Background

In 1914, Schottmueller wrote, Septicemia is a state of microbial invasion from a portal of entry into the blood stream which causes sign of illness. The definition did not change much over the years, because the terms sepsis and septicemia referred to several ill-defined clinical conditions present in a patient with bacteremia. In practice, the terms often were used interchangeably; however, fewer than half the patients with signs and symptoms of sepsis have positive results on blood culture. Furthermore, not all patients with bacteremia have signs of sepsis; therefore, sepsis and septicemia are not identical. In the past few decades, the discovery of endogenous mediators of the host response has led to the recognition that the clinical syndrome of sepsis is the result of excessive activation of host defense mechanisms rather than the direct effect of microorganisms. Sepsis and its sequelae represent a continuum of clinical and pathophysiologic severity. Serious bacterial infections at any body site, with or without bacteremia, are usually associated with important changes in the function of every organ system in the body. These changes are mediated mostly by elements of the host immune system against infection. Shock is deemed present when volume replacement fails to increase blood pressure to acceptable levels and associated clinical evidence indicates inadequate perfusion of major organ systems, with progressive failure of organ system functions. Multiple organ dysfunctions, the extreme end of the continuum, are incremental degrees of physiologic derangements in individual organs (ie, processes rather than events). Alteration in organ function can vary widely from a mild degree of organ dysfunction to frank organ failure. This article does not cover sepsis of the neonate or infant. Special consideration must be given to neonates, infants, and small children with regard to fluid resuscitation, appropriate antibiotic coverage, intravenous (IV) access, and vasopressor therapy. See Neonatal Sepsis for complete information on this topic.

Classification of shock
Shock is identified in most patients by hypotension and inadequate organ perfusion, which may be caused by either low cardiac output or low systemic vascular resistance. Circulatory shock can be subdivided into 4 distinct classes on the basis of underlying mechanism and characteristic hemodynamics, as follows: Hypovolemic shock Obstructive shock Distributive shock Cardiogenic shock These classes of shock should be considered and systemically differentiated before establishing a definitive diagnosis of septic shock. Hypovolemic shock results from the loss of blood volume caused by such conditions as gastrointestinal (GI) bleeding, extravasation of plasma, major surgery, trauma, and severe burns. The patient demonstrates tachycardia, cool clammy extremities, hypotension, dry skin and mucus membranes, and poor turgor. Obstructive shock results from impedance of circulation by an intrinsic or extrinsic obstruction. Pulmonary embolism and pericardial tamponade both result in obstructive shock. Distributive shock is caused by such conditions as direct arteriovenous shunting and is characterized by decreased resistance or increased venous capacity from the vasomotor dysfunction. These patients have

high cardiac output, hypotension, large pulse pressure, a low diastolic pressure, and warm extremities with a good capillary refill. These findings on physical examination strongly suggest a working diagnosis of septic shock. Cardiogenic shock is characterized by primary myocardial dysfunction, resulting in the inability of the heart to maintain adequate cardiac output. These patients demonstrate clinical signs of low cardiac output, while evidence exists of adequate intravascular volume. The patients have cool clammy extremities, poor capillary refill, tachycardia, narrow pulse pressure, and a low urine output.

Definitions of key terms

The basis of sepsis is the presence of infection associated with a systemic inflammatory response that results in physiologic alterations at the capillary endothelial level. The difficulty in diagnosis comes in knowing when a localized infection has become systemic and requires more aggressive hemodynamic support. No criterion standard exists for the diagnosis of endothelial dysfunction, and patients with sepsis may not initially present with frank hypotension and overt shock. Clinicians often use the terms sepsis, severe sepsis, and septic shock without a commonly understood definition. In 1991, the American College of Chest Physicians (ACCP) and the Society of Critical Care Medicine (SCCM) convened a consensus conference to establish definitions of these and related terms. [1,
2]

Systemic inflammatory response syndrome (SIRS) is a term that was developed in an attempt to describe the clinical manifestations that result from the systemic response to infection. Criteria for SIRS are considered to be met if at least 2 of the following 4 clinical findings are present: Temperature greater than 38C (100.4F) or less than 36C (96.8F) Heart rate (HR) greater than 90 beats per minute (bpm) Respiratory rate (RR) greater than 20 breaths per minute or arterial carbon dioxide tension (PaCO 2) lower than 32 mm Hg White blood cell (WBC) count higher than 12,000/L or lower than 4000/L, or 10% immature (band) forms Of course, a patient can have either severe sepsis or septic shock without meeting SIRS criteria, and conversely, SIRS criteria may be present in the setting of many other illnesses (see the image below).

Venn diagram showing the overlap of infection, bacteremia, sepsis, systemic inflammatory response syndrome (SIRS), and multiorgan dysfunction.

In 2001, as a follow-up to the original ACCP/SCCM conference, an International Sepsis Definitions Conference was convened, with representation not only from the ACCP and the SCCM but also from the European Society of Intensive Care Medicine (ESICM), the American Thoracic Society (ATS), and the Surgical Infection Society (SIS). The following definitions of sepsis syndromes were published in order to clarify the terminology used to describe the spectrum of disease that results from severe infection. [3] Sepsis is defined as the presence of infection in association with SIRS. The presence of SIRS is, of course, not limited to sepsis, but in the presence of infection, an increase in the number of SIRS criteria observed should alert the clinician to the possibility of endothelial dysfunction, developing organ dysfunction, and the need for aggressive therapy. Certain biomarkers have been associated with the endothelial dysfunction of sepsis; however, the use of sepsis-specific biomarkers has not yet translated to establishing a clinical diagnosis of sepsis in the emergency department (ED).

