173. Wu JJ, Huang MS, Tang GJ, et al: Hemodynamic response of modified fluid gelatin compared with lactated ringer's solution for volume expansion in emergency resuscitation of hypovolemic shock patients: Preliminary report of a prospective, randomized trial.  World J Surg  2001; 25:598-602.

174. Lamke LO, Liljedahl SO: Plasma volume changes after infusion of various plasma expanders.  Resuscitation  1976; 5:93-102.

175. Ernest D, Belzberg AS, Dodek PM: Distribution of normal saline and 5% albumin infusions in septic patients.  Crit Care Med  1999; 27:46-50.

176. Rackow EC, Falk JL, Fein IA, et al: Fluid resuscitation in circulatory shock: a comparison of the cardiorespiratory effects of albumin, hetastarch, and saline solutions in patients with hypovolemic and septic shock.  Crit Care Med  1983; 11:839-850.

177. Dart RC, Sanders AB: Oxygen free radicals and myocardial reperfusion injury.  Ann Emerg Med  1988; 17:53-58.

178. Tait AR, Larson LO: Resuscitation fluids for the treatment of hemorrhagic shock in dogs: Effects on myocardial blood flow and oxygen transport.  Crit Care Med  1991; 19:1560-1565.

179. Tabuchi N, de Haan J, Gallandat Huet RC, et al: Gelatin use impairs platelet adhesion during cardiac surgery.  Thromb Haemost  1995; 74:1447-1451.

180. Messmer KF: The use of plasma substitutes with special attention to their side effects.  World J Surg  1987; 11:69-74.

181. Laxenaire MC, Charpentier C, Feldman L: [Anaphylactoid reactions to colloid plasma substitutes: incidence, risk factors, mechanisms. A French multicenter prospective study].  Ann Fr Anesth Reanim  1994; 13:301-310.

182. Dahn MS, Lucas CE, Ledgerwood AM, et al: Negative inotropic effect of albumin resuscitation for shock.  Surgery  1979; 86:235-241.

183. Lucas CE, Ledgerwood AM, Higgins RF: Impaired salt and water excretion after albumin resuscitation for hypovolemic shock.  Surgery  1979; 86:544-549.

184. Schortgen F, Lacherade JC, Bruneel F, et al: Effects of hydroxyethylstarch and gelatin on renal function in severe sepsis: A multicentre randomised study.  Lancet  2001; 357:911-916.

185. Wade CE, Kramer GC, Grady JJ, et al: Efficacy of hypertonic 7.5% saline and 6% dextran-70 in treating trauma: A meta-analysis of controlled clinical studies.  Surgery  1997; 122:609-616.

186. Choi PT, Yip G, Quinonez LG, et al: Crystalloids vs. colloids in fluid resuscitation: A systematic review.  Crit Care Med  1999; 27:200-210.

187. Wade CE, Grady JJ, Kramer GC, et al: Individual patient cohort analysis of the efficacy of hypertonic saline/dextran in patients with traumatic brain injury and hypotension.  J Trauma  1997; 42(Suppl 5):S60-S65.

188. Velanovich V: Crystalloid versus colloid fluid resuscitation: a meta-analysis of mortality.  Surgery  1989; 105:65-71.

189. Bisonni RS, Holtgrave DR, Lawler F, et al: Colloids versus crystalloids in fluid resuscitation: An analysis of randomized controlled trials.  J Fam Pract  1991; 32:387-390.

190. Schierhout G, Roberts I: Fluid resuscitation with colloid or crystalloid solutions in critically ill patients: A systematic review of randomised trials.  BMJ  1998; 316:960-964.

191. Cochrane Injuries Group Albumin Reviewers : Human albumin administration in critically ill patients: Systematic review of randomized controlled trials.  BMJ  1998; 317:235-240.

192. Finfer S, Bellomo R, Boyce N, et al: A comparison of albumin and saline for fluid resuscitation in the intensive care unit.  N Engl J Med  2004; 350:2247-2256.

193. Sloan EP, Koenigsberg M, Gens D, et al: Diaspirin cross-linked hemoglobin (DCLHb) in the treatment of severe traumatic hemorrhagic shock: a randomized controlled efficacy trial.  JAMA  1999; 282:1857-1864.

194. Morales D, Madigan J, Cullinane S, et al: Reversal by vasopressin of intractable hypotension in the late phase of hemorrhagic shock.  Circulation  1999; 100:226-229.

195. Voelckel WG, Lurie KG, Lindner KH, et al: Vasopressin improves survival after cardiac arrest in hypovolemic shock.  Anesth Analg  2000; 91:627-634.

196. Malay MB, Ashton Jr RC, Landry DW, et al: Low-dose vasopressin in the treatment of vasodilatory septic shock.  J Trauma  1999; 47:699-703.discussion 703-695

197. Jackson Jr WL, Shorr AF: Vasopressin and cardiac performance.  Chest  2002; 121:1723-1724.discussion 1724

198. Krismer AC, Wenzel V, Voelckel WG, et al: Employing vasopressin as an adjunct vasopressor in uncontrolled traumatic hemorrhagic shock. Three cases and a brief analysis of the literature.  Anaesthesist  2005; 54:220-224.

199. Sharma RM, Setlur R: Vasopressin in hemorrhagic shock.  Anesth Analg  2005; 101:833-834.table of contents

200. Poole-Wilson PA, Langer GA: Effect of pH on ionic exchange and function in rat and rabbit myocardium.  Am J Physiol  1975; 229:570-581.

201. Shapiro JI: Functional and metabolic responses of isolated hearts to acidosis: Effects of sodium bicarbonate and Carbicarb.  Am J Physiol  1990; 258:H1835-H1839.

202. Forsythe SM, Schmidt GA: Sodium bicarbonate for the treatment of lactic acidosis.  Chest  2000; 117:260-267.

203. Arieff AI, Leach W, Park R, et al: Systemic effects of NaHCO3 in experimental lactic acidosis in dogs.  Am J Physiol  1982; 242:F586-F591.

204. Graf H, Leach W, Arieff AI: Metabolic effects of sodium bicarbonate in hypoxic lactic acidosis in dogs.  Am J Physiol  1985; 249:F630-F635.

205. Tanaka M, Nishikawa T: Acute haemodynamic effects of sodium bicarbonate administration in respiratory and metabolic acidosis in anaesthetized dogs.  Anaesth Intensive Care  1997; 25:615-620.

206. Cooper DJ, Walley KR, Wiggs BR, et al: Bicarbonate does not improve hemodynamics in critically ill patients who have lactic acidosis. A prospective, controlled clinical study.  Ann Intern Med  1990; 112:492-498.

207. Mathieu D, Neviere R, Billard V, et al: Effects of bicarbonate therapy on hemodynamics and tissue oxygenation in patients with lactic acidosis: A prospective, controlled clinical study.  Crit Care Med  1991; 19:1352-1356.

208. Beech JS, Williams SC, Iles RA, et al: Haemodynamic and metabolic effects in diabetic ketoacidosis in rats of treatment with sodium bicarbonate or a mixture of sodium bicarbonate and sodium carbonate.  Diabetologia  1995; 38:889-898.

209. Nakashima K, Yamashita T, Kashiwagi S, et al: The effect of sodium bicarbonate on CBF and intracellular pH in man: Stable Xe-CT and 31P-MRS.  Acta Neurol Scand Suppl  1996; 166:96-98.

210. Bjerneroth G, Sammeli O, Li YC, et al: Effects of alkaline buffers on cytoplasmic pH in lymphocytes.  Crit Care Med  1994; 22:1550-1556.

211. Ritter JM, Doktor HS, Benjamin N: Paradoxical effect of bicarbonate on cytoplasmic pH.  Lancet  1990; 335:1243-1246.

212. Thompson CH, Syme PD, Williams EM, et al: Effect of bicarbonate administration on skeletal muscle intracellular pH in the rat: Implications for acute administration of bicarbonate in man.  Clin Sci (Lond)  1992; 82:559-564.

213. Bollaert PE, Robin-Lherbier B, Mallie JP, et al: Effects of sodium bicarbonate on striated muscle metabolism and intracellular pH during endotoxic shock.  Shock  1994; 1:196-200.

214. Bellingham AJ, Detter JC, Lenfant C: Regulatory mechanisms of hemoglobin oxygen affinity in acidosis and alkalosis.  J Clin Invest  1971; 50:700-706.

215. Vatner SF: Effects of hemorrhage on regional blood flow distribution in dogs and primates.  J Clin Invest  1974; 54:225-235.

216. Lieberthal W, McGarry AE, Sheils J, et al: Nitric oxide inhibition in rats improves blood pressure and renal function during hypovolemic shock.  Am J Physiol  1991; 261:F868-F872.

217. Bahrami S, Yao YM, Leichtfried G, et al: Significance of TNF in hemorrhage-related hemodynamic alterations, organ injury, and mortality in rats.  Am J Physiol  1997; 272:H2219-H2226.

