You are on page 1of 23

Sepsis syndromes in adults: Epidemiology, definitions, clinical

presentation, diagnosis, and prognosis


Author
Remi Neviere, MD
Literature review current through: May 2016. | This topic last updated: May 20, 2016.
INTRODUCTION
Sepsis is a clinical syndrome that has physiologic, biologic, and biochemical abnormalities caused by a
dysregulated inflammatory response to infection. Sepsis and the inflammatory response that ensues can lead
to multiple organ dysfunction syndrome and death.
EPIDEMIOLOGY
Incidence
In the late 1970s, it was estimated that 164,000 cases of sepsis occurred in the United States (US) each year.
Since then, rates of sepsis in the US and elsewhere have dramatically increased as supported by the following
studies: One national database analysis of discharge records from hospitals in the US estimated an annual
rate of more than 1,665,000 cases of sepsis between 1979 and 2000. Another retrospective population-based
analysis reported increased rates of sepsis and septic shock from 13 to 78 cases per 100,000 between 1998
and 2009. A retrospective analysis of an international database reported a global incidence of 437 per 100,000
person-years for sepsis and 270 per 100,000 person-years for severe sepsis between the years 1995 and
2015, although this rate was not reflective of contributions from low- and middle-income countries.
The increased rate of sepsis is thought to be a consequence of advancing age, immunosuppression, and
multidrug-resistant infection. It is also likely to be due to the increased detection of early sepsis from
aggressive sepsis education and awareness campaigns, although this hypothesis is unproven. The incidence
of sepsis varies among the different racial and ethnic groups, but appears to be highest among AfricanAmerican males. The incidence is also greatest during the winter, probably due to the increased prevalence of
respiratory infections. Older patients 65 years of age account for the majority (60 to 85 percent) of all
episodes of sepsis; with an increasing aging population, it is likely that the incidence of sepsis will continue to
increase in the future.
Pathogens
The contribution of various infectious organisms to the burden of sepsis has changed over time. Gram positive
bacteria are most frequently identified in patients with sepsis in the United States, although the number of
cases of Gram negative sepsis remains substantial. The incidence of fungal sepsis has increased over the
past decade, but remains lower than bacterial sepsis.
Disease severity
The severity of disease appears to be increasing. In one retrospective analysis, the proportion of patients with
sepsis who also had at least one dysfunctional organ increased from 26 to 44 percent between 1993 and
2003. The most common manifestations of severe organ dysfunction were acute respiratory distress
syndrome, acute renal failure, and disseminated intravascular coagulation. However, it is unclear as to whether
the rising incidence of severe sepsis and septic shock reflects the overall increased incidence of sepsis or
altered definitions of sepsis over time.
PATHOPHYSIOLOGY
The normal host response to infection is a complex process that localizes and controls bacterial invasion, while
initiating the repair of injured tissue. It involves the activation of circulating and fixed phagocytic cells, as well
as the generation of proinflammatory and antiinflammatory mediators. Sepsis results when the response to
infection becomes generalized and involves normal tissues remote from the site of injury or infection.
Normal response to infection
The host response to an infection is initiated when innate immune cells, particularly macrophages, recognize
and bind to microbial components. This may occur by several pathways:

Pattern recognition receptors (PRRs) on the surface of host immune cells may recognize and bind to
the pathogen-associated molecular patterns (PAMPs) of microorganisms. There are three families of

PRRs: toll-like receptors (TLRs), nucleotide-oligomerization domain (NOD) leucine-rich repeat


proteins, and retinoic-acid-inducible gene I (RIG-I)-like helicases. Examples include the peptidoglycan
of Gram-positive bacteria binding to TLR-2 on host immune cells, as well as the lipopolysaccharide of
Gram-negative bacteria binding to TLR-4 and/or lipopolysaccharide-binding protein (CD14 complex)
on host immune cells.

The triggering receptor expressed on myeloid cell (TREM-1) and the myeloid DAP12-associating
lectin (MDL-1) receptors on host immune cells may recognize and bind to microbial components
The binding of immune cell surface receptors to microbial components has multiple effects:

The engagement of TLRs elicits a signaling cascade via the activation of cytosolic nuclear factor-kb
(NF-kb). Activated NF-kb moves from the cytoplasm to the nucleus, binds to transcription sites, and
induces activation of a large set of genes involved in the host inflammatory response, such as
proinflammatory cytokines (tumor necrosis factor alpha [TNFa], interleukin-1 [IL-1]), chemokines
(intercellular adhesion molecule-1 [ICAM-1], vascular cell adhesion molecule-1 [VCAM-1]), and nitric
oxide.

Polymorphonuclear leukocytes (PMNs) become activated and express adhesion molecules that
cause their aggregation and margination to the vascular endothelium. This is facilitated by the
endothelium expressing adherence molecules to attract leukocytes. The PMNs then go through a
series of steps (rolling, adhesion, diapedesis, and chemotaxis) to migrate to the site of injury. The
release of mediators by PMNs at the site of infection is responsible for the cardinal signs of local
inflammation: warmth and erythema due to local vasodilation and hyperemia, and protein-rich edema
due to increased microvascular permeability.
This process is highly regulated by a mixture of proinflammatory and antiinflammatory mediators secreted by
macrophages, which have been triggered and activated by the invasion of tissue by bacteria:

Proinflammatory mediators Important proinflammatory cytokines include tumor necrosis factor-alpha


(TNFa) and interleukin-1 (IL-1), which share a remarkable array of biological effects. The release of
TNFa is self-sustaining (ie, autocrine secretion), while non-TNF cytokines and mediators (eg, Il-1, IL2, IL-6, IL-8, IL-10, platelet activating factor, interferon, and eicosanoids) increase the levels of other
mediators (ie, paracrine secretion). The proinflammatory milieu leads to the recruitment of more
PMNs and macrophages.

Antiinflammatory mediators Cytokines that inhibit the production of TNFa and IL-1 are considered
antiinflammatory cytokines. Such antiinflammatory mediators suppress the immune system by
inhibiting cytokine production by mononuclear cells and monocyte-dependent T helper cells. However,
their effects may not be universally antiinflammatory. As examples, IL-10 and IL-6 both enhance B cell
function (proliferation, immunoglobulin secretion) and encourage the development of cytotoxic T cells.
The balance of proinflammatory and antiinflammatory mediators regulates the inflammatory processes,
including adherence, chemotaxis, phagocytosis of invading bacteria, bacterial killing, and phagocytosis of
debris from injured tissue. If the mediators balance each other and the initial infectious insult is overcome,
homeostasis will be restored. The end result will be tissue repair and healing.

Transition to sepsis
Sepsis occurs when the release of proinflammatory mediators in response to an infection exceeds the
boundaries of the local environment, leading to a more generalized response (algorithm 1). When a similar
process occurs in response to a noninfectious condition (eg, pancreatitis, trauma), the process is referred to as

systemic inflammatory response syndrome (SIRS). The focus of our review is on sepsis, but much of our
discussion is applicable to SIRS
Sepsis can be conceptualized as malignant intravascular inflammation.

Malignant because it is uncontrolled, unregulated, and self-sustaining

Intravascular because the blood spreads mediators that are usually confined to cell-to-cell interactions
within the interstitial space

Inflammatory because all characteristics of the septic response are exaggerations of the normal
inflammatory response
It is uncertain why immune responses that usually remain localized sometimes spread beyond the local
environment causing sepsis. The cause is likely multifactorial and may include the direct effects of the invading
microorganisms or their toxic products, release of large quantities of proinflammatory mediators, and
complement activation. In addition, some individuals may be genetically susceptible to developing sepsis.
Effects of microorganisms
Bacterial cell wall components (endotoxin, peptidoglycan, muramyl dipeptide, and lipoteichoic acid) and
bacterial products (staphylococcal enterotoxin B, toxic shock syndrome toxin-1, Pseudomonas exotoxin A, and
M protein of hemolytic group A streptococci) may contribute to the progression of a local infection to sepsis.
This is supported by the following observations regarding endotoxin, a lipopolysaccharide found in the cell wall
of gram negative bacteria:

Endotoxin is detectable in the blood of septic patients.

Elevated plasma levels of endotoxin are associated with shock and multiple organ dysfunction

Endotoxin reproduces many of the features of sepsis when it is infused into humans, including
activation of the complement, coagulation, and fibrinolytic systems. These effects may lead to
microvascular thrombosis and the production of vasoactive products, such as bradykinin.
Excess proinflammatory mediators
Large quantities of proinflammatory cytokines released in patients with sepsis may spill into the bloodstream,
contributing to the progression of a local infection to sepsis. These include tumor necrosis factor-alpha (TNFa)
and interleukin-1 (IL-1), whose plasma levels peak early and eventually decrease to undetectable levels. Both
cytokines can cause fever, hypotension, leukocytosis, induction of other proinflammatory cytokines, and the
simultaneous activation of coagulation and fibrinolysis. The evidence indicating that TNFa has an important
role in sepsis is particularly strong. It includes the following: circulating levels of TNFa are higher in septic
patients than non-septic patients with shock, infusion of TNFa produces symptoms similar to those observed in
septic shock, and anti-TNFa antibodies protect animals from lethal challenge with endotoxin. The high levels of
TNFa in sepsis are due in part to the binding of endotoxin to lipopolysaccharide (LPS)-binding protein and its
subsequent transfer to CD14 on macrophages, which stimulates TNFa release.
Complement activation
The complement system is a protein cascade that helps clear pathogens from an organism. It is described in
detail separately. There is evidence that activation of the complement system plays an important role in sepsis;
most notably, inhibition of the complement cascade decreases inflammation and improves mortality in animal
models:

In a rodent model of sepsis, a complement fragment 5a receptor (C5aR) antagonist decreased


mortality, inflammation, and vascular permeability. In contrast, increased production of complement
fragment 5a (C5a) and increased expression of C5aR enhanced neutrophil trafficking.

In several animal models of sepsis (lipopolysaccharide injection in mice and rats, Escherichia coli
infusion in dogs and baboons, and cecal ligation and puncture in mice), a complement fragment 1
(C1) inhibitor decreased mortality, inflammation, and vascular permeability.
Genetic susceptibility
The single nucleotide polymorphism (SNP) is the most common form of genetic variation. SNPs are stable
substitutions of a single base that have a frequency of more than one percent in at least one population and
are strewn throughout the genome, including promoters and intergenic regions. At most, only 2 to 3 percent
alter the function or expression of a gene. The total number of common SNPs in the human genome is
estimated to be more than 10 million. SNPs are used as genetic markers. Various SNPs are associated with
increased susceptibility to infection and poor outcomes. They include SNPs of genes that encode cytokines
(eg, TNF, lymphotoxin-alpha, IL-10, IL-18, IL-1 receptor antagonist, IL-6, and interferon gamma), cell surface
receptors (eg, CD14, MD2, toll-like receptors 2 and 4, and Fc-gamma receptors II and III), lipopolysaccharide
ligands (lipopolysaccharide binding protein, bactericidal permeability increasing protein), mannose-binding
lectin, heat shock protein 70, angiotensin I-converting enzyme, plasminogen activator inhibitor, and caspase12.

