Keywords

Introduction

Circulatory shock is a clinical syndrome characterized by inadequate tissue perfusion of oxygen and other nutrients, resulting in first reversible and then, if prolonged, irreversible cellular injury. The resulting deficit in tissue oxygenation leads to cellular hypoxia and anaerobic metabolism manifested systemically as lactic acidosis. The magnitude of the oxygen debt correlates with the lactate level, which may be used to quantify the severity of shock [1]. If not interrupted, the cascade of cell death, end-organ damage, and multisystem organ dysfunction can cause significant morbidity and death.

Accordingly, the clinician must recognize the early manifestations of shock, and resuscitation must proceed expeditiously and simultaneously with efforts to identify the specific etiology of the shock state.

Pathophysiology and Monitoring

During the past several decades, understanding the biology of shock at the cellular and subcellular levels has exploded (Fig. 1.1a, b). While an awareness of this complex biology is important, especially in sepsis, the focus of this chapter will be on the clinical issues surrounding oxygen delivery, various methods of recognizing oxygen deficits, and clinical approaches to restoring oxygen delivery and maintaining tissue perfusion.

Fig. 1.1
figure 1

(a, b) The pathogenesis of shock is complex and involves extensive interactions of various mediators in the plasma, blood, vasculature, and organs, which can lead to apoptosis, refractory hypotension, multiple organ failure, and death if not interrupted. (Courtesy of Dr. Joseph Parrillo)

Full size image

The presence of hypotension has traditionally been the clinical hallmark of circulatory shock. Current thinking recognizes the fact that patients may remain normotensive despite the presence of systemic hypoperfusion, due to compensatory mechanisms, a situation sometimes referred to as “cryptic shock” [2]. Patients who are hypotensive in the emergency department are at substantially increased risk of end-organ dysfunction and death compared to normotensive patients, even if the hypotension is transient [3].

Adequate perfusion pressure is therefore necessary, but not sufficient to ensure that organs are indeed well perfused. Occasionally, a patient may be hypotensive but NOT in shock (e.g., due to vasodilatation with normal intravascular volume and cardiac output and maintained tissue perfusion); however, the presence of hypotension should garner concern for shock first. Accordingly, monitoring of blood pressure remains a fundamental priority in caring for the critically ill.

Blood pressure may be measured noninvasively by a variety of techniques or directly by the insertion of an arterial catheter. Auscultatory and oscillometric methods rely on vibrations in the arterial wall caused by pulsatile flow through the compressed arterial segments as pressure in the cuff is released. These vibrations are diminished in patients with severe arteriosclerosis, vasoconstriction, or low flow states. Given that these conditions are prevalent among critically ill patients, placement of an arterial catheter in critically ill patients may be preferable to relying on noninvasive measurements. Indeed, Cohn demonstrated that among patients with low cardiac output and high systemic vascular resistance, cuff pressures grossly underestimated true intra-arterial pressures [4]. More recently, Low et al. confirmed these findings in unstable helicopter transport patients [5]. Arterial lines enable the accurate titration of therapies aimed at maintaining mean arterial pressure (MAP) in the recommended ranges (i.e., MAP >65 mmHg in septic shock and cardiogenic shock patients) [6]. The establishment of an arterial line also enables safe and frequent sampling of arterial blood to measure blood gases and lactate.

Tissue perfusion is determined by several factors, including blood pressure (BP), cardiac output, and vascular tone both at the arteriole and venule levels. Under normal circumstances, 85% of the circulating blood volume is housed in the venous capacitance vessels.

Blood pressure (perfusion pressure) is determined by the interplay between cardiac output (CO) and systemic vascular resistance (SVR) (Fig. 1.2):

$$ {\displaystyle \begin{array}{l}\mathrm{BP}=\mathrm{CO}\times \mathrm{SVR}\\ {}\mathrm{SV}\times \mathrm{HR}\end{array}} $$
Fig. 1.2
figure 2

Cardiac output is determined by the product of heart rate and stoke volume (the amount of blood pumped out of the heart with each beat). Stroke volume is affected by the filling of the heart (preload), the pressure against which heart must overcome to eject blood (afterload), and the intrinsic pumping capacity of the heart (contractility)

Full size image

Physiologic compensatory mechanisms defend against hypotension. During hemorrhage, for example, as blood volume decreases, resulting in decreased stroke volume (SV); heart rate increases, maintaining CO and BP. Further bleeding results in decreased CO, for which the body compensates by increasing SVR to maintain blood pressure. Accordingly, patients may have normal or even elevated blood pressure in the face of substantial systemic hypoperfusion. This makes blood pressure an inadequate monitor of perfusion. Rather than relying on hypotension to define it, shock is best conceptualized as an imbalance between tissue oxygen supply and tissue oxygen demand (Fig. 1.3). Regardless, most authorities recommend maintaining a mean arterial pressure (MAP) of >65 mmHg in patients in shock, which may require the use of vasoactive medications [6].

Fig. 1.3
figure 3

Circulatory shock results from on imbalance between tissue oxygen supply and tissue oxygen demands

Full size image

In the overwhelming majority of patients in shock, decreased oxygen supply is the primary cause of this imbalance. Factors that increase oxygen demands such as increased work of breathing, fever, seizures, and shivering may tip the scale toward hypoperfusion and should be addressed in these patients.

To enable intelligent therapeutic intervention in shock patients, clinicians must fully understand the determinants of oxygen delivery. Oxygen delivery (DO2) is determined by the amount of oxygen contained in the blood (arterial oxygen content) and the total systemic flow (CO) (Fig. 1.4).

Fig. 1.4
figure 4

Total systemic oxygen delivery (DO2), hemoglobin in grams (Hg); 1.34 is a constant that represents the amount of oxygen in cubic centimeters carried on 1 g of Hg when fully saturated; .0031 is the solubility coefficient of oxygen in plasma; Falk et al. [60]

Full size image

Arterial oxygen content comprises oxygen both carried on hemoglobin and dissolved in the plasma. Because the solubility coefficient of oxygen in plasma is very low (0.0031), for clinical purposes, when calculating the oxygen content of blood, the dissolved amount of oxygen is so small that it may be ignored. Conversely, the PaO2 is critically important because it determines the saturation of hemoglobin (Hgb) (Fig. 1.5).

