DEFINITIONS (See Table 265-1) Animals mount both local and systemic responses to microbes that traverse epithelial

barriers and invade underlying tissues. Fever or hypothermia, leukocytosis or leukopenia, tachypnea, and tachycardia are the cardinal signs of the systemic response often called the systemic inflammatory response syndrome (SIRS). SIRS may have an infectious or a noninfectious etiology. If infection is suspected or proven, a patient ith SIRS is said to have sepsis. !hen sepsis is associated ith dysfunction of organs distant from the site of infection, the patient has severe sepsis. Severe sepsis may be accompanied by hypotension or evidence of hypoperfusion. !hen hypotension cannot be corrected by infusing fluids, the diagnosis is septic shock. "hese definitions ere proposed by consensus conference committees in #$$% and %&&# and are no idely used' there is evidence that the different stages form a continuum. As sepsis progresses to septic shock, the risk of dying increases substantially. Sepsis is usually reversible, hereas patients ith septic shock often succumb despite aggressive therapy. ETIOLOGY Severe sepsis can be a response to any class of microorganism. (icrobial invasion of the bloodstream is not essential for the development of severe sepsis, since local inflammation can also elicit distant organ dysfunction and hypotension. In fact, blood cultures yield bacteria or fungi in only )%&*+&, of cases of severe sepsis and +&*-&, of cases of septic shock. Individual gram.negative or gram.positive bacteria account for )-&, of these isolates' the remainder are fungi or a mi/ture of microorganisms (Table 265-2). In patients hose blood cultures are negative, the etiologic agent is often established by culture or microscopic e/amination of infected material from a local site. In some case series, a ma0ority of patients ith a clinical picture of severe sepsis or septic shock have had negative microbiologic data. EPIDEMIOLOGY "he septic response is a contributing factor in 1%&&,&&& deaths per year in the 2nited States. "he incidence of severe sepsis and septic shock has increased over the past %& years, and the annual number of cases is no 1-&&,&&& ()3 per #&&& population). Appro/imately t othirds of the cases occur in patients ith significant underlying illness. Sepsis.related incidence and mortality rates increase ith age and pree/isting comorbidity. "he rising incidence of severe sepsis in the 2nited States is attributable to the aging of the population, the increasing longevity of patients ith chronic diseases, and the relatively high fre4uency ith hich sepsis develops in patients ith AI5S. "he idespread use of antimicrobial agents, immunosuppressive drugs, ind elling catheters and mechanical devices, and mechanical ventilation also plays a role. Invasive bacterial infections are prominent causes of death around the orld, particularly among young children. In sub.

Saharan Africa, for e/ample, careful screening for positive blood cultures found that community.ac4uired bacteremia accounted for at least one.fourth of deaths of children 1# year of age. 6ontyphoidal Salmonella species, Streptococcus pneumoniae, Haemophilus influenzae, and Escherichia coli ere the most commonly isolated bacteria. 7acteremic children often had 8I9 infection or ere severely malnourished. 7acteremia :resence of bacteria in blood, as evidenced by positive blood cultures Septicemia :resence of microbes or their to/ins in blood Systemic inflammatory response syndrome (SIRS) Two or more of t e follow!"# $o"%!t!o"&' (#) fever (oral temperature 13;<=) or hypothermia (>3?<=)' (%) tachypnea (1%+ breaths@min)' (3) tachycardia (heart rate 1$& beats@min)' (+) leukocytosis (1#%,&&&@<A), leukopenia (>+,&&&@<A), or 1#&, bands' may have a noninfectious etiology Sepsis SIRS that has a proven or suspected microbial etiology Severe sepsis (similar to Bsepsis syndromeC) Sepsis ith one or more signs of organ dysfunctionDfor e/ampleE #. Cardiovascular: Arterial systolic blood pressure F$& mm8g or mean arterial pressure F-& mm8g that responds to administration of intravenous fluid %. Renal: 2rine output >&.G mA@kg per hour for # h despite ade4uate fluid resuscitation 3. Respiratory: :aH%@FIH% F%G& or, if the lung is the only dysfunctional organ, F%&& +. Hematologic: :latelet count >;&,&&&@<A or G&, decrease in platelet count from highest value recorded over previous 3 days G. Unexplained metabolic acidosis: A p8 F-.3& or a base deficit IG.& mJ4@A and a plasma lactate level 1#.G times upper limit of normal for reporting lab ?. de!uate fluid resuscitation: :ulmonary artery edge pressure I#% mm8g or central venous pressure I; mm8g

Septic shock Sepsis ith hypotension (arterial blood pressure >$& mm8g systolic, or +& mm8g less than patientKs normal blood pressure) for at least # h despite ade4uate fluid resuscitation' or 6eed for vasopressors to maintain systolic blood pressure I$& mm8g or mean arterial pressure I-& mm8g Refractory septic shock Septic shock that lasts for 1# h and does not respond to fluid or pressor administration (ultiple.organ dysfunction syndrome ((H5S) 5ysfunction of more than one organ, re4uiring intervention to maintain homeostasis P(T)OP)YSIOLOGY (ost cases of severe sepsis are triggered by bacteria or fungi that do not ordinarily cause systemic disease in immunocompetent hