Chapter 21. Shock

• Shock is present when there is evidence of multisystem organ hypoperfusion; it often presents as decreased mean blood pressure.
• Initial resuscitation aims to establish adequate airway, breathing, and circulation. Rapid initial resusciation (usefully driven by protocol) is fundamental for improved outcome, since “time is tissue.”
• A working diagnosis or clinical hypothesis of the cause of inadequate circulation should always be made immediately, while treatment is initiated, based on clinical presentation, physical examination, and by observing the response to therapy.
• Drug and/or definitive therapy for specific causes of shock must be considered and implemented early (e.g., thrombolysis for myocardial infarction, hemostasis for hemorrhage, appropropriate antibiotics, and activated protein C for severe sepsis, etc.).
• The most common causes of shock are high cardiac output hypotension, or septic shock; reduced pump function of the heart, or cardiogenic shock; and reduced venous return despite normal pump function, or hypovolemic shock. Overlapping etiologies can confuse the diagnosis, as can a short list of other less common etiologies, which are often separated by echocardiography and pulmonary artery catheterization.
• Shock has a hemodynamic component, which is the focus of the initial resuscitation, but shock also has a systemic inflammatory component (ameliorated by rapid initial resuscitation) that leads to adverse sequelae including subsequent organ system dysfunction.

This chapter discusses shock with respect to the bedside approach: first with an early working diagnosis, then an approach to urgent resuscitation that confirms or changes the working diagnosis, followed by a pause to ponder the broader differential diagnosis of the types of shock and the pathophysiology of shock leading to potential adverse sequelae. Effective initial diagnosis and treatment at a rapid pace depend in large part on understanding cardiovascular pathophysiology.

Definition of Shock

Shock is present if evidence of multisystem organ hypoperfusion is apparent. Evidence of hypoperfusion includes tachycardia, tachypnea, low mean blood pressure, diaphoresis, poorly perfused skin and extremities, altered mental status, and decreased urine output. Hypotension has special importance because it commonly occurs during shock, because blood pressure is easily measured, and because extreme hypotension always results in shock. Important caveats are: (1) relatively low blood pressure is normal in some healthy individuals and (2) systolic blood pressure may be preserved in some patients in shock by excessive sympathetic tone. In the latter case, it is important to anticipate that sedation will unmask hypotension. Further, cuff blood pressure measurements may markedly underestimate central blood pressure in low flow states.1 The focus of initial resuscitation is reversing the hemodynamic component of shock, which leads to tissue hypoxia and lactic acidosis. However, all types of shock are also associated with a systemic inflammatory component that is a key contributor to subsequent multisystem organ failure and death. The development of the systemic inflammatory component is minimized by rapid and adequate (usefully driven by protocol) initial resuscitation.2

A Questioning Approach to the Initial Clinical Examination

Mean blood pressure is the product of cardiac output and systemic vascular resistance (SVR). Accordingly, hypotension may be caused by reduced cardiac output or reduced SVR. Therefore, initial examination of the hypotensive patient seeks to answer the question: Is cardiac output reduced? High cardiac output hypotension is most often signaled by a high pulse pressure, a low diastolic pressure, warm extremities with good nail bed return, fever (or hypothermia), leukocytosis (or leukopenia), and other evidence of infection; these clinical findings strongly suggest a working diagnosis of septic shock (Table 21-1), the initial treatment for which is thoughtful antibiosis combined with rapid expansion of the vascular volume and subsequent vasopressors, inotropes, and blood transfusion as necessary to achieve an adequate mean arterial pressure (MAP) and central venous oxygen saturation (ScvO2; see below).

In contrast, low cardiac output is signaled by a low pulse pressure, mottled cyanotic skin, and cool extremities with poor nail bed return. In this case, clinical examination turns to a second question: Is the heart too full? A heart that is too full in a hypotensive patient is signaled by elevated jugular venous pressure, peripheral edema, crepitations on lung auscultation, a laterally displaced precordial apical impulse with extra heart sounds (S3, S4), chest pain, ischemic changes on the electrocardiogram, and a chest radiograph showing a large heart with dilated upper lobe vessels and pulmonary edema.3 These findings suggest cardiogenic shock, most often caused by ischemic heart disease, and are generally absent when low cardiac output results from hypovolemia (see Table 21-1). Clinical examination then shows manifestations of blood loss (hematemesis, tarry stools, abdominal distention, reduced hematocrit, or trauma) or manifestations of dehydration (reduced tissue turgor, vomiting or diarrhea, or negative fluid balance). This distinction between cardiogenic and hypovolemic shock allows initial therapy to focus on vasoactive drugs and on volume infusions, respectively.

Whenever the clinical formulation is not obvious after answering the first two questions, ask a third: What does not fit? Most often, the answer is that the hypotension is due to overlap of two or more of these common etiologies of shock: septic shock complicated by myocardial dysfunction or hypovolemia, cardiogenic shock complicated by hypovolemia or sepsis, and hypovolemic shock masking sepsis or ischemic heart disease. At this time, more data are frequently needed, especially aided by echocardiography and pulmonary artery catheterization. Interpretations of the data and response to initial therapy frequently confirm the multiple etiologies or lead to a broader differential diagnosis of the etiologies of shock (see below). A short list of common etiologies other than septic, cardiogenic, or hypovolemic shock can be grouped as they present (see Table 21-1): high cardiac output hypotension that does not appear to be caused by sepsis, high right atrial pressure hypotension not caused by left ventricular ischemia, and poorly responsive hypovolemic shock.

Primary Survey

Early institution of aggressive resuscitation improves a patient's chances of survival.2,4 To improve efficiency at the necessarily rapid tempo, a systematic approach to initial evaluation and resuscitation is useful as it is during cardiac emergencies (advanced cardiac life support [ACLS]) and trauma (advanced trauma life support [ATLS]). In analogy to these systematic “ABC” approaches, a primary survey includes establishing an airway (airway), choosing a ventilator mode and small tidal volumes that minimize ventilator-induced lung injury (breathing), rapid (usefully protocol driven) resuscitation of the inadequate circulation (circulation), and drugs/definitive therapy consisting of early consideration and implementation of definitive therapy for specific causes of shock (e.g., thrombolysis or angioplasty for myocardial infarction; hemostasis for hemorrhage; appropriate antibiotics, surgical drainage of abscess, activated protein C, and steroids for severe sepsis; etc.).

Airway

Almost all patients in shock have one or more indications for airway intubation and mechanical ventilation (Table 21-2), which should be instituted early. Significant hypoxemia (based on blood-gas analysis, pulse oximetry, or high clinical suspicion) is one indication for airway intubation because external masks and other devices do not reliably deliver an adequate fraction of inspired O2 (FiO2). Initially, a high FiO2 (100%) is used until blood-gas analysis or reliable pulse oximetry allows titration of the FiO2 down toward less toxic concentrations.

Ventilatory failure is another indication for airway intubation and mechanical ventilation. Elevated and rising partial pressure of CO2 in arterial blood reliably establishes the diagnosis of ventilatory failure but is often a late finding. In particular, young, previously healthy patients are able to defend partial pressure of CO2 (PCO2) and pH up until a precipitous respiratory arrest. Therefore, clinical signs of respiratory muscle fatigue or subtle evidence of inadequate ventilation are more important early indicators.5 Evidence of respiratory muscle fatigue, including labored breathing precluding more than rudimentary verbal responses, tachypnea greater than 40/minute or an inappropriately low and decreasing respiratory rate, abdominal paradoxical respiratory motion, accessory muscle use, and other manifestations of ventilatory failure such as inadequately compensated acidemia should lead to early elective intubation and ventilation of the patient in shock (see Chap. 31).

Obtundation, due to shock or other causes, resulting in inadequate airway protection is an important indication for intubation. In shock, airway intubation and mechanical ventilation should precede other complicated procedures, such as central venous catheterization, or complicated tests that require transportation of the patient because these procedures and tests restrict the medical staff's ability to continuously assess the airway and ensure adequacy of ventilation.