Systemic Effects Of Sepsis


Widespread cellular injury may occur when the immune response becomes generalized; cellular injury is the
precursor to organ dysfunction. The precise mechanism of cellular injury is not understood, but its occurrence
is indisputable as autopsy studies have shown widespread endothelial and parenchymal cell injury.
Mechanisms proposed to explain the cellular injury include: tissue ischemia (insufficient oxygen relative to
oxygen need), cytopathic injury (direct cell injury by proinflammatory mediators and/or other products of
inflammation), and an altered rate of apoptosis (programmed cell death).
Tissue ischemia
Significant derangement in metabolic autoregulation, the process that matches oxygen availability to changing
tissue oxygen needs, is typical of sepsis. In addition, microcirculatory and endothelial lesions frequently
develop during sepsis. These lesions reduce the cross-sectional area available for tissue oxygen exchange,
disrupting tissue oxygenation and causing tissue ischemia and cellular injury:

Microcirculatory lesions The microcirculatory lesions may be the result of imbalances in the
coagulation and fibrinolytic systems, both of which are activated during sepsis.

Endothelial lesions The endothelial lesions may be a consequence of interactions between


endothelial cells and activated polymorphonuclear leukocytes (PMNs). The increase in receptormediated neutrophil-endothelial cell adherence induces the secretion of reactive oxygen species, lytic
enzymes, and vasoactive substances (nitric oxide, endothelin, platelet-derived growth factor, and
platelet activating factor) into the extracellular milieu, which may injure the endothelial cells. LPS may
also induce cytoskeleton disruption and microvascular endothelial barrier integrity, in part, through
NOS, RhoA, and NF-B activation.
Another factor contributing to tissue ischemia in sepsis is that erythrocytes lose their normal ability to deform
within the systemic microcirculation. Rigid erythrocytes have difficulty navigating the microcirculation during
sepsis, causing excessive heterogeneity in the microcirculatory blood flow and depressed tissue oxygen flux.
Cytopathic injury
Proinflammatory mediators and/or other products of inflammation may cause sepsis induced mitochondrial
dysfunction (eg, impaired mitochondrial electron transport) via a variety of mechanisms, including direct
inhibition of respiratory enzyme complexes, oxidative stress damage, and mitochondrial DNA breakdown. Such
mitochondrial injury leads to cytotoxicity. There are several lines of evidence that support this belief:

Cell culture experiments have shown that endotoxin, TNFa, and nitric oxide cause
destruction and/or dysfunction of inner membrane and matrix mitochondrial proteins, followed by
degeneration of the mitochondrial ultrastructure. These changes are followed by measurable changes
in other cellular organelles by several hours. The end result is functional impairment of mitochondrial
electron transport, disordered energy metabolism, and cytotoxicity.

Studies using various animal models have found normal or supranormal oxygen tension in organs
during sepsis, suggesting impaired oxygen utilization at the mitochondrial level. As examples, a study
in resuscitated endotoxemic pigs found a supranormal ileomucosal oxygen tension, while a study in
endotoxemic rats found an elevated oxygen tension in the bladder epithelium.
The clinical relevance of mitochondrial dysfunction in septic shock was suggested by a study of 28 critically ill
septic patients who underwent skeletal muscle biopsy within 24 hours of admission to the ICU. Skeletal muscle
ATP concentrations, a marker of mitochondrial oxidative phosphorylation, were significantly lower in the 12
patients who died of sepsis than in 16 survivors. In addition, there was an association between nitric oxide
overproduction, antioxidant depletion, and severity of clinical outcome. Thus, cell injury and death in sepsis
may be explained by cytopathic (or histotoxic) anoxia, which is an inability to utilize oxygen even when present.
Mitochondria can be repaired or regenerated by a process called biogenesis. Mitochondrial biogenesis may
prove to be an important therapeutic target, potentially accelerating organ dysfunction and recovery from
sepsis.
Apoptosis
Apoptosis (also called programmed cell death) describes a set of regulated physiologic and morphologic
cellular changes leading to cell death. This is the principal mechanism by which senescent or dysfunctional
cells are normally eliminated and the dominant process by which inflammation is terminated once an infection
has subsided. During sepsis, proinflammatory cytokines may delay apoptosis in activated macrophages and
neutrophils, thereby prolonging or augmenting the inflammatory response and contributing to the development
of multiple organ failure. Sepsis also induces extensive lymphocyte and dendritic cell apoptosis, which alters
the immune response efficacy and results in decreased clearance of invading microorganisms. Apoptosis of
lymphocytes has been observed at autopsies in both animal and human sepsis. The extent of lymphocyte
apoptosis correlates with and the severity of the septic syndrome and the level of immunosuppression.

Apoptosis has been also observed in parenchymal cells, endothelial, and epithelial cells. Several experiments
studies show that inhibiting apoptosis protect animal from organ dysfunction and lethality.
Mitochondrial dysfunction in sepsis-induced multiple organ failure
In patients dying from sepsis, light and electron microscopy as well as immunohistochemical staining for
markers of cellular injury and stress, revealed that cell death was rare in sepsis-induced heart and renal
dysfunction. Moreover, the degree of cell injury death did not account for severity of sepsis-induced organ
dysfunction. The presence of subtle mitochondrial morphological changes could indicate that mitochondrial
energetic crisis (metabolic substrate utilization and mitochondrial OxPhos machinery perturbations) may be
involved in organ dysfunction, in the absence of cell death.
Immunosuppression
Clinical observations and animal studies suggest that the excess inflammation of sepsis may be followed by
immunosuppression. Among the evidence supporting this hypothesis, an observational study removed the
spleens and lungs from 40 patients who died with active severe sepsis and then compared them with the
spleens from 29 control patients and the lungs from 30 control patients. The median duration of sepsis was
four days. The secretion of proinflammatory cytokines (ie, tumor necrosis factor, interferon gamma, interleukin6, and interleukin-10) from the splenocytes of patients with severe sepsis was generally less than 10 percent
that of controls, following stimulation with either anti-CD3/anti-CD28 or lipopolysaccharide. Moreover, the cells
from the lungs and spleens of patients with severe sepsis exhibited increased expression of inhibitory
receptors and ligands, as well as expansion of suppressor cell populations, compared with cells from control
patients. The inability to secrete proinflammatory cytokines combined with enhanced expression of inhibitory
receptors and ligands suggests clinically relevant immunosuppression.
Organ-Specific Effects Of Sepsis
The cellular injury described above, accompanied by the release of proinflammatory and antiinflammatory
mediators, often progresses to organ dysfunction. No organ system is protected from the consequences of
sepsis; those listed included in this section are the organ systems that are most often involved. Multiple organ
dysfunction is common.
Circulation
Hypotension due to diffuse vasodilation is the most severe expression of circulatory dysfunction in sepsis. It is
probably an unintended consequence of the release of vasoactive mediators, whose purpose is to improve
metabolic autoregulation (the process that matches oxygen availability to changing tissue oxygen needs) by
inducing appropriate vasodilation. Mediators include the vasodilators prostacyclin and nitric oxide (NO), which
are produced by endothelial cells. NO is believed to play a central role in the vasodilation accompanying septic
shock, since NO synthase can be induced by incubating vascular endothelium and smooth muscle with
endotoxin. When NO reaches the systemic circulation, it depresses metabolic autoregulation at all of the
central, regional, and microregional levels of the circulation. In addition, NO may trigger an injury in the central
nervous system that is localized to areas that regulate autonomic control. Another factor that may contribute to
the persistence of vasodilation during sepsis is impaired compensatory secretion of antidiuretic hormone
(vasopressin). This hypothesis is supported by a study that found that plasma vasopressin levels were lower in
patients with septic shock than in patients with cardiogenic shock (3.1 versus 22.7 pg/mL), even though the
groups had similar systemic blood pressure.Vasodilation is not the only cause of hypotension during sepsis.
Hypotension may also be due to redistribution of intravascular fluid. This is a consequence of both increased
endothelial permeability and reduced arterial vascular tone leading to increased capillary pressure. In addition
to these diffuse effects of sepsis on the circulation, there are also localized effects:

In the central circulation (ie, heart and large vessels), decreased systolic and diastolic ventricular
performance due to the release of myocardial depressant substances is an early manifestation of
sepsis. Despite this, ventricular function may still be able to use the Frank Starling mechanism to
increase cardiac output, which is necessary to maintain the blood pressure in the presence of
systemic vasodilation. Patients with preexisting cardiac disease (eg, elderly patients) are often unable
to increase their cardiac output appropriately.

In the regional circulation (ie, small vessels leading to and within the organs), vascular
hyporesponsiveness (ie, inability to appropriately vasoconstrict) leads to an inability to appropriately
distribute systemic blood flow among organ systems. As an example, sepsis interferes with the
redistribution of blood flow from the splanchnic organs to the core organs (heart and brain) when
oxygen delivery is depressed.

The microcirculation (ie, capillaries) may be the most important target in sepsis. Sepsis is associated
with a decrease in the number of functional capillaries, which causes an inability to extract oxygen
maximally. Techniques including reflectance spectrophotometry and orthogonal polarization spectral
imaging have allowed in vivo visualization of the sublingual and gastric microvasculature. Compared
to normal controls or critically ill patients without sepsis, patients with severe sepsis have decreased

capillary density. This may be due to extrinsic compression of the capillaries by tissue edema,
endothelial swelling,and/or plugging of the capillary lumen by leukocytes or red blood cells (which lose
their normal deformability properties in sepsis).

At the level of the endothelium, sepsis induces phenotypic changes to endothelial cells. This occurs
through direct and indirect interactions between the endothelial cells and components of the bacterial
wall. These phenotypic changes may cause endothelial dysfunction, which is associated with
coagulation abnormalities reduced leukocytes, decreased red blood cell deformability, upregulation of
adhesion molecules, adherence of platelets and leukocytes, and degradation of the glycocalyx
structure. Diffuse endothelial activation leads to widespread tissue edema, which is rich in protein.
Microparticles from circulating and vascular cells also participate in the deleterious effects of sepsis-induced
intravascular inflammation.
Lung
Endothelial injury in the pulmonary vasculature during sepsis disturbs capillary blood flow and enhances
microvascular permeability, resulting in interstitial and alveolar pulmonary edema. Neutrophil entrapment within
the lung's microcirculation initiates and/or amplifies the injury in the alveolocapillary membrane. The result is
pulmonary edema, which creates ventilation-perfusion mismatch and leads to hypoxemia. Such lung injury is
prominent during sepsis, likely reflecting the lung's large microvascular surface area. Acute respiratory distress
syndrome is a manifestation of these effects.
Gastrointestinal tract
The circulatory abnormalities typical of sepsis may depress the gut's normal barrier function, allowing
translocation of bacteria and endotoxin into the systemic circulation (possibly via lymphatics, rather than the
portal vein) and extending the septic response. This is supported by animal models of sepsis, as well as a
prospective cohort study that found that increased intestinal permeability (determined from the urinary
excretion of orally administered lactulose and mannose) was predictive of the development of multiple organ
dysfunction syndrome.
Liver
The reticuloendothelial system of the liver normally acts as the first line of defense in clearing bacteria and
bacteria-derived products that have entered the portal system from the gut. Liver dysfunction can prevent the
elimination of enteric-derived endotoxin and bacteria-derived products, which precludes the appropriate local
cytokine response and permits direct spillover of these potentially injurious products into the systemic
circulation.
Kidney
Sepsis is often accompanied by acute renal failure. The mechanisms by which sepsis and endotoxemia lead to
acute renal failure are incompletely understood. Acute tubular necrosis due to hypoperfusion and/or hypoxemia
is one mechanism. However, systemic hypotension, direct renal vasoconstriction, release of cytokines (eg,
tumor necrosis factor), and activation of neutrophils by endotoxin and FMLP (a three amino acid [fMet-LeuPhe] chemotactic peptide in bacterial cell walls) may also contribute to renal injury. The likelihood of death is
increased in patients with sepsis who develop renal failure. It is not well understood why this occurs. One
factor may be the release of proinflammatory mediators as a result of leukocyte-dialysis membrane
interactions when hemodialysis is necessary. Use of biocompatible membranes can prevent these interactions
and may improve survival and the recovery of renal function In critically ill patients, it is thought that one
modality of renal replacement therapy, hemofiltration, removes circulating nephrogenic toxins that may play a
role in the pathogenesis of sepsis-induced acute renal failure. In this context, initial studies suggested that the
high volume mode of hemofiltration (high volume hemofiltration [HVHF]) offers benefit over conventional
dialysis (hemodialysis, peritoneal dialysis, continuous renal replacement therapy) for the treatment of acute
renal failure in sepsis. However, results from prospective, multicenter randomized studies suggest that there is
insufficient evidence for the routine use of HVHF over other forms of renal replacement therapy for critically ill
patients with septic shock
Nervous system
Central nervous system (CNS) complications occur frequently in septic patients, often before the failure of
other organs. The most common CNS complications are an altered sensorium (encephalopathy). The
pathogenesis of the encephalopathy is poorly defined. A high incidence of brain microabscesses was noted in
one study, but the significance of hematogenous infection as the principal mechanism remains uncertain
because of the heterogeneity of the observed pathology. CNS dysfunction has been attributed to changes in
metabolism and alterations in cell signalling due to inflammatory mediators. Dysfunction of the blood brain
barrier probably contributes, allowing increased leukocyte infiltration, exposure to toxic mediators, and active
transport of cytokines across the barrier. Mitochondrial dysfunction and microvascular failure both precede

functional CNS changes, as measured through somatosensory evoked potentials. In addition to these
neurological consequences of sepsis, there is growing recognition that the parasympathetic nervous system
may be a mediator of systemic inflammation during sepsis. This is supported by numerous observations in
various animal models. Afferent vagus nerve stimulation during sepsis increases the secretion of corticotropinreleasing hormone (CRH), ACTH, and cortisol; the last effect can be suppressed by subdiaphragmatic
vagotomy. Parasympathetic tone affects thermoregulation, as experimental vagotomy attenuates the
hyperthermic response to IL-1. Efferent parasympathetic activity, mediated by acetylcholine, has an
antiinflammatory effect on the cytokine profile, with decreased in vitro expression of the proinflammatory
cytokines TNF, IL-1, IL-6 and IL-18. And, external vagal stimulation prevents the onset of shock following
vagotomy, while an acetylcholine receptor agonist diminishes the pathologic response to sepsis
DEFINITIONS
Sepsis exists on a continuum of severity ranging from infection and bacteremia to sepsis and septic shock,
which can lead to multiple organ dysfunction syndrome (MODS) and death. The definitions of sepsis and septic
shock have rapidly evolved since the early 1990s. The systemic inflammatory response syndrome (SIRS) is no
longer included in the definition since it is not always caused by infection. The definitions for sepsis that we
provide below reflect expert opinion from task forces generated by national societies including the Society of
Critical Care Medicine (SCCM), the European Society of Intensive Care Medicine (ESICM), the American
Thoracic Society (ATS), the American College of Chest physicians (ACCP), and the Surgical Infection Society
(SIS). Importantly, such definitions are not diagnostic of sepsis since they do not comprehensively include
specific criteria for the identification of infection.
Early sepsis Infection and bacteremia may be early forms of infection that can progress to sepsis.
However, there is no formal definition of early sepsis. Nonetheless, despite the lack of definition, monitoring
those suspected of having sepsis is critical for its prevention.
Infection and bacteremia All patients with infection or bacteremia are at risk of developing sepsis and
represent early phases in the continuum of sepsis severity:
Infection is defined as the invasion of normally sterile tissue by organisms resulting in infectious
pathology.
Bacteremia is the presence of viable bacteria in the blood.
Identification of early sepsis
Societal guidelines place emphasis on the early identification of infected patients who may go on to develop
sepsis as a way to decrease sepsis-associated mortality. The 2016 SCCM/ESICM task force have described
an assessment score for patients outside the intensive care unit as a way to facilitate the identification of
patients potentially at risk of dying from sepsis [25-27]. This score is a modified version of the Sequential
(Sepsis-related) Organ Failure Assessment score (SOFA) called the quickSOFA (qSOFA). The qSOFA only
has three components that are each allocated one point: respiratory rate 22/minute, altered mentation, and
systolic blood pressure 100 mmHg. A score 2 is associated with poor outcomes due to sepsis. However, the
ability of qSOFA to predict death from sepsis requires prospective evaluation before it can be routinely used for
this purpose. Importantly, this qSOFA score is different from the full SOFA score which is part of the
2016 SCCM/EISCM definition of sepsis, the details of which are described separately.
Sepsis A 2016 SCCM/EISCM task force has defined sepsis as life-threatening organ dysfunction caused by
a dysregulated host response to infection:
Organ dysfunction Organ dysfunction is defined by the 2016 SCCM/ESICM task force as an
increase of two or more points in the SOFA score. The validity of this score was derived from critically-ill
patients with suspected sepsis by interrogating over a million intensive care unit (ICU) electronic health
record encounters from ICUs both inside and outside the United States [25-27]. ICU patients were
suspected as having infection if body fluids were cultured and they received antibiotics. Predictive scores
(SOFA, systemic inflammatory response syndrome [SIRS], and logistic Organ Dysfunction System
[LODS]) were compared for their ability to predict mortality. Among critically ill patients with suspected
sepsis, the predictive validity of the SOFA score for in-hospital mortality was superior to that for the SIRS
criteria (area under the receiver operating characteristic curve 0.74 versus 0.64). Patients who fulfill
these criteria have a predicted mortality of 10 percent. Although the predictive capacity of SOFA and
LODS were similar, SOFA is considered easier to calculate, and was therefore recommended by the task
force.
Importantly, the SOFA score is an organ dysfunction score. It is not diagnostic of sepsis nor does it
identify those whose organ dysfunction is truly due to infection but rather helps identify patients who
potentially have a high risk of dying from infection. In addition, it does not determine individual treatment

strategies nor does it predict mortality based upon demographics (eg, age) or underlying condition (eg,
stem cell transplant recipient versus postoperative patient). SOFA and other predictive scores are
discussed separately. (See "Predictive scoring systems in the intensive care unit", section on 'Sequential
(sepsis-related) Organ Failure Assessment (SOFA)'.)
Infection There are no clear guidelines to help the clinician identify the presence of infection or to
causally link an identified organism with sepsis. In our experience, for this component of the diagnosis,
the clinician is reliant upon clinical suspicion derived from the signs and symptoms of infection as well as
supporting radiologic and microbiologic data and response to therapy.
The term severe sepsis, which originally referred to sepsis that was associated with tissue hypoperfusion (eg,
elevated lactate, oliguria) or organ dysfunction (eg, elevated creatinine, coagulopathy) [14,23], is no longer
used since the 2016 sepsis and septic shock definitions include patients with evidence of tissue hypoperfusion
and organ dysfunction.
Septic shock Septic shock is a type of vasodilatory or distributive shock. Septic shock is defined as sepsis
that has circulatory, cellular, and metabolic abnormalities that are associated with a greater risk of mortality
than sepsis alone [25]. Clinically, this includes patients who fulfill the criteria for sepsis (see 'Sepsis' above)
who, despite adequate fluid resuscitation, require vasopressors to maintain a mean arterial pressure (MAP)
65 mmHg and have a lactate >2 mmol/L (>18 mg/dL). Per predictions from the SOFA score, patients who
fulfill these criteria for septic shock have a higher mortality than those who do not (40 versus 10 percent).
Others Multiple organ dysfunction syndrome (MODS) and systemic inflammatory response syndrome
(SIRS) are terms frequently used in practice that need to be distinguished from sepsis.
Multiple organ dysfunction syndrome Multiple organ dysfunction syndrome (MODS) refers to progressive
organ dysfunction in an acutely ill patient, such that homeostasis cannot be maintained without intervention. It
is at the severe end of the severity of illness spectrum of both infectious (sepsis, septic shock) and
noninfectious conditions (eg, SIRS from pancreatitis). MODS can be classified as primary or secondary:
Primary MODS is the result of a well-defined insult in which organ dysfunction occurs early and can be
directly attributable to the insult itself (eg, renal failure due to rhabdomyolysis).
Secondary MODS is organ failure that is not in direct response to the insult itself, but is a consequence
of the host's response (eg, acute respiratory distress syndrome in patients with pancreatitis).
There are no universally accepted criteria for individual organ dysfunction in MODS. However, progressive
abnormalities of the following organ-specific parameters are commonly used to diagnose MODS and are also
used in scoring systems (eg, SOFA or LODS) to predict ICU mortality :
Respiratory Partial pressure of arterial oxygen (PaO2)/fraction of inspired oxygen (FiO2) ratio
Hematology Platelet count
Liver Serum bilirubin
Renal Serum creatinine (or urine output)
Brain Glasgow coma score
Cardiovascular Hypotension and vasopressor requirement
In general, the greater the number of organ failures, the higher the mortality, with the greatest risk being
associated with respiratory failure requiring mechanical ventilation.
Systemic inflammatory response syndrome The use of systemic inflammatory response syndrome
(SIRS) criteria to identify those with sepsis has fallen out of favor since it is considered by many experts that
SIRS criteria are present in many hospitalized patients who do not develop infection, and their ability to predict
death is poor when compared with other scores such as the SOFA score [27,31,32]. SIRS is considered a
clinical syndrome that is a form of dysregulated inflammation. It was previously defined as two or more
abnormalities in temperature, heart rate, respiration, or white blood cell count [23]. SIRS may occur in several
conditions related, or not, to infection. Noninfectious conditions classically associated with SIRS include
autoimmune disorders, pancreatitis, vasculitis, thromboembolism, burns, or surgery.
RISK FACTORS
The importance of identifying risk factors for sepsis was highlighted in one epidemiologic study that reported
that risk factors for septic shock were the fifth leading cause of years of productive life lost due to premature
mortality [33]. Risk factors for sepsis include the following:
Intensive care unit admission Approximately 50 percent of intensive care unit (ICU) patients have a
nosocomial infection and are, therefore, intrinsically at high risk for sepsis [44].
Bacteremia Patients with bacteremia often develop systemic consequences of infection. In a study of
270 blood cultures, 95 percent of positive blood cultures were associated with sepsis, severe sepsis, or
septic shock [39].