Fig. 1.5
figure 5

The oxyhemoglobin dissociation curve indicates that at venous levels of PO2 (40 mmHg), hemoglobin is 75% saturated. When arterial PO2 falls below 60 mmHg, hemoglobin saturation falls rapidly. At arterial PO2 levels above 80 mmHg, hemoglobin is nearly fully saturated

Full size image

The PaO2 should be maintained at 80–85 mmHg, keeping Hgb nearly fully saturated. Interventions aimed at further increasing the PaO2 do little to increase oxygen delivery because the Hgb is already near full saturation. When the PaO2 decreases below 60 mmHg, Hgb rapidly desaturates and DO2 is compromised. When Hgb levels fall, DO2 may also be compromised. This decrease may be ameliorated by interventions that increase CO or via transfusion of red blood cells. (Transfusions will be discussed further in Chap. 33.)

Manipulating the loading conditions of the heart can increase CO. Fluid resuscitation increases preload, which by the Starling mechanism increases SV and CO. In patients with high SVR, vasodilators can increase SV and CO by decreasing afterload. Inotropic and chronotropic drugs can increase contractility and heart rate to increase CO, but do so at the expense of increasing myocardial oxygen demands.

Optimizing oxygen delivery, therefore, should include respiratory therapy techniques that ensure that hemoglobin is fully saturated. Hemoglobin levels should be maintained with red cell transfusion, and cardiac output should be optimized beginning with the restoration of adequate preload and subsequently with the judicious use of vasoactive medication to maintain both perfusion and perfusion pressure (BP) (Fig. 1.6).

Fig. 1.6
figure 6

The fraction of inspired oxygen (FIO2), positive end-expiratory pressure (PEEP), continuous positive airway pressure (CPAP). In the face of systemic hypoperfusion, clinicians can increase oxygen delivery by manipulating the determinants of oxygen delivery in the biologically most cost-effective manner

Full size image

Humans live in a state of oxygen excess, using only 25% of the oxygen delivered to the body each minute. While at rest, this is evidenced by the fact that blood returning to the right heart contains Hgb that is 75% saturated. If illness results in only a modest reduction in oxygen delivery, there is little physiologic impact. As delivery is further reduced, tissues will extract more of the available oxygen and mixed venous saturation will decrease (oxygen extraction ratio increases), maintaining tissue PO2. As these mechanisms are overwhelmed, tissue PO2 falls and glucose can no longer be metabolized in the mitochondria in the citric acid (Krebs) cycle. Rather, glucose metabolism is shunted to the cytoplasm. This produces far fewer ATP molecules per moles of glucose metabolized (2 ATP in anaerobic metabolism versus 38 ATP in aerobic metabolism), and the byproduct of this process is lactate (Fig. 1.7).

Fig. 1.7
figure 7

Adenosine triphosphate molecules (ATP), adenosine diphosphate molecules (ADP). In the face of tissue hypoxia, anaerobic metabolism results in fewer moles of ATP produced per mole of glucose metabolized this way. Lactate is a byproduct of anaerobic metabolism

Full size image

Lactate has been recognized as a monitor of systemic hypoperfusion in animal models and patients for over 50 years [7]. Arterial lactate concentration remains an excellent tool with which to monitor the presence and severity of oxygen debt in shock patients and its level correlates with the likelihood of survival [8]. (Severe Sepsis and Septic Shock will be discussed further in Chap. 19.)

In classic studies, Weil and Afifi showed that patients in an ICU carried an 80% risk of death when arterial lactate was >10 mm/dl, while that risk was <20% when lactate was <2 mm/dl. Patients with lactates around 5 mm/dl had a 50% mortality rate. More recently, Shapiro and colleagues demonstrated that in patients with serious infections presenting to the emergency department, lactate level could risk stratify patients [9]. Those with arterial lactate levels greater than 4.0 mg/dl had 3-day and 28-day mortality rates of 22.4% and 28.4%, respectively. This mortality rate was significantly higher than among patients with lactates between 2.5 and 3.9 mg/dl (4.6% 3-day mortality and 9.0% 28-day mortality) and those between 0 and 2.4 mg/dl (1.5% 3-day mortality and 4.9% 28-day mortality).

Perhaps, more striking is the ability of lactate to identify patients with systemic hypoperfusion but normal or high blood pressure. Abou-Khalil and Scalea [10] studied seriously injured patients requiring resuscitation and room transfusions at a level one trauma center. At 1 hour into the resuscitation, lactate was 7.7 +/− 1.2 mg/dl among patients who died and 4.1 +/− 0.6 mg/dl among survivors (P = .001). Mean arterial pressures in both groups were 106 mmHg, and neither BP, HR, CVP, PAWP, Hct, CO nor oxygen consumption was different between the groups. Thus, lactate level was superior to traditional vital signs at identifying patients with hypoperfusion, and higher levels of lactate predicted increased mortality risk.

Similarly, in a subset of septic patients from the Early Goal-Directed Therapy Study (EGDT) reported by Rivers [11], Nguyen found that 23/133 control and 25/130 EGDT patients were normotensive with MAP ≥100 mmHg. At study entry, lactates were 9.6 mm/L and 8.4 mm/L in the control and EGDT groups, respectively. In hospital, mortality was 60.9% in the control group and 20% in the EGDT group. These data substantiate the importance of lactate as an invaluable bedside tool in identifying critically hypoperfused patients [12]. In a study examining lactate levels and survival utilizing data from the Surviving Sepsis Campaign database of 28,150 septic subjects from 218 international sites, Casserly and colleagues demonstrated that elevated lactate levels are highly associated with in hospital mortality. Both hypotensive and normotensive patients who presented with lactate levels greater than 4 mmol/L were demonstrated to have significantly higher risk than those with intermediate levels (2–3 and 3–4 mmo/L) [13].

Peripheral venous lactate may be used as a screening tool that obviates the need for arterial puncture. Younger and Falk [14] demonstrated that peripheral venous lactate was 100% sensitive in predicting arterial hyperlactatemia. Falsely elevated venous lactate levels occasionally occur, most commonly due to specimen-collection issues (e.g., prolonged tourniquet application, long interval between sampling and testing). If concern exists, an arterial sample can be obtained to confirm an elevated venous lactate level, although this is usually not necessary. A normal venous lactate level reliably predicts normal arterial lactate.

Serial lactate measurements can guide ongoing therapeutic interventions. Falk and colleagues demonstrated that patients surviving an episode of septic shock progressively cleared their lactate levels during the first 24 hours following fluid resuscitation, while among patients who expired, lactate levels failed to decrease or increase [15]. Nguyen and colleagues studied septic patients in the emergency setting and determined that mortality was lower among patients with more rapid clearance of lactate. Patients who cleared lactate at a rate greater than or equal to 10% per hour experienced a 60-day mortality rate significantly lower than those who did not [16].

These data suggest that as patients are being resuscitated, serial lactate measurements can be used to help guide resuscitative measures. Therapies can be titrated to maintain lactate clearance at or above 10% per hour. If lactate clearance is slower or if lactate levels are increasing, then further intervention to increase oxygen delivery and decrease oxygen demands is indicated. The choice of intervention should be guided by a firm understanding of the determinants of oxygen delivery. The clinician must choose the intervention that promises the best chance for improving the oxygen delivery/demand balance at the lowest biological cost.

While lactate is an excellent metabolic marker of shock that predicts severity and mortality, it has limitations. It takes time for lactate to accumulate and especially to clear. Regional hypoperfusion (such as the splanchnic bed) may be missed, as blood mixes centrally [17]. Increased sympathetic stimulation may result in increased lactate production without hypoxia at the cellular level. Lactate levels under these circumstances are generally very modest and do not affect the utility of lactate as a monitor. Other causes of elevated lactate are occasionally present; elevated lactate without evidence of systemic hypoperfusion is Type B lactic acidosis. It can be caused by regional hypoperfusion, liver disease, diabetes mellitus (especially with metformin therapy), alcoholism, malignancy, HIV and antiretroviral therapy, thiamine deficiency, mitochondrial dysfunction, poisoning, and other mechanisms. (Type A lactic acidosis refers to lactic acidosis due to systemic hypoperfusion, as discussed earlier.)

Monitoring of mixed venous (SVO2) or central venous oxygen saturation (SCVO2) may complement lactate monitoring and has the advantage of responding to physiologic changes in real time. In the face of systemic hypoperfusion, tissues will extract more of the available oxygen and venous oxygen saturation will decrease. SCVO2 has been shown to be closely correlated with SVO2, allowing for either continuous or intermittent sampling of SCVO2 for monitoring purposes without the need to place a pulmonary artery catheter [18].

When SVO2 or SCVO2 is low, tissue hypoxia is present and measures to increase oxygen delivery are indicated. If this situation occurred acutely, lactate levels may not have had time to increase. It is crucial to understand that normal SVO2 or SCVO2 does not preclude the presence of hypoperfusion, especially in septic patients. Sepsis can result in disproportionate perfusion of metabolically relatively inactive tissues, such as the skin, while flow is shunted away from critically hypoperfused areas such as the splanchnic bed. Desaturated blood returning from the splanchnic bed and mixing with saturated blood returning from the skin may not show SVO2 or SCVO2 desaturation [17]. True arteriovenous shunts or “metabolic block” may also contribute to this observation. Under these circumstances, lactate levels will likely be elevated.

Recently, there has been a debate in the literature regarding which shock monitor is preferable as a target for ongoing care during early sepsis therapy. Jones et al. compared lactate and SCVO2 as targets and found no mortality difference between the groups [19]. We would argue that this choice is a false dichotomy. The key issue in caring for these patients is that the clinician must have a firm understanding of the physiology and should use all the available monitoring tools to their best advantage. Venous saturation monitoring and lactate levels should be viewed as complimentary, each providing useful and potentially critical information.

Capnography has emerged as a very useful monitoring tool over the past 25 years. It has been embraced as an essential method to verify endotracheal tube placement in the prehospital and ED settings following the publications by Katz and Falk and Silvestri and colleagues [20, 21]. It has been shown, as well, to be an indicator of the effectiveness of closed chest massage during CPR and the earliest indication that spontaneous circulation has been re-established [22].

Recently, ETCO2 has been used to noninvasively identify patients with metabolic acidosis such as diabetic ketoacidosis [23]. Patients with metabolic acidosis hyperventilate to compensate and would be expected to have low ETCO2, assuming adequate cardiopulmonary reserve. We examined the relationship between ETCO2, lactate, and mortality among patients suspected to be septic. As expected, there was an inverse correlation between lactate and ETCO2. The sickest patients had the highest lactates, the lowest ETCO2, and the highest mortality [24]. Accordingly, we believe a spot ETCO2 may serve as a useful adjunct to SIRS criteria, shock index, and lactate as a rapid, noninvasive screening tool when assessing patients in the emergency department.

Classification of Shock

The Shubin/Weil Classification of Shock, first described in the 1960s, remains a most useful framework for clinicians at the bedside. It recognizes four broad categories of shock: hypovolemic, cardiogenic, obstructive, and distributive (Fig. 1.8). Multiple factors may contribute to the shock state. As the syndrome progresses, common pathways of inflammatory and hormonal mediators are activated, and if unchecked, it results in cellular dysfunction, refractory hypotension, multiple organ dysfunction, and death.

Fig. 1.8
figure 8

Shubin/Weil classification of shock. The drawings represent components of the circulatory system: the heart, resistance vessels (arteries and arterioles), capacitance vessels (veins), the capillary beds, and arteriovenous shunts. Weil Critical Care Research Institute

Full size image

Hypovolemic Shock

Hypovolemic shock is characterized by intravascular volume loss. Acute hemorrhage from trauma, gastrointestinal bleeding, ruptured abdominal aortic aneurysm, ectopic pregnancy, or other causes is the most common etiology of hypovolemic shock. Gastrointestinal disorders that result in substantial fluid losses (i.e., cholera), decreased oral intake, or excessive diuresis as well as third space losses (i.e., pancreatitis) can also cause hypovolemic shock. Insufficient intravascular volume results in decreased preload, decreased cardiac output, and decreased oxygen delivery. Compensatory mechanisms such as tachycardia and arteriolar vasoconstriction may maintain perfusion in some patients for limited periods of time. Patients who are not yet frankly hypotensive may demonstrate postural hypotension, although autonomic dysfunction and medications such as antihypertensive agents, especially among the elderly, may be alternative reasons for this finding. Patients with intra-abdominal bleeding may have paradoxical bradycardia, resulting from vagal stimulation by the irritated peritoneum, most commonly seen in ruptured ectopic pregnancy [25].