218. Chaudry IH, Ayala A, Ertel W, et al: Hemorrhage and resuscitation: Immunological aspects.  Am J Physiol  1990; 259:R663-R678.

219. Spain DA, Fruchterman TM, Matheson PJ, et al: Complement activation mediates intestinal injury after resuscitation from hemorrhagic shock.  J Trauma  1999; 46:224-233.

220. Salvo I, de Cian W, Musicco M, et al: The Italian SEPSIS study: Preliminary results on the incidence and evolution of SIRS, sepsis, severe sepsis and septic shock.  Intensive Care Med  1995; 21(Suppl 2):S244-S249.

221. Balk RA: Severe sepsis and septic shock. Definitions, epidemiology, and clinical manifestations.  Crit Care Clin  2000; 16:179-192.

222. Centers for Disease Control : Increase in national hospital discharge survey rates for septicemia—United States 1979-1987.  JAMA  1990; 263:937-938.

223. Bone RC, Balk RA, Cerra FB, et al: Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine.  Chest  1992; 101:1644-1655.

224. Rangel-Frausto MS, Pittet D, Costigan M, et al: The natural history of the systemic inflammatory response syndrome (SIRS). A prospective study.  JAMA  1995; 273:117-123.

225. Kreger BE, Craven DE, Carling PC, et al: Gram-negative bacteremia. III. Reassessment of etiology, epidemiology and ecology in 612 patients.  Am J Med  1980; 68:332-343.

226. Marik PE, Varon J: Sepsis: State of the art.  Dis Mon  2001; 47:465-532.

227. Glauser MP: Pathophysiologic basis of sepsis: Considerations for future strategies of intervention.  Crit Care Med  2000; 28(Suppl S):4-8.

228. Heumann D, Glauser MP: Pathogenesis of sepsis.  Sci Am Sci Med  1994; 1:28-37.

229. Heumann D, Glauser MP, Calandra T: Molecular basis of host-pathogen interaction in septic shock.  Curr Opin Microbiol  1998; 1:49-55.

230. Natanson C, Hoffman WD, Suffredini AF, et al: Selected treatment strategies for septic shock based on proposed mechanisms of pathogenesis.  Ann Intern Med  1994; 120:771-783.

231. Freeman BD, Natanson C: Clinical trials in sepsis and septic shock.  Curr Opin Crit Care  1995; 1:349.

232. Zeni F, Freeman B, Natanson C: Anti-inflammatory therapies to treat sepsis and septic shock: A reassessment.  Crit Care Med  1997; 25:1095-1100.

233. Tavares-Murta BM, Zaparoli M, Ferreira RB, et al: Failure of neutrophil chemotactic function in septic patients.  Crit Care Med  2002; 30:1056-1061.

234. Manhart N, Oismuller C, Lassnig A, et al: Receptor and non-receptor mediated production of superoxide anion and hydrogen peroxide in neutrophils of intensive care patients.  Wiener Klinische Wochenschrift  1998; 110(22):796-801.

235. Natanson C, Eichenholz PW, Danner RL, et al: Endotoxin and tumor necrosis factor challenges in dogs simulate the cardiovascular profile of human septic shock.  J Exp Med  1989; 169:823-832.

236. Waage A, Espevik T: Interlekin-1 potentiates the lethal effects of tumor necrosis factor in mice.  J Exp Med  1988; 169:823-832.

237. Okusawa S, Gelfand JA, Ikejima T, et al: Interleukin 1 induces a shock-like state in rabbits. Synergism with tumor necrosis factor and the effect of cyclooxygenase inhibition.  J Clin Invest  1988; 81:1162-1172.

238. Ohlsson K, Bjork P, Bergenfeldt M, et al: Interleukin-1 receptor antagonist reduces mortality from endotoxin shock.  Nature  1990; 348:550-552.

239. Wakabayashi G, Gelfand JA, Burke JF, et al: A specific receptor antagonist for interleukin 1 prevents Escherichia coli-induced shock in rabbits.  FASEB J  1991; 5:338-343.

240. Beutler B, Milsark IW, Cerami AC: Passive immunization against cachectin/tumor necrosis factor protects mice from lethal effect of endotoxin.  Science  1985; 229:869-871.

241. Petrak RA, Balk RA, Bone RC: Prostaglandins, cyclo-oxygenase inhibitors, and thromboxane synthetase inhibitors in the pathogenesis of multiple systems organ failure.  Crit Care Clin  1989; 5:303-314.

242. Mathiak G, Szewczyk D, Abdullah F, et al: Platelet-activating factor (PAF) in experimental and clinical sepsis.  Shock  1997; 7:391-404.

243. Siebenlist U, Franzoso G, Brown K: Structure, regulation and function of NF-kappa B.  Annu Rev Cell Biol  1994; 10:405-455.

244. Karin M: The beginning of the end: IkappaB kinase (IKK) and NF-kappaB activation.  J Biol Chem  1999; 274:27339-27342.

245. Senftleben U, Karin M: The IKK/NF-kappa B pathway.  Crit Care Med  2002; 30(Suppl 1):S18-S26.

246. Lorente JA, Garcia-Frade LJ, Landin L, et al: Time course of hemostatic abnormalities in sepsis and its relation to outcome.  Chest  1993; 103:1536-1542.

247. Taylor Jr FB, Chang A, Esmon CT, et al: Protein C prevents the coagulopathic and lethal effects of Escherichia coli infusion in the baboon.  J Clin Invest  1987; 79:918-925.

248. Fourrier F, Chopin C, Goudemand J, et al: Septic shock, multiple organ failure, and disseminated intravascular coagulation. Compared patterns of antithrombin III, protein C, and protein S deficiencies.  Chest  1992; 101:816-823.

249. Ince C, Sinaasappel M: Microcirculatory oxygenation and shunting in sepsis and shock.  Crit Care Med  1999; 27:1369-1377.

250. Akira S: Toll-like receptor signaling.  J Biol Chem  2003; 278:38105-38108.

251. Poltorak A, He X, Smirnova I, et al: Defective LPS signaling in C3H/HeJ and C57BL/10 ScCr mice: Mutations in Tlr4 gene.  Science  1998; 282:2085-2088.

252. Cunningham PN, Wang Y, Guo R, et al: Role of Toll-like receptor 4 in endotoxin-induced acute renal failure.  J Immunol  2004; 172:2629-2635.

253. Kreger BE, Craven DE, McCabe WR: Gram-negative bacteremia. IV. Re-evaluation of clinical features and treatment in 612 patients.  Am J Med  1980; 68:344-355.

254. Kilbourn RG, Gross SS, Jubran A, et al: NG-methyl-L-arginine inhibits tumor necrosis factor-induced hypotension: Implications for the involvement of nitric oxide.  Proc Natl Acad Sci U S A  1990; 87:3629-3632.

255. Parrillo JE, Parker MM, Natanson C, et al: Septic shock in humans. Advances in the understanding of pathogenesis, cardiovascular dysfunction, and therapy.  Ann Intern Med  1990; 113:227-242.

256. Ellrodt AG, Riedinger MS, Kimchi A, et al: Left ventricular performance in septic shock: Reversible segmental and global abnormalities.  Am Heart J  1985; 110:402-409.

257. Monsalve F, Rucabado L, Salvador A, et al: Myocardial depression in septic shock caused by meningococcal infection.  Crit Care Med  1984; 12:1021-1023.

258. Kumar A, Haery C, Parrillo JE: Myocardial dysfunction in septic shock.  Crit Care Clin  2000; 16:251-287.

259. Paulus WJ, Vantrimpont PJ, Shah AM: Acute effects of nitric oxide on left ventricular relaxation and diastolic distensibility in humans. Assessment by bicoronary sodium nitroprusside infusion.  Circulation  1994; 89:2070-2078.

260. Doyle RL, Szaflarski N, Modin GW, et al: Identification of patients with acute lung injury. Predictors of mortality.  Am J Respir Crit Care Med  1995; 152:1818-1824.

261. Balk RA, Bone RC: The septic syndrome. Definition and clinical implications.  Crit Care Clin  1989; 5:1-8.

262. Pinsky MR, Vincent JL, Deviere J, et al: Serum cytokine levels in human septic shock. Relation to multiple-system organ failure and mortality.  Chest  1993; 103:565-575.

263. Meduri GU, Headley S, Kohler G, et al: Persistent elevation of inflammatory cytokines predicts a poor outcome in ARDS. Plasma IL-1 beta and IL-6 levels are consistent and efficient predictors of outcome over time.  Chest  1995; 107:1062-1073.

264. Soni A, Pepper GM, Wyrwinski PM, et al: Adrenal insufficiency occurring during septic shock: Incidence, outcome, and relationship to peripheral cytokine levels.  Am J Med  1995; 98:266-271.