Advanced age (65 years) The incidence of sepsis is disproportionately increased in older adult
patients and age is an independent predictor of mortality due to sepsis. Moreover, older adult nonsurvivors tend to die earlier during hospitalization and older adult survivors more frequently require
skilled nursing or rehabilitation after hospitalization [40].
Immunosuppression Comorbidities that depress host-defense (eg, neoplasms, renal failure, hepatic
failure, AIDS, asplenism) and immunosuppressant medications are common among patients with sepsis,
severe sepsis, or septic shock.
Diabetes and cancer Diabetes and some cancers may alter the immune system, result in an
elevated risk for developing sepsis, and increase the risk of nosocomial sepsis.
Community acquired pneumonia Severe sepsis and septic shock develop in approximately 48 and
5 percent, respectively, of patients hospitalized with community-acquired pneumonia [41].
Previous hospitalization Hospitalization is thought to induce an altered human microbiome,
particularly in patients who are treated with antibiotics. Previous hospitalization has been associated with
a three-fold increased risk of developing severe sepsis in the subsequent 90 days [42]. Patients with
hospitalizations for infection-related conditions, especially Clostridium difficile infection, are at greatest
risk.
Genetic factors Both experimental and clinical studies have confirmed that genetic factors can
increase the risk of infection. In few cases, monogenic defects underlie vulnerability to specific infection,
but genetic factors are typically genetic polymorphisms. Genetic studies of susceptibility to infection have
initially focused on defects of antibody production, or a lack of T cells, phagocytes, natural killer cells, or
complement. Recently, genetic defects have been identified that impair recognition of pathogens by the
innate immune system, increasing susceptibility to specific classes of microorganisms [43].
CLINICAL PRESENTATION
Patients with suspected or documented sepsis typically present with hypotension, tachycardia, fever, and
leukocytosis. As severity worsens, signs of shock (eg, cool skin and cyanosis) and organ dysfunction develop
(eg, oliguria, acute kidney injury, altered mental status) [14,23]. Importantly, the presentation is nonspecific
such that many other conditions (eg, pancreatitis, acute respiratory distress syndrome) may present similarly.
Detailed discussion of the clinical features of shock are discussed separately.
Symptoms and signs The symptoms and signs of sepsis are nonspecific but may include the following:
Symptoms and signs specific to an infectious source (eg, cough dyspnea may suggest pneumonia,
pain and purulent exudate in a surgical wound may suggest an underlying abscess)
Arterial hypotension (eg, systolic blood pressure [SBP] <90 mmHg, mean arterial pressure [MAP] <70
mmHg, an SBP decrease >40 mmHg, or less than two standard deviations below normal for age)
Temperature >38.3 or <36C
Heart rate >90 beats/min or more than two standard deviations above the normal value for age
Tachypnea, respiratory rate >20 breaths/min
Altered mental status
Ileus (absent bowel sounds; often an end-stage sign of hypoperfusion)
Decreased capillary refill, cyanosis, or mottling (may indicate shock)
Laboratory signs Similarly, laboratory features are nonspecific and may be associated with abnormalities
due to the underlying cause of sepsis or to tissue hypoperfusion or organ dysfunction from sepsis. They
include the following:
Leukocytosis (white blood cell [WBC] count >12,000 microL1) or leukopenia (WBC count <4000
microL1)
Normal WBC count with greater than 10 percent immature forms
Hyperglycemia (plasma glucose >140 mg/dL or 7.7 mmol/L) in the absence of diabetes
Plasma C-reactive protein more than two standard deviations above the normal value
Plasma procalcitonin more than two standard deviations above the normal value (not routinely
performed in many centers)
Arterial hypoxemia (arterial oxygen tension [PaO2]/fraction of inspired oxygen [FiO2] <300)
Acute oliguria (urine output <0.5 mL/kg/hour for at least two hours despite adequate fluid resuscitation)
Creatinine increase >0.5 mg/dL or 44.2 micromol/L
Coagulation abnormalities (international normalized ratio [INR] >1.5 or activated partial thromboplastin
time [aPTT] >60 seconds)
Thrombocytopenia (platelet count <100,000 microL1)
Hyperbilirubinemia (plasma total bilirubin >4 mg/dL or 70 micromol/L)
Hyperlactatemia (higher than the laboratory upper limit of normal)

Adrenal dysfunction (eg, hypernatremia, hypokalemia), and the euthyroid sick syndrome can also be
found in sepsis
Imaging There are no radiologic signs that are specific to the identification of sepsis other than those
associated with infection in a specific site (eg, pneumonia on chest radiography, fluid collection on computed
tomography of the abdomen).
Microbiology The identification of an organism in culture in a patient who fulfills the definition of sepsis
(see 'Sepsis' above) is highly supportive of the diagnosis of sepsis but is not necessary. The rationale behind
its lack of inclusion in the diagnostic criteria for sepsis is that a culprit organism is frequently not identified in up
to 50 percent of patients who present with sepsis nor is a positive culture required to make a decision
regarding treatment with empiric antibiotics [45].
DIAGNOSIS
A limitation of the definitions above is that they cannot identify patients whose organ dysfunction is truly
secondary to an underlying infection. Thus, a constellation of clinical, laboratory, radiologic, physiologic, and
microbiologic data is typically required for the diagnosis of sepsis and septic shock. The diagnosis is often
made empirically at the bedside upon presentation, or retrospectively when followup data returns (eg, positive
blood cultures in a patient with endocarditis) or a response to antibiotics is evident. Importantly, the
identification of a culprit organism, although preferred, is not always feasible since in many patients no
organism is ever identified. In some patients this may be because they have been partially treated with
antibiotics before cultures are obtained.
Although septic shock has a specific hemodynamic profile on pulmonary artery catheterization (PAC) (table 1),
PACs are difficult to interpret and rarely placed in patients with suspected sepsis.
PROGNOSIS
In-hospital morbidity and mortality Sepsis has a high mortality rate. Rates depend upon how the data are
collected but estimates range from 10 to 52 percent. Data derived from death certificates report that sepsis is
responsible for 6 percent of all deaths while administrative claims data suggest higher rates]. Mortality rates
increase linearly according to the disease severity of sepsis [31]. In one study, the mortality rates of SIRS,
sepsis, severe sepsis, and septic shock were 7, 16, 20, and 46 percent, respectively [21]. In another study, the
mortality associated with sepsis was 10 percent while that associated with septic shock was 40 percent [25].
Mortality appears to be lower in younger patients (<44 years) without comorbidities (<10 percent) [4].
Several studies have reported decreasing mortality rates over time. As an example, a 12-year study of 101,064
patients with severe sepsis and septic shock from 171 intensive care units (ICUs) in Australia and New
Zealand reported a 50 percent risk reduction (from 35 to 18 percent) in in-hospital mortality from 2000 to 2012
[4]. This persisted after adjusting for multiple variables including underlying disease severity, comorbidities,
age, and the rise in incidence of sepsis over time. This suggested that the reduction in mortality observed in
this study was less likely due to the increased detection of early sepsis and possibly due to improved
therapeutic strategies for sepsis. However, despite improved compliance with practice guidelines for the
treatment of sepsis (also known as sepsis bundles), compliance rates vary and there is conflicting evidence as
to whether sepsis bundles truly improve mortality [
During hospital admission, sepsis may increase the risk of acquiring a subsequent hospital-related infection.
One prospective observational study of 3329 admissions to the ICU reported that ICU-acquired infections
occurred in 13.5 percent admissions of patients with sepsis compared with 15 percent of non-sepsis ICU
admissions [63]. Patients admitted with sepsis also developed more ICU-acquired infections including infection
with opportunistic pathogens, hinting at possible immune suppression. In patients with a sepsis admission
diagnosis, secondary infections were mostly catheter-related blood stream infections (26 percent), pneumonia
(25 percent), or abdominal infections (16 percent), compared with patients with non-sepsis admission where
pneumonia was the most common ICU-acquired infection (48 percent). In both groups, patients who
developed ICU-acquired infection were more severely ill on admission (eg, higher Acute Physiologic and
Chronic Health Evaluation [APACHE] IV and Sequential Organ Failure Assessment scores and more shock on
admission) and had higher mortality at day 60. However, the contribution of developing a secondary infection
was small.
Long-term prognosis Following discharge from the hospital, sepsis carries an increased risk of death (up
to 20 percent) as well as an increased risk of further sepsis and recurrent hospital admissions (up to 10
percent are readmitted). Most deaths occur within the first six months but the risk remains elevated at two
years [64-71]. Patients who survive sepsis are more likely to be admitted to acute care and/or long term care
facilities in the first year after the initial hospitalization, and also appear to have a persistent decrement in their

quality of life [49,66-68]. The most common diagnoses associated with readmission at 90 days in one database
analysis of 3494 hospital admissions included heart failure, pneumonia, acute exacerbations of chronic
obstructive pulmonary disease, and urinary tract infections [69]. Higher rates of readmission with subsequent
infection and sepsis may be associated with previous hospitalization for an infection, particularly infection
with clostridium difficile [42,72].
Prognostic factors Clinical characteristics that impact the severity of sepsis and, therefore, the outcome
include the host's response to infection, the site and type of infection, and the timing and type of antimicrobial
therapy.
Host-related Anomalies in the host's inflammatory response may indicate increased susceptibility to severe
disease and mortality. As examples, the failure to develop a fever (or hypothermia) and the development of
leukopenia, thrombocytopenia, hyperchloremia, a patient's comorbidities, age, and hypocoagulability have all
been associated with poor outcomes [73-78]. Failure to develop a fever (defined as a temperature below
35.5C) was more common among non-survivors of sepsis than survivors (17 versus 5 percent) in one study of
519 patients with sepsis [73]. Leukopenia (a white blood cell count less than 4000/mm3) was similarly more
frequent among non-survivors than survivors (15 versus 7 percent) in a study of 612 patients with Gram
negative sepsis [75] and a platelet count <100,000/mm3 was found to be an early prognostic marker of 28-day
mortality in another study of 1486 patients with septic shock [78]. In another retrospective analysis of critically
ill septic patients, hyperchloremia (Cl 110 mEq/L) at 72 hours after ICU admission was independently
associated with an increase in all-cause hospital mortality [77].A patient's comorbidities and functional health
status are also important determinants of outcome in sepsis [73]. Risk factors for mortality include new-onset
atrial fibrillation [79], an age above 40 years [12], and comorbidities such as AIDS [80], liver disease [81],
cancer [82], alcohol dependence [81], and/or immune suppression [80,83]. Age is probably a risk factor for
mortality because of its association with comorbid illnesses, impaired immunologic responses, malnutrition,
increased exposure to potentially resistant pathogens in nursing homes, and increased utilization of medical
devices, such as indwelling catheters and central venous lines [1,12,84]. Inability to clot has also been
associated with increased mortality. In one prospective study of 260 patients with severe sepsis, indicators of
hypocoagulability using standard and functional levels of fibrinogen, were associated with a six-fold increase in
the risk of death, particularly in patients treated with hydroxyethyl starch [76].
Site of infection The site of infection in patients with sepsis may be an important determinant of outcome,
with sepsis from a urinary tract infection generally being associated with the lowest mortality rates [73,85]. One
study found that mortality from sepsis was 50 to 55 percent when the source of infection was unknown,
gastrointestinal, or pulmonary, compared with only 30 percent when the source of infection was the urinary
tract [85]. Another retrospective, multicenter cohort study of nearly 8000 patients with septic shock reported
similar results with the highest mortality in those with sepsis from ischemic bowel (78 percent) and the lowest
rates in those with obstructive uropathy-associated urinary tract infection (26 percent) [54]. Approximately 50
percent of patients with severe sepsis are bacteremic at the time of diagnosis according to one study [86]. This
is consistent with a study of 85,750 hospital admissions, which found that the incidence of positive blood
cultures increased along a continuum, ranging from 17 percent of patients with sepsis to 69 percent with septic
shock [87]. However, the presence or absence of a positive blood culture does not appear to influence the
outcome, suggesting that prognosis is more closely related to the severity of sepsis than the severity of the
underlying infection [87,88].
Type of infection Sepsis due to nosocomial pathogens has a higher mortality than sepsis due to
community-acquired pathogens [89,90]. Increased mortality is associated with bloodstream infections due to
methicillin-resistant staphylococcus aureus (odds ratio 2.70, 95% CI 2.03-3.58), non-candidal fungus (odds
ratio 2.66, 95% CI 1.27-5.58), candida (odds ratio 2.32 95% CI 1.21-4.45), methicillin-sensitive staphylococcus
aureus (odds ratio 1.9, 95% CI 1.53-2.36), and pseudomonas (odds ratio 1.6, 95% CI 1.04-2.47), as well as
polymicrobial infections (odds ratio 1.69, 95% CI 1.24-2.30) [89,91]. When bloodstream infections become
severe (ie, severe sepsis or septic shock), the outcome is similar regardless of whether the pathogens are
Gram-negative or Gram-positive bacteria [35,92].
Antimicrobial therapy Studies have shown that the early administration of appropriate antibiotic therapy
(ie, antibiotics to which the pathogen is sensitive) has a beneficial impact on bacteremic sepsis [75,88]. In one
report, early institution of adequate antibiotic therapy was associated with a 50 percent reduction in the
mortality rate compared to antibiotic therapy to which the infecting organisms were resistant [75]. In contrast,
prior antibiotic therapy (ie, antibiotics within the past 90 days) may be associated with increased mortality, at
least among patients with Gram negative sepsis [93]. This is probably because patients who have received
prior antibiotic therapy are more likely to have higher rates of antibiotic resistance, making it less likely that