Typically, compensatory vasoconstriction produces cool, pale skin and delayed capillary refill. In dehydrated patients, skin and axillae may be dry and the skin may have reduced turgor; in patients with acute hemorrhage, diaphoresis is present. Patients who have hemorrhage may also have pale mucous membranes and conjunctiva. Patients can be in hemorrhagic shock with initial hemoglobin levels that are not dramatically reduced because there has not been sufficient time for transcapillary refill to occur. As asanguineous fluid resuscitation is instituted, hemoglobin level drops dramatically. Recent animal and clinical studies have emphasized the need to stop the bleeding as soon as possible. Pepe and Mattock found that patients with penetrating truncal trauma had improved survival when crystalloid resuscitation was restricted until surgical intervention could be accomplished [26]. Similar findings have been described in patients with gastrointestinal hemorrhage [27]. The notion that restoring blood pressure in hypotensive hemorrhagic shock patients by infusing large volumes of crystalloids may dislodge clots and exacerbate bleeding is supported by animal studies [28]. Accordingly, in hemorrhagic shock patients, the goal of fluid resuscitation should be to achieve an acceptable perfusion pressure (MAP 60–65 mmHg, systolic 100 mmHg; some authors have suggested a systolic goal of 70 mmHg or MAP of 50 mmHg) [26, 29, 30], and once accomplished, crystalloid infusions should be minimized until control of the bleeding is accomplished.

Patients with hypovolemic shock from other fluid losses must have careful monitoring of their electrolytes to correct perturbations in a safe and thoughtful manner. Multiple electrolyte abnormalities may be present in these patients. To avoid the devastating complication of central pontine myelinolysis, hyponatremia must not be corrected too rapidly [31].

Obstructive Shock

Obstructive shock results from an extracardiac process that mechanically obstructs either the filling or emptying of the heart. Stroke volume (SV) is diminished and cardiac output falls. Common etiologies include tension pneumothorax, cardiac tamponade, massive pulmonary embolus, and SVC syndrome. Less common causes are dissecting aortic aneurysm, severe pulmonary hypertension, and constrictive pericarditis. Relieving the obstruction of the circulation is the priority, while other therapies such as fluid administration and vasopressors are temporizing maneuvers.

The presence of diminished breath sounds, tracheal deviation, and distended neck veins in a hypotensive patient are the hallmarks of tension pneumothorax, and immediate needle decompression is indicated. These findings, in a trauma patient, require immediate tube thoracostomy. Muffled heart sounds, distended neck veins, and hypotension with good breath sounds bilaterally suggest cardiac tamponade. (Trauma will be discussed further in Chap. 24.) Immediate bedside echocardiography can demonstrate the effusion as well as tamponade physiology, allowing for immediate bedside pericardiocentesis. Patients with tamponade physiology may maintain their blood pressure through compensatory mechanisms. Jugular venous distention (JVD) may not be present in dehydrated patients until plasma volume is restored with fluid resuscitation.

Massive pulmonary embolism resulting in obstructive shock is a most challenging clinical situation. (Massive PE will be discussed further in Chap. 7.) Historical features suggesting the risk for venous thromboembolic disease (malignancy, travel, recent surgery, etc.) combined with symptoms such as pleuritic chest pain and dyspnea suggest the diagnosis. Syncope may be the only presenting complaint, especially among the elderly. Hypotension, tachycardia, distended neck veins, clear lung fields, and hypoxemia should result in emergent echocardiography. Right ventricular distention with shift of the interventricular septum to the left is diagnostic. Relieving the right ventricular outflow tract obstruction by reducing clot burden is imperative, but exactly how to do that remains controversial [32]. Surgical thrombectomy, catheter retrieval, and local or systemic thrombolysis with or without the use of venoarterial ECMO are techniques that are available in many centers. Maintaining right ventricular perfusion can be accomplished with the liberal use of norepinephrine while avoiding overzealous fluid resuscitation that may exacerbate the over distention of the right ventricle, further compromising left ventricular filling. Intubation may be necessary in these patients, but induction hypotension may impair RV perfusion, and positive-pressure ventilation can increase pulmonary vascular resistance and reduce preload. Accordingly, in patients unstable enough to require intubation, plans for emergent mechanical or pharmacologic embolectomy should be initiated. Central line placement in these patients should anticipate the subsequent need for systemic thrombolysis and/or anticoagulation and should, therefore, be carefully placed by the most experienced operator under ultrasound guidance [32].

In its most advanced stages, obstructive shock may present as a pulseless electrical activity (PEA) cardiac arrest. Accordingly, in patients with PEA arrests, causes of obstructive shock must be sought and treated. Aspiration of as little as 30 mL of pericardial fluid may restore the circulation in patients arresting from tamponade, while a simple needle thoracostomy may do so in patients in PEA arrest from tension pneumothorax.

Cardiogenic Shock

Cardiogenic shock most commonly results from acute myocardial infarction (AMI). When 30–40% of the left ventricular muscle mass is infarcted, the patient may develop cardiogenic shock. This can occur from a single episode or a after a number of smaller infarcts (ischemic cardiomyopathy). The best treatment for cardiogenic shock resulting from AMI is to prevent it by minimizing infarct size through aggressive early intervention. Dramatic progress has been made in this regard by the institution of STEMI alert protocols with reduced door to balloon times [33,34,35,36]. (AMI will be discussed further in Chap. 8.) Nonischemic cardiomyopathies (postpartum, viral, infiltrative, etc.) less commonly cause cardiogenic shock. Mechanical causes of cardiogenic shock include acute valvular dysfunction, ventricular septal defects, ruptured ventricular free wall, and blunt cardiac injury. Acute valvular dysfunction can result from slowly progressive syndromes, such as aortic or mitral stenosis, or from abrupt processes, such as severe mitral or tricuspid regurgitation from papillary muscle rupture, or endocarditis. These syndromes are typically associated with characteristic murmurs. Bedside echocardiography is diagnostic and emergent surgical correction is required.