Breathing

Initially, mechanical ventilation with sedation and, if necessary, paralysis are instituted to remove work of breathing as a confounding factor from the initial resuscitation and diagnostic pathway and to redistribute limited blood flow to vital organs.6 The change from spontaneous breathing (negative intrathoracic pressure ventilation) to mechanical ventilation (positive intrathoracic pressure ventilation) leads to reduced venous return so that additional volume resuscitation must be anticipated when hypovolemia contributes to shock. Application of positive end-expiratory pressure (increases intrathoracic pressure) and administration of sedative or narcotic drugs (increases venous capacitance) similarly should be expected to reduce venous return and highlight the importance of aggressive volume resuscitation at the time of intubation and institution of mechanical ventilation in hypovolemic patients. Conversely, when hypovolemia is not a problem (e.g., cardiogenic shock), application of positive intrathoracic pressure may improve cardiac output and blood pressure.

A relatively small tidal volume (6 to 8 mL/kg) should be selected to minimize hypotension due to high intrathoracic pressures and, more importantly, to reduce ventilator-induced lung injury.7 When arterial hypoxemia due to acute lung injury or frank adult respiratory distress syndrome (ARDS) complicates shock, adherence to tidal volumes of 6 mL/kg ideal body weight significantly decreases mortality rate and number of days on a ventilator in the intensive care unit.8

Circulation

Just as low tidal volumes limit ongoing lung inflammation and injury, rapid resuscitation of the circulation limits ongoing generation of a systemic inflammatory response and multiple organ injury. Hence, rapid protocol-driven approaches with defined end points improve shock outcome.4,9 For all types of shock, “time is tissue.” Thus, for cardiogenic shock secondary to acute myocardial infarction, the early goal is immediate thrombolysis, angioplasty, or surgical revascularization.3 For hypovolemic shock due to hemorrhage, the early goal is immediate hemostasis and rapid volume resuscitation. The early goals of volume resuscitation in hypovolemic or septic shock are incorporated in the Early Goal-Directed Therapy algorithm (Fig. 21-1), which was initially designed to aid resuscitation of septic shock.2 This requires immediate monitoring (even before formal admission to the intensive care unit) of MAP (goal >65 mm Hg), central venous pressure (CVP; goal 8 to 12 mm Hg), and ScvO2 (goal >70%).

Volume

Aggressive volume resuscitation up to the point of a heart that is too full is the first step in resuscitation of the circulation. The rate and composition of volume expanders must be adjusted in accord with the working diagnosis. The Early Goal-Directed Therapy algorithm for resuscitation of septic shock calls for 500 mL saline every 30 minutes, but this is much too slow in hypovolemic patients in whom 1 L every 10 minutes, or faster, is initially required. During volume resuscitation, infusions must be sufficient to test the clinical hypothesis that the patient is hypovolemic by effecting a short-term end point indicating benefit (increased blood pressure and pulse pressure and decreased heart rate) or complication (increased jugular venous pressure and pulmonary edema). Absence of either response indicates an inadequate challenge, so the volume administered in the next interval must be greater than the previous one. In obvious hemorrhagic shock, immediate hemostasis is essential10; blood must be obtained early, warmed and filtered; blood substitutes are administered in large amounts (crystalloid or colloid solutions) until blood pressure increases or the heart becomes too full. At the other extreme, a working diagnosis of cardiogenic shock without obvious fluid overload requires a smaller volume challenge (250 mL NaCl in 20 minutes). In each case, and in all other types of shock, the next volume challenge depends on the response to the first; it should proceed soon after the first so that the physician does not miss the diagnostic clues evident only to the examining critical care team at the bedside during this urgent resuscitation (Table 21-3).

Role of Red Blood Cell Transfusion during Initial Resuscitation

Transfusion of red blood cells is a component of the initial volume resuscitation of shock when severe or ongoing blood loss contributes to shock. In addition, when anemia contributes to inadequate oxygen delivery so that mixed venous oxygen saturation or its surrogate ScvO2 <70% despite an adequate CVP (8 to 12 mm Hg) and an adequate MAP (>65 mm Hg), then transfusion of red blood cells to hematocrit greater than 30% is a reasonable component of Early Goal-Directed Therapy and improves outcome.2 After initial resuscitation and stabilization, transfusion of red blood cells to maintain a hemoglobin above 90 g/L is no more beneficial than maintaining a hemoglobin level above 70 g/L and only incurs additional transfusion risk.11

Is There a Role for Delayed Resuscitation of Hypovolemia?

During brisk ongoing hemorrhage, massive crystalloid or colloid resuscitation increases blood pressure and the rate of hemorrhage, so patient outcome may be worse.12 This does not mean that resuscitation is detrimental; rather, control of active bleeding is more important than volume replacement. Preventing blood loss conserves warm, oxygen-carrying, protein-containing, biocompatible intravascular volume and is therefore far superior to replacing ongoing losses with fluids deficient in one or more of these areas. I and my colleagues believe that delayed or inadequate volume resuscitation, after blood loss is controlled, is a significant error that will have a detrimental effect on patient outcome.10

This approach of aggressive volume resuscitation avoids persistent hypovolemia as the cause for prolonged hypotension at the risk of causing pulmonary edema or aggravating the pumping dysfunction of the ischemic myocardium. We are not cavalier about fluid overload; just as soon as early resuscitation goals are met and as the heart becomes evaluated as too full, we shift goals to aim for the lowest circulating volume that provides adequate perfusion and O2 delivery, but we emphasize that rational diagnosis and early resuscitation from shock require an adequate circulating volume before vasoactive drugs can be effective. Of course, the discerning intensivist is aware that the initial evaluation of some hypotensive patients demonstrates a heart that is too full, so vasoactive therapy starts immediately.

Vasopressors

Volume-resuscitated septic shock stands out as a challenging hemodynamic problem. Here the inflammatory component of shock is prominent; after vigorous volume resuscitation, right atrial pressure and cardiac output may be high, but MAP may be distressingly low and evidence of organ system hypoperfusion may persist (oliguria, impaired mentation, and lactic acidosis). Here there is a role for pressor agents. Whereas adequate cardiac output is more important than blood pressure (because adequate tissue oxygen delivery is the underlying issue), effective distribution of flow by the vascular system depends on an adequate pressure head. At pressures below an autoregulatory limit, normal flow distribution mechanisms are lost, so significant vital organ system hypoperfusion may persist in the face of elevated cardiac output due to maldistribution of blood flow. In this case, where inadequate pressure is the dominant problem, an assessment of organ system perfusion is made (urine output, mentation, and lactic acid concentration), and then a vasopressor agent such as noradrenaline is initiated to raise MAP.13 The increased afterload will decrease cardiac output, so this intervention as single therapy is appropriate only when cardiac output is high. If cardiac output and oxygen delivery are inadequate, then combination of vasopressor therapy with inotropic agents should be considered (see below).

Too often, the blanket coverage of “pressor agents” is used early in all types of shock. Because some positive inotropic drugs can cause venoconstriction to increase cardiac output by endogenous volume shifts (dopamine and epinephrine), there is a rationale (like that of raising the patient's legs or applying military antishock trousers) for the common practice of starting these agents in some hypotensive patients while the volume resuscitation proceeds. However, this strategy often confounds the determination of an adequate circulating volume and the diagnosis of the etiology of shock. Thus, vasopressor use as part of Early Goal-Directed Therapy must be reassessed during ongoing volume resuscitation. Even when the numerical MAP and CVP goals have been attained, additional rapid volume challenge generally should be used to test for further clinical improvement (increased MAP, decreased heart rate, increased ScvO2, and increased urine output) and to determine whether this will allow titration of vasopressor use down or off.

Assessment of organ system perfusion is the most important component of vasopressor therapy; increase in blood pressure by itself is insufficient and can distract from careful reassessment of adequacy of oxygen delivery. If urine output increases, mentation improves, and lactate levels decrease, then vasopressor therapy has achieved its goals, and there is no need to increase MAP further even if the MAP that reverses these signs of hypoperfusion is 55 mm Hg. If the measures of organ system perfusion are not improved by vasopressor therapy, then arbitrarily driving MAP much above 70 mm Hg is rarely useful in septic shock and usually detrimental because cardiac output will decrease further and excessive vasoconstrictor tone will impair blood flow distribution. If evidence of hypoperfusion persists, then inadequate volume resuscitation, cardiac output, hemoglobin, and oxygen saturation are more likely problems.