appropriate antibiotic therapy will be chosen empirically. Empiric antibiotic regimens for patients with suspected
sepsis are discussed separately.
Restoration of perfusion Failure to aggressively try to restore perfusion early (ie, failure to initiate early
goal-directed therapy) may also be associated with mortality [94]. A severely elevated lactate (>4 mmol/L) is
associated with a poor prognosis in patients with sepsis with one study reporting a mortality of 78 percent in a
population of critically ill patients, a third of whom had sepsis [95]. Restoration of perfusion is discussed in
detail separately.
SUMMARY AND RECOMMENDATIONS
Sepsis is the consequence of a dysregulated inflammatory response to an infectious insult. The
severity and rates of sepsis have dramatically increased with reports suggesting rates as high as 437
and 270 per 100,000 person-years for sepsis and severe sepsis, respectively. Gram positive bacteria are
the pathogens that are most commonly isolated from patients with sepsis. (See 'Introduction' above
and 'Epidemiology' above.)
Sepsis exists on a continuum of severity ranging from infection (invasion of sterile tissue by organisms)
and bacteremia (bacteria in the blood) to sepsis and septic shock, which can lead to multiple organ
dysfunction syndrome (MODS) and death. A 2016 task force from the Society of Critical Care Medicine
and European Society of Intensive Care Medicine (SCCM/EISCM) define sepsis and septic shock as the
following (see 'Definitions' above):
Sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to
infection; organ dysfunction is defined as an increase of two or more points in the sequential
(sepsis-related) organ failure assessment (SOFA) score. The systemic inflammatory response
syndrome (SIRS) criteria are no longer used to identify those with sepsis.
Septic shock is defined as sepsis that has circulatory, cellular, and metabolic abnormalities that
are associated with a greater risk of mortality than sepsis alone; these abnormalities can be
clinically identified as patients who fulfill the criteria for sepsis who, despite adequate fluid
resuscitation, require vasopressors to maintain a mean arterial pressure (MAP) 65 mmHg and
have a lactate >2 mmol/L (>18 mg/dL).
Risk factors for sepsis include intensive care unit (ICU) admission, a nosocomial infection, bacteremia,
advanced age, immunosuppression, previous hospitalization (in particular hospitalization associated with
infection), and community-acquired pneumonia. Genetic defects have also been identified that may
increase susceptibility to specific classes of microorganisms. (See 'Risk factors' above.)
Patients with suspected or documented sepsis typically present with hypotension, tachycardia, fever,
and leukocytosis. As severity worsens, signs of shock (eg, cool skin and cyanosis) and organ dysfunction
develop (eg, oliguria, acute kidney injury, altered mental status) [14,23]. Importantly, the presentation is
nonspecific such that many other conditions (eg, pancreatitis, acute respiratory distress syndrome) may
present similarly. (See 'Clinical presentation' above.)
A constellation of clinical, laboratory, radiologic, physiologic, and microbiologic data is typically required
for the diagnosis of sepsis and septic shock. The diagnosis is often made empirically at the bedside upon
presentation, or retrospectively when follow-up data return or a response to antibiotics is evident.
Importantly, the identification of a culprit organism, although preferred, is not always feasible since many
patients have been partially treated with antibiotics before cultures are obtained. (See 'Diagnosis' above.)
Sepsis has a high mortality rate that appears to be decreasing. Estimates range from 10 to 52 percent
with rates increasing linearly according to the disease severity of sepsis. Following discharge from the
hospital, sepsis carries an increased risk of death as well as an increased risk of further sepsis and
recurrent hospital admissions. Poor prognostic factors include the inability to mount a fever, leukopenia,
age >40 years, certain comorbidities (eg, AIDS, hepatic failure, cirrhosis, cancer, alcohol dependence,
immunosuppression), a non-urinary source of infection, a nosocomial source of infection, and
inappropriate or late antibiotic coverage. (See 'Prognosis' above.)

EVALUATION AND MANAGEMENT OF SUSPECTED SEPSIS AND SEPTIC SHOCK


IN ADULTS
INTRODUCTION
Sepsis is a clinical syndrome characterized by systemic inflammation due to infection. There is a
continuum of severity ranging from sepsis to severe sepsis and septic shock. Over 1,665,000
cases of sepsis occur in the United States each year, with a mortality rate up to 50 percent. Even
with optimal treatment, mortality due to severe sepsis or septic shock is approximately 40
percent and can exceed 50 percent in the sickest patients. In this topic review, the management
of severe sepsis and septic shock is discussed. Definitions, diagnosis, pathophysiology, and
investigational therapies for sepsis, as well as management of sepsis in the asplenic patient are
reviewed separately.

THERAPEUTIC PRIORITIES
The early administration of fluids and antibiotics is the cornerstone of management for patients
with severe sepsis and septic shock.
Therapeutic priorities for patients with severe sepsis or septic shock include:

Early initiation of supportive care to correct physiologic abnormalities, such as


hypoxemia and hypotension

Distinguishing sepsis from systemic inflammatory response syndrome (SIRS) (table 1)


because, if an infection exists, it must be identified and treated as soon as possible (table
2). This may require appropriate antibiotics as well as a surgical procedure (eg,
drainage).

EARLY MANAGEMENT
The first priority in any patient with severe sepsis or septic shock is stabilization of their airway
and breathing. Next, perfusion to the peripheral tissues should be restored and antibiotics
administered.
Stabilize respiration Supplemental oxygen should be supplied to all patients with sepsis and
oxygenation should be monitored continuously with pulse oximetry. Intubation and mechanical
ventilation may be required to support the increased work of breathing that typically
accompanies sepsis, or for airway protection since encephalopathy and a depressed level of
consciousness frequently complicate sepsis [11,12].
The choice and use of sedative and induction agents (eg, etomidate, ketamine) used to intubate
patients with severe sepsis or septic shock are discussed separately. Other aspects of intubation
and mechanical ventilation are similarly described elsewhere. Chest radiographs and arterial
blood gas analysis should be obtained following initial stabilization. These studies are used in
combination with other clinical parameters to diagnose acute respiratory distress syndrome
(ARDS), which frequently complicates sepsis.

Assess perfusion Once the patient's respiratory status has been stabilized, the adequacy of
perfusion should be assessed. Hypotension is the most common sign but critical hypoperfusion
can also occur in the absence of hypotension, especially during early sepsis. Clinical signs of
impaired perfusion include the following:

Hypotension Hypotension is the most common indicator that perfusion is inadequate


(eg, systolic blood pressure [SBP] <90 mmHg, mean arterial pressure <70 mmHg,
decrease in SBP >40 mmHg). Therefore, it is important that the blood pressure be
assessed early and often. Because a sphygmomanometer may be unreliable in
hypotensive patients, an arterial catheter may be inserted if blood pressure is labile or
restoration of arterial perfusion pressures is expected to be a protracted process [8].
Attempts to insert an arterial line should not delay the prompt management of shock.

Signs of poor end-organ perfusion Warm, flushed skin may be present in the early
phases of sepsis. As sepsis progresses to shock, the skin may become cool due to
redirection of blood flow to core organs. Additional signs of hypoperfusion include
tachycardia >90 per min, obtundation or restlessness, and oliguria or anuria. These
findings may be modified by preexisting disease or medications. As examples, older
patients, diabetic patients, and patients who take beta-blockers may not exhibit an
appropriate tachycardia as blood pressure falls. In contrast, younger patients frequently
develop a severe and prolonged tachycardia and fail to become hypotensive until acute
decompensation later occurs, often suddenly. Patients with chronic hypertension may
develop critical hypoperfusion at a higher blood pressure than healthy patients (ie,
relative hypotension).

Elevated lactate An elevated serum lactate (eg, >2 mmol/L) can be a manifestation
of organ hypoperfusion in the presence or absence of hypotension and is an important
component of the initial evaluation, since elevated lactate is associated with poor
prognosis [9,13-15]. A serum lactate level 4 mmol/L is consistent with, but not
diagnostic of, severe sepsis. Additional laboratory studies that help characterize the
severity of sepsis include a low platelet count, and elevated international normalized
ratio, creatinine, and bilirubin. Values for laboratory parameters that suggest severe
sepsis are described separately.

Other Tests that combine output from many organs (eg, arterial lactate) may obscure
the presence of significant ischemia in an individual organ [16]. Gastric tonometry
indirectly measures perfusion to the gut by estimating the gastric mucosal PCO 2. It can
be used to detect gut hypoxia by calculating the gastric to arterial PCO 2 gap [16-18]. But,
gastric tonometry is not widely available and it is uncertain whether it can successfully
guide therapy. Additional studies and clinical experience are needed.
Establish venous access Venous access should be established as soon as possible in
patients with suspected sepsis. While peripheral venous access may be sufficient in some
patients, particularly for initial resuscitation, the majority will require central venous access at
some point during their course. A central venous catheter (CVC) can be used to infuse
intravenous fluids, medications (particularly vasopressors), and blood products, as well as to
draw blood for frequent laboratory studies. In addition, this access can be used for hemodynamic
monitoring by measuring the central venous pressure (CVP) and the central venous
oxyhemoglobin saturation (ScvO2). While in the past, a major purpose of a CVC was the
measurement of ScVO2 and CVP, evidence from randomized trials on the value these targets to
follow therapeutic effect is conflicting [19-21]. We believe that pulmonary artery catheters (PACs)
should not be used in the routine management of patients with severe sepsis or septic shock.
PACs can measure the pulmonary artery occlusion pressure (PAOP) and mixed venous
oxyhemoglobin saturation (SvO2). In theory, this may be helpful to guide circulatory resuscitation.
However, the PAOP has proven to be a poor predictor of fluid responsiveness in sepsis and the
SvO2 is similar to the ScvO2, which can be obtained from a CVC [22,23]. PACs increase
complications and have not been shown to improve outcome [24-26].
Interventions to restore perfusion The rapid restoration of perfusion is predominantly
achieved by the administration of intravenous fluids, usually crystalloids. Modalities such as
vasopressor therapy, inotropic therapy, and blood transfusion are added, depending on the
response to fluid resuscitation, evidence for myocardial dysfunction, and presence of anemia.