With sepsis, at least 1 of the following manifestations of inadequate organ function/perfusion is typically included: Alteration in mental state Hypoxemia (arterial oxygen tension [PaO2] < 72 mm Hg at fraction of inspired oxygen [FiO 2] 0.21; overt pulmonary disease not the direct cause of hypoxemia) Elevated plasma lactate level Oliguria (urine output < 30 mL or 0.5 mL/kg for at least 1 h) Severe sepsis is defined as sepsis complicated by end-organ dysfunction, as signaled by altered mental status, an episode of hypotension, elevated creatinine concentration, or evidence of disseminated intravascular coagulopathy (DIC). Septic shock is defined as a state of acute circulatory failure characterized by persistent arterial hypotension despite adequate fluid resuscitation or by tissue hypoperfusion (manifested by a lactate concentration greater than 4 mg/dL) unexplained by other causes. Patients receiving inotropic or vasopressor agents may not be hypotensive by the time that they manifest hypoperfusion abnormalities or organ dysfunction. Bacteremia is defined as the presence of viable bacteria within the liquid component of blood. It may be primary (without an identifiable focus of infection) or, more often, secondary (with an intravascular or extravascular focus of infection). Although sepsis is commonly associated with bacterial infection, bacteremia is not a necessary ingredient in the activation of the inflammatory response that results in severe sepsis. In fact, septic shock is associated with culture-positive bacteremia in only 30-50% of cases.[4, 5, 6, 7] Multiple organ dysfunction syndrome (MODS) is defined as the presence of altered organ function in a patient who is acutely ill and in whom homeostasis cannot be maintained without intervention. The American-European Consensus Conference on ARDS agreed upon the following definitions of acute lung injury (ALI) and acute respiratory distress syndrome (ARDS). [8] The criteria for ALI include the following: An oxygenation abnormality with a PaO2/FiO2 ratio less than 300 Bilateral opacities on chest radiograph compatible with pulmonary edema Pulmonary artery occlusion pressure less than 18 mm Hg or no clinical evidence of left atrial hypertension if PaO2 is not available ARDS is a more severe form of ALI and is defined similarly, except that the PaO 2/FiO2 ratio is 200 or less. See the following articles for more information: Pediatric Sepsis Bacterial Sepsis Toxic Shock Syndrome Pediatric Toxic Shock Syndrome

Pathophysiology
The pathophysiology of septic shock is not precisely understood, but it involves a complex interaction between the pathogen and the hosts immune system. The normal physiologic response to localized infection includes the activation of host defense mechanisms that result in the influx of activated neutrophils and monocytes, the release of inflammatory mediators, local vasodilation, increased endothelial permeability, and activation of coagulation pathways. These mechanisms are in play during septic shock, but on a systemic scale, leading to diffuse endothelial disruption, vascular permeability, vasodilation, and thrombosis of end-organ capillaries. Endothelial damage itself can further activate inflammatory and coagulation cascades, creating in effect a positive feedback loop, and leading to further endothelial and end-organ damage.

Mediator-induced cellular injury

The evidence that sepsis results from an exaggerated systemic inflammatory response induced by infecting organisms is compelling; inflammatory mediators are the key players in the pathogenesis (see the table below). Table 1. Mediators of Sepsis (Open Table in a new window)
Type Cellular mediators Mediator Lipopolysaccharide Lipoteichoic acid Peptidoglycan Superantigens Endotoxin Humoral mediators Cytokines TNF-alpha and IL-1 Neutrophil chemotactic factor IL-8 Acts as pyrogen, stimulates B and T lymphocyte proliferation, inhibits cytokine production, induces immunosuppression IL-6 Activation and degranulation of neutrophils IL-10 Cytotoxic, augments vascular permeability, contributes to shock Potent proinflammatory effect Activity Activation of macrophages, neutrophils, platelets, and endothelium releases various cytokines and other mediators

MIF

Involved in hemodynamic alterations of septic shock

G-CSF

Promote neutrophil and macrophage, platelet activation and chemotaxis, other proinflammatory effects

Complement Nitric oxide Lipid mediators

Enhance vascular permeability and contributes to lung injury

Enhance neutrophil-endothelial cell interaction, regulate leukocyte migration and adhesion, and play a role in pathogenesis of sepsis

Phospholipase A2

PAF

Eicosanoids

Arachidonic acid metabolites Adhesion molecules

Selectins

Leukocyte integrins

G-CSF = Granulocyte colony-stimulating factor; IL = interleukin; MIF = macrophage inhibitory factor; PAF = platelet-activating factor; TNF = tumor necrosis factor.