Patients in cardiogenic shock are typically hypotensive and vasoconstricted, cold and clammy, with jugular venous distention (JVD) and pulmonary edema. A subset of patients in cardiogenic shock may be hypovolemic for a variety of reasons. Extravasation of fluid from the vascular space into the pulmonary interstitium (pulmonary edema fluid) may occur as left-sided pressures abruptly increase due to ischemia. Patients may continue to have taken diuretic medications while having been anorexic from ischemia leading up to the event. Accordingly, a modest fluid challenge is warranted in hypotensive AMI patients [37, 38]. Maintaining coronary perfusion pressure is imperative in caring for cardiogenic shock patients, resulting from AMI. Diseased coronary arteries cannot autoregulate, so coronary perfusion is dependent on diastolic pressure. Norepinephrine is the catecholamine agent of choice because it increases blood pressure with less increase in heart rate than agents with more balanced alpha and beta effects (dopamine, epinephrine). Mechanical support with an intra-aortic balloon pump can sustain patients while they await emergent revascularization. Coronary artery bypass surgery has been shown to result in better survival rates than angioplasty in this setting [39]. (CHF and cardiogenic shock will be discussed further in Chap. 12.)

In contradistinction to patients with left ventricular infarcts, patients with right ventricular (RV) infarcts have JVD but clear lungs, and hypotension resulting from decreased left ventricular filling. The electrocardiogram is helpful in evaluating these patients. RV infarcts are most commonly seen in patients with inferior wall MIs. Right-sided chest leads can confirm the diagnosis.

Severe brady or tachydysrhythmias can result in very low cardiac output, mimicking cardiogenic shock. Rapid treatment with transcutaneous electrical pacing, cardioversion, or pharmacotherapy can restore perfusion. (Dysrhythmias will be discussed further in Chap. 9.)

Distributive Shock

Distributive shock is caused by systemic vasodilation, which decreases venous return and preload. In septic shock, vasodilation is often accompanied by leaking capillary membranes, which produces intravascular volume depletion and further decreases preload. Sepsis is the predominant cause of distributive shock. All infective agents can cause the syndrome, but gram-negative and gram-positive bacteria are the predominant pathogens. Distributive shock can also be caused by anaphylaxis, high spinal cord trauma, adrenal insufficiency, and the systemic immune response syndrome (SIRS), resulting from major burns, pancreatitis, or polytrauma.

Distributive shock is typically associated with a compensatory elevation in cardiac output, but this hyperdynamic state requires that intravascular volume be maintained. In severely hypovolemic patients, cardiac output is decreased and resultant vasoconstriction can make them clinically indistinguishable from cardiogenic shock patients. Because capillary leak occurs, patients may have decreased intravascular volume and excessive interstitial fluid (edema). Inflammatory mediators associated with sepsis often produce myocardial depression. Accordingly, patients with distributive shock may have low, normal, or high cardiac output, depending on the interaction among these multiple factors. When cardiac output is high, preferential perfusion of the skin resulting from vasodilation results in these patients typically having warm pink skin (“warm shock”). Patients with neurogenic shock may have inappropriately normal HR or frank bradycardia due to loss of sympathetic cardiac stimulation.

Clinical Assessment

Patients in circulatory shock may demonstrate a variety of vital sign abnormalities. Febrile patients in shock are likely to be septic and should be evaluated carefully for the source of infection. Septic patients may present with hypothermia (<95° F). Noninfectious sources of fever include pulmonary embolism, pancreatitis, and other causes of SIRS.

Bradycardia (typically HR <50) may result in low cardiac output and poor perfusion. Tachycardia may be present as a compensatory mechanism in patients with hypovolemic, distributive, or obstructive shock. Tachydysrhythmias may also result in low cardiac output because ventricular filling is compromised at very rapid heart rates. The clinician must determine if the tachycardia is the primary problem or a compensatory mechanism. Elderly patients frequently present in atrial fibrillation with rapid ventricular response under conditions in which younger patients would have sinus tachycardia. Many patients, especially the elderly, who may have chronotropic incompetence from heart disease, and those who take beta-blockers, calcium-channel blockers, or other medications that limit increases in heart rate may have inappropriately normal heart rates in the face of circulatory shock.

Tachypnea may be the earliest manifestation of sepsis and the most sensitive vital sign abnormality among patients in shock. Sadly, respiratory rate is the one vital sign most likely to be charted incorrectly among hospitalized patients. Normal respiratory rate is 12–16 breaths per minute in adults. Breaths should be counted for a minimum of 30 seconds and reported accurately. Respiratory alkalosis may be an early sign of sepsis, poisoning (Aspirin), and encephalopathy. Tachypnea is an important marker of metabolic acidosis in shock, and ETCO2 measurement may assist in defining the presence of compensatory respiratory alkalosis (low ETCO2). Profound tachypnea with its associated increased work of breathing may also be a source of increased metabolic demands. This has led to the approach of early intubation with controlled mechanical ventilation in shock patients suffering from concurrent respiratory compromise. Reducing the work of breathing can improve systemic perfusion in these patients. Patients with primary pulmonary pathology (pneumonia, pulmonary embolism, COPD exacerbations, asthma, etc.) may, of course, also be tachypneic in the absence of shock. Accordingly, tachypnea is a very sensitive but nonspecific marker of shock.

Low oxygen saturation as measured by pulse oximetry may indicate a pulmonary etiology of shock, such as pneumothorax, pneumonia, or pulmonary embolus. Poor perfusion especially in the cold, vasoconstricted patient in shock or on vasopressors may result in the inability to achieve an adequate pulse oximetry waveform and yield falsely low values. Under these circumstances, arterial blood gas sampling is required.

Physical Examination

Physical examination can provide important clues to the presence of and etiology of shock states. Mild confusion may be an early warning sign of impaired cerebral perfusion. Elderly patients presenting with even subtle mental status changes from baseline should be evaluated for a potential source of infection. Significant shock may be associated with more profound mental status changes, progressing from confusion to delirium to obtundation.

Oliguria (defined as urine output less than 0.5 mg/kg/h) indicates impaired renal perfusion from either intravascular volume depletion, low cardiac output, or shunting of renal blood flow to other vital organs. Urine output is an important monitor of the success of therapeutic interventions and should be monitored closely in all shock patients.