Intravenous fluids In patients with sepsis, intravascular hypovolemia is typical and may be
severe, requiring rapid fluid resuscitation.
Volume The optimal volume of resuscitative fluid is unknown. Several studies of early goal
directed therapy reported intravenous fluid infusions targeted to physiologic endpoints and
resulted in volumes ranging from 3 to 5 liters [19-21]. The volume of fluid that was administered
within the initial six hours of presentation was targeted to set physiologic endpoints (eg, mean
arterial pressure). While an early study of early goal-directed therapy (EGDT) reported mean
infusion volume in the first six hours of 3 to 5 liters [19], later trials reporting mean infusion
volumes of 2 to 3 liters [20,21]. Thus, rapid, large volume infusions of intravenous fluids are
indicated as initial therapy for severe sepsis or septic shock, unless there is coexisting clinical or
radiographic evidence of heart failure. Suggested targets for fluid resuscitation are discussed
separately.
Fluid therapy should be administered in well-defined (eg, 500 mL), rapidly infused boluses [9].
Volume status, tissue perfusion, blood pressure, and the presence or absence of pulmonary
edema must be assessed before and after each bolus. Intravenous fluid challenges can be
repeated until blood pressure and tissue perfusion are acceptable, pulmonary edema ensues, or
fluid fails to augment perfusion.
Careful monitoring is essential because patients with sepsis may develop noncardiogenic
pulmonary edema (ie, acute respiratory distress syndrome [ARDS]). Once patients with ARDS
have been fluid resuscitated a liberal approach to intravenous fluid administration has been
shown to prolong the duration of mechanical ventilation, compared to a more restrictive
approach that also typically requires large doses of furosemide [27]. In addition, small
retrospective studies have reported that fluid overload is common in patients with sepsis and is
associated with the increased performance of medical interventions (eg, diuresis, thoracentesis);
the effect of fluid overload and such interventions on mortality and functional recovery is unclear
[28-30]. Thus, while the early, aggressive fluid therapy is appropriate in severe sepsis and septic
shock, fluids may be unhelpful or harmful when the circulation is no longer fluid-responsive.
Choice of fluid Evidence from randomized trials and meta-analyses have found no convincing
difference between using albumin solutions and crystalloid solutions (eg, normal saline, Ringers
lactate) in the treatment of severe sepsis or septic shock, but they have identified potential harm
from using pentastarch or hydroxyethyl starch rather than a crystalloid solution [31-38]:

Crystalloid versus albumin: In the Saline versus Albumin Fluid Evaluation (SAFE) trial,
6997 critically ill patients were randomly assigned to receive 4 percentalbumin
solution or normal saline for up to 28 days [31]. There were no differences between
groups for any endpoint, including the primary endpoint, mortality. Among the patients
with severe sepsis (18 percent of the total group), there were also no differences in
outcome. In another multicenter open-label randomized trial of patients with severe
sepsis or septic shock, the addition of albumin to crystalloid did not improve survival
compared to crystalloid alone (31 versus 32 percent) [32].

Crystalloid versus hydroxyethyl starch: In the Scandinavian Starch for Severe Sepsis
and Septic Shock (6S) trial, 804 patients with severe sepsis were randomly assigned to
receive either 6 percent hydroxyethyl starch or Ringers acetate at a volume of up to
33 mL/kg of ideal body weight per day [33]. When assessed 90 days after randomization,
mortality was increased in the hydroxyethyl starch group (51 versus 43 percent) and
more patients in the hydroxyethyl starch group had required renal replacement therapy
at some time during their illness (22 versus 16 percent).

Crystalloid versus pentastarch: The Efficacy of Volume Substitution and Insulin


Therapy in Severe Sepsis (VISEP) trial compared pentastarch to modified Ringer's lactate
in patients with severe sepsis and found no difference in 28-day mortality [34]. The trial
was stopped early because there was a trend toward increased 90-day mortality among
patients who received pentastarch.
In our clinical practice, we generally use a crystalloid solution instead of albumin
solution because of the lack of clear benefit and higher cost of albumin. We believe that giving a
sufficient quantity of intravenous fluids rapidly and targeting appropriate goals is more important
than the type of fluid chosen. We do not use hydroxyethyl starch or pentastarch. These choices
are consistent with the Society of Critical Care Medicine guidelines

Vasopressors Vasopressors are second line agents in the treatment of severe sepsis and
septic shock; we prefer intravenous fluids as long as they increase perfusion without seriously
impairing gas exchange [39]. However, intravenous vasopressors are useful in patients who
remain hypotensive despite adequate fluid resuscitation or who develop cardiogenic pulmonary
edema. In most patients with severe septic shock, we prefer to use norepinephrine (table 3)
[7,9,40]. However, we find phenylephrine (a pure alpha-adrenergic agonist) to be useful when
tachycardia or arrhythmias preclude the use of agents with beta-adrenergic activity (eg,
norepinephrine). Choosing a vasopressor agent is discussed in greater detail elsewhere.
Additional therapies There is conflicting evidence on the use of additional therapies, such
as inotropic therapy or red blood cell transfusion. Such therapies are targeted at increasing the
cardiac output to improve tissue perfusion and thereby raise the central venous (superior vena
cava) oxyhemoglobin saturation toward normal (ScvO 2 70 percent). We prefer that their use be
limited to those with refractory shock in whom the ScvO 2 remains <70 percent after optimization
of intravenous fluid and vasopressor therapy.
Inotropic therapy A trial of inotropic therapy may be warranted in patients who have
refractory shock who also have diminished cardiac output [7,8,19,41,42]. Inotropic therapy
should not be used to increase the cardiac index to supranormal levels [7]. Dobutamine is the
usual inotropic agent [9]. At low doses, dobutamine may cause the blood pressure to decrease
because its peripheral effects can dilate the systemic arteries. However, as the dose is increased,
blood pressure usually rises because cardiac output increases out of proportion to the fall in
peripheral vascular resistance.
Red blood cell transfusions Based upon clinical experience, randomized studies, and
guidelines on transfusion of blood products in critically ill patients, we typically reserve red blood
cell transfusion for patients with a hemoglobin level 7 g per deciliter. Exceptions include
suspicion of concurrent hemorrhagic shock or active myocardial ischemia.
Support for a restrictive transfusion strategy (goal hemoglobin >7 g/dL) is derived from direct
and indirect evidence from randomized studies of patients with septic shock:

One multicenter randomized study of 998 patients with septic shock reported no
difference in 28 day mortality between patients who were transfused when the
hemoglobin was 7 g/dL (restrictive strategy) and patients who were transfused when
the hemoglobin was 9 g/dL (liberal strategy) [43]. The restrictive strategy resulted in 50
percent fewer red blood cell transfusions (1545 versus 3088 transfusions) and did not
have any adverse effect on the rate of ischemic events (7 versus 8 percent).

Data from randomized studies of EGDT that use red blood cell transfusion as part of the
protocol for treating patients with sepsis are conflicting. While one trial initially reported
a mortality benefit from EGDT that included transfusing patients to a goal hematocrit
>30 (hemoglobin level 10 g/dL) [19], two similarly designed studies published since then
reported no benefit to this strategy [20,21]. These studies are discussed below.
(See 'Protocol-directed therapy' below.)
In further support of a restrictive approach to transfusion in patients with septic shock is the
consensus among experts that transfusing to a goal of >7 g/dL is also preferred in critically ill
patients without sepsis [44-46]. The use of blood transfusions in critically-ill patients is discussed
in detail separately. (
Goals of initial resuscitation The goal of fluid resuscitation is early restoration of perfusion
to prevent or limit multiple organ dysfunction, as well as to reduce mortality. The term "early
goal-directed therapy" (EGDT) refers to the administration of intravenous fluids within the first
six hours of presentation using physiologic targets to guide fluid management. EGDT has
gained widespread acceptance in clinical practice but the optimal targets are unknown.
Early goal-directed therapy targets Although evidence is conflicting regarding the routine
measurement of early goal-directed therapy targets, we suggest measuring the following targets
for fluid management in patients with sepsis:

Mean arterial pressure (MAP) 65 mmHg (MAP = [(2 x diastolic) + systolic]/3) (calculator
1)

Urine output 0.5 mL/kg/hour

Static or dynamic predictors of fluid responsiveness, eg, CVP 8 to 12 mmHg when central
access is available (static measurement) or respiratory changes in the radial artery pulse
pressure (dynamic measurement).

Central venous (superior vena cava) oxyhemoglobin saturation (ScvO 2) 70 percent


(when central access is available) or mixed venous oxyhemoglobin saturation (SvO 2) 65
percent (if a pulmonary artery catheter is being used).
Lactate clearance should be followed as a target in patients with sepsis to ensure a trend that
demonstrates adequate clearance with therapy. Newer point of care analyzers are commercially
available that may allow clinicians to follow lactate levels at the bedside more readily [47-49].
The optimal physiologic target(s) of EGDT is unknown. There is also conflicting evidence on the
value of measuring such targets, particularly CVP and ScVO 2, which require central catheter
placement [19-21,50]. In addition, the generalizability of a standard targeted approach to both
resource-poor and resource-rich facilities is unknown. We prefer measuring MAP and urine output
as universal targets that can be readily measured in all patients with sepsis, with the addition of
CVP and/or ScVO2 in those in whom central access is otherwise required. This approach differs
slightly from that of The Surviving Sepsis Campaign guidelines that recommend central venous
access for CVP/ScvO2 measurement together with MAP and urine output in all patients with
severe sepsis [9]. However, these guidelines were created before the results of three major
randomized trials (ProCESS, ARISE, ProMISe), that showed no mortality benefit to an EGDT-based
approach, were published [20,21,50,51].
Evidence that supports the use of EGDT targets is described below:

CVP, MAP and urine output CVP 8 to 12 mmHg, MAP 65 mmHg (calculator 1), and
urine output 0.5 mL/kg per hour are common EGDT targets used in clinical practice.
Support for their use is derived from clinical experience and their use in the single
randomized trial that studied them with and without ScvO 2 [19]. They have not been
compared to each other nor have they been proven to be superior to any other target or
to clinical assessment. The ideal targets for MAP, CVP, and urine output are unknown.
One trial that randomized patients to a target MAP of 65 to 70 mmHg (low target MAP) or
80 to 85 mmHg (high target MAP) reported no mortality benefit to targeting a higher MAP
[52,53]. Patients with a higher MAP had a greater incidence of atrial fibrillation (7 versus
3 percent), suggesting that targeting a MAP >80 mmHg is potentially harmful.

ScvO2 Evidence from randomized trials that study the value of ScvO 2) report mixed
results. While one early trial of patients with septic shock reported a mortality benefit to
targeting ScvO2 70 percent in a protocol-based therapy, trials published since then
(ProCESS, ARISE, ProMISe) have reported no mortality benefit [19-21,50].

Lactate clearance Although the optimal frequency is unknown, we follow serum


lactate (eg, every six hours), as an additional EGDT target in patients with sepsis until
the lactate value has clearly fallen. The lactate clearance is defined by the equation
[(initial lactate - lactate >2 hours later)/initial lactate] x 100. The lactate clearance and
interval change in lactate over the first 12 hours of resuscitation has been evaluated as a
potential marker for effective resuscitation [54,55]. One trial randomly assigned 300
patients with severe sepsis to undergo resuscitation targeting either a lactate clearance
10 percent or an ScvO2 70 percent (other than these targets, the resuscitation
protocols that included MAP, CVP, and urine output targets were identical) [54]. There
was no difference in hospital mortality, length of stay, ventilator-free days, or incidence
of multiorgan failure, suggesting that lactate clearance criteria may be an acceptable
alternative to ScvO2 criteria. After the restoration of perfusion, lactate is a poor marker of
tissue perfusion [56]. As a result, lactate values are generally unhelpful following
restoration of perfusion, with one exception that a rising lactate level should prompt
reevaluation of perfusion.

Other Dynamic indices have been studied as a potential target to guide fluid
management in sepsis. Respiratory changes in the vena caval diameter, radial artery
pulse pressure, aortic blood flow peak velocity, and brachial artery blood flow velocity are
considered dynamic hemodynamic measures, whereas CVP, MAP, ScvO 2and pulmonary
artery occlusion pressure are considered static hemodynamic measures [57,58]. There is
increasing evidence that dynamic measures are more accurate predictors of fluid
responsiveness than static measures, as long as the patients are in sinus rhythm and
passively ventilated with a sufficient tidal volume [22,59,60]. For actively breathing
patients or those with irregular cardiac rhythms, an increase in the cardiac output in
response to a passive leg-raising maneuver (measured by echocardiography, arterial

pulse waveform analysis, or pulmonary artery catheterization) is a sensitive and specific


predictor of fluid responsiveness [61]. Large randomized studies will be needed proving
the efficacy of assessing dynamic measurement in response to intravenous fluids before
they can be routinely applied to patients for the management of sepsis.
Protocol-directed therapy Protocols targeted at the use of a combination of physiologic
endpoints to guide fluid management in patients with severe sepsis and septic shock are
common practice [19-21,50,51,62-64]. Typically, they combine the EGDT targets (ScvO 2, CVP,
MAP (calculator 1) and urine output, lactate) for fluid management with early administration of
antibiotics, both within the first six hours of presentation.
There is conflicting evidence regarding the value of protocol-based therapy for sepsis [1921,50,51,64-66]:

One single center randomized trial of 263 patients with severe sepsis or septic shock
compared a protocol that included targeting ScvO 2 70 percent, CVP 8 to 12 mmHg, MAP
65 mmHg, and urine output 0.5 mL/kg/hour to conventional therapy that targeted
CVP, MAP, and urine output only [19]. Both groups initiated therapy (including antibiotics)
within six hours of presentation. Mortality was lower in the group where all four targets
were used (31 versus 47 percent), suggesting that targeting ScvO 2, CVP, MAP, and urine
output was a superior strategy. There was a heavy emphasis on the use of red cell
transfusion (for a hematocrit >30) anddobutamine in order to reach the ScvO 2 target in
this trial. In addition, the results of this trial may not be generalizable due to the inclusion
of a significant number of sick patients with liver and heart disease that may have
potentially biased the outcome favorably.

A multicenter randomized trial (ProCESS) of 1341 patients with septic shock reported no
mortality benefit with protocol-based therapies [20]. A protocol-based therapy that used
all of the EGDT targets (ScvO 2, CVP, MAP and urine output; protocol-based EGDT; central
access required) was compared to a protocol that used some of the EGDT targets (MAP
and urine output; protocol-based standard therapy; central access not required) and to
usual care (no protocol used to direct fluid management). There were no differences in
60-day mortality between the groups (21 versus 18 versus 19 percent).
Two similarly designed multicenter randomized trials of 1600 (ARISE) and 2160 (ProMISe)
patients with septic shock also reported no mortality benefit from EGDT [21,50]. In ARISE,
compared to usual care, the 90 day mortality of 19 percent was similar in patients who received
EGDT using the traditional targets outlined in prior studies [21]. Similarly, in ProMISe, the 90 day
mortality was no different (29 percent) between the EGDT and usual care groups [50].
One explanation for the apparent negative results from these three trials may be that central line
placement was common (>50 percent) in patients receiving protocol-based standard therapy and
usual care; it is likely that CVP and ScvO 2 were measured and targeted in these patients as well.
Lack of benefit may also be attributed to overall better outcomes in these studies, perhaps due
to early administration of antibiotics (70 to 100 percent before randomization) in all groups, and
to improved clinical performance by highly trained clinicians in academic centers during an era
that follows an aggressive sepsis education and management campaign.
Timing and duration The early administration of fluid appears to be more important than
volume or type of fluid in reducing mortality associated with sepsis. Based upon evidence from
randomized studies and meta-analyses, we favor the initiation of fluid resuscitation
within six hours of presentation. Once the targets of resuscitation are met and perfusion is
restored, fluids can be reduced or stopped, and occasionally patients can be diuresed, when
necessary. Resolution of severe sepsis and septic shock can take as little as a few hours or can be
protracted to days or weeks.
A 2008 meta-analysis of randomized trials that initiated resuscitation targeting specific
physiologic endpoints reported that compared to standard care, only trials that initiated
resuscitation within 24 hours of the onset of sepsis showed a mortality benefit (39 versus 57
percent, odds ratio 0.50, 95% CI 0.37-0.69) [67]. In contrast, analysis of randomized trials that
initiated therapy more than 24 hours after the onset of sepsis found no difference in mortality
(64 versus 58 percent for standard resuscitation, odds ratio 1.16, 95% CI 0.60-2.22).
There are two possible outcomes following the interventions described above:

Inadequate perfusion Despite aggressive therapy, the patient may have persistent
hypoperfusion and progressive organ failure. This should prompt reassessment of the

adequacy of the above therapies, antimicrobial regimen, and control of the septic focus,
as well as the accuracy of the diagnosis and the possibility that unexpected
complications or coexisting problems have intervened (eg, pneumothorax following CVC
insertion).
Adequate perfusion Patients who respond to therapy should have the rate of fluid
administration reduced or stopped, and vasopressor support weaned. Patients should
also continue to have their clinical and laboratory parameters followed closely. These
include blood pressure, arterial lactate, urine output, creatinine, platelet count, Glasgow
coma scale score, serum bilirubin, liver enzymes, oxygenation (ie, arterial oxygen tension
or oxyhemoglobin saturation), and gut function (table 4). Reevaluation is indicated if any
of these parameters worsen or fail to improve.

CONTROL OF THE SEPTIC FOCUS Prompt identification and treatment of the primary site or
sites of infection are essential [68-70]. This is the primary therapeutic intervention, with most
other interventions being purely supportive. Antibiotics should be administered within the first six
hours of presentation or earlier.
Identification of the septic focus A careful history and physical examination may yield
clues to the source of sepsis and help guide microbiologic evaluation (table 5). As an example,
sepsis arising after trauma or surgery is often due to infection at the site of injury or surgery. The
presence of a urinary or vascular catheter increases the chances that these are the source of
sepsis.
Gram stain of material from sites of possible infection may give early clues to the etiology of
infection while cultures are incubating. As examples, urine should be routinely analyzed via
dipstick for leukocyte esterase, Gram stained, and cultured; sputum should be examined in a
patient with a productive cough; and an intra-abdominal collection in a postoperative patient
should be percutaneously sampled under ultrasound or other radiologic guidance.
Blood should be drawn from two distinct venipuncture sites and inoculated into standard blood
culture media (aerobic and anaerobic). For patients with a vascular catheter, blood should be
obtained both through the catheter and from another site [9].
If invasive candida or aspergillus infection is suspected, serologic assays for 1,3 beta-D-glucan,
galactomannan, and anti-mannan antibodies, if available, may provide early evidence of these
fungal infections [9]. The limitations of these assays and their role in the diagnosis of fungal
infection are discussed separately.
There is no single test that immediately confirms the diagnosis of severe sepsis or septic shock.
However, several laboratory tests, all of which are still investigational, have been studied as
diagnostic markers of active bacterial infection [6]:

Elevated serum procalcitonin levels are associated with bacterial infection and sepsis
[71-73]. Despite this, a meta-analysis of 18 studies found that procalcitonin did not
readily distinguish sepsis from nonseptic systemic inflammation (sensitivity of 71 percent
and specificity of 71 percent) [72]. An additional randomized trial and another metaanalysis found that using clinical algorithms based upon procalcitonin levels did not
affect mortality or duration of antibiotic treatment [74,75].

The plasma concentration of soluble TREM-1 (triggering receptor expressed on myeloid


cells), a member of the immunoglobulin superfamily that is specifically upregulated in
the presence of bacterial products, is increased in patients with sepsis [76-78]. In a small
trial, increased TREM-1 levels were both sensitive and specific for the diagnosis of
bacterial sepsis (96 and 89 percent, respectively) [76]. However, a subsequent
prospective cohort study found that increased TREM-1 levels predicted sepsis with a
sensitivity and specificity of only 53 and 86 percent, respectively [79]. Serial monitoring
of TREM-1 may also provide prognostic information in patients with established sepsis
[77,78].

Increased expression of CD64 on polymorphonuclear leukocytes indicates cellular


activation and has been shown to occur in patients with sepsis [80,81]. In a prospective
cohort study of 300 consecutive critically ill patients, increased CD64 expression
predicted sepsis with a sensitivity of 84 percent and a specificity of 95 percent [79]. In
this study, the sensitivity and specificity of increased CD64 expression were superior to
that of increased procalcitonin or TREM-1 levels.
The combination of procalcitonin levels, TREM-1 levels, and CD64 expression appears to be
superior to the use of any of these markers alone. However, evaluation of the clinical usefulness

of such biomarkers is still in its early stages and should be considered preliminary. Until
additional clinical investigations have been performed, we do not suggest the routine use of such
biomarkers to identify sepsis.
Eradication of infection Prompt and effective treatment of the active infection is essential to
the successful treatment of severe sepsis and septic shock [9]. Source control (physical
measures undertaken to eradicate a focus of infection and eliminate or treat ongoing microbial
proliferation and infection) should be undertaken since undrained foci of infection may not
respond to antibiotics alone (table 2). As examples, potentially infected foreign bodies (eg,
vascular access devices) should be removed when possible, and abscesses should undergo
percutaneous or surgical drainage. Some patients require extensive soft tissue debridement or
amputation; in severe cases, fulminant Clostridium difficile-associated colitis may necessitate
colectomy [82].
Antimicrobial regimen Intravenous antibiotic therapy should be initiated within the first six
hours or earlier (eg, within one hour), after obtaining appropriate cultures, since early initiation of
antibiotic therapy is associated with lower mortality [7,83]. The choice of antibiotics can be
complex and should consider the patient's history (eg, recent antibiotics received [84]),
comorbidities, clinical context (eg, community- or hospital-acquired), Gram stain data, and local
resistance patterns [7,85,86].
Poor outcomes are associated with inadequate or inappropriate antimicrobial therapy (ie,
treatment with antibiotics to which the pathogen was later shown to be resistant in vitro) [87-93].
They are also associated with delays in initiating antimicrobial therapy, even short delays (eg, an
hour).

A prospective cohort study of 2124 patients demonstrated that inappropriate antibiotic


selection was surprisingly common (32 percent) [90]. Mortality was markedly increased
in these patients compared to those who had received appropriate antibiotics (34 versus
18 percent).

A retrospective analysis of 2731 patients with septic shock demonstrated that the time
to initiation of appropriate antimicrobial therapy was the strongest predictor of mortality
[91].

When the potential pathogen or infection source is not immediately obvious, we favor
broad-spectrum antibiotic coverage directed against both gram-positive and gramnegative bacteria. Few guidelines exist for the initial selection of empiric antibiotics in
severe sepsis or septic shock.