An initial step in the activation of innate immunity is the synthesis de novo of small polypeptides, called cytokines, that induce protean manifestations on most cell types, from immune effector cells to vascular smooth muscle and parenchymal cells. Several cytokines are induced, including tumor necrosis factor (TNF) and interleukins (ILs), especially IL-1. Both of these factors also help to keep infections localized, but, once the infection becomes systemic, the effects can also be detrimental. Circulating levels of IL-6 correlate well with outcome. High levels of IL-6 are associated with mortality, but its role in pathogenesis is not clear. IL-8 is an important regulator of neutrophil function, synthesized and released in significant amounts during sepsis. IL-8 contributes to the lung injury and dysfunction of other organs. The chemokines (monocyte chemoattractant protein1) orchestrate the migration of leukocytes during endotoxemia and sepsis. The other cytokines that have a supposed role in sepsis are IL-10, interferon gamma, IL-12, macrophage migration inhibition factor, granulocyte colony-stimulating factor (G-CSF), and granulocyte macrophage colony-stimulating factor (GM-CSF). In addition, cytokines activate the coagulation pathway, resulting in capillary microthrombi and end-organ ischemia.[9, 10, 11] (See Abnormalities of coagulation and fibrinolysis.) Gram-positive and gram-negative bacteria induce a variety of proinflammatory mediators, including the cytokines just mentioned, which play a pivotal role in initiating sepsis and shock. Various bacterial cell wall components are known to release the cytokines, including lipopolysaccharide (gram-negative bacteria), peptidoglycan (gram-positive and gram-negative bacteria), and lipoteichoic acid (gram-positive bacteria). Several of the harmful effects of bacteria are mediated by proinflammatory cytokines induced in host cells (macrophages/monocytes and neutrophils) by the bacterial cell wall component. The most toxic component of the gram-negative bacteria is the lipid A moiety of lipopolysaccharide. The gram-positive bacteria cell wall leads to cytokine induction via lipoteichoic acid. Additionally, gram-positive bacteria may secrete the superantigen cytotoxins that bind directly to the major histocompatibility complex (MHC) molecules and T-cell receptors, leading to massive cytokine production. The complement system is activated and contributes to the clearance of the infecting microorganisms but probably also enhances the tissue damage. The contact systems become activated; consequently, bradykinin is generated. Hypotension, the cardinal manifestation of sepsis, occurs via induction of nitric oxide (NO). NO plays a major role in the hemodynamic alterations of septic shock, which is a hyperdynamic form of shock.

A dual role exists for neutrophils; they are necessary for defense against microorganisms but also may become toxic inflammatory mediators contributing to tissue damage and organ dysfunction. The lipid mediators (eicosanoids), platelet-activating factor (PAF), and phospholipase A2 are generated during sepsis, but their contributions to the sepsis syndrome remain to be established.

Abnormalities of coagulation and fibrinolysis

An imbalance of homeostatic mechanisms leads to disseminated intravascular coagulopathy (DIC) and microvascular thrombosis, causing organ dysfunction and death.[12] Inflammatory mediators instigate direct injury to the vascular endothelium; the endothelial cells release tissue factor (TF), triggering the extrinsic coagulation cascade and accelerating production of thrombin. Plasma levels of endothelial activation biomarkers are higher in patients whose hypotension is the result of sepsis than in patients with hypotension of other causes.[13] The coagulation factors are activated as a result of endothelial damage. The process is initiated via binding of factor XII to the subendothelial surface. This activates factor XII, and then factor XI and eventually factor X are activated by a complex of factor IX, factor VIII, calcium, and phospholipid. The final product of the coagulation pathway is the production of thrombin, which converts soluble fibrinogen to fibrin. The insoluble fibrin, along with aggregated platelets, forms intravascular clots. Inflammatory cytokines, such as IL-1, IL-1, and TNF-alpha, initiate coagulation by activating TF. TF interacts with factor VIIa to form factor VIIa-TF complex, which activates factors X and IX. Activation of coagulation in sepsis has been confirmed by marked increases in thrombin-antithrombin complex and the presence of D-dimer in plasma, indicating activation of the clotting system and fibrinolysis.[14, 15]Tissue plasminogen activator (t-PA) facilitates conversion of plasminogen to plasmin, a natural fibrinolytic. Endotoxins increase the activity of inhibitors of fibrinolysisnamely, plasminogen activator inhibitor (PAI1) and thrombin activatable fibrinolysis inhibitor (TAFI). The levels of protein C and endogenous activated protein C (APC) are also decreased in sepsis. Endogenous APC is an important proteolytic inhibitor of coagulation cofactors Va and VIIa. Thrombin, via thrombomodulin, activates protein C, which then functions as an antithrombotic in the microvasculature. Endogenous APC also enhances fibrinolysis by neutralizing PAI-1 and by accelerating t-PAdependent clot lysis. The imbalance among inflammation, coagulation, and fibrinolysis results in widespread coagulopathy and microvascular thrombosis and suppressed fibrinolysis, ultimately leading to multiple organ dysfunction and death. The insidious nature of sepsis is such that microcirculatory dysfunction can occur while global hemodynamic parameters such as blood pressure may remain normal.[16]