Inotropes

If evidence of inadequate perfusion persists (assessed by clinical indicators, by direct measurement of cardiac output, or by ScvO2, etc.) despite adequate circulating volume (Early Goal-Directed Therapy goal: CVP 8 to 12 mm Hg) and vasopressors (Early Goal-Directed Therapy goal: >65 mm Hg), then inotropic agents are indicated.13 Inotropes are not effective when volume resuscitation is incomplete. In this case, the arterial vasodilating properties of inotropes such as dobutamine and milrinone result in a drop in arterial pressure that is not countered by an increase in cardiac output because venous return is still limited by the inadequate volume resuscitation. The corollary is, if initiation of inotropes results in a significant drop in blood pressure, then adequate volume resuscitation is not complete.

The objective of inotrope use is to increase cardiac output to achieve adequate oxygen delivery to all tissues. Of the many clinical and laboratory indicators that should be measured, mixed venous O2 saturation (when a pulmonary artery catheter is placed) or ScvO2 are useful surrogate measures of adequacy of O2 delivery.14 Rapidly achieving a goal ScvO2 greater than 70% results in a substantial improvement in survival and limits the systemic inflammatory response so that the subsequent need for further volume, red blood cell transfusion, vasopressor use, and mechanical ventilation is reduced.2

Accordingly, we encourage early aggressive volume resuscitation, judicious use of vasopressors to aid in rapidly achieving an adequate MAP in septic shock, titration of a positive inotropic agent (dobutamine 2 to 20 μg/kg per minute) to enhance myocardial contractility, and blood transfusion to hematocrit greater than 30, if necessary, to achieve ScvO2 greater than 70%. Resuscitation of the circulation is often usefully driven by protocol. When numerical resuscitation goals are met, additional rapid volume infusions can be used to determine whether volume resuscitation is complete and to help titrate vasopressor agents down and off. After a rapid and adequate resuscitation, timely reduction in vasopressor and inotrope use should be anticipated and is the next goal.

Drugs/Definitive Therapy

During the rapid initial assessment of the patient in shock and initial resuscitation aimed at supporting respiration and circulation, it is important to consider early institution of other definitive therapy for specific causes of shock and early input from consultant experts. When myocardial infarction is the cause of cardiogenic shock, immediate thrombolysis or angioplasty is considered, using intraaortic balloon pump support and coronary artery bypass surgery when necessary3 (see Chap. 25). During resuscitation of hypovolemic shock, continuous and early application of techniques to anticipate, prevent, or correct hypothermia prevents secondary coagulopathy, coma, and nonresponsiveness to volume and pharmacologic resuscitation. Hemostasis is the immediate goal for hemorrhage10 because it removes the cause of hypovolemic shock and lessens the need for further volume expanders, none of which are as effective as keeping the patient's own blood intravascular. Emergent radiologic and surgical consultation and intervention may be required. Similarly, when septic shock is secondary to a perforated viscus, an undrained abscess, or rapid spread of infection in devitalized tissue or in tissue planes (gas gangrene, necrotizing fasciitis, etc.), then immediate surgical intervention is fundamental to survival. Early institution of appropriate antibiotics has a profound effect on patient survival from septic shock.15 Activated protein C decreases mortality for severe sepsis and is most effective when given early.16 Low-dose steroid therapy of septic shock in patients who are unresponsive to corticotropin (ACTH) similarly improves survival in patients having septic shock.17 Thus, an ACTH stimulation test should be conducted and low-dose steroid therapy should be initiated immediately thereafter. Steroid therapy can be discontinued as soon as the results of the ACTH stimulation test demonstrate adrenal responsiveness.17

Catheters and Frequent Measurements during Initial Resuscitation

After an airway is established and breathing ensured, correction of the circulatory abnormality always requires good intravenous access. For large-volume administration, two peripheral intravenous catheters of gauge 16 or larger or large-bore central venous access is required. Early Goal-Directed Therapy mandates immediate placement of a central venous catheter. Electrocardiographic monitoring is easily accomplished and usefully measures heart rate and rhythm for early detection and, hence, rational treatment of tachyarrhythmias or bradyarrhythmias aggravating the low-flow state.

The urinary bladder should be catheterized to measure urine output and to facilitate urine sampling. A nasogastric or orogastric tube to decompress the stomach and later to deliver medication and nutrition is generally required in the intubated patient. Measuring arterial pressure with a peripheral arterial or femoral arterial catheter is useful because, in the patient in shock with low cardiac output or low blood pressure, cuff pressures may be inaccurate.1 Arterial blood-gas and other blood samples are then readily obtained.

Early echocardiography is a useful adjunct to ScvO2 to distinguish poor ventricular pumping function from hypovolemia; a good study can exclude or confirm tamponade, right heart failure, pulmonary hypertension possibly due to pulmonary embolism, or significant valve dysfunction, all of which influence therapy, and can replace more invasive pulmonary artery catheterization.

Effective use and interpretation of ScvO2 and echocardiography often obviate pulmonary artery catheterization when the clinical hypothesis of cardiogenic, hypovolemic, or septic shock is confirmed and corrected by initial therapeutic intervention. There is no role for pulmonary artery catheterization for routine monitoring or management of uncomplicated shock states.18,19 Use of pulmonary artery catheterization should be restricted to circumstances in which the derived measurements will alter management or direct therapeutic interventions. For example, when confirmed septic shock does not respond fully, or as expected, to volume resuscitation and vasopressor use (e.g., persistent narrow pulse pressure, low ScvO2, and elevated lactate), then myocardial dysfunction contributes. When echocardiography excludes other causes (see above), then subsequent pulmonary artery catheterization guides the balance between vasopressor (to maintain an adequate MAP for blood flow distribution to vital organs) and inotrope use (to achieve an adequate cardiac output and oxygen delivery to maintain vital organ function).

Tempo

One of the most important contributions the intensivist can make to the care of a shock patient is to establish an appropriately rapid management tempo. Rapid initial resuscitation improves survival (“time is tissue”). In many instances, resuscitation driven by protocol can achieve adequate resuscitation faster. Effective protocol-driven resuscitation requires significant preliminary discussion, buy in, and training of emergency room physicians, house staff, nurses, respiratory therapists, and others.

The mirror image of urgent implementation is rapid liberation of the resuscitated patient from excessive therapy. It is not uncommon for the patient with hypovolemic or septic shock to stabilize hemodynamically on positive-pressure ventilation with high circulating volume and several vasoactive drugs infusing at a high rate. Too often, hours or days of “weaning” pass, when a trial of spontaneous breathing, diuresis, and sequential reduction of the drug dose by half each 10 minutes can return the patient to a much less treated, stable state within the hour. Of course, this rapid discontinuation may be limited by intercurrent hemodynamic or other instability, but defining each limit and justifying ongoing or new therapy is the essence of titrated care in this postresuscitation period.

We use cardiac function curves and venous return relations in the following discussion to compare and contrast cardiovascular mechanisms responsible for cardiogenic shock (Fig. 21-2), hypovolemic shock (Fig. 21-3), and septic shock (Fig. 21-4). The goal is to link pathophysiology of the circulation to the broader differential diagnosis of the types of shock presented in Table 21-4 to facilitate the accurate etiologic diagnosis and management based on additional hemodynamic measures as required by the response to urgent resuscitation.