265. Bouachour G, Tirot P, Gouello JP, et al: Adrenocortical function during septic shock.  Intensive Care Med  1995; 21:57-62.

266. Marik PE, Kiminyo K, Alexo S, et al: Occult adrenal insufficiency in critically ill patients: An underdiagnosed entity.  Crit Care Med  1999; 27:A141.

267. Zaloga GP: Sepsis-induced adrenal deficiency syndrome.  Crit Care Med  2001; 29:688-690.

268. Zaloga GP, Marik P: Hypothalamic-pituitary-adrenal insufficiency.  Crit Care Clin  2001; 17:25-41.

269. Levi M: Pathogenesis and treatment of disseminated intravascular coagulation in the septic patient.  J Crit Care  2001; 16:167-177.

270. Bolton CF: Sepsis and the systemic inflammatory response syndrome: Neuromuscular manifestations.  Crit Care Med  1996; 24:1408-1416.

271. Bleck TP, Smith MC, Pierre-Louis SJ, et al: Neurologic complications of critical medical illnesses.  Crit Care Med  1993; 21:98-103.

272. Papadopoulos MC, Davies DC, Moss RF, et al: Pathophysiology of septic encephalopathy: A review.  Crit Care Med  2000; 28:3019-3024.

273. Schor N: Acute renal failure and the sepsis syndrome.  Kidney Int  2002; 61:764-776.

274. Groeneveld AB, Tran DD, van der Meulen J, et al: Acute renal failure in the medical intensive care unit: predisposing, complicating factors and outcome.  Nephron  1991; 59:602-610.

275. Badr KF, Kelley VE, Rennke HG, et al: Roles for thromboxane A2 and leukotrienes in endotoxin-induced acute renal failure.  Kidney Int  1986; 30:474-480.

276. Kikeri D, Pennell JP, Hwang KH, et al: Endotoxemic acute renal failure in awake rats.  Am J Physiol  1986; 250:F1098-F1106.

277. Brenner M, Schaer GL, Mallory DL, et al: Detection of renal blood flow abnormalities in septic and critically ill patients using a newly designed indwelling thermodilution renal vein catheter.  Chest  1990; 98:170-179.

278. Ravikant T, Lucas CE: Renal blood flow distribution in septic hyperdynamic pigs.  J Surg Res  1977; 22:294-298.

279. Messmer UK, Briner VA, Pfeilschifter J: Tumor necrosis factor-alpha and lipopolysaccharide induce apoptotic cell death in bovine glomerular endothelial cells.  Kidney Int  1999; 55:2322-2337.

280. Klenzak J, Himmelfarb J: Sepsis and the kidney.  Crit Care Clin  2005; 21:211-222.

281. Mariano F, Guida G, Donati D, et al: Production of platelet-activating factor in patients with sepsis-associated acute renal failure.  Nephrol Dial Transplant  1999; 14:1150-1157.

282. Dos Santos OF, Boim MA, Barros EJ, et al: Role of platelet activating factor in gentamicin and cisplatin nephrotoxicity.  Kidney Int  1991; 40:742-747.

283. Kohan DE: Production of endothelin-1 by rat mesangial cells: Regulation by tumor necrosis factor.  J Lab Clin Med  1992; 119:477-484.

284. Marsden PA, Dorfman DM, Collins T, et al: Regulated expression of endothelin 1 in glomerular capillary endothelial cells.  Am J Physiol  1991; 261:F117-F125.

285. Katoh T, Chang H, Uchida S, et al: Direct effects of endothelin in the rat kidney.  Am J Physiol  1990; 258:F397-F402.

286. Yoshitomi K, Naruse M, Uchida S, et al: Endothelin inhibits luminal sodium channel in rabbit cortical collecting duct.  J Am Soc Nephrol  1991; 2:423.

287. Oishi R, Nonoguchi H, Tomita K, et al: Endothelin-1 inhibits AVP-stimulated osmotic water permeability in rat inner medullary collecting duct.  Am J Physiol  1991; 261:F951-F956.

288. Klahr S: Role of arachidonic acid metabolites in acute renal failure and sepsis.  Nephrol Dial Transplant  1994; 9(Suppl 4):52-56.

289. Bremm KD, Konig W, Spur B, et al: Generation of slow-reacting substance (leukotrienes) by endotoxin and lipid A from human polymorphonuclear granulocytes.  Immunology  1984; 53:299-305.

290. Balk RA: Pathogenesis and management of multiple organ dysfunction or failure in severe sepsis and septic shock.  Crit Care Clin  2000; 16:337-352.vii

291. Fink MP, Evans TW: Mechanisms of organ dysfunction in critical illness: Report from a Round Table Conference held in Brussels.  Intensive Care Med  2002; 28:369-375.

292. Rybak MJ, McGrath BJ: Combination antimicrobial therapy for bacterial infections. Guidelines for the clinician.  Drugs  1996; 52:390-405.

293. Wheeler AP, Bernard GR: Treating patients with severe sepsis.  N Engl J Med  1999; 340:207-214.

294. Ognibene FP: Hemodynamic support during sepsis.  Clin Chest Med  1996; 17:279-287.

295. Rivers E, Nguyen B, Havstad S, et al: Early goal-directed therapy in the treatment of severe sepsis and septic shock.  N Engl J Med  2001; 345:1368-1377.

296. Dellinger RP, Carlet JM, Masur H, et al: Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock.  Crit Care Med  2004; 32:858-873.

297. Levy B, Bollaert PE, Charpentier C, et al: Comparison of norepinephrine and dobutamine to epinephrine for hemodynamics, lactate metabolism, and gastric tonometric variables in septic shock: A prospective, randomized study.  Intensive Care Med  1997; 23:282-287.

298. Arnauld E, Czernichow P, Fumoux F, et al: The effects of hypotension and hypovolaemia on the liberation of vasopressin during haemorrhage in the unanaesthetized monkey (Macaca mulatta).  Pflugers Arch  1977; 371:193-200.

299. Landry DW, Levin HR, Gallant EM, et al: Vasopressin deficiency contributes to the vasodilation of septic shock.  Circulation  1997; 95:1122-1125.

300. Patel BM, Chittock DR, Russell JA, et al: Beneficial effects of short-term vasopressin infusion during severe septic shock.  Anesthesiology  2002; 96:576-582.

301. Schrier RW, Wang W: Acute renal failure and sepsis.  N Engl J Med  2004; 351:159-169.

302. Gattinoni L, Brazzi L, Pelosi P, et al: A trial of goal-oriented hemodynamic therapy in critically ill patients. SvO2 Collaborative Group.  N Engl J Med  1995; 333:1025-1032.

303. Bernard GR, Vincent JL, Laterre PF, et al: Efficacy and safety of recombinant human activated protein C for severe sepsis.  N Engl J Med  2001; 344(10):699-709.

304.   Xigris: Drotrocogin alfa (activated): PV 3420 AMP. 2001.

305. Abraham E, Laterre PF, Garg R, et al: Drotrecogin alfa (activated) for adults with severe sepsis and a low risk of death.  N Engl J Med  2005; 353:1332-1341.

306. Balk RA, Bedrossian C, McCormick L, et al: Prospective double blind placebo controlled trial of antithrombin III substitution in sepsis (Abstr).  Int Care Med  1995; 21(suppl 1):S17.

307. Baudo F, Caimi TM, de Cataldo F, et al: Antithrombin III (ATIII) replacement therapy in patients with sepsis and/or postsurgical complications: A controlled double-blind, randomized, multicenter study.  Intensive Care Med  1998; 24:336-342.

308. Eisele B, Lamy M, Thijs LG, et al: Antithrombin III in patients with severe sepsis. A randomized, placebo-controlled, double-blind multicenter trial plus a meta-analysis on all randomized, placebo-controlled, double-blind trials with antithrombin III in severe sepsis.  Intensive Care Med  1998; 24:663-672.

309. Fourrier F, Chopin C, Huart JJ, et al: Double-blind, placebo-controlled trial of antithrombin III concentrates in septic shock with disseminated intravascular coagulation.  Chest  1993; 104:882-888.

310. Creasey AA, Chang AC, Feigen L, et al: Tissue factor pathway inhibitor reduces mortality from Escherichia coli septic shock.  J Clin Invest  1993; 91:2850-2856.

311. Carr C, Bild GS, Chang AC, et al: Recombinant E. coli-derived tissue factor pathway inhibitor reduces coagulopathic and lethal effects in the baboon gram-negative model of septic shock.  Circ Shock  1994; 44:126-137.

312. Creasey AA, Reinhart K: Tissue factor pathway inhibitor activity in severe sepsis.  Crit Care Med  2001; 29(Suppl S):126-129.

313. Welty-Wolf KE, Carraway MS, Miller DL, et al: Coagulation blockade prevents sepsis-induced respiratory and renal failure in baboons.  Am J Respir Crit Care Med  2001; 164:1988-1996.