Circulatory abnormalities
As noted (see Background), septic shock falls under the category of distributive shock, which is characterized by pathologic vasodilation and shunting of blood from vital organ to nonvital tissues such as skin, skeletal muscle, and adipose. The endothelial dysfunction and vascular maldistribution characteristic of distributive shock result in global tissue hypoxia or inadequate delivery of oxygen to vital tissues. In addition, mitochondria can become dysfunctional, thus compromising oxygen utilization at the tissue level. The predominant hemodynamic feature of septic shock is arterial vasodilation. The mechanisms implicated in this pathologic vasodilation are multifactorial, but the primary factors are thought to be (1) activation of adenosine triphosphate (ATP)-sensitive potassium channels in vascular smooth muscle cells and (2) activation of NO synthase. The potassium-ATP channels are directly activated by lactic acidosis. NO also activates potassium channels. Potassium efflux from cells results in hyperpolarization, inhibition of calcium influx, and vascular smooth muscle relaxation.[17] The resulting vasodilation can be refractory to endogenous vasoactive hormones (eg, norepinephrine and epinephrine) that are released during shock.

Diminished peripheral arterial vascular tone may result in dependency of blood pressure on cardiac output, causing vasodilation to result in hypotension and shock if insufficiently compensated by a rise in cardiac output. Early in septic shock, the rise in cardiac output often is limited by hypovolemia and a fall in preload because of low cardiac filling pressures. When intravascular volume is augmented, the cardiac output usually is elevated (the hyperdynamic phase of sepsis and shock). Even though cardiac output is elevated, the performance of the heart, reflected by stroke work as calculated from stroke volume and blood pressure, usually is depressed. Factors responsible for myocardial depression of sepsis are myocardial depressant substances, coronary blood flow abnormalities, pulmonary hypertension, various cytokines, NO, and beta-receptor down-regulation. Although an elevation of cardiac output occurs, the arterial-mixed venous oxygen difference usually is narrow, and the blood lactate level is elevated. This implies that low global tissue oxygen extraction is the mechanism that may limit total body oxygen uptake in septic shock. The basic pathophysiologic problem seems to be a disparity between the uptake and oxygen demand in the tissues, which may be more pronounced in some areas than in others. This disparity is termed maldistribution of blood flow, either between or within organs, with a resultant defect in capacity to extract oxygen locally. During a fall in oxygen supply, cardiac output becomes distributed so that most vital organs, such as the heart and brain, remain relatively better perfused than nonvital organs are. However, sepsis leads to regional changes in oxygen demand and regional alteration in blood flow of various organs. The peripheral blood flow abnormalities result from the balance between local regulation of arterial tone and the activity of central mechanisms (eg, the autonomic nervous system). The regional regulation and the release of vasodilating substances (eg, NO, prostacyclin) and vasoconstricting substances (eg, endothelin) affect regional blood flow. Increased systemic microvascular permeability also develops, remote from the infectious focus, and contributes to edema of various organs, particularly the lung microcirculation, and to the development of ARDS. In patients experiencing septic shock, the delivery of oxygen is relatively high, but the global oxygen extraction ratio is relatively low. The oxygen uptake increases with a rise in body temperature despite a fall in oxygen extraction. In patients with sepsis who have low oxygen extraction and elevated arterial blood lactate levels, oxygen uptake depends on oxygen supply over a much wider range than normal. Therefore, oxygen extraction may be too low for tissue needs at a given oxygen supply, and oxygen uptake may increase with a boost in oxygen supplya phenomenon termed oxygen uptake supply dependence or pathologic supply dependence. However, this concept is controversial, because other investigators argue that supply dependence is artifactual rather than a real phenomenon. Maldistribution of blood flow, disturbances in the microcirculation, and, consequently, peripheral shunting of oxygen are responsible for diminished oxygen extraction and uptake, pathologic supply dependency of oxygen, and lactate acidemia in patients experiencing septic shock.

Mechanisms of organ dysfunction

Sepsis is described as an autodestructive process that permits the extension of the normal pathophysiologic response to infection (involving otherwise normal tissues), resulting in multiple organ dysfunction syndrome. Organ dysfunction or organ failure may be the first clinical sign of sepsis, and no organ system is immune to the consequences of the inflammatory excesses of sepsis. The precise mechanisms of cell injury and resulting organ dysfunction in patients with sepsis are not fully understood. MODS is associated with widespread endothelial and parenchymal cell injury because of the following proposed mechanisms:

Hypoxic hypoxia - The septic circulatory lesion disrupts tissue oxygenation, alters the metabolic regulation of tissue oxygen delivery, and contributes to organ dysfunction. Microvascular and endothelial abnormalities contribute to the septic microcirculatory defect in sepsis. Reactive oxygen species, lytic enzymes, vasoactive substances (eg, NO), and endothelial growth factors lead to microcirculatory injury, which is compounded by the inability of the erythrocytes to navigate the septic microcirculation. Direct cytotoxicity - Endotoxin, TNF-alpha, and NO may cause damage to mitochondrial electron transport, leading to disordered energy metabolism. This is called cytopathic or histotoxic anoxiathat is, an inability to use oxygen even when it is present. Apoptosis (programmed cell death) - This is the principal mechanism by which dysfunctional cells normally are eliminated. The proinflammatory cytokines may delay apoptosis in activated macrophages and neutrophils, but other tissues, such as the gut epithelium, may undergo accelerated apoptosis. Therefore, derangement of apoptosis plays a critical role in tissue injury in patients with sepsis. Immunosuppression - The interaction between proinflammatory and anti-inflammatory mediators may lead to an imbalance and an inflammatory reaction, or immunodeficiency may predominate, or both may occur. Coagulopathy - Subclinical coagulopathy signified by mild elevation of the thrombin time or activated partial thromboplastin time or by a moderate reduction in platelet count is extremely common, but overt DIC is rare. Coagulopathy is caused by deficiencies of coagulation system proteins, including protein C, antithrombin III, and TF inhibitors.