Shock caused by decreased pump function of the heart is commonly cardiogenic shock resulting from left ventricular ischemia. For high right atrial pressure hypotension not obviously caused by left ventricular ischemia, greatly elevated pulmonary arterial pressure (most commonly pulmonary embolism), right ventricular ischemia, and dysfunction of heart valves must be excluded. Of course, right atrial pressure may be increased by abnormal pressures surrounding the heart in the absence of ventricular dysfunction. Because of this clinical presentation, I have included cardiac tamponade in the category of high right atrial pressure hypotension. Tamponade and pulmonary embolism might have been classified as decreased venous return due to obstruction of the circulation (see Table 21-1). Decreased venous return despite normal pump function is most commonly due to hemorrhagic or dehydration hypovolemia, but I emphasize other mechanisms including decreased venous tone caused by drugs, neurologic injury, and adrenal insufficiency, particularly for nonresponsive hypovolemia. Septic shock is the most common cause of high cardiac output hypotension resulting from abnormal arterial tone and blood flow distribution, although other causes such as severe liver failure, severe pancreatitis, trauma with tissue damage eliciting a significant inflammatory response, anaphylactic shock, thyroid storm, Paget disease, and other peripheral shunts share this mechanism. Defining shock as anaerobic metabolism of multiple organ systems, often signaled by lactic acidosis, allows classification of the shock state associated with metabolic poisons such as carbon monoxide that results in histotoxic hypoxia caused by inadequate uptake of oxygen by the mitochondria (see Table 21-4). Other shock states, in particular septic shock, may also involve an inability of tissues to extract delivered oxygen, but the lactic acidosis of sepsis is not necessarily caused by anaerobic metabolism.

Decreased Pump Function—Cardiogenic Shock

Pump function is measured as the output of a pump for a given input. The diagnosis of decreased pump function as the cause of shock is made by finding evidence of inappropriately low output (cardiac output) despite normal or high input (right atrial pressure). Cardiac output is the most important “output” of the heart and is clinically assessed in the same way that perfusion was assessed during the urgent initial examination. Better estimates are later obtained by thermodilution measurement using a pulmonary artery catheter, nuclear medicine scans, and Doppler echocardiographic techniques. Right atrial pressure or CVP is the most easily measured “input” of the whole heart and is initially assessed by examination of jugular veins. After catheter insertion, CVP can be measured accurately. Other outputs, such as stroke work or left ventricular ejection fraction, and other inputs, such as left ventricular end-diastolic pressure (LVEDP) or volume (LVEDV), are useful to determine the specific cause of decreased pump function. Left and right ventricular dysfunction can be caused by decreased systolic contractility, increased diastolic stiffness, greatly increased afterload (including obstruction), valvular dysfunction, or abnormal heart rate and rhythm.

Causes of Left Ventricular Failure

Acute or acute-on-chronic left ventricular failure resulting in shock is the classic example of cardiogenic shock and is identified as a subset of decreased pump function by evidence of a low cardiac output in relation to high left ventricular filling pressures. Clinical findings of low cardiac output and increased left ventricular filling pressures include, in addition to assessment of perfusion, pulmonary crackles in dependent lung regions, a laterally displaced and diffuse precordial apical impulse, elevated jugular veins, and presence of a third heart sound.20 These findings are not always present or unambiguous. Therefore, echocardiography or pulmonary artery catheterization is helpful and often essential in establishing the diagnosis and titrating therapy. Cardiogenic shock then is usually associated with a cardiac index lower than 2.2 L/m2 per minute when the pulmonary artery occlusion pressure has been raised above 18 mm Hg.21

Systolic Dysfunction

Figure 21-2 illustrates the pathophysiologic abnormalities of cardiogenic shock resulting from decreased left ventricular contractility. The primary abnormality is that the relation of end-systolic pressure to volume is shifted down and to the right (see Fig. 21-2, upper panel) so that, at the same afterload, the ventricle cannot eject as far (decreased contractility). It follows that pump function is also impaired, indicated by a shift down and to the right (see Fig. 21-2, lower panel) so that at similar preloads cardiac output is reduced. Three mechanisms that counter the decrease in cardiac output are illustrated. The diastolic ventricle becomes more compliant, possibly from stress relaxation of the pericardium and myocardium, so that stroke volume increases at the same end-diastolic pressure (rightward shift of the diastolic pressure-volume relation in the upper panel). Afterload decreases, resulting in increased stroke volume (see Fig. 21-2, upper panel). Mean systemic pressure rises (see Fig. 21-2, lower panel), aided by avid fluid retention by the kidneys and by increased venous tone mediated by the sympathetic nervous system. Thus, the Frank-Starling mechanism of increasing cardiac output by increasing diastolic filling is used.

As a result of a decrease in contractility, the patient presents with elevated left and right ventricular filling pressures and a low cardiac output. Mixed venous oxygen saturation may be well below 50% because cardiac output is low. In the presence of physiologic pulmonary shunt that accompanies pulmonary edema, the low saturation of mixed venous blood shunting by the lung contributes to substantial arterial desaturation. Accordingly, arterial desaturation aggravates the low oxygen delivery due to reduced cardiac output, as does intercurrent anemia.

Acute myocardial infarction or ischemia is the most common cause of left ventricular failure leading to shock. The principal effect of myocardial infarction is to depress systolic contractility, which in completely infarcted areas becomes zero or even negative (paradoxical regional wall motion). Earlier series described shock occurring in 10% to 20% of patients with transmural myocardial infarction.22 However, the recent use of fibrinolytic therapy and early angioplasty or surgical revascularization has reduced the incidence of cardiogenic shock to less than 5%.23 Infarction greater than 40% of the myocardium is often associated with cardiogenic shock;24 anterior infarction is 20 times more likely to lead to shock than is inferior or posterior infarction.25 Details of the diagnosis and management of ischemic heart disease are discussed in Chap. 25; other causes of decreased left ventricular contractility in critical illness are discussed in more detail in Chaps. 22 and 23, and each may contribute to shock.

Diastolic Dysfunction

Increased left ventricular diastolic chamber stiffness contributing to cardiogenic shock occurs during myocardial ischemia and in a range of less common disorders including late stages of hypovolemic shock and septic shock (see Table 21-4); all causes of tamponade listed in Table 21-4 need to be considered in a systematic review of causes of diastolic dysfunction.26 Cardiac function is depressed because stroke volume is decreased by decreased end-diastolic volume caused by increased diastolic chamber stiffness. Diastolic dysfunction in a hypotensive patient with low cardiac output and high filling pressures is often identified by a small (rather than large) LVEDV by bedside echocardiography. Conditions resulting in increased diastolic stiffness are particularly detrimental when systolic contractility is decreased because decreased diastolic stiffness (increased compliance; see Fig. 21-2, upper panel) is normally a compensatory mechanism. Increased diastolic chamber stiffness contributing to hypotension in patients with low cardiac output and high ventricular diastolic pressures is best identified echocardiographically by small diastolic volumes.

Treatment of increased diastolic stiffness is approached by first considering the potentially contributing reversible causes. Acute reversible causes include ischemia and the many causes of tamponade physiology listed in Table 21-4. Fluid infusion results in large increases in diastolic pressure without much increase in diastolic volume. Positive inotropic agents and afterload reduction are generally not helpful and may decrease blood pressure further. If conventional therapy of cardiogenic shock aimed at improving systolic function is ineffective, then increased diastolic stiffness should be strongly considered as the cause of decreased pump function. Cardiac output responsiveness to heart rate is another subtle clue suggesting impaired diastolic filling. Heart rate does not normally alter cardiac output (which is normally set by and equal to venous return) except at very low heart rates (maximally filled ventricle before end diastole) or at very high heart rates (incomplete ventricular relaxation and filling). However, if diastolic filling is limited by tamponade or a stiff ventricle, then very little further filling occurs late in diastole. In this case, increasing heart rate from 80 to 100 or 110 beats/min may result in a significant increase in cardiac output, which may be therapeutically beneficial and also a diagnostic clue.

Valvular Dysfunction

Acute mitral regurgitation, due to cordal or papillary muscle rupture or papillary muscle dysfunction, most commonly is caused by ischemic injury. The characteristic murmur and the presence of large V waves on the pulmonary artery occlusion pressure trace suggest significant mitral regurgitation, which is quantified by Doppler echocardiographic examination. Rupture of the ventricular septum with left-to-right shunt is detected by Doppler echocardiographic examination or by observing a step-up in oxygen saturation of blood from the right atrium to the pulmonary artery.27 Rarely, acute obstruction of the mitral valve by left atrial thrombus or myxoma may also result in cardiogenic shock. These conditions are generally surgical emergencies.