314. Bone RC, Fisher Jr CJ, Clemmer TP, et al: A controlled clinical trial of high-dose methylprednisolone in the treatment of severe sepsis and septic shock.  N Engl J Med  1987; 317:653-658.

315. Cronin L, Cook DJ, Carlet J, et al: Corticosteroid treatment for sepsis: A critical appraisal and meta-analysis of the literature.  Crit Care Med  1995; 23:1430-1439.

316. Bollaert PE, Charpentier C, Levy B, et al: Reversal of late septic shock with supraphysiologic doses of hydrocortisone.  Crit Care Med  1998; 26:4.645-650

317. Briegel J, Forst H, Haller M, et al: Stress doses of hydrocortisone reverse hyperdynamic septic shock: A prospective, randomized, double-blind, single-center study.  Crit Care Med  1999; 27:723-732.

318. Annane D, Sebille V, Charpentier C, et al: Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock.  JAMA  2002; 288:862-871.

319. Reinhart K, Menges T, Gardlund B, et al: Randomized, placebo-controlled trial of the anti-tumor necrosis factor antibody fragment afelimomab in hyperinflammatory response during severe sepsis: The RAMSES Study.  Crit Care Med  2001; 29:765-769.

320. Clark MA, Plank LD, Connolly AB, et al: Effect of a chimeric antibody to tumor necrosis factor-alpha on cytokine and physiologic responses in patients with severe sepsis—a randomized, clinical trial.  Crit Care Med  1998; 26:1650-1659.

321. Panacek E, Marcshall J, Fischkoff S, et al: Neutralization of tumor necrosis factor by a monoclonal antibody improves survival and reduces organ dysfunction in human sepsis: Results of the MONARCS trial (abstract).  Chest  2000; 118(suppl 4):883.

322. Bone RC, Balk RA, Fein AM, et al: A second large controlled clinical study of E5, a monoclonal antibody to endotoxin: Results of a prospective, multicenter, randomized, controlled trial. The E5 Sepsis Study Group.  Crit Care Med  1995; 23:994-1006.

323. Vincent JL, Spapen H, Bakker J, et al: Phase II multicenter clinical study of the platelet-activating factor receptor antagonist BB-882 in the treatment of sepsis.  Crit Care Med  2000; 28:638-642.

324. Fein AM, Bernard GR, Criner GJ, et al: Treatment of severe systemic inflammatory response syndrome and sepsis with a novel bradykinin antagonist, deltibant (CP-0127). Results of a randomized, double-blind, placebo-controlled trial. CP-0127 SIRS and Sepsis Study Group.  JAMA  1997; 277:482-487.

325. Bernard GR, Wheeler AP, Russell JA, et al: The effects of ibuprofen on the physiology and survival of patients with sepsis. The Ibuprofen in Sepsis Study Group.  N Engl J Med  1997; 336:912-918.

326. Vincent JL, Brase R, Santman F, et al: A multi-centre, double-blind, placebo-controlled study of liposomal prostaglandin E1 (TLC C-53) in patients with acute respiratory distress syndrome.  Intensive Care Med  2001; 27:1578-1583.

327. Opal SM, Fisher Jr CJ, Dhainaut JF, et al: Confirmatory interleukin-1 receptor antagonist trial in severe sepsis: A phase III, randomized, double-blind, placebo-controlled, multicenter trial. The Interleukin-1 Receptor Antagonist Sepsis Investigator Group.  Crit Care Med  1997; 25:1115-1124.

328. Grover R, Lopez A, Lorente J: Multicenter, randomized, placebo-controlled, double blind study of the nitric oxide synthetase inhibitor 546C88: Effect on survival in patients with septic shock.  Crit Care Med  1999; 27(suppl 1):A33.

329. Peake SL, Moran JL, Leppard PI: N-acetyl-L-cysteine depresses cardiac performance in patients with septic shock.  Crit Care Med  1996; 24:1302-1310.

330. Vincent JL, Sun Q, Dubois MJ: Clinical trials of immunomodulatory therapies in severe sepsis and septic shock.  Clin Infect Dis  2002; 34:1084-1093.

331. Werdan K: Supplemental immune globulins in sepsis.  Clin Chem Lab Med  1999; 37:341-349.

332. Caliezi C, Zeerleder S, Redondo M, et al: C1-inhibitor in patients with severe sepsis and septic shock: Beneficial effect on renal dysfunction.  Crit Care Med  2002; 30:1722-1728.

333. Elliott D, Wiles 3rd CE, Reynolds HN: Removal of cytokines in septic patients using continuous veno-venous hemodiafiltration.  Crit Care Med  1994; 22:718-719.discussion 719-721

334. Sander A, Armbruster W, Sander B, et al: The influence of continuous hemofiltration on cytokine elimination and the cardiovascular stability in the early phase of sepsis.  Contrib Nephrol  1995; 116:99-103.

335. van Bommel EF, Hesse CJ, Jutte NH, et al: Cytokine kinetics (TNF-alpha, IL-1 beta, IL-6) during continuous hemofiltration: a laboratory and clinical study.  Contrib Nephrol  1995; 116:62-75.

336. Barrera P, Janssen EM, Demacker PN, et al: Removal of interleukin-1 beta and tumor necrosis factor from human plasma by in vitro dialysis with polyacrylonitrile membranes.  Lymphokine Cytokine Res  1992; 11:99-104.

337. Bellomo R, Tipping P, Boyce N: Interleukin-6 and interleukin-8 extraction during continuous venovenous hemodiafiltration in septic acute renal failure.  Ren Fail  1995; 17:457-466.

338. Kellum JA, Johnson JP, Kramer D, et al: Diffusive vs. convective therapy: Effects on mediators of inflammation in patient with severe systemic inflammatory response syndrome.  Crit Care Med  1998; 26:1995-2000.

339. Wakabayashi Y, Kamijou Y, Soma K, et al: Removal of circulating cytokines by continuous haemofiltration in patients with systemic inflammatory response syndrome or multiple organ dysfunction syndrome.  Br J Surg  1996; 83:393-394.

340. Sander A, Armbruster W, Sander B, et al: Hemofiltration increases IL-6 clearance in early systemic inflammatory response syndrome but does not alter IL-6 and TNF alpha plasma concentrations.  Intensive Care Med  1997; 23:878-884.

341. Hoffmann JN, Hartl WH, Deppisch R, et al: Effect of hemofiltration on hemodynamics and systemic concentrations of anaphylatoxins and cytokines in human sepsis.  Intensive Care Med  1996; 22:1360-1367.

342. van Bommel EF, Hesse CJ, Jutte NH, et al: Impact of continuous hemofiltration on cytokines and cytokine inhibitors in oliguric patients suffering from systemic inflammatory response syndrome.  Ren Fail  1997; 19:443-454.

343. Heering P, Morgera S, Schmitz FJ, et al: Cytokine removal and cardiovascular hemodynamics in septic patients with continuous venovenous hemofiltration.  Intensive Care Med  1997; 23:288-296.

344. Hoffmann JN, Werdan K, Hartl WH, et al: Hemofiltrate from patients with severe sepsis and depressed left ventricular contractility contains cardiotoxic compounds.  Shock  1999; 12:174-180.

345. Cole L, Bellomo R, Hart G, et al: A phase II randomized, controlled trial of continuous hemofiltration in sepsis.  Crit Care Med  2002; 30:100-106.

346. Kamijo Y, Soma K, Sugimoto K, et al: The effect of a hemofilter during extracorporeal circulation on hemodynamics in patients with SIRS.  Intensive Care Med  2000; 26:1355-1359.

347. De Vriese AS, Colardyn FA, Philippe JJ, et al: Cytokine removal during continuous hemofiltration in septic patients.  J Am Soc Nephrol  1999; 10:846-853.

348. Grootendorst AF, van Bommel EF, van Leengoed LA, et al: High volume hemofiltration improves hemodynamics and survival of pigs exposed to gut ischemia and reperfusion.  Shock  1994; 2:72-78.

349. Lee PA, Matson JR, Pryor RW, et al: Continuous arteriovenous hemofiltration therapy for Staphylococcus aureus-induced septicemia in immature swine.  Crit Care Med  1993; 21:914-924.

350. Heidemann SM, Ofenstein JP, Sarnaik AP: Efficacy of continuous arteriovenous hemofiltration in endotoxic shock.  Circ Shock  1994; 44:183-187.

351. Freeman BD, Yatsiv I, Natanson C, et al: Continuous arteriovenous hemofiltration does not improve survival in a canine model of septic shock.  J Am Coll Surg  1995; 180:286-292.

352. Riegel W, Ziegenfuss T, Rose S, et al: Influence of venovenous hemofiltration on posttraumatic inflammation and hemodynamics.  Contrib Nephrol  1995; 116:56-61.

353. Braun N, Rosenfeld S, Giolai M, et al: Effect of continuous hemodiafiltration on IL-6, TNF-alpha, C3a, and TCC in patients with SIRS/septic shock using two different membranes.  Contrib Nephrol  1995; 116:89-98.