Cardiovascular dysfunction
Significant derangement in the autoregulation of the circulatory system is typical in patients with sepsis. Vasoactive mediators cause vasodilatation and increase the microvascular permeability at the site of infection. NO plays a central role in the vasodilatation of septic shock. Impaired secretion of vasopressin also may occur, which may permit the persistence of vasodilatation. Changes in both systolic and diastolic ventricular performance occur in patients with sepsis. Through the use of the Frank-Starling mechanism, cardiac output is often increased to maintain BP in the presence of systemic vasodilatation. Patients with preexisting cardiac disease are unable to increase their cardiac output appropriately. Sepsis interferes with the normal distribution of systemic blood flow to organ systems; therefore, core organs may not receive appropriate oxygen delivery. The microcirculation is the key target organ for injury in patients with sepsis syndrome. A decrease in the number of functional capillaries leads to an inability to extract oxygen maximally; this inability is caused by intrinsic and extrinsic compression of capillaries and plugging of the capillary lumen by blood cells. Increased endothelial permeability leads to widespread tissue edema involving protein-rich fluid. Hypotension is caused by the redistribution of intravascular fluid volume resulting from reduced arterial vascular tone, diminished venous return from venous dilation, and release of myocardial depressant substances.

Pulmonary dysfunction
The pathogenesis of sepsis-induced ARDS is a pulmonary manifestation of SIRS. A complex interaction between humoral and cellular mediators, inflammatory cytokines and chemokines, is involved in this process. A direct or indirect injury to the endothelial and epithelial cells of the lung increases alveolar capillary permeability, causing ensuing alveolar edema. The edema fluid is protein rich; the ratio of alveolar fluid edema to plasma is 0.75-1.0, compared with patients with hydrostatic cardiogenic pulmonary edema, in whom the ratio is less than 0.65. Injury to type II pneumocytes decreases surfactant production; furthermore, the plasma proteins in alveolar fluid inactivate the surfactant previously manufactured. These enhance the surface tension at the air-fluid interfaces, producing diffuse microatelectasis.

Neutrophil entrapment within the pulmonary microcirculation initiates and amplifies the injury to alveolar capillary membrane. ARDS is a frequent manifestation of these effects. As many as 40% of patients with severe sepsis develop ALI. ALI is a type of pulmonary dysfunction secondary to parenchymal cellular damage that is characterized by endothelial cell injury and destruction, deposition of platelet and leukocyte aggregates, destruction of type I alveolar pneumocytes, an acute inflammatory response through all the phases of injury, and repair and hyperplasia of type II pneumocytes. Migration of macrophages and neutrophils into the interstitium and alveoli produces many different mediators, which contribute to the alveolar and epithelial cell damage. If addressed at an early stage, ALI may be reversible, but in many cases, the host response is uncontrolled, and ALI progresses to ARDS. Continued infiltration occurs with neutrophils and mononuclear cells, lymphocytes, and fibroblasts. An alveolar inflammatory exudate persists, and type II pneumocyte proliferation is evident. If this process can be halted, complete resolution may occur. In other patients, a progressive respiratory failure and pulmonary fibrosis develop. The central pathologic finding in ARDS is severe injury to the alveolocapillary unit. Following initial extravasation of intravascular fluid, inflammation and fibrosis of pulmonary parenchyma develops into a morphologic picture, termed diffuse alveolar damage (DAD). The clinical and pathological evolution can be categorized into the following 3 overlapping phases (Katzenstein, 1986): (1) the exudative phase of edema and hemorrhage, (2) the proliferative phase of organization and repair, and (3) the fibrotic phase of end stage fibrosis. The exudative phase occurs in the first week and is dominated by alveolar edema and hemorrhage. The other histological features include dense eosinophilic hyaline membranes and disruption of the capillary membranes. Necrosis of endothelial cells and type I pneumocytes occur, along with leukoagglutination and deposition of platelet fibrin thrombi.

Acute respiratory distress syndrome (ARDS), commonly observed in septic shock as a part of multiorgan failure syndrome, is pathologically diffuse alveolar damage (DAD). This photomicrograph shows an

early stage (exudative stage) of DAD. Acute respiratory distress syndrome (ARDS), commonly observed in septic shock as a part of multiorgan failure syndrome, is pathologically diffuse alveolar damage (DAD). This is a high-powered photomicrograph of an early stage (exudative stage) of DAD.

The proliferative phase is prominent in the second and third week following onset of ARDS but may begin as early as the third day. Organization of the intra-alveolar and interstitial exudate, infiltration with chronic inflammatory cells, parenchymal necrosis, and interstitial myofibroblast reaction occur. Proliferation of type II cells and fibroblasts, which convert the exudate to cellular granulation tissue, occurs; excessive collagen deposition, transforming into fibrous tissue, occurs.