More commonly, valve dysfunction aggravates other primary etiologies of shock. Aortic and mitral regurgitation reduces forward flow and raises LVEDP, and this regurgitation is ameliorated by effective arteriolar dilation and by nitroprusside infusion. Vasodilator therapy can effect large increases in cardiac output without much change in mean blood pressure, pulse pressure, or diastolic pressure, so pulmonary artery catheterization and repeat echocardiography, to confirm increased cardiac output and reduced valvular regurgitation, are essential to titrating effective vasodilator doses. In contrast, occasional patients develop decreased blood pressure and cardiac output on inotropic drugs such as dobutamine; in this case, excluding dynamic ventricular outflow tract obstruction by echocardiography or treating it by increasing preload, afterload, and end-systolic volume are essential.

Cardiac Arrhythmias

Not infrequently, arrhythmias aggravate hypoperfusion in other shock states. Ventricular tachyarrhythmias are often associated with cardiogenic shock; sinus tachycardia and atrial tachyarrhythmias are often observed with hypovolemic and septic shock. Specific therapy of tachyarrhythmias depends on the specific diagnosis, as discussed in Chap. 24. Inadequately treated pain and unsuspected drug withdrawal should be included in the intensive care unit differential diagnosis of tachyarrhythmias; whatever their etiology, the reduced ventricular filling time can reduce cardiac output and aggravate shock. Bradyarrhythmias contributing to shock may respond acutely to atropine or isoproterenol infusion and then pacing; hypoxia or myocardial infarction as the cause should be sought and treated. Symptomatic hypoperfusion resulting from bradyarrhythmias, even in the absence of myocardial infarction or high-degree atrioventricular block, is an important indication for temporary pacemaker placement that is sometimes overlooked.

Treatment of Left Ventricular Failure

After initial resuscitation, which includes consideration of early institution of thrombolytic therapy in acute coronary thrombosis and revascularization or surgical correction of other anatomic abnormalities where appropriate,3 management of patients with cardiogenic shock requires repeated testing of the hypothesis of “too little versus too much.” Clinical examination is not accurate enough; when the response to initial treatment of cardiogenic shock is inadequate, a pulmonary artery catheter may be required. Therapy for cardiogenic shock follows from consideration of the pathophysiology illustrated in Fig. 21-2 and includes optimizing filling pressures, increasing contractility by improving the ratio of myocardial oxygen supply to demand, or by using inotropic drugs, and optimizing afterload. Temporary mechanical support using an intraaortic balloon pump is often extremely useful in cardiogenic shock and should be considered early as a support device in patients who may benefit from later surgical therapy.28 Cardiac transplantation and mechanical heart implantation are considered when other therapy fails.

Filling pressures are optimized to improve cardiac output but avoid pulmonary edema. Depending on the initial presentation, cardiogenic shock frequently spans the spectrum of hypovolemia (so fluid infusion helps) to hypervolemia with pulmonary edema (where reduction in intravascular volume results in substantial improvement). If gross fluid overload is not present, then a rapid fluid bolus should be given. In contrast to patients with hypovolemic or septic shock, a smaller bolus (250 mL) of crystalloid solution should be infused as quickly as possible. Immediately after infusion, the patient's circulatory status should be reassessed. If there is improvement but hypoperfusion persists, then further infusion with repeat examination is indicated to attain an adequate cardiac output and oxygen delivery while seeking the lowest filling pressure needed to accomplish this goal. If there is no improvement in oxygen delivery and evidence of worsened pulmonary edema or gas exchange, then the limit of initial fluid resuscitation has been defined. Crystalloid solutions are used particularly if the initial evaluation is uncertain because crystalloid solutions rapidly distribute to the entire extracellular fluid compartment. Therefore, after a brief period only one-fourth to one-third remains in the intravascular compartment, and evidence of intravascular fluid overload rapidly subsides.

Contractility increases if ischemia can be relieved by decreasing myocardial oxygen demand, by improving myocardial oxygen supply by increasing coronary blood flow (coronary vasodilators, thrombolytic therapy, surgical revascularization, or intra-aortic balloon pump counterpulsation), or by increasing the oxygen content of arterial blood. Inotropic drug infusion attempts to correct the physiologic abnormality by increasing contractility (see Fig. 21-2). However, this occurs at the expense of increased myocardial oxygen demand. Afterload is optimized to maintain arterial pressures high enough to perfuse vital organs (including the heart) but low enough to maximize systolic ejection. When systolic function is reduced, vasodilator therapy may improve systolic ejection and increase perfusion, even to the extent that blood pressure rises.29 In patients with very high blood pressure, end-systolic volume increases considerably so that stroke volume and cardiac output decrease unless LVEDV and LVEDP are greatly increased; this sequence is reversed by judicious afterload reduction.

Diagnosis and Management of Right Ventricular Failure

Right ventricular failure as a cause of cardiogenic shock is often identified by elevated right atrial pressure and low cardiac output not explained by left ventricular failure or cardiac tamponade. The most common causes of shock owing to right ventricular failure are right ventricular infarction and pulmonary embolism resulting in greatly increased right ventricular afterload.

Right ventricular infarction is found in approximately half of inferior myocardial infarctions and is complicated by shock only 10% to 20% of the time.30 Isolated right ventricular infarction with shock is uncommon and has a mortality rate ∼50% comparable to left ventricular infarction shock.25 The hemodynamic findings of right ventricular infarction must be distinguished from cardiac tamponade and constrictive pericarditis and include Kussmaul sign, low cardiac output, high filling pressures, and often equalization of right atrial, right ventricular diastolic, pulmonary artery diastolic, and pulmonary artery occlusion pressures. Pulmonary crackles are classically absent. Early recognition of right versus left ventricular infarction as the cause of shock is important, so potentially dangerous therapy, including vasodilators, morphine, and β blockers, are avoided. Therapy includes infusion of dobutamine and volume expansion, although excessive volume can aggravate shock by shifting the intraventricular septum from right to left.31 Because bradyarrhythmias are common and atrioventricular conduction is frequently abnormal, atrioventricular sequential pacing may preserve right ventricular synchrony and often dramatically improves cardiac output and blood pressure in shock caused by right ventricular infarction.31 Afterload reduction using balloon counterpulsation may also be useful,25 as are early fibrinolytic therapy and angioplasty when indicated (see Chap. 25).

Right ventricular ischemia, with or without coronary artery disease, probably is a more important cause of right ventricular dysfunction than generally recognized. In shock states systemic arterial pressure is often low, and right ventricular afterload (pulmonary artery pressure) may be high owing to emboli, hypoxemic pulmonary vasoconstriction, acidemic pulmonary vasoconstriction, sepsis, or ARDS. Therefore, right ventricular perfusion pressure is low leading to right ventricular ischemia and decreased contractility, which, in the face of normal or high right ventricular afterload, results in right ventricular dilation. Subsequent right-to-left shift of the interventricular septum limits left ventricular filling. Cardiac output is then limited by right ventricular systolic ejection and left ventricular diastolic filling.

Therapy of right ventricular failure caused by decreased right ventricular perfusion and increased afterload is evolving. Animal studies suggest that, acutely in right ventricular shock caused by pulmonary embolism, interventions such as norepinephrine infusion may increase systemic arterial pressure more than pulmonary arterial pressure, resulting in improved right ventricular perfusion. Improved right ventricular function and total cardiac function may result. This approach has not been carefully tested in patients in shock owing to right ventricular failure.32 Established approaches include verifying that pulmonary emboli are present and initiating therapy with anticoagulation, fibrinolytic agents for submassive pulmonary embolism or shock, or surgical embolectomy as necessary.33 Hypoxic pulmonary vasoconstriction may be reduced by improving alveolar and mixed venous oxygenation. More aggressive correction of acidemia should be considered in this setting. Pulmonary vasodilator therapy may be useful in some patients if pulmonary artery pressures can be lowered without significantly lowering systemic arterial pressures. Inhaled nitric oxide, prostaglandin E1, and many other agents have been variably successful. Measurements of pulmonary artery pressure, systemic pressure, cardiac output, and oxygen delivery before and after a trial of a specific potential pulmonary vasodilator are essential (see Chap. 26).