Vasoconstricted patients with cool, clammy, sometimes cyanotic skin often have low cardiac output and increased SVR; vasodilated patients with low SVR (as in distributive shock) may have warm skin. Skin examination should look for evidence of cutaneous abscess, infected joints, decubitus ulcers, and cutaneous signs of endocarditis. Weak or absent pulses indicate poor perfusion.

Auscultation of the heart and lungs may reveal signs of tension pneumothorax (decreased or absent breath sounds unilaterally), pneumonia (crackles), pericardial effusion/tamponade (muffled heart sounds, rubs), heart failure (lung crackles, gallops), or valvular dysfunction (murmurs). Pulmonary edema should raise concern for a cardiogenic cause of shock. Tracheal deviation is a cardinal finding in tension pneumothorax. JVD may be seen in obstructive and cardiogenic shock, while flat neck veins may be present in distributive or hypovolemic shock.

Abdominal and pelvic examinations are crucial in the assessment of shock patients, although the presence of shock may result in less obvious findings than in normally perfused patients. A tender or rigid abdomen indicates the likelihood of a surgical emergency. Bowel perforations, volvulus, intra-abdominal abscesses, appendicitis, ruptured diverticulitis, and pancreatitis may all present this way. Vascular problems, such as ruptured or leaking AAA, or intestinal ischemia must be considered. A rectal examination looking for frank or occult blood should be performed if gastrointestinal bleeding is suspected. Intra-abdominal bleeding from ruptured ectopic pregnancy or retroperitoneal bleeding must be considered in hemorrhagic shock patients without other sources such as gastrointestinal bleeding. Urogenital sources of infection (UTI, gynecologic, Fournier’s gangrene, perirectal abscess) should be sought.

In patients with trauma in whom hemorrhagic shock is suspected, careful attention should be paid to abdominal, thoracic, pelvic, and extremity sources of bleeding, as well as hemorrhage from open wounds. Signs of cervical or thoracic spinal trauma may indicate neurogenic shock. (Head and Spinal Cord Injuries will be discussed further in Chap. 23.)

Diagnostic Testing

Laboratory testing routinely includes serum electrolytes, renal function testing, complete blood count, troponin, liver function tests, coagulation profile, and urinalysis. D-dimer may be measured in patients with suspected venous thromboembolism. Amylase and/or lipase should be measured in patients with a suspected intra-abdominal source of infection. Cultures of blood and urine should be obtained in cases of suspected sepsis. Measurements of arterial blood gases provide critical information about oxygenation and acid–base status as well as the presence or absence of carboxyhemoglobin or methemoglobin. Point of care testing that enables very rapid results at the bedside is extraordinarily helpful to the emergency physician when confronted with a critically ill patient. Blood gases, lactate, troponin, sodium, potassium, glucose, and beta HCG are among the most useful available point of care tests.

In critically ill patients, diagnostic imaging should be used liberally. CT scanning of the head, thorax, abdomen, and pelvis has become essential in evaluating critically ill patients, and concerns regarding radiation exposure should not preclude the use of this technology in these patients. The decision of how and when to transport patients away from a resuscitation area for imaging and who should accompany the patient are critical decisions that need to be made by a seasoned clinician in conjunction with nurses and respiratory therapists. Whenever possible, imaging that can be performed at the bedside in the resuscitation area, such as portable x-rays and ultrasound, should be the first option.

Increasingly, bedside ultrasound can assist in the rapid diagnosis of causes of shock; this may be especially useful in patients who are too unstable for other types of imaging:

  • FAST: Identify pneumothorax, traumatic cardiac tamponade, and intra-abdominal fluid (which may represent hemorrhage).

  • Abdominal/pelvic ultrasound: Recognition of AAA, intraperitoneal/retroperitoneal hemorrhage, ectopic pregnancy.

  • Cardiac ultrasound: Identify pericardial effusion and tamponade physiology, estimate preload via IVC measurement, estimate ventricle size and cardiac output, identify RV dilatation associated with massive PE. (Ultrasound will be discussed further in Chap. 35.)

Initial Stabilization

After establishing the ABCs, initial stabilization of the patient suspected to be in shock must proceed rapidly, with the emergency physician orchestrating the priorities in the sequencing of multiple pressing imperatives. While the physical examination primary survey is being conducted, adequate peripheral venous access, electrocardiographic monitoring, noninvasive blood pressure monitoring, and pulse oximetry should be established before bedside chest radiography and 12-lead EKG are attempted. Immediate life threats identified in the primary survey are addressed as they are discovered (i.e., applying pressure to bleeding wounds, tube thoracostomy in tension pneumothorax). As data become available, specific therapies may be initiated, such as emergent cardiac catheterization in STEMI patient or intramuscular epinephrine injection in anaphylactic patients (wheezing, hives, hypotension).

Intravascular Volume Resuscitation

Fluid resuscitation is the mainstay of treatment for most patients in circulatory shock. Fluid resuscitation expands intravascular volume, increases venous return to the heart, and thereby increases preload, resulting in increased stroke volume and cardiac output. This results in increased oxygen delivery, which in hypoperfused (shock) patients will increase oxygen delivery and begin to reverse lactic acidosis (Fig. 1.9). The amount of fluid required varies with the type of shock and individual patient. Patients in septic shock and other forms of distributive shock generally require large volumes of resuscitative fluids. Hemorrhagic shock patients need blood products and should not receive large volumes of asanguineous fluids. Patients in nonhemorrhagic, hypovolemic shock may need substantial fluid resuscitation, but the amounts needed vary considerably from patient to patient, making careful monitoring imperative. Patients with AMI and hypotension, even with pulmonary edema, may benefit from judicious fluid challenge. Patients with obstructive shock should receive emergent mechanical intervention and, generally, should not receive large volumes of resuscitative fluids.