Staphylococcus aureus is associated with significant morbidity if not treated early in the
course of infection [94]. There is growing recognition that methicillin-resistant S. aureus
(MRSA) is a cause of sepsis not only in hospitalized patients, but also in community
dwelling individuals without recent hospitalization [95,96]. For these reasons, we
recommend that severely ill patients presenting with sepsis of unclear etiology be
treated with intravenous vancomycin (adjusted for renal function) until the possibility of
MRSA sepsis has been excluded. Potential alternative agents to vancomycin
(eg, daptomycin for non-pulmonary MRSA, linezolid, ceftaroline) should be considered for
patients with refractory or virulent MRSA, or a contraindication to vancomycin. These
agents are discussed separately.
In our practice, if Pseudomonas is an unlikely pathogen, we favor combining vancomycin with
one of the following:

Cephalosporin, 3rd generation (eg, ceftriaxone or cefotaxime) or 4th generation


(cefepime), or

Beta-lactam/beta-lactamase inhibitor (eg, piperacillin-tazobactam, ticarcillin-clavulanate),


or

Carbapenem (eg, imipenem or meropenem)


Alternatively, if Pseudomonas is a possible pathogen, we favor combining vancomycin with two
of the following:

Antipseudomonal cephalosporin (eg, ceftazidime, cefepime), or

Antipseudomonal carbapenem (eg, imipenem, meropenem), or

Antipseudomonal beta-lactam/beta-lactamase inhibitor


(eg, piperacillintazobactam, ticarcillin-clavulanate), or

Fluoroquinolone with good anti-pseudomonal activity (eg, ciprofloxacin), or


Aminoglycoside (eg, gentamicin, amikacin), or

Monobactam (eg, aztreonam)


Selection of two agents from the same class, for example, two beta-lactams, should be avoided.
We emphasize the importance of considering local susceptibility patterns when choosing an
empiric antibiotic regimen.
After culture results and antimicrobial susceptibility data return, we recommend that therapy be
pathogen- and susceptibility-directed, even if there has been clinical improvement while on the
initial antimicrobial regimen. Gram-negative pathogens have historically been covered with two
agents from different antibiotic classes. However, several clinical trials and two meta-analyses
have failed to demonstrate superior overall efficacy of combination therapy compared to
monotherapy with a third generation cephalosporin or a carbapenem [90,97-101]. Furthermore,
one meta-analysis found double coverage that included an aminoglycoside was associated with
an increased incidence of adverse events (nephrotoxicity) [100,101]. For this reason, in patients
with gram negative pathogens, we recommend use of a single agent with proven efficacy and
the least possible toxicity, except in patients who are either neutropenic or whose severe sepsis
is due to a known or suspected Pseudomonas infection [7,99].
Regardless of the antibiotic regimen selected, patients should be observed closely for toxicity,
evidence of response, and the development of nosocomial superinfection [102]. There are no
published randomized controlled trials testing safety of de-escalation of antibiotic therapy in
adult patients with sepsis or septic shock [103]. The duration of therapy is typically 7 to 10 days,
although longer courses may be appropriate in patients who have a slow clinical response, an
undrainable focus of infection, or immunologic deficiencies [7]. In patients who are neutropenic,
antibiotic treatment should continue until the neutropenia has resolved or the planned antibiotic
course is complete, whichever is longer. In non-neutropenic patients in whom infection is
thoroughly excluded, antibiotics should be discontinued to minimize colonization or infection with
drug-resistant microorganisms and superinfection with other pathogens.
Other agents Invasive fungal infections occasionally complicate the course of critical illness
in non-neutropenic patients, especially when the following risk factors are present: surgery,
parenteral nutrition, prolonged antimicrobial treatment, severe sepsis, or multisite colonization
with Candida spp. To limit the risk of candida-related mortality empirical anti-fungal treatments
have been proposed. In a meta-analysis of 22 studies (most often comparing fluconazole to
placebo, but also usingketoconazole, anidulafungin, caspofungin, micafungin, and amphotericin
B), untargeted empiric antifungal therapy possibly reduced fungal colonization and the risk of
invasive fungal infection but did not reduce all-cause mortality [104]. Similarly, in a study of
critically-ill patients ventilated at least five days, empiric antifungal treatment (mostly
fluconazole) was not associated with a decreased risk of mortality or occurrence of invasive
candidiasis [105]. Thus, the routine administration of empirical antifungal therapy is not
warranted in non-neutropenic critically-ill patients.
ADDITIONAL THERAPIES
Glucocorticoids Glucocorticoids have long been investigated as therapeutic agents in sepsis
because the pathogenesis of sepsis involves an intense and potentially deleterious host
inflammatory response. Evidence from randomized trials suggest that corticosteroid therapy is
most likely to be beneficial in patients who have severe septic shock (defined as a systolic blood
pressure <90 mmHg) that is unresponsive to adequate fluid resuscitation and vasopressor
administration. Data from ongoing clinical trials are needed to confirm that benefit. This topic is
discussed in detail separately.
Nutrition There is consensus that nutritional support improves nutritional outcomes in
critically ill patients, such as body weight and mid-arm muscle mass. However, it is uncertain
whether nutritional support improves important clinical outcomes (eg, duration of mechanical
ventilation, length of stay, mortality), or whether there is a validated role for specific
supplements (eg, immune modulators). The principles of nutritional support in patients with
sepsis should parallel that in critically ill patients, which is reviewed in detail elsewhere.
Venous thromboembolism prophylaxis Patients with sepsis and septic shock are at
increased risk for venous thromboembolism such that patients should receive
thromboprophylaxis [106], the details of which are discussed separately.
Intensive insulin therapy Hyperglycemia and insulin resistance are common in critically ill
patients, independent of a history of diabetes mellitus [107]. The optimal blood glucose range is

controversial. Most clinicians target blood glucose levels between 140 and 180 mg/dL (7.7 to
10 mmol/L). This topic is discussed separately.
External cooling or antipyretics Controlling fever during severe sepsis and septic shock
has potential benefits and adverse effects, the net effects of which are uncertain.
A trial was performed to compare the effects of external cooling with no external cooling.
External cooling consists of using either an automatic cooling blanket, or ice-cold bed sheets and
ice packs, to achieve a core body temperature of 36.5 to 37C for 48 hours. It decreases the time
to fever control without exposing the patient to potential adverse effects of antipyretic drugs.
The trial randomly assigned 200 patients with septic shock (the patients were requiring
vasopressors, mechanically ventilated, and sedated) to receive either external cooling or no
external cooling [108]. Patients in the external cooling group had lower 14-day mortality (19
versus 34 percent) and were more likely to have their vasopressor dose lowered by 50 percent
(54 versus 20 percent) and their shock reversed during their ICU stay (86 versus 73 percent). No
antipyretic agents were received during the trial.
While these results are promising, we believe that the results need to be confirmed before
external cooling is adopted as routine clinical practice. Among the limits of the trial, patients in
the external cooling group may have been less severely ill (ie, they required a lower baseline
vasopressor dose), the trial was not blinded so co-interventions cannot be excluded, and there
were relatively few events (ie, deaths, patients with a 50 percent vasopressor dose decrease, and
patients with shock reversal), which lowers confidence in the accuracy of the estimated effects.
Moreover, the results suggest that external cooling is preferable to no cooling, but they do not
provide guidance about whether external cooling is preferable to antipyretic medications.
The role of antipyretics for fever control in critically ill patients is also of uncertain benefit and is
discussed separately. (See "Fever in the intensive care unit", section on 'Management'.)
Investigational therapies A variety of investigational therapies including cytokine and toxin
inactivation, as well as hemofiltration, statins, and beta blockade are discussed in detail
separately. (See "Investigational and ineffective therapies for sepsis".)
INFORMATION FOR PATIENTS UpToDate offers two types of patient education materials,
The Basics and Beyond the Basics. The Basics patient education pieces are written in plain
language, at the 5th to 6th grade reading level, and they answer the four or five key questions a
patient might have about a given condition. These articles are best for patients who want a
general overview and who prefer short, easy-to-read materials. Beyond the Basics patient
education pieces are longer, more sophisticated, and more detailed. These articles are written at
the 10th to 12th grade reading level and are best for patients who want in-depth information and
are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print
or e-mail these topics to your patients. (You can also locate patient education articles on a
variety of subjects by searching on patient info and the keyword(s) of interest.)
Basics topic (see "Patient information: Sepsis in adults (The Basics)")
SUMMARY AND RECOMMENDATIONS
Therapeutic priorities for patients with sepsis and septic shock include securing the airway,
correcting hypoxemia, and administering fluids and antibiotics. Intubation and mechanical
ventilation are required in some patients. (See 'Therapeutic priorities' above and 'Stabilize
respiration' above.)
The adequacy of perfusion should be assessed in patients with suspected severe sepsis and
septic shock. Hypotension is the most common indicator of inadequate perfusion. However,
critical hypoperfusion can also occur in the absence of hypotension, especially during early
sepsis. Common signs of hypoperfusion include warm, vasodilated skin in early sepsis that
progresses to cool, vasoconstricted skin in late sepsis, tachycardia >90 per min, obtundation or
restlessness, oliguria or anuria, and lactic acidosis. (See 'Assess perfusion' above.)
For patients with severe sepsis and septic shock, we recommend intravenous fluids, rather than
vasopressors, inotropes, or red blood cell transfusions as first-line therapy for the restoration of
tissue perfusion (Grade 1B). Therapy should be initiated as early as possible, within six hours of
presentation. Fluid boluses are the preferred method of administration and should be repeated
until blood pressure and tissue perfusion are acceptable, pulmonary edema ensues, or there is
no further response. These parameters should be assessed before and after each fluid bolus
(See 'Interventions to restore perfusion' above.)
For initial fluid replacement, we suggest using a crystalloid solution rather than albumincontaining solution (Grade 2B) and recommend that a hyperoncotic starch solution NOT be
administered (Grade 1A). (See 'Choice of fluid' above and "Treatment of severe hypovolemia or
hypovolemic shock in adults", section on 'Choice of replacement fluid'.)

For patients who remain hypotensive following intravascular volume repletion, we recommend
vasopressors (Grade 1B); the preferred initial agent isnorepinephrine. (See 'Vasopressors' above
and "Use of vasopressors and inotropes", section on 'Choice of agent in septic shock'.)
For patients with severe sepsis and septic shock that are refractory to intravenous fluid and
vasopressor therapy, additional therapies, such as inotropic therapy and blood transfusions, are
administered based on individual assessment. We typically reserve red blood cell transfusion for
patients with a hemoglobin level <7 g per deciliter. (See 'Additional therapies' above and "Use of
vasopressors and inotropes", section on 'Choice of agent in septic shock'.)
For most patients with sepsis and septic shock, we suggest fluid management be guided using
specific targets (early goal-directed therapy [EGDT]), rather than being managed without specific
therapeutic targets. The optimal target to guide fluid management is unknown. For most
patients, we target mean arterial pressure 65 mmHg (calculator 1) and urine output
0.5 mL/kg/hour and integrate it with static measures of determining adequacy of fluid
administration (eg, central venous pressure [CVP] 8 to 12 mmHg), or dynamic predictors of fluid
responsiveness (eg, respiratory changes in the radial artery pulse pressure) or central venous
oxygen saturation 70 percent. In addition, we follow serum lactate (eg, every six hours), until
there is a clear clinical response. (See 'Goals of initial resuscitation' above.)
Prompt identification and treatment of the site of infection are essential. Sputum and urine
should be collected for Gram stain and culture. Intra-abdominal fluid collections should be
percutaneously sampled. Blood should be taken from two distinct venipuncture sites and from
indwelling vascular access devices and cultured aerobically and anaerobically. (See 'Identification
of the septic focus' above.)
Antibiotics should be administered within six hours of presentation, preferably after appropriate
cultures have been obtained. We recommend empiric broad spectrum antibiotics when a definite
source of infection cannot be identified (Grade 1B). The routine administration of antifungal
therapy is not warranted in non-neutropenic patients. (See 'Antimicrobial regimen' above.)
Potentially infected vascular access devices should be removed (if possible), abscesses should
be drained, and extensive soft tissue infections should be debrided or amputated (table 2).
(See 'Eradication of infection' above.)
Glucocorticoid therapy, nutritional support, glucose control, and investigational therapies are
additional considerations in the management of patients with severe sepsis or septic shock. Each
is discussed separately.