Compression of the Heart by Surrounding Structures

Compression of the heart (cardiac tamponade) limits diastolic filling and can result in shock with inadequate cardiac output despite very high right atrial pressures. Diagnosis of cardiac tamponade is made physiologically by using pulmonary artery catheterization to demonstrate a low cardiac output in addition to elevated and approximately equal right atrial, right ventricular diastolic, pulmonary artery diastolic, and pulmonary artery occlusion pressures (particularly their waveforms). The diagnosis is often best confirmed anatomically by using echocardiographic examination to demonstrate pericardial fluid, diastolic collapse of the atria and right ventricle, and right-to-left septal shift during inspiration. Septal shift during inspiration and increased afterload that accompany decreased intrathoracic pressure during inspiration account for the clinically observed pulsus paradoxus. Although pericardial tamponade by accumulation of pericardial fluid is the most common cause of cardiac tamponade, other structures surrounding the heart may also produce tamponade. Tension pneumothorax, massive pleural effusion, pneumopericardium (rarely), and greatly elevated abdominal pressures may also impair diastolic filling.

Decreasing the pressure of the tamponading chamber by needle drainage of the pericardium, pleural space, and peritoneum can rapidly and dramatically improve venous return, blood pressure, and organ system perfusion. Therefore, the goal of therapy is to accomplish this decompression as rapidly and safely as possible. In patients who are hemodynamically stable, fluid infusion is a temporizing therapy that increases mean systemic pressure so that venous return increases even though right atrial pressure is high. In hemodynamically stable patients, if it is safe to take the time needed to get ultrasonic guidance for needle aspiration or surgical drainage, then this should be done. Otherwise, in an emergency, blind needle drainage is necessary.

Decreased Venous Return—Hypovolemic Shock

The pressure driving venous return to the right atrium is described as mean systemic pressure minus right atrial pressure, where the mean systemic pressure is determined by the vascular volume and by the unstressed volume and capacitance of the systemic vessels. Venous return to the heart when right atrial pressure is not elevated may be inadequate owing to decreased intravascular volume (hypovolemic shock), to decreased tone of the venous capacitance bed so that mean systemic pressure is low (e.g., drugs, neurogenic shock), and occasionally to increased resistance to venous return (e.g., obstruction of the inferior vena cava by abdominal compartment syndrome). In the presence of shock, decreased venous return is determined to be a contributor to shock by finding low left and right ventricular diastolic pressures, often in an appropriate clinical setting such as trauma or massive gastrointestinal hemorrhage.

Hypovolemic Shock

Hypovolemia is the most common cause of shock caused by decreased venous return and is illustrated in Fig. 21-3. Intravascular volume is decreased, so the venous capacitance bed is not filled, leading to a decreased pressure driving venous return back to the heart. This is seen as a left shift of the venous return curve in Fig. 21-3, lower panel, so that cardiac output decreases at a low end-diastolic pressure (intersection of the venous return curve and cardiac function curve). Endogenous catecholamines attempt to compensate by constricting the venous capacitance bed, thereby raising the pressure driving venous return back to the heart, so that 25% reductions in intravascular volume are nearly completely compensated for. Orthostatic decrease in blood pressure by 10 mm Hg or an increase in heart rate of more than 30 beats/min34 may detect this level of intravascular volume reduction. When approximately 40% of the intravascular volume is lost, sympathetic stimulation can no longer maintain mean systemic pressure, resulting in decreased venous return and clinical shock.

After sufficient time (>2 hours) and severity (>40% loss of intravascular volume), patients often cannot be resuscitated from hypovolemic shock.35 This observation highlights the urgency with which patients should be resuscitated. A “no reflow” phenomenon is described in microvascular beds, gut ischemia with systemic release of inflammatory mediators,36 and increased diastolic stiffness (see Fig. 21-3) contribute to the pathophysiology.37

Shock after trauma is a form of hypovolemic shock in which a significant systemic inflammatory response, in addition to intravascular volume depletion, is present. Intravascular volume may be decreased because of loss of blood and significant redistribution of intravascular volume to other compartments, i.e., “third spacing.” Release of inflammatory mediators may result in pathophysiologic abnormalities resembling septic shock. Cardiac dysfunction may be depressed from direct damage from myocardial contusion, from increased diastolic stiffness, from right heart failure, or even from circulating myocardial depressant substances. Shock related to burns similarly is multifactorial with a significant component of intravascular hypovolemia and a systemic inflammatory response (see Chaps. 98, 99, and 100).

Other causes of shock caused by decreased venous return include severe neurologic damage or drug ingestion resulting in hypotension caused by loss of venous tone. As a result of decreased venous tone, mean systemic pressure decreases, thereby reducing the pressure gradient driving blood flow back to the heart so that cardiac output and blood pressure decrease. Obstruction of veins owing to compression, thrombus formation, or tumor invasion increases the resistance to venous return and occasionally may result in shock.

The principal therapy of hypovolemic shock and other forms of shock caused by decreased venous return is rapid initial fluid resuscitation. Warmed crystalloid solutions are readily available. Colloid-containing solutions result in a more sustained increase in intravascular volume. However, in the setting of demonstrated or potential leaking endothelial surfaces (e.g., ARDS), the colloid rapidly redistributes into the entire extravascular water compartment. Pulmonary edema and tissue edema may be aggravated. Overall, no benefit of colloid over crystalloid has been convincingly demonstrated. The role of hypertonic saline and other resuscitation solutions is currently uncertain. Alternatively, transfusion of packed red blood cells increases oxygen-carrying capacity and expands the intravascular volume and is therefore a doubly useful therapy. In an emergency, initial transfusion often begins with type-specific blood before a complete cross-match is available. During initial resuscitation, the Early Goal-Directed Protocol suggests that achieving a hematocrit greater than 30% may be beneficial when ScvO2 is less than 70%. However, after initial resuscitation, maintaining hemoglobin above 90 g/L does not appear to be better than maintaining hemoglobin above 70 g/L. After a large stored red blood cell transfusion, clotting factors, platelets, and serum ionized calcium decrease and therefore should be measured and replaced if necessary (see Chap. 68).

Recognizing inadequate venous return as the primary abnormality of hypovolemic shock alerts the physician to several commonly encountered and potentially lethal complications of therapy. Airway intubation and mechanical ventilation increase negative intrathoracic pressures to positive values and thus raise right atrial pressure. The already low pressure gradient driving venous return to the heart worsens, resulting in marked reduction in cardiac output and blood pressure. However, ventilation treats shock by reducing the work of respiratory muscle, so ventilation should be implemented early with adequate volume expansion. Sedatives and analgesics are often administered at the time of airway intubation, resulting in reduced venous tone because of a direct relaxing effect on the venous capacitance bed or because of a decrease in circulating catecholamines. Thus, the pressure gradient driving venous return decreases. Therefore, in the hypovolemic patient, these medications may markedly reduce cardiac output and blood pressure and should be used with caution and with ongoing volume expansion.

High Cardiac Output Hypotension—Septic Shock

Septic shock is the most common example of shock that may be caused primarily by reduced arterial vascular tone and reactivity, often associated with abnormal distribution of blood flow. Gram-negative bacilli account for half of all cases of sepsis and approximately 50% of these patients develop septic shock.38 In contrast, shock accompanies only 5% to 10% of gram-positive or fungal bloodstream infections,38 although candida infections are emerging as an important cause of attributable mortality.39 Evidence of end-organ hypoperfusion and dysfunction may be present at low, normal, and high cardiac outputs and oxygen deliveries. During evaluation and resuscitation, normal or increased cardiac output with low SVR hypotension is manifested by a high pulse pressure, warm extremities, good nail bed capillary filling, and low diastolic and mean blood pressures. This high cardiac output hypotension is often accompanied by an abnormal temperature and white blood cell count and differential and an evident site of sepsis. However, the diagnosis is sometimes initially unclear when septic shock is combined with cardiogenic or hypovolemic shock, which limit the usual increase in cardiac output, oxygen delivery, and mixed-venous oxygen saturation.