354. Bellomo R, Baldwin I, Cole L, et al: Preliminary experience with high-volume hemofiltration in human septic shock.  Kidney Int Suppl  1998; 66:S182-S185.

355. Garcia-Fernandez N, Lavilla FJ, Rocha E, et al: Haemostatic changes in systemic inflammatory response syndrome during continuous renal replacement therapy.  J Nephrol  2000; 13:282-289.

356. Honore PM, Jamez J, Wauthier M, et al: Prospective evaluation of short-term, high-volume isovolemic hemofiltration on the hemodynamic course and outcome in patients with intractable circulatory failure resulting from septic shock.  Crit Care Med  2000; 28:3581-3587.

357. Tetta C, Bellomo R, Brendolan A, et al: Use of adsorptive mechanisms in continuous renal replacement therapies in the critically ill.  Kidney Int Suppl  1999; 72:S15-S19.

358. Forrester JS, Diamond G, Chatterjee K, et al: Medical therapy of acute myocardial infarction by application of hemodynamic subsets (first of two parts).  N Engl J Med  1976; 295:1356-1362.

359. Hasdai D, Holmes Jr DR, Topol EJ, et al: Frequency and clinical outcome of cardiogenic shock during acute myocardial infarction among patients receiving reteplase or alteplase. Results from GUSTO-III. Global Use of Strategies to Open Occluded Coronary Arteries.  Eur Heart J  1999; 20:128-135.

360. Holmes Jr DR, Berger PB, Hochman JS, et al: Cardiogenic shock in patients with acute ischemic syndromes with and without ST-segment elevation.  Circulation  1999; 100:2067-2073.

361. Holmes Jr DR, Bates ER, Kleiman NS, et al: Contemporary reperfusion therapy for cardiogenic shock: the GUSTO-I trial experience. The GUSTO-I Investigators. Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries.  J Am Coll Cardiol  1995; 26:668-674.

362. Killip 3rd T, Kimball JT: Treatment of myocardial infarction in a coronary care unit. A two year experience with 250 patients.  Am J Cardiol  1967; 20:457-464.

363. Hochman JS, Buller CE, Sleeper LA, et al: Cardiogenic shock complicating acute myocardial infarction—etiologies, management and outcome: A report from the SHOCK Trial Registry. SHould we emergently revascularize Occluded Coronaries for cardiogenic shocK?.  J Am Coll Cardiol  2000; 36(Suppl A):1063-1070.

364. Babaev A, Frederick PD, Pasta DJ, et al: Trends in management and outcomes of patients with acute myocardial infarction complicated by cardiogenic shock.  JAMA  2005; 294:448-454.

365. Hollenberg SM: Cardiogenic shock.  Crit Care Clin  2001; 17:391-410.

366. Page DL, Caulfield JB, Kastor JA, et al: Myocardial changes associated with cardiogenic shock.  N Engl J Med  1971; 285:133-137.

367. Baliga RR: Apoptosis in myocardial ischemia, infarction, and altered myocardial states.  Cardiol Clin  2001; 19:91-112.

368. Menon V, White H, LeJemtel T, et al: The clinical profile of patients with suspected cardiogenic shock due to predominant left ventricular failure: A report from the SHOCK Trial Registry. SHould we emergently revascularize Occluded Coronaries in cardiogenic shocK?.  J Am Coll Cardiol  2000; 36(Suppl A):1071-1076.

369. Geppert A, Steiner A, Zorn G, et al: Multiple organ failure in patients with cardiogenic shock is associated with high plasma levels of interleukin-6.  Crit Care Med  2002; 30:1987-1994.

370. Kohsaka S, Menon V, Lowe AM, et al: Systemic inflammatory response syndrome after acute myocardial infarction complicated by cardiogenic shock.  Arch Intern Med  2005; 165:1643-1650.

371. Lim N, Dubois MJ, De Backer D, et al: Do all nonsurvivors of cardiogenic shock die with a low cardiac index?.  Chest  2003; 124:1885-1891.

372. Chiolero RL, Revelly JP, Leverve X, et al: Effects of cardiogenic shock on lactate and glucose metabolism after heart surgery.  Crit Care Med  2000; 28:3784-3791.

373. Hama N, Itoh H, Shirakami G, et al: Rapid ventricular induction of brain natriuretic peptide gene expression in experimental acute myocardial infarction.  Circulation  1995; 92:1558-1564.

374. Maisel A: B-type natriuretic peptide in the diagnosis and management of congestive heart failure.  Cardiol Clin  2001; 19:557-571.

375. Kazanegra R, Cheng V, Garcia A, et al: A rapid test for B-type natriuretic peptide correlates with falling wedge pressures in patients treated for decompensated heart failure: A pilot study.  J Card Fail  2001; 7:21-29.

376. Bersten AD, Holt AW, Vedig AE, et al: Treatment of severe cardiogenic pulmonary edema with continuous positive airway pressure delivered by face mask.  N Engl J Med  1991; 325:1825-1830.

377. Shah PK, Francis GS: Pump failure, shock and cardiac rupture in acute myocardial infarction.   In: Francis GS, Alpert JS, ed. Modern Coronary Care,  Boston: Little Brown; 1990:295-315.

378. Tisdale JE, Patel R, Webb CR, et al: Electrophysiologic and proarrhythmic effects of intravenous inotropic agents.  Prog Cardiovasc Dis  1995; 38:167-180.

379. Jolly S, Newton G, Horlick E, et al: Effect of vasopressin on hemodynamics in patients with refractory cardiogenic shock complicating acute myocardial infarction.  Am J Cardiol  2005; 96:1617-1620.

380. Baim DS: Effect of phosphodiesterase inhibition on myocardial oxygen consumption and coronary blood flow.  Am J Cardiol  1989; 63:23A-226A.

381. Mukae S, Yanagishita T, Geshi E, et al: The effects of dopamine, dobutamine and amrinone on mitochondrial function in cardiogenic shock.  Jpn Heart J  1997; 38:515-529.

382. Thackray S, Easthaugh J, Freemantle N, et al: The effectiveness and relative effectiveness of intravenous inotropic drugs acting through the adrenergic pathway in patients with heart failure—a meta-regression analysis.  Eur J Heart Fail  2002; 4:515-529.

383. Colucci WS, Elkayam U, Horton DP, et al: Intravenous nesiritide, a natriuretic peptide, in the treatment of decompensated congestive heart failure. Nesiritide Study Group.  N Engl J Med  2000; 343:246-253.

384. Marcus LS, Hart D, Packer M, et al: Hemodynamic and renal excretory effects of human brain natriuretic peptide infusion in patients with congestive heart failure. A double-blind, placebo-controlled, randomized crossover trial.  Circulation  1996; 94:3184-3189.

385. Sackner-Bernstein JD, Skopicki HA, Aaronson KD: Risk of worsening renal function with nesiritide in patients with acutely decompensated heart failure.  Circulation  2005; 111:1487-1491.

386. Wang DJ, Dowling TC, Meadows D, et al: Nesiritide does not improve renal function in patients with chronic heart failure and worsening serum creatinine.  Circulation  2004; 110:1620-1625.

387. Cotter G, Kaluski E, Blatt A, et al: L-NMMA (a nitric oxide synthase inhibitor) is effective in the treatment of cardiogenic shock.  Circulation  2000; 101:1358-1361.

388. Cotter G, Kaluski E, Milo O, et al: LINCS: L-NAME (a NO synthase inhibitor) in the treatment of refractory cardiogenic shock: A prospective randomized study.  Eur Heart J  2003; 24:1287-1295.

389. Mueller H, Ayres SM, Conklin EF, et al: The effects of intra-aortic counterpulsation on cardiac performance and metabolism in shock associated with acute myocardial infarction.  J Clin Invest  1971; 50:1885-1900.

390. Kern MJ, Aguirre F, Bach R, et al: Augmentation of coronary blood flow by intra-aortic balloon pumping in patients after coronary angioplasty.  Circulation  1993; 87:500-511.

391. Anderson RD, Ohman EM, Holmes Jr DR, et al: Use of intraaortic balloon counterpulsation in patients presenting with cardiogenic shock: observations from the GUSTO-I Study. Global Utilization of Streptokinase and TPA for Occluded Coronary Arteries.  J Am Coll Cardiol  1997; 30:708-715.

392. Sanborn TA, Sleeper LA, Bates ER, et al: Impact of thrombolysis, intra-aortic balloon pump counterpulsation, and their combination in cardiogenic shock complicating acute myocardial infarction: A report from the SHOCK Trial Registry. SHould we emergently revascularize Occluded Coronaries for cardiogenic shocK?.  J Am Coll Cardiol  2000; 36(Suppl A):1123-1129.