This photomicrograph shows a delayed stage (proliferative or organizing stage) of diffuse alveolar damage (DAD). Proliferation of type II pneumocytes has occurred, hyaline membranes are present, and

collagen and fibroblasts are present. This photomicrograph shows a delayed stage (proliferative or organizing stage) of diffuse alveolar damage (DAD). The fibrin stain showing collagenous tissue, which may develop into the fibrotic stage of DAD.

The fibrotic phase occurs by the third or fourth week of the onset, though the process may begin in the first week. The collagenous fibrosis completely remodels the lung, the air spaces are irregularly enlarged, and alveolar duct fibrosis is apparent. Lung collagen deposition increases, microcystic honeycomb formation, and traction bronchiectasis follows.

Gastrointestinal dysfunction
The GI tract may help to propagate the injury of sepsis. Overgrowth of bacteria in the upper GI tract may aspirate into the lungs and produce nosocomial pneumonia. The guts normal barrier function may be affected, thereby allowing translocation of bacteria and endotoxin into the systemic circulation and extending the septic response. Septic shock usually causes ileus, and the use of narcotics and sedatives delays the institution of enteral feeding. The optimal level of nutritional intake is interfered with in the face of high protein and energy requirements. Glutamine is necessary for normal enterocyte functioning. Its absence in commercial total parenteral nutrition (TPN) formulations leads to a breakdown of the intestinal barrier and to translocation of the gut flora into the circulation. This may be one of the factors that drives sepsis. In addition to inadequate glutamine levels, this may lessen the immune response by decreasing leukocyte and natural killer cell counts, as well as total B-cell and T-cell counts.[18]

Hepatic dysfunction
By virtue of the livers role in host defense, the abnormal synthetic functions caused by liver dysfunction can contribute to both the initiation and progression of sepsis. The reticuloendothelial system of the liver acts as a first line of defense in clearing bacteria and their products; liver dysfunction leads to a spillover of these products into the systemic circulation.

Renal dysfunction
Acute renal failure (ARF) caused by acute tubular necrosis often accompanies sepsis. The mechanism involves systemic hypotension, direct renal vasoconstriction, release of cytokines (eg, TNF), and activation of neutrophils by endotoxins and other peptides, which contribute to renal injury.

Central nervous system dysfunction

Central nervous system (CNS) involvement in sepsis produces encephalopathy and peripheral neuropathy. The pathogenesis is poorly defined.

Etiology
Most patients who develop sepsis and septic shock have underlying circumstances that interfere with the local or systemic host defense mechanisms. Sepsis is seen most frequently in elderly persons and in those with comorbid conditions that predispose to infection, such as diabetes or any immunocompromising disease. The most common disease states predisposing to sepsis are malignancies, diabetes mellitus, chronic liver disease, chronic renal failure, and the use of immunosuppressive agents. In addition, sepsis also is a common complication after major surgery, trauma, and extensive burns. Patients with indwelling catheters or devices are also at high risk. In most patients with sepsis, a source of infection can be identified, with the exception of patients who are immunocompromised with neutropenia, where an obvious source often is not found. Multiple sites of infection may occur in 6-15% of patients.

Causative microorganisms
Before the introduction of antibiotics in clinical practice, gram-positive bacteria were the principal organisms causing sepsis. More recently, gram-negative bacteria have become the key pathogens causing severe sepsis and septic shock. Anaerobic pathogens are becoming less important as a cause of sepsis. In one institution, the incidence of anaerobic bacteremia declined by 45% over a 15-year period. Fungal infections are the cause of sepsis in 0.8-10.2% of patients with sepsis, and their incidence appears to be increasing (see the image below).

An 8-year-old boy developed septic shock secondary to Blastomycosis pneumonia. Fungal infections are a rare cause of septic shock.

Respiratory tract infection and urinary tract infection are the most frequent causes of sepsis, followed by abdominal and soft tissue infections. Each organ system tends to be infected by a particular set of pathogens (see below). Lower respiratory tract infections are the cause of septic shock in 25% of patients, and the following are the common pathogens: Streptococcus pneumoniae Klebsiella pneumoniae Staphylococcus aureus Escherichia coli Legionella species Haemophilus species Anaerobes

Gram-negative bacteria Fungi Urinary tract infections are the cause of septic shock in 25% of patients, and the following are the common pathogens: E coli Proteus species Klebsiella species Pseudomonas species Enterobacter species Serratia species Soft tissue infections are the cause of septic shock in 15% of patients, and the following are the common pathogens: S aureus Staphylococcus epidermidis Streptococci Clostridia Gram-negative bacteria Anaerobes GI tract infections are the cause of septic shock in 15% all patients, and the following are the common pathogens: E coli Streptococcus faecalis Bacteroides fragilis Acinetobacter species Pseudomonas species Enterobacter species Salmonella species Infections of the male and female reproductive systems are the cause of septic shock in 10% of patients, and the following are the common pathogens: Neisseria gonorrhoeae Gram-negative bacteria Streptococci Anaerobes Foreign bodies leading to infections are the cause of septic shock in 5% of patients, and S aureus, S epidermidis, and fungi/yeasts (eg, Candida species) are the common pathogens. Miscellaneous infections are the cause of septic shock in 5% of patients, and Neisseria meningitidis is the most common cause of such infections (see the image below).