Several pathophysiologic mechanisms contribute to inadequate organ system perfusion in septic shock. There may be abnormal distribution of blood flow at the organ system level, within individual organs, and even at the capillary bed level. The result is inadequate oxygen delivery in some tissue beds.

The cardiovascular abnormalities of septic shock (see Fig. 21-4) are extensive and include systolic and diastolic abnormalities of the heart, abnormal arterial tone, decreased venous tone, and abnormal distribution of capillary flow leading to regions of tissue hypoxia. In addition, there may be a cellular defect in metabolism so that even cells exposed to adequate oxygen delivery may not maintain normal aerobic metabolism. Depressed systolic contractility illustrated as a rightward shift of the end-systolic pressure-volume relation in Fig. 21-4, upper panel, occurs in septic shock40 due to circulating proinflammatory cytokines and other mediators, nitric oxide production by activated endothelial cells and cardiomyocytes, and reactive oxygen intermediate generation by retained leukocytes and other cells.41 Decreased systolic contractility associated with septic shock is reversible over 5 to 10 days as the patient recovers.40 Systolic and diastolic dysfunctions during sepsis that progress to the point that high cardiac output (hyperdynamic circulation) is no longer maintained (normal or low cardiac output is observed) are associated with poor outcome.40

Decreased arterial resistance is almost always observed in septic shock. Early in septic shock, a high cardiac output state exists with normal or low blood pressure. The low arterial resistance is associated with impaired arterial and precapillary autoregulation and may be due to increased endothelial nitric oxide production and opening of potassium adenosine triphosphate channels on vascular smooth muscle cells. Redistribution of blood flow to low-resistance, short time-constant vascular beds (such as skeletal muscle) results in decreased resistance to venous return, as illustrated in Fig. 21-4 (lower panel) by a steeper venous return curve. As a result, cardiac output may be increased even when cardiac function is decreased (see Fig. 21-4, lower panel) because of decreased contractility (see Fig. 21-4, upper panel). Hypovolemia, caused by redistribution of fluid out of the intravascular compartment and to decreased venous tone, impedes venous return during septic shock.

Early institution of appropriate antibiotic therapy and surgical drainage of abscesses or excision of devitalized and infected tissue is central to successful therapy. Activated protein C and low-dose steroids therapy in patients with inadequate adrenal responses to ACTH stimulation improve outcome. Many other anticytokine and anti-inflammatory therapies and inhibition of nitric oxide production have not been successful in improving outcome.

Other Types of Shock

As detailed in Table 21-4, there are many less common etiologies of shock, and the diagnosis and management of several causes of high right atrial pressure hypotension are discussed elsewhere in this book (see Chaps. 22, 26, and 28). A few other types of hypovolemic shock merit early identification by their characteristic features and lack of response to volume resuscitation including neurogenic shock and adrenal insufficiency. Anaphylactic shock results from the effects of histamine and other mediators of anaphylaxis on the heart, circulation, and the peripheral tissues (see Chap. 106). Despite increased circulating catecholamines and the positive inotropic effect of cardiac H2 receptors, histamine may depress systolic contractility via H1 stimulation and other mediators of anaphylaxis. Marked arterial vasodilation results in hypotension even at normal or increased cardiac output. Like septic shock, blood flow is redistributed to short time-constant vascular beds. The endothelium becomes more permeable, so fluid may shift out of the vascular compartment into the extravascular and intracellular compartments, resulting in intravascular hypovolemia. Venous tone and therefore venous return are reduced, so the mainstay of therapy of anaphylactic shock is fluid resuscitation of the intravascular compartment and includes epinephrine and antihistamines as adjunctive therapy.42

Neurogenic shock is uncommon. In general, in a patient with neurologic damage that may be extensive, the cause of shock is usually associated with blood loss. Patients with neurogenic shock develop decreased vascular tone, particularly of the venous capacitance bed, which results in pooling of blood in the periphery. Therapy with fluid will increase mean systemic pressure. Catecholamine infusion also will increase mean systemic pressure, and stimulation of α receptors will increase arterial resistance, but these are rarely needed once circulation volume is repleted.

Several endocrinologic conditions may result in shock. Adrenal insufficiency (Addison disease, adrenal hemorrhage and infarction, Waterhouse-Friderichsen syndrome, adrenal insufficiency of sepsis, and systemic inflammation) or other disorders with inadequate catecholamine response may result in shock or may be important contributors to other forms of shock.17 Whenever inadequate catecholamine response is suspected, diagnosis should be established by measuring serum cortisol and conducting an ACTH stimulation test, whereas presumptive therapy proceeds using dexamethasone (see Chap. 79). Hypothyroidism and hyperthyroidism may in extreme cases result in shock; thyroid storm is an emergency requiring urgent therapy with propylthiouracil or other antithyroid drug, steroids, propranolol, fluid resuscitation, and identification of the precipitating cause43 (see Chap. 80). Pheochromocytoma may lead to shock by markedly increasing afterload and by redistributing intravascular volume into extravascular compartments.44 In general, the therapeutic approach involves treating the underlying metabolic abnormality, resuscitating with fluid to produce an adequate cardiac output at the lowest adequate filling pressure, and infusing inotropic drugs, if necessary, to improve ventricular contractility if it is decreased. Details of diagnosis and therapy of shock associated with poisons (carbon monoxide, cyanide) are discussed in Chap. 102.

Inflammatory Component of Shock

Shock has a hemodynamic component that has been the focus of much of the preceding discussion. In addition, shock is invariably associated with some degree of inflammatory response, although this component of shock varies greatly. A severe systemic inflammatory response (e.g., sepsis) can result primarily in shock. Conversely, shock results in an inflammatory response because ischemia-reperfusion injury45 will be triggered to some extent after successful hemodynamic resuscitation of shock of any kind. Ischemia-reperfusion causes release of proinflammatory mediators, chemotactic cytokines, and activation of endothelial cells and leukocytes. Because of the multiorgan system involvement of shock, the inflammatory response of ischemia-reperfusion involves many organ systems. Rapid hemodynamic correction of hypovolemic or cardiogenic shock may result in a minimal systemic inflammatory response. However, trauma with significant tissue injury or prolonged hypoperfusion states usually elicit marked systemic inflammatory responses. Because the resolution and repair phases of the inflammatory response are complex and take time, this component of shock is important to recognize and characterize clinically because it has prognostic value with profound effects on the subsequent clinical course.

A systemic inflammatory response results in elevated levels of circulating proinflammatory mediators (tumor necrosis alpha, interleukins, prostaglandins, etc.) that activate endothelial cells and leukocytes. Subsequent production of nitric oxide by activated vascular endothelial cells via inducible nitric oxide synthase results in substantial vasodilation. Products of the arachidonic acid pathway generated during the systemic inflammatory response contribute to systemic vasodilation (prostaglandin I2) and pulmonary hypertension (thromboxane A2). Activated endothelial cells and leukocytes upregulate expression of cellular adhesion molecules and their corresponding ligands, resulting in accumulation of activated leukocytes in pulmonary and systemic capillaries and postcapillary venules. Expression of chemotactic cytokines by endothelial and parenchymal cells contributes to flow of activated leukocytes into the lungs and systemic tissues. Activated leukocytes release destructive oxygen free radicals, resulting in further microvascular and tissue damage. Damaged and edematous endothelial cells, retained leukocytes, and fibrin and platelet plugs associated with activation of the complement and coagulation cascades block capillary beds in a patchy manner, leading to increased heterogeneity of microvascular blood flow. As a result of the significant damage to the microvasculature, oxygen uptake by metabolizing tissues is further impaired.46 A severe systemic inflammatory response leads to very high levels of circulating proinflammatory mediators, leukopenia, and thrombocytopenia owing to uptake in excess of production, disseminated intravascular coagulation owing to excessive activation of the coagulation cascades, diffuse capillary leak, marked vasodilation that may be quite unresponsive to high doses of vasopressors, and generalized organ system dysfunction.