393. Barron HV, Every NR, Parsons LS, et al: The use of intra-aortic balloon counterpulsation in patients with cardiogenic shock complicating acute myocardial infarction: Data from the National Registry of Myocardial Infarction 2.  Am Heart J  2001; 141:933-939.

394. Prewitt RM, Gu S, Schick U, et al: Intraaortic balloon counterpulsation enhances coronary thrombolysis induced by intravenous administration of a thrombolytic agent.  J Am Coll Cardiol  1994; 23:794-798.

395. Ferguson 3rd JJ, Cohen M, Freedman Jr RJ, et al: The current practice of intra-aortic balloon counterpulsation: Results from the Benchmark Registry.  J Am Coll Cardiol  2001; 38:1456-1462.

396. Cohen M, Dawson MS, Kopistansky C, et al: Sex and other predictors of intra-aortic balloon counterpulsation-related complications: Prospective study of 1119 consecutive patients.  Am Heart J  2000; 139:282-287.

397. Crystal E, Borer A, Gilad J, et al: Incidence and clinical significance of bacteremia and sepsis among cardiac patients treated with intra-aortic balloon counterpulsation pump.  Am J Cardiol  2000; 86:1281-1284.A1289

398. Bengtson JR, Kaplan AJ, Pieper KS, et al: Prognosis in cardiogenic shock after acute myocardial infarction in the interventional era.  J Am Coll Cardiol  1992; 20:1482-1489.

399. Gruppo Italiano per lo studio della streptochinasi nell infarto miocardico : Effectiveness of intravenous thrombolytic therapy after acute myocardial infarction.  Drugs  1996; 52:589-605.

400. Hochman JS, Sleeper LA, Webb JG, et al: Early revascularization in acute myocardial infarction complicated by cardiogenic shock. SHOCK Investigators. Should We Emergently Revascularize Occluded Coronaries for Cardiogenic Shock.  N Engl J Med  1999; 341:625-634.

401. Hochman JS, Sleeper LA, White HD, et al: One-year survival following early revascularization for cardiogenic shock.  JAMA  2001; 285:190-192.

402. Dzavik V, Sleeper LA, Cocke TP, et al: Early revascularization is associated with improved survival in elderly patients with acute myocardial infarction complicated by cardiogenic shock: A report from the SHOCK Trial Registry.  Eur Heart J  2003; 24:828-837.

403. Dauerman HL, Ryan Jr TJ, Piper WD, et al: Outcomes of percutaneous coronary intervention among elderly patients in cardiogenic shock: A multicenter, decade-long experience.  J Invasive Cardiol  2003; 15:380-384.

404. Chan AW, Chew DP, Bhatt DL, et al: Long-term mortality benefit with the combination of stents and abciximab for cardiogenic shock complicating acute myocardial infarction.  Am J Cardiol  2002; 89:132-136.

405. Giri S, Mitchel J, Azar RR, et al: Results of primary percutaneous transluminal coronary angioplasty plus abciximab with or without stenting for acute myocardial infarction complicated by cardiogenic shock.  Am J Cardiol  2002; 89:126-131.

406. Young JB: Healing the heart with ventricular assist device therapy: Mechanisms of cardiac recovery.  Ann Thorac Surg  2001; 71(Suppl 3):210-219.

407. Sharma A, Hermann DD, Mehta RL: Clinical benefit and approach of ultrafiltration in acute heart failure.  Cardiology  2001; 96:144-154.

408. Packer M: Neurohormonal interactions and adaptations in congestive heart failure.  Circulation  1988; 77:721-730.

409. Kumar A, Brownjohn AM, Turney JH: Congestive heart failure—can the nephrologist help?.  J R Coll Physicians Lond  2000; 34:36-37.

410. Marenzi G, Grazi S, Giraldi F, et al: Interrelation of humoral factors, hemodynamics, and fluid and salt metabolism in congestive heart failure: Effects of extracorporeal ultrafiltration.  Am J Med  1993; 94:49-56.

411. Blake P, Hasegawa Y, Khosla MC, et al: Isolation of “myocardial depressant factor(s)” from the ultrafiltrate of heart failure patients with acute renal failure.  ASAIO J  1996; 42:M911-M915.

412. Marenzi G, Lauri G, Grazi M, et al: Circulatory response to fluid overload removal by extracorporeal ultrafiltration in refractory congestive heart failure.  J Am Coll Cardiol  2001; 38:963-968.

413. Canaud B, Leblanc M, Leray-Moragues H, et al: Slow continuous and daily ultrafiltration for refractory congestive heart failure.  Nephrol Dial Transplant  1998; 13(Suppl 4):51-55.

414. Coraim FI, Wolner E: Continuous hemofiltration for the failing heart.  New Horiz  1995; 3:725-731.

415. Jaski BE, Ha J, Denys BG, et al: Peripherally inserted veno-venous ultrafiltration for rapid treatment of volume overloaded patients.  J Card Fail  2003; 9:227-231.

416. Susini G, Zucchetti M, Bortone F, et al: Isolated ultrafiltration in cardiogenic pulmonary edema.  Crit Care Med  1990; 18:14-17.

417. Dormans TP, Huige RM, Gerlag PG: Chronic intermittent haemofiltration and haemodialysis in end stage chronic heart failure with oedema refractory to high dose frusemide.  Heart  1996; 75:349-351.

418. Liang K, Williams A, Karon B, et al: Initial clinical experience with a novel ultrafiltration device as a treatment strategy for diuretic resistant, refractory heart failure.  J Am Coll Cardiol  2005; 45:169A.

419. Bernuau J, Goudeau A, Poynard T, et al: Multivariate analysis of prognostic factors in fulminant hepatitis B.  Hepatology  1986; 6:648-651.

420. O'Grady JG, Schalm W, Williams R: Acute liver failure: Redefining the syndromes.  Lancet  1993; 342:273-275.

421. Williams R: Classification, etiology, and considerations of outcome in acute liver failure.  Semin Liver Dis  1996; 16:343-348.

422. Bernstein D, Tripodi J: Fulminant hepatic failure.  Crit Care Clin  1998; 14:181-197.

423. Sinclair S, Wakefield A, Levy G: Fulminant hepatitis.  Springer Semin Immunopathol  1990; 12:33-45.

424. Sergi C, Jundt K, Seipp S, et al: The distribution of HBV, HCV and HGV among livers with fulminant hepatic failure of different aetiology.  J Hepatol  1998; 29:860-871.

425. Williams R, Wendon J: Indications for orthotopic liver transplantation in fulminant liver failure.  Hepatology  1994; 20:S5-10S.

426. Feray C, Gigou M, Samuel D, et al: Hepatitis C virus RNA and hepatitis B virus DNA in serum and liver of patients with fulminant hepatitis.  Gastroenterology  1993; 104:549-555.

427. Worm HC, van der Poel WH, Brandstatter G: Hepatitis E: An overview.  Microbes Infect  2002; 4:657-666.

428. Saracco G, Macagno S, Rosina F, et al: Serologic markers with fulminant hepatitis in persons positive for hepatitis B surface antigen. A worldwide epidemiologic and clinical survey.  Ann Intern Med  1988; 108:380-383.

429. Carmichael Jr GP, Zahradnik JM, Moyer GH, et al: Adenovirus hepatitis in an immunosuppressed adult patient.  Am J Clin Pathol  1979; 71:352-355.

430. Sobue R, Miyazaki H, Okamoto M, et al: Fulminant hepatitis in primary human herpesvirus-6 infection.  N Engl J Med  1991; 324:1290.

431. Stanley TV: Acute liver failure following influenza B infection.  N Z Med J  1994; 107:382.

432. Wieland T, Faulstich H: Amatoxins, phallotoxins, phallolysin, and antamanide: The biologically active components of poisonous Amanita mushrooms.  CRC Crit Rev Biochem  1978; 5:185-260.

433. Sallie R, Katsiyiannakis L, Baldwin D, et al: Failure of simple biochemical indexes to reliably differentiate fulminant Wilson's disease from other causes of fulminant liver failure.  Hepatology  1992; 16:1206-1211.

434. Butterworth RF: The neurobiology of hepatic encephalopathy.  Semin Liver Dis  1996; 16:235-244.

435. Colquhoun SD, Lipkin C, Connelly CA: The pathophysiology, diagnosis, and management of acute hepatic encephalopathy.  Adv Intern Med  2001; 46:155-176.

436. Jones EA, Basile AS, Skolnick P: Hepatic encephalopathy, GABA-ergic neurotransmission and benzodiazepine receptor ligands.  Adv Exp Med Biol  1990; 272:121-134.

437. Ellis A, Wendon J: Circulatory, respiratory, cerebral, and renal derangements in acute liver failure: Pathophysiology and management.  Semin Liver Dis  1996; 16:379-388.

438. Basile AS, Harrison PM, Hughes RD, et al: Relationship between plasma benzodiazepine receptor ligand concentrations and severity of hepatic encephalopathy.  Hepatology  1994; 19:112-121.

439. Lidofsky SD, Bass NM, Prager MC, et al: Intracranial pressure monitoring and liver transplantation for fulminant hepatic failure.  Hepatology  1992; 16:1-7.

440. Ede RJ, Williams RW: Hepatic encephalopathy and cerebral edema.  Semin Liver Dis  1986; 6:107-118.

441. McClung HJ, Sloan HR, Powers P, et al: Early changes in the permeability of the blood-brain barrier produced by toxins associated with liver failure.  Pediatr Res  1990; 28:227-231.

442. Blei AT: Cerebral edema and intracranial hypertension in acute liver failure: Distinct aspects of the same problem.  Hepatology  1991; 13:376-379.

443. Blei AT, Olafsson S, Webster S, et al: Complications of intracranial pressure monitoring in fulminant hepatic failure.  Lancet  1993; 341:157-158.

444. Bass NM: Monitoring and treatment of intracranial hypertension.  Liver Transpl  2000; 6(Suppl 1):S21-S26.

445. Inagaki M, Shaw B, Schafer D, et al: Advantages of intracranial pressure monitoring in patients with fulminant hepatic failure (abstract).  Gastroenterology  1992; 102:A826.

446. Herrera JL: Management of acute liver failure.  Dig Dis  1998; 16:274-283.

447. Schafer DF, Shaw Jr BW: Fulminant hepatic failure and orthotopic liver transplantation.  Semin Liver Dis  1989; 9:189-194.

448. O'Brien CJ, Wise RJ, O'Grady JG, et al: Neurological sequelae in patients recovered from fulminant hepatic failure.  Gut  1987; 28:93-95.

449. Ganger DR: Liver failure in critical care medicine: Principles of diagnosis and management in the adult.   In: Parillo JE, Dillinger RP, ed. Critical Care Medicine,  Philadelphia: Mosby; 2001:1355-1369.

450. Pereira SP, Langley PG, Williams R: The management of abnormalities of hemostasis in acute liver failure.  Semin Liver Dis  1996; 16:403-414.

451. Gill RQ, Sterling RK: Acute liver failure.  J Clin Gastroenterol  2001; 33:191-198.

452. Rolando N, Harvey F, Brahm J, et al: Prospective study of bacterial infection in acute liver failure: An analysis of fifty patients.  Hepatology  1990; 11:49-53.

453. Almasio PL, Hughes RD, Williams R: Characterization of the molecular forms of fibronectin in fulminant hepatic failure.  Hepatology  1986; 6:1340-1345.

454. Bailey RJ, Woolf IL, Cullens H, et al: Metabolic inhibition of polymorphonuclear leucocytes in fulminant hepatic failure.  Lancet  1976; 1:1162-1163.

455. Rolando N, Gimson A, Wade J, et al: Prospective controlled trial of selective parenteral and enteral antimicrobial regimen in fulminant liver failure.  Hepatology  1993; 17:196-201.

456. Moore K: Renal failure in acute liver failure.  Eur J Gastroenterol Hepatol  1999; 11:967-975.

457. Davenport A, Will EJ, Davison AM: Early changes in intracranial pressure during haemofiltration treatment in patients with grade 4 hepatic encephalopathy and acute oliguric renal failure.  Nephrol Dial Transplant  1990; 5:192-198.

458. Davenport A, Will EJ, Davison AM: Continuous vs. intermittent forms of haemofiltration and/or dialysis in the management of acute renal failure in patients with defective cerebral autoregulation at risk of cerebral oedema.  Contrib Nephrol  1991; 93:225-233.

459. Meng HC, Lin HC, Huang CC, et al: Transjugular liver biopsy: Comparison with percutaneous liver biopsy.  J Gastroenterol Hepatol  1994; 9:457-461.

460. Harrison PM, Keays R, Bray GP, et al: Improved outcome of paracetamol-induced fulminant hepatic failure by late administration of acetylcysteine.  Lancet  1990; 335:1572-1573.

461. Santantonio T, Mazzola M, Pastore G: Lamivudine is safe and effective in fulminant hepatitis B.  J Hepatol  1999; 30:551.

462. Broussard CN, Aggarwal A, Lacey SR, et al: Mushroom poisoning—from diarrhea to liver transplantation.  Am J Gastroenterol  2001; 96:3195-3198.

463. Wellington K, Jarvis B: Silymarin: A review of its clinical properties in the management of hepatic disorders.  BioDrugs  2001; 15:465-489.

464. Lee WM: Management of acute liver failure.  Semin Liver Dis  1996; 16:369-378.

465. Harrison PM, Wendon JA, Gimson AE, et al: Improvement by acetylcysteine of hemodynamics and oxygen transport in fulminant hepatic failure.  N Engl J Med  1991; 324:1852-1857.

466. Walsh TS, Hopton P, Philips BJ, et al: The effect of N-acetylcysteine on oxygen transport and uptake in patients with fulminant hepatic failure.  Hepatology  1998; 27:1332-1340.

467. Wendon JA, Harrison PM, Keays R, et al: Effects of vasopressor agents and epoprostenol on systemic hemodynamics and oxygen transport in fulminant hepatic failure.  Hepatology  1992; 15:1067-1071.

468. Davenport A, Will EJ, Davison AM: Adverse effects on cerebral perfusion of prostacyclin administered directly into patients with fulminant hepatic failure and acute renal failure.  Nephron  1991; 59:449-454.

469. Munoz SJ: Difficult management problems in fulminant hepatic failure.  Semin Liver Dis  1993; 13:395-413.

470. Strauss E, Tramote R, Silva EP, et al: Double-blind randomized clinical trial comparing neomycin and placebo in the treatment of exogenous hepatic encephalopathy.  Hepatogastroenterology  1992; 39:542-545.

471. Sterling RK, Shiffman ML, Schubert ML: Flumazenil for hepatic coma: The elusive wake-up call?.  Gastroenterology  1994; 107:1204-1205.

472. Ede RJ, Gimson AE, Bihari D, et al: Controlled hyperventilation in the prevention of cerebral oedema in fulminant hepatic failure.  J Hepatol  1986; 2:43-51.

473. Wendon JA, Harrison PM, Keays R, et al: Cerebral blood flow and metabolism in fulminant liver failure.  Hepatology  1994; 19:1407-1413.

474. Jalan R, Olde Damink SW, Deutz NE, et al: Moderate hypothermia in patients with acute liver failure and uncontrolled intracranial hypertension.  Gastroenterology  2004; 127:1338-1346.

475. Woolf GM, Redeker AG: Treatment of fulminant hepatic failure with insulin and glucagon. A randomized, controlled trial.  Dig Dis Sci  1991; 36:92-96.

476. Rakela J, Mosley JW, Edwards VM, et al: A double-blinded, randomized trial of hydrocortisone in acute hepatic failure. The Acute Hepatic Failure Study Group.  Dig Dis Sci  1991; 36:1223-1228.

477. Brunner G: Benefits and dangers of plasma exchange in patients with fulminant hepatic failure (abstract).  Artif Organs  1988; 12:165.

478. Lepore MJ, Martel AJ: Plasmapheresis in hepatic coma.  Lancet  1967; 2:771-772.

479. Freeman JG, Matthewson K, Record CO: Plasmapheresis in acute liver failure.  Int J Artif Organs  1986; 9:433-438.

480. Strom SC, Fisher RA, Thompson MT, et al: Hepatocyte transplantation as a bridge to orthotopic liver transplantation in terminal liver failure.  Transplantation  1997; 63:559-569.

481. Miwa S, Hashikura Y, Mita A, et al: Living-related liver transplantation for patients with fulminant and subfulminant hepatic failure.  Hepatology  1999; 30:1521-1526.

482. Marcos A, Ham JM, Fisher RA, et al: Emergency adult to adult living donor liver transplantation for fulminant hepatic failure.  Transplantation  2000; 69:2202-2205.

483. Ash SR: Extracorporeal blood detoxification by sorbents in treatment of hepatic encephalopathy.  Adv Renal Replace Ther  2002; 9:3-18.

484. Ash SR, Kuczek T, Risler T, et al: Randomized clinical trial of liver dialysis in treatment of hepatic failure and hepatorenal failure.  Ther Apher  2000; 4:421.

485. Stange J, Mitzner SR, Risler T, et al: Molecular adsorbent recycling system (MARS): Clinical results of a new membrane-based blood purification system for bioartificial liver support.  Artif Organs  1999; 23:319-330.

486. Mitzner SR, Stange J, Klammt S, et al: Improvement of hepatorenal syndrome with extracorporeal albumin dialysis MARS: Results of a prospective, randomized, controlled clinical trial.  Liver Transpl  2000; 6:277-286.

487. Watanabe FD, Mullon CJ, Hewitt WR, et al: Clinical experience with a bioartificial liver in the treatment of severe liver failure. A phase I clinical trial.  Ann Surg  1997; 225:484-491.discussion 491-484

488. Demetriou AA, Brown Jr RS, Busuttil RW, et al: Prospective, randomized, multicenter, controlled trial of a bioartificial liver in treating acute liver failure.  Ann Surg  2004; 239:660-667.discussion 667-670

489. Sussman NL, Gislason GT, Kelly JH: Extracorporeal liver support. Application to fulminant hepatic failure.  J Clin Gastroenterol  1994; 18:320-324.

490. Ellis AJ, Hughes RD, Wendon JA, et al: Pilot-controlled trial of the extracorporeal liver assist device in acute liver failure.  Hepatology  1996; 24:1446-1451.

491. Clermont G, Acker CG, Angus DC, et al: Renal failure in the ICU: comparison of the impact of acute renal failure and end-stage renal disease on ICU outcomes.  Kidney Int  2002; 62:986-996.

492. Metnitz PG, Krenn CG, Steltzer H, et al: Effect of acute renal failure requiring renal replacement therapy on outcome in critically ill patients.  Crit Care Med  2002; 30:2051-2058.

493. Kellum JA: The Acute Dialysis Quality Initiative: Methodology.  Adv Ren Replace Ther  2002; 9:245-247.

494. Clark WR, Mueller BA, Alaka KJ, et al: A comparison of metabolic control by continuous and intermittent therapies in acute renal failure.  J Am Soc Nephrol  1994; 4:1413-1420.

495. Chima CS, Meyer L, Hummell AC, et al: Protein catabolic rate in patients with acute renal failure on continuous arteriovenous hemofiltration and total parenteral nutrition.  J Am Soc Nephrol  1993; 3:1516-1521.

496. Mehta RL: Indications for dialysis in the ICU: Renal replacement vs. renal support.  Blood Purif  2001; 19:227-232.

497. Macias WL, Mueller BA, Scarim SK, et al: Continuous venovenous hemofiltration: An alternative to continuous arteriovenous hemofiltration and hemodiafiltration in acute renal failure.  Am J Kidney Dis  1991; 18:451-458.

498. Manns M, Sigler MH, Teehan BP: Intradialytic renal haemodynamics—potential consequences for the management of the patient with acute renal failure.  Nephrol Dial Transplant  1997; 12:870-872.

499. McDonald BR, Mehta RL: Decreased mortality in patients with acute renal failure undergoing continuous arteriovenous hemodialysis.  Contrib Nephrol  1991; 93:51-56.

500. Gettings LG, Reynolds HN, Scalea T: Outcome in post-traumatic acute renal failure when continuous renal replacement therapy is applied early vs. late.  Intensive Care Med  1999; 25:805-813.

501. III. NKF-K/DOQI Clinical Practice Guidelines for Vascular Access: Update 2000.  Am J Kidney Dis  2001; 37(Suppl 1):S137-S181.

502. Trerotola SO, Shah H, Johnson M, et al: Randomized comparison of high-flow versus conventional hemodialysis catheters.  J Vasc Interv Radiol  1999; 10:1032-1038.

503. Trerotola SO, Kraus M, Shah H, et al: Randomized comparison of split tip versus step tip high-flow hemodialysis catheters.  Kidney Int  2002; 62:282-289.

504. Leblanc M, Bosc JY, Vaussenat F, et al: Effective blood flow and recirculation rates in internal jugular vein twin catheters: Measurement by ultrasound velocity dilution.  Am J Kidney Dis  1998; 31:87-92.

505. Kraus MA, McKusky M, Clark WR: Temporary hemodialysis catheter recirculation studies, a comparison of design and site selection.  Blood Purification  1995; 13:385-401.

506. Cimochowski GE, Worley E, Rutherford WE, et al: Superiority of the internal jugular over the subclavian access for temporary dialysis.  Nephron  1990; 54:154-161.

507. Pisoni RL, Young EW, Dykstra DM, et al: Vascular access use in Europe and the United States: Results from the DOPPS.  Kidney Int  2002; 61:305-316.

508. Twardowski ZJ, Van Stone JC, Jones ME, et al: Blood recirculation in intravenous catheters for hemodialysis.  J Am Soc Nephrol  1993; 3:1978-1981.

509. Oliver M, Edwards LJ, Treleaven DJ, et al: Randomized study of temporary hemodialysis catheters.  Int J Artif Organs  2002; 25:40-44.

510. Oliver MJ, Callery SM, Thorpe KE, et al: Risk of bacteremia from temporary hemodialysis catheters by site of insertion and duration of use: A prospective study.  Kidney Int  2000; 58:2543-2545.

511. Vanholder R, Van Biesen W, Lameire N: What is the renal replacement method of first choice for intensive care patients?.  J Am Soc Nephrol  2001; 12(Suppl 17):S40-S43.

512. Tominaga GT, Ingegno M, Ceraldi C, et al: Vascular complications of continuous arteriovenous hemofiltration in trauma patients.  J Trauma  1993; 35:285-288.discussion 288-289

513. Brunet S, Leblanc M, Geadah D, et al: Diffusive and convective solute clearances during continuous renal replacement therapy at various dialysate and ultrafiltration flow rates.  Am J Kidney Dis  1999; 34:486-492.

514. Uchino S, Fealy N, Baldwin I, et al: Pre-dilution vs. post-dilution during continuous veno-venous hemofiltration: Impact on filter life and azotemic control.  Nephron Clin Pract  2003; 94:c94-c98.

515. Palsson R, Niles JL: Regional citrate anticoagulation in continuous venovenous hemofiltration in critically ill patients with a high risk of bleeding.  Kidney Int  1999; 55:1991-1997.

516. Tolwani AJ, Campbell RC, Schenk MB, et al: Simplified citrate anticoagulation for continuous renal replacement therapy.  Kidney Int  2001; 60:370-374.

517. Evanson JA, Himmelfarb J, Wingard R, et al: Prescribed versus delivered dialysis in acute renal failure patients.  Am J Kidney Dis  1998; 32:731-738.

518. Schiffl H, Lang SM, Fischer R: Daily hemodialysis and the outcome of acute renal failure.  N Engl J Med  2002; 346:305-310.

519. Hakim RM, Wingard RL, Parker RA: Effect of the dialysis membrane in the treatment of patients with acute renal failure.  N Engl J Med  1994; 331:1338-1342.

520. Marshall MR, Golper TA, Shaver MJ, et al: Sustained low-efficiency dialysis for critically ill patients requiring renal replacement therapy.  Kidney Int  2001; 60:777-785.

521. Ash SR: Peritoneal dialysis in acute renal failure of adults: The safe, effective, and low-cost modality.  Contrib Nephrol  2001; 132:210-221.

522. Phu NH, Hien TT, Mai NT, et al: Hemofiltration and peritoneal dialysis in infection-associated acute renal failure in Vietnam.  N Engl J Med  2002; 347:895-902.

523. Liao Z, Huang Z, Cui X, et al: Dose capabilities of renal replacement therapies in acute renal failure (ARF).  J Am Soc Nephrol  2002; 13:237A.

524. Venkataraman R, Kellum JA, Palevsky P: Dosing patterns for continuous renal replacement therapy at a large academic medical center in the United States.  J Crit Care  2002; 17:246-250.

525. Marshall MR, Alam MG, Chatoth DK: Prescribed versus delivered dose of sustained low efficiency dialysis (SLED).  J Am Soc Nephrol  2000; 11:324A.

526. Clark WR, Turk JE, Kraus MA, et al: Dose determinants in continuous renal replacement therapy.  Artif Organs  2003; 27:815-820.

527. Clark WR, Kraus MA: Dialysis in acute renal failure: Is more better?.  Int J Artif Organs  2002; 25:1119-1122.

528. Gillum DM, Dixon BS, Yanover MJ, et al: The role of intensive dialysis in acute renal failure.  Clin Nephrol  1986; 25:249-255.

529. Paganini EP, Halstenberg WK, Goormastic M: Risk modeling in acute renal failure requiring dialysis: The introduction of a new model.  Clin Nephrol  1996; 46:206-211.

530. Storck M, Hartl WH, Zimmerer E, et al: Comparison of pump-driven and spontaneous continuous haemofiltration in postoperative acute renal failure.  Lancet  1991; 337:8739.452-455

531. Ronco C, Bellomo R, Homel P, et al: Effects of different doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: A prospective randomised trial.  Lancet  2000; 356:26-30.

532. Swartz RD, Messana JM, Orzol S, et al: Comparing continuous hemofiltration with hemodialysis in patients with severe acute renal failure.  Am J Kidney Dis  1999; 34:424-432.

533. Mehta RL, McDonald B, Gabbai FB, et al: A randomized clinical trial of continuous versus intermittent dialysis for acute renal failure.  Kidney Int  2001; 60:1154-1163.

534. Kellum JA, Angus DC, Johnson JP, et al: Continuous versus intermittent renal replacement therapy: A meta-analysis.  Intensive Care Med  2002; 28:29-37.