Fig. 1.9
figure 9

5% dextrose in water (D5W), normal saline (NS), lactated Ringer’s solution (LR), albumin (ALB), hydroxyethyl starch (HES), hypertonic saline (HYPER). The figure depicts the impact of infusions of the various fluids on the three fluid compartments in the body

Full size image

Types of Fluid

Characteristics of resuscitative fluids will determine where they are distributed within the three body compartments (intravascular, interstitial, and intracellular) after infusion into the veins (Fig. 1.10). D5W, lacking any solutes, is distributed throughout the total body water. The intracellular compartment is by far the largest of the three. Accordingly, at the end of an infusion, very little D5W remains in the vascular space. It should not be considered a resuscitative fluid and should not be used for this purpose. Crystalloid fluids, such as normal saline (NS) and lactated Ringer’s solution (LR), contain solutes, which are relatively impermeable to the cell membranes. Accordingly, they are distributed only to the vascular and interstitial compartments. Because the interstitial compartment is so much larger than the vascular space, at the end of a 1-L infusion of these isotonic crystalloids, only approximately 200 mL remains in the vascular space. The remainder crosses the vascular membrane and enters the interstitial space. This is why large volume fluid resuscitation with these fluids results in edema formation, with its potential complications. Recent studies indicate that buffered, balanced solutions such as LR may be preferable to NS because NS may result in a hyperchloremic metabolic acidosis when infused in large quantities [40]. Colloid-containing fluids such as 5% albumin (ALB) and 6% hydroxyethyl starch (HES) create oncotic pressure and are retained in the vascular space for longer periods than crystalloids. Under conditions of “leaky capillaries” seen in sepsis and potentially other forms of shock, colloids may also become permeable to the vascular membranes. The starches have been associated with coagulopathies, immune suppression, and AKI with need for RRT and should no longer be used. Hypertonic saline can expand the plasma volume by pulling fluid out of the intracellular and interstitial compartments. This makes them potentially useful in head injured or burned patients where edema can be life threatening.

Fig. 1.10
figure 10

Fluid resuscitation remains the cornerstone for increasing oxygen delivery in shock patients

Full size image

Although much larger volumes of crystalloid fluids are required to provide the same intravascular volume expansion compared to colloids, they appear to be equally effective, as demonstrated by several studies [41,42,43]. However, in the SAFE trial [44], subset analysis showed a trend toward improved mortality for septic patients treated with colloids. Enthusiasm for colloid-containing fluids, especially 5% albumin, is re-emerging in Europe. Hypo-oncotic patients, especially those requiring surgery, are more prone to complications and death. While it has not been definitively established that albumin infusions can reverse this effect, many clinicians infuse albumin-containing fluids to maintain the serum albumin level at 3 mg/dl. Given the lack of proven mortality benefit and the increased expense of albumin, isotonic crystalloid solutions remain the recommended choice for patients in shock. The “colloid/crystalloid” debate rages on.

Goals for Fluid Resuscitation

CVP

The goal of fluid resuscitation is to restore adequate preload to optimize cardiac output (Fig. 1.9). Central venous pressure (CVP) has been used as a tool to estimate preload and guide fluid resuscitation. Multiple studies have demonstrated the lack of correlation between plasma volume and CVP [45]. Nonetheless, a target CVP of 8–12 mmHg has been recommended to optimize fluid administration in septic shock patients and is often used in patients with other types of shock. (Even higher CVPs may be necessary for patients with mechanical ventilation, especially with high PEEP.) It is reasonable to suspect that patients with very low CVP (<4) measurements need additional volume administration, and that patients with very high measurements (>20) are at risk for volume overload. It remains unclear if CVP is routinely useful in patients undergoing fluid resuscitation in shock and if a CVP in the 8–12 mmHg range indicates that further fluid administration will not further increase stroke volume and cardiac output. Additionally, CVP measurement requires placement of a subclavian or internal jugular catheter, which is not without risk.

Fluid Responsiveness

Fluid responsiveness refers to the ability of fluid resuscitation to increase stroke volume and cardiac output and, thereby, improve organ perfusion. Dynamic indices, such as radial artery pulse pressure variation and aortic blood flow peak velocity, may predict fluid responsiveness accurately, with some limitations [46]. These dynamic measurements estimate preload reserve by measuring the variability in predicted stroke volume with respiratory change or estimate stroke volume changes after fluid challenge or simulated fluid challenge by the passive leg-raising maneuver. Ultrasound examination of the vena cava and the presence of collapse during the ventilatory cycle can provide similar data [47]. There is increasing evidence that these dynamic indices are sensitive and specific predictors of fluid responsiveness; however, additional trials are needed before their routine use can be recommended in the management of patients in shock. In addition, many of these methods require a patient to be in sinus rhythm and passively mechanically ventilated, thus limiting widespread use.

Blood Products

For patients with hemorrhage, blood products are the logical choice to replace intravascular volume. In addition to volume expansion (red cells remain in the vascular space as does FFP and platelets), red blood cell transfusion provides additional oxygen-carrying capacity and supports oxygen delivery to tissues. Old banked PRBCs may not effectively carry oxygen until 2,3 DPG is regenerated, which can take many hours. Blood products may also be indicated in patients with preexisting anemia.

In hemorrhagic shock patients, initial resuscitation usually begins with crystalloid administration. For patients with large blood loss, or ongoing bleeding, or those who require more than 30 mL/kg crystalloid, immediate transfusion of packed-red blood cells is indicated. Emergency-release blood may be necessary if cross-matched blood is not yet available.

Patients who receive large amounts of transfused PRBCs are at risk for transfusion-related coagulopathy that may worsen hemorrhage. Early transfusion of FFP and platelets is increasingly recognized as an important component of blood product administration, especially in trauma patients [48, 49]. Evidence for the benefits of massive transfusion protocols, which automatically provide FFP and platelets along with PRBCs, suggests a morbidity and mortality benefit. Patients who are expected to receive >10U PRBCs in 24 hours are good candidates for massive transfusion protocols. (Blood Products will be discussed further in Chap. 33.)

Hemorrhagic shock in trauma represents a special circumstance for clinicians who seek to balance adequate perfusion and tissue oxygenation with limiting ongoing hemorrhage while awaiting operative control. There is evidence to support a strategy that limits initial volume replacement and permits some degree of hypotension (to a systolic BP of 70 mmHg) in some patients until bleeding can be controlled. One prospective trial in patients with penetrating truncal trauma demonstrated survival benefit for patients in whom fluid resuscitation was delayed [26]. Other studies have supported a similar strategy and have shown no evidence of harm associated with permissive hypotension [29, 30]. Patients with coexisting brain injury require adequate cerebral perfusion pressure and are not candidates for permissive hypotension. Results of these limited studies should not be applied to a broad population of trauma patients without further research.