Gram stain of blood showing the presence of Neisseria meningitidis.

Risk factors
Risk factors for severe sepsis and septic shock include the following:

Extremes of age ( < 10 y and >70 y) Primary diseases (eg, liver cirrhosis, alcoholism, diabetes mellitus, cardiopulmonary diseases, solid malignancy, hematologic malignancy) Immunosuppression (eg, neutropenia, immunosuppressive therapy, corticosteroid therapy, IV drug abuse [see the image below], complement deficiencies, asplenia) Major surgery, trauma, burns Invasive procedures (eg, catheters, intravascular devices, prosthetic devices, hemodialysis and peritoneal dialysis catheters, endotracheal tubes) Previous antibiotic treatment Prolonged hospitalization

Other factors, such as childbirth, abortion, and malnutrition

A 28-year-old woman who was a previous intravenous drug user (human immunodeficiency virus [HIV] status: negative) developed septic shock secondary to bilateral pneumococcal pneumonia.

Epidemiology
United States statistics
Since the 1930s, studies have shown an increasing incidence of sepsis. In the United States, 200,000 cases of septic shock and 100,000 deaths per year occur from this disease. In 1 study, the incidence of bacteremic sepsis (both gram-positive and gram-negative) increased from 3.8 cases per 1000 admissions in 1970 to 8.7 per 1000 in 1987. Between 1980 and 1992, the incidence of nosocomial blood stream infection in 1 institution increased from 6.7 cases per 1000 discharges to 18.4 per 1000. The increase in the number of patients who are immunocompromised and an increasing use of invasive diagnostic and therapeutic devices predisposing to infection are major reasons for the increase in incidences of sepsis. A 2001 article reported the incidence, cost, and outcome of severe sepsis in the United States. [19] Analysis of a large sample from the major centers reported the incidence of severe sepsis as 3 cases per 1000 population and 2.26 cases per 100 hospital discharges. Out of these cases, 51.1% were admitted to an intensive care unit (ICU), and an additional 17.3% were cared for in an intermediate care or coronary care unit. Incidence ranged from 0.2 cases per 1000 admissions in children to 26.2 per 1000 in individuals older than 85 years. The mortality rate was 28.6% and ranged from 10% in children to 38.4% in elderly people. Severe sepsis resulted in an average cost of $2200 per case, with an annual total cost of $16.7 billion nationally.[19] The National Center for Health Statistics published a large retrospective analysis using the National Hospital Discharge Survey of 500 nonfederal US hospitals, which included more than 10 million cases of sepsis over a 22-year period. Septicemia accounted for 1.3% of all hospitalizations, and the incidence of sepsis increased 3-fold between 1979 and 2000, from 83 cases per 100,000 population per year to 240 per 100,000.[20]

The reasons for this growing incidence likely include an increasingly elderly population, increased recognition of disease, increased performance of invasive procedures and organ transplantation, increased use of immunosuppressive agents and chemotherapy, increased use of indwelling lines and devices, and increase in chronic diseases such as end-stage renal disease and HIV. Of note, in 1987, gram-positive organisms surpassed gram-negative organisms as the most common cause of sepsis, a position they still hold today.[20] Angus et al published linked data from several sources related to hospital discharge from all hospitals from 7 large states.[19] Hospital billing codes were used to identify patients with infection and organ dysfunction consistent with the definition of severe sepsis. This method yielded 300 annual cases per 100,000 population, 2.3% of hospital discharges, or an estimated 750,000 cases annually in the United States.[19] A more recent large survey of ED visits showed that severe sepsis accounts for more than 500,000 such visits annually (0.7% of total visits), that the majority of patients presented to EDs without an academic affiliation, and that the mean length of stay in the ED is approximately 5 hours.[21] ARDS has a reported incidence ranging from 1.5-8.4 cases per 100,000 population per year.[22] Subsequent studies report a higher incidence: 12.6 cases per 100,000 population per year for ARDS and 18.9 cases per 100,000 population per year for acute lung injury. The mortality rate from ARDS has been documented at approximately 50% in most studies.

International statistics
A Dutch surveillance study examined the incidence, causes, and outcome of sepsis in patients admitted to a university hospital. The investigators reported that the incidences of sepsis syndrome and septic shock were, respectively, 13.6 and 4.6 cases per 1000 persons.[23]

Age distribution for septic shock

Sepsis and septic shock occur at all ages. However, a strong correlation exists between advanced age and the incidence of septic shock, with a sharp increase in the number of cases in patients older than 50 years.[19, 24] At present, most sepsis episodes are observed in patients older than 60 years. Advanced age is a risk factor for acquiring nosocomial blood stream infection in the development of severe forms of sepsis. Compared with younger patients, elderly patients are more susceptible to sepsis, have less physiologic reserve to tolerate the insult from infection, and are more likely to have underlying diseases; all of these factors adversely affect survival. In addition, elderly patients are more likely to have atypical or nonspecific presentations with sepsis.