Whereas the hemodynamic component of shock is often rapidly reversible, the resolution and repair phases of an inflammatory response follow a frustratingly slow time course: recruitment of adequate and appropriate leukocyte populations, walling off or control of the initial inciting stimuli, modulation of the subsequent inflammatory response toward clearance with apoptosis of inflammatory and damaged cells (T-helper 1 type of response), or, when the inflammatory stimulus is not as easily cleared, toward a more chronic response with recruitment of new populations of mononuclear leukocytes and fibrin and collagen deposition (T-helper 2 type of response). During this repair and resolution phase, current therapy involves vigilant supportive care of the patient to prevent and avoid the common multiple complications associated with multiple organ system dysfunction and mechanical ventilation.

Individual Organ Systems

Altered mental status, ranging from mild confusion to coma, is a frequently observed effect of shock on neurologic function, when brain blood flow decreases by approximately 50%.47 Autoregulation of cerebral blood flow is maximal, and decreased neurologic function ensues, when MAP decreases to below 50 to 60 mm Hg in normal individuals. Elevated PCO2 transiently dilates and decreased PCO2 transiently constricts cerebral vessels. Profound hypoxia also results in markedly decreased cerebral vascular resistance. Patients recovering from shock infrequently suffer frank neurologic deficit unless they have concomitant cerebrovascular disease. However, subsequent subtle neurocognitive dysfunction is now increasingly recognized.

Systolic and diastolic myocardial dysfunction during shock have been discussed above. Myocardial oxygen extraction is impaired during sepsis and myocardial perfusion is redistributed away from the endocardium. This maldistribution is further aggravated by circulating catecholamines. Segmental and global myocardial dysfunction occur with ST and T-wave changes apparent on the electrocardiogram, and elevations in creatine kinase and troponin concentrations may be observed48 in the absence of true myocardial infarction. In addition, the metabolic substrate for myocardial metabolism changes so that free fatty acids are no longer the prime substrate and more lactic acid and endogenous fuels are metabolized.

More than any other organ system, the lungs are involved in the inflammatory component of shock. ARDS is the acronym given to lung injury caused by the effect of the systemic inflammatory response on the lung and has aptly been called “shock lung.” Inflammatory mediators and activated leukocytes in the venous effluent of any organ promptly affect the pulmonary capillary bed, leading to activation of pulmonary vascular endothelium and plugging of pulmonary capillaries with leukocytes. Ventilation perfusion matching is impaired and shunt increases. High tidal volume ventilation induces a further intrapulmonary inflammatory response and lung damage. Increased ventilation associated with shock results in increased work of breathing to the extent that a disproportionate amount of blood flow is diverted to fatiguing ventilatory muscles.

Early in shock, increased catecholamines, glucagon, and glucocorticoids increase hepatic gluconeogenesis leading to hyperglycemia. Later, when synthetic function fails, hypoglycemia occurs. Clearance of metabolites and immunologic function of the liver are also impaired during hypoperfusion. Typically, centrilobular hepatic necrosis leads to release of transaminases as the predominant biochemical evidence of hepatic damage, and bilirubin levels may be high. Shock may lead to gut ischemia before other organ systems become ischemic, even in the absence of mesenteric vascular disease. Mucosal edema, submucosal hemorrhage, and hemorrhagic necrosis of the gut may occur. Hypoperfusion of the gut has been proposed as a key link in the development of multisystem organ failure after shock, particularly when ARDS precedes sepsis49; that is, loss of gut barrier function results in entrance of enteric organisms and toxins into lymphatics and the portal circulation. Because the immunologic function of the liver is impaired, bacteria and their toxic products, particularly from portal venous blood, are not adequately cleared. These substances and inflammatory mediators produced by hepatic reticuloendothelial cells are released into the systemic circulation and may be an important initiating event of a diffuse systemic inflammatory process that leads to multisystem organ failure or to the high cardiac output hypotension of endotoxemia. Decreased hepatic function during shock impairs normal clearance of drugs such as narcotics and benzodiazepines, lactic acid, and other metabolites that may adversely affect cardiovascular function. In addition, pancreatic ischemic damage may result in the systemic release of a number of toxic substances including a myocardial depressant factor.

The glomerular filtration rate decreases as renal cortical blood flow is reduced by decreased arterial perfusion pressures and by afferent arteriolar vasoconstriction owing to increased sympathetic tone, catecholamines, and angiotensin. The ratio of renal cortical to medullary blood flow decreases. Renal hypoperfusion may lead to ischemic damage with acute tubular necrosis, and debris and surrounding tissue edema obstruct tubules. Loss of tubular function is compounded by loss of concentrating ability because medullary hypertonicity decreases. Impaired renal function or renal failure leads to worsened metabolic acidosis, hyperkalemia, impaired clearance of drugs and other substances; all contribute to the poor outcome of patients in shock with renal failure.

Shock impairs reticuloendothelial system function, leading to impaired immunologic function. Coagulation abnormalities and thrombocytopenia are common hematologic effects of shock. Disseminated intravascular coagulation occurs in approximately 10% of patients with hypovolemic and septic shock. Shock combined with impaired hematopoietic and immunologic function seen with hematologic malignancies or after chemotherapy is nearly uniformly lethal. Endocrine disorders, from insufficient or ineffective insulin secretion to adrenal insufficiency, adversely affect cardiac and other organ system function. Conceivably, impaired parathyroid function is unable to maintain calcium homeostasis. As a result, ionized hypocalcemia is observed during lactic acidosis or its treatment with sodium bicarbonate infusion.50

Shock and Therapeutic Interventions

Hypoperfusion alters the efficacy of drug therapy by slowing delivery of drugs, altering pharmacokinetics once delivered, and decreasing the clearance of drugs. For example, subcutaneous injection of medications may fail to deliver useful quantities of a drug in the setting of decreased perfusion. When adequate perfusion is re-established, the drug may be delivered in an unpredictable way at an inappropriate time. Thus, parenteral medications should be given intravenously to patients with evidence of hypoperfusion. In marked hypoperfusion states, peripheral intravenous infusion also may be ineffective, and central venous administration may be necessary to effectively deliver medications. Once the drug is delivered to its site of action, it may not have the same effect in the setting of shock. For example, catecholamines may be less effective in an acidotic or septic state. Because there may be significant renal and hepatic hypoperfusion, drug clearance is frequently greatly impaired. With these observations in mind, it is appropriate to consider, for each drug, necessary changes in route, dose, and interval of administration in shock patients.

Bicarbonate therapy of metabolic acidosis associated with shock may have adverse consequences.50 Bicarbonate decreases ionized calcium levels further, with a potentially detrimental effect on myocardial contractility. Because bicarbonate and acid reversibly form carbon dioxide and water, a high PCO2 is observed. Particularly during bolus infusion, acidotic blood containing bicarbonate may have a very high PCO2, which readily diffuses into cells, resulting in marked intracellular acidosis; recall that hypoperfusion increases tissue PCO2 by carrying off the tissue CO2 production at a higher mixed venous PCO2 owing to reduced blood flow. Intracellular acidosis results in decreased myocardial contractility. These adverse consequences of bicarbonate therapy may account in part for the lack of benefit observed with bicarbonate therapy of metabolic acidosis.50

Untreated, shock leads to death. Even with rapid, appropriate resuscitation, shock is associated with a high initial mortality rate, and tissue damage sustained during shock may lead to delayed sequelae. Several studies have identified important predictors. For cardiogenic shock, 85% of the predictive information is contained in age, systolic blood pressure, heart rate, and presenting Killip class.3 A blood lactic acid level in excess of 5 mmol/L is associated with a 90% mortality rate in cardiogenic shock51 and a high mortality rate in other shock states. These mortality rates have decreased during the past decade of interventional cardiology and aggressive antibiosis (see Chaps. 25 and 46). In septic shock, decreasing cardiac output predicts death, and high concentrations of bacteria in blood and a failure to mount a febrile response predict a poor outcome. Age and pre-existing illness are important determinants of outcome. Multisystem organ failure is an important adverse outcome, leading to a mortality rate in excess of 60%.