Nonhemorrhagic Shock

There is conflicting evidence on the optimal transfusion threshold for patients with anemia and other (nonhemorrhagic) types of shock. In stable patients, a restrictive transfusion strategy, in which patients do not receive blood products until their hemoglobin is <7 mg/dL, was found to be superior to a more liberal transfusion strategy [50]. RBC transfusion carries risk for infection, SIRS, ARDS, and multiorgan failure and has been associated with increased complications and mortality.

Vasopressors

Vasopressors should be used in patients who cannot maintain MAP (>65) despite adequate volume resuscitation. Norepinephrine has emerged as the vasopressor of choice for the treatment of undifferentiated shock and septic shock patients [51]. (Vasopressors will be discussed further in Chap. 32.)

Norepinephrine acts on alpha-1 and beta-1 receptors, producing potent vasoconstriction and a modest increase in cardiac output. Its vasoconstrictor effect on venous capacitance vessels has the added benefit of moving volume into the active circulation from the venous capacitance bed that ordinarily houses 85% of the circulating blood volume. The chronotropic effect created by beta-1 stimulation is usually modest and may be offset by the reflex bradycardia that occurs in response to increased MAP. Compared to other vasopressor agents such as dopamine and epinephrine, its beta effects are far less prominent. Accordingly, perfusion pressure is improved without the deleterious effects of tachycardia and arrhythmia. Under circumstances in which further increases in cardiac output are required after blood pressure is restored by adequate fluid loading and norepinephrine, an inotropic agent such as dobutamine or milrinone may be added. Because these agents are inotropic vasodilators, careful monitoring of blood pressure is required to be sure that they do not produce hypotension. Under these circumstances, epinephrine alone or added, in small doses, to norepinephrine may accomplish the goals of maintaining perfusion pressure and enhancing contractility. The cost is in increased myocardial oxygen demand and the potential for ischemia.

The physiologic effects of dopamine depend on the dose at which it is administered: at 1–2 mcg/kg/min, it stimulates primarily renal dopamine-1 receptors; at 2–5 mcg/kg/min, it stimulates beta-1 receptors, increasing cardiac output by increasing primarily stroke volume and to a lesser degree, heart rate. At 5–10 mcg/kg/min, it stimulates alpha-1 receptors and produces vasoconstriction and increased SVR. Use of dopamine is primarily limited by the risk for tachycardia and arrhythmia. The ability of low-dose dopamine to convert oliguric renal failure into nonoliguric renal failure has been debunked [52].

Phenylephrine is a purely alpha-adrenergic agent that increases SVR and produces direct vasoconstriction. Reflex bradycardia is common. It is often a second-line agent after norepinephrine. It is available in prefilled syringes and can be given in small bolus doses in critical, time-sensitive situations while preparing a norepinephrine drip [53].

Epinephrine is a potent beta-1 agonist with moderate beta-2 and alpha-1 effects, producing increased inotropy and chronotropy and increased cardiac output. At high doses, predominately alpha-1 effects produce increased SVR. Epinephrine is the preferred agent in anaphylactic shock. Disadvantages include increased splanchnic vasoconstriction and risk for arrhythmia.

Cardiogenic Shock

Norepinephrine is the preferred vasopressor for treating patients in cardiogenic shock with profound hypotension [54]. However, these patients have low cardiac output and intrinsically elevated SVR: Further increases in afterload may limit improvement in cardiac output. This effect may be ameliorated by the concurrent increase in coronary perfusion pressure. Agents with beta-adrenergic activity provide inotropic (and often chronotropic) effects that increase cardiac output; however, this is not desirable in patients with acute ischemia as it increases myocardial work and myocardial oxygen demand. Accordingly, norepinephrine is the first-line agent for patients with severe hypotension to maintain coronary perfusion pressure while arranging for mechanical support and /or revascularization.

Septic Shock

There is evidence to support norepinephrine as the first-line vasopressor in septic shock. In meta-analyses comparing dopamine and norepinephrine in patients with septic shock, increased mortality was seen with dopamine, along with a twofold increase in arrhythmia [55]. Vasopressin may be beneficial when added to norepinephrine or other agents. Some patients with septic shock have myocardial depression and diminished cardiac output: dobutamine may restore cardiac output and improve tissue perfusion and oxygen delivery. Dobutamine may cause peripheral vasodilatation and decreased blood pressure and should be used cautiously. Inotropic therapy is controversial for septic shock patients without clear evidence of myocardial depression. It was used successfully as part of the EGDT algorithm of Rivers et al. [11] when employed in the face of continued evidence of tissue hypoxia after fluids, blood, and vasopressors. More recent trials failed to show mortality benefit to its use in similar, though less critically ill, septic shock patients [56, 57].

Airway Management

Most therapies for shock focus on increasing oxygen delivery to eliminate oxygen debt and anaerobic metabolism. However, many patients in shock have increased work of breathing. Respiratory muscle use can account for significant oxygen consumption, and under these circumstances, perfusion to the diaphragm and other respiratory muscles can rob nutrient flow from other vital organs. Intubation and passive mechanical ventilation (controlled mandatory ventilation) to decrease respiratory muscle work can reduce oxygen demand and improve oxygen debt. Positive-pressure ventilation decreases venous return to the heart. In hypovolemic patients, hypotension and cardiovascular collapse can occur. Appropriate fluid resuscitation prior to intubation can avert this. A low-tidal volume ventilation strategy is recommended for patients in shock with ARDS [58, 59].

Protocol-Directed Therapy

Protocols that combine various physiologic endpoints to guide resuscitation in patients with septic shock have been previously recommended and are commonly used as part of bundled therapy for sepsis. Typically, CVP, MAP, ScVO2, and sometimes UOP and lactate are stepwise targets for fluids and vasoactive agents. Two recent trials were unable to demonstrate a mortality difference between protocolized and standard emergency department care in septic patients [56, 57]. However, in all of the patients, early recognition, attention to adequate volume resuscitation (2 L prior to study entry), and restoration of mean arterial pressure were accomplished. Early antibiotic administration, a most important aspect of care in this group of patients, was also accomplished. The authors did question the need for the routine use of central venous pressure monitoring, SCVO2 monitoring, or ongoing monitoring of lactate. It seems clear that the most important issue is that the physicians directing the care of these patients have a firm understanding of the pathophysiology and current concepts of resuscitation and understand how to monitor the progress of the resuscitative efforts.