Sex distribution for septic shock

Epidemiologic data have shown that the age-adjusted incidence and mortality of septic shock are consistently greater in men. The percentage of male patients varies from 52% to 66%.However, it is not clear whether this difference can be attributed to an underlying higher prevalence of comorbid conditions, a higher incidence of lung infection in men, or whether women are inherently protected against the inflammatory injury that occurs in severe sepsis.[20, 19]

Incidence of septic shock by race

One large epidemiologic study showed that the risk of septicemia in the nonwhite population is almost twice that in the white population, with the highest risk accruing to black men. Potential reasons for this include issues relating to decreased access to health care and increased prevalence of underlying medical conditions.[20] A more recent large epidemiologic study tied the increased incidence of septic shock in the black population to increased rates of infection necessitating hospitalization and increased development of organ dysfunction.[25]

In this study, black patients with septic shock had a higher incidence of underlying diabetes and renal disease, which might explain the higher rates of infection. However, development of acute organ dysfunction was independent of comorbidities. Furthermore, the incidence of septic shock and severe invasive infection was higher in the young, healthy black population, which suggests a possible genetic predisposition to developing septic shock.

Prognosis
The mortality rate of severe sepsis and septic shock is frequently quoted as anywhere from 20% to 50%. In some studies, the mortality rate specifically caused by the septic episode itself is specified and is 14.320%. In recent years, mortality rates seem to have decreased. The National Center for Health Statistics study showed a reduction in hospital mortality rates from 28% to 18% for septicemia over the years; however, more overall deaths occurred due to the increased incidence of sepsis. The study by Angus et al, which likely more accurately reflects the incidence of severe sepsis and septic shock, reported a mortality rate of about 30%.[19] Given that there is a spectrum of disease from sepsis to severe sepsis to septic shock, mortality varies depending on the degree of illness. The following clinical characteristics are related to the severity of sepsis: An abnormal host response to infection Site and type of infection Timing and type of antimicrobial therapy Offending organism Development of shock Any underlying disease Patients long-term health condition Location of the patient at the time of septic shock Factors consistently associated with increased mortality in sepsis include advanced age, comorbid conditions, and clinical evidence of organ dysfunction.[19, 24] One study found that in the setting of suspected infection, just meeting SIRS criteria without evidence of organ dysfunction did not predict increased mortality; this emphasizes the importance of identifying organ dysfunction over the presence of SIRS criteria.[24] However, there is evidence to suggest that meeting increasing numbers of SIRS criteria is associated with increased mortality.[26] In patients with septic shock, several clinical trials have documented a mortality rate of 40-75%. The poor prognostic factors are advanced age, infection with a resistant organism, impaired host immune status, poor prior functional status, and continued need for vasopressors past 24 hours. Development of sequential organ failure, despite adequate supportive measures and antimicrobial therapy, is a harbinger of poor outcome. The mortality rates were 7% with SIRS, 16% with sepsis, 20% with severe sepsis, and 46% with septic shock.[27] A link between impaired adrenal function and higher septic shock mortality has been suggested. The adrenal gland is enlarged in patients with septic shock compared with controls. A study by Jung et al found that an absence of this enlargement, indicated by total adrenal volume of less than10 cm 3, was associated with increased 28-day mortality in patients with septic shock.[28] In 1995, a multicenter prospective study published by Brun-Buisson (1995) reported a mortality rate of 56% during ICU stays and 59% during hospital stays.[4]Twenty-seven percent of all deaths occurred within 2 days of the onset of severe sepsis, and 77% of all deaths occurred within the first 14 days. The risk factors for early mortality in this study were higher severity of illness score, the presence of 2 or more acute organ failures at the time of sepsis, shock, and a low blood pH (< 7.3). Studies have shown that appropriate antibiotic administration (ie, antibiotics that are effective against the organism that is ultimately identified) has a significant influence on mortality. For this reason, initiating

broad-spectrum coverage until the specific organism is cultured and antibiotic sensitivities are determined is important. The long-term use of statins appears to have a significant protective effect on sepsis, bacteremia, and pneumonia.[29] End-organ failure is a major contributor to mortality in sepsis and septic shock. The complications with the greatest adverse effect on survival are ARDS, DIC, and ARF. (See Clinical Presentation.) The frequency of ARDS in sepsis has been reported from 18-38%, the highest with gram-negative sepsis, ranging from 18-25%. Sepsis and multiorgan failure are the most common cause of death in ARDS patients. Approximately 16% of patients with ARDS died from irreversible respiratory failure. Most patients who showed improvement achieved maximal recovery by 6 months, with the lung function improving to 80-90% of predicted values. Controversy exists over the use of etomidate as an induction agent for patients with sepsis, with debate centered on its association with adrenal insufficiency. Sprung et al, in the CORTICUS study, reported that patients who received etomidate had a significantly higher mortality rate than those who did not receive etomidate.[30] However, the authors did not address the fact that those patients receiving etomidate required orotracheal intubation and thus were a sicker subset. There have been no studies to date that have prospectively evaluated the effect of single-dose etomidate on the mortality of septic shock. Although sepsis mortality is known to be high, its effect on the quality of life of survivors was previously not well characterized. New evidence shows that septic shock in elderly persons leads to significant longterm cognitive and functional disability compared with those hospitalized with nonsepsis conditions. Septic shock is often a major sentinel event that has lasting effects on the patients indep endence, reliance on family support, and need for chronic nursing home or institutionalized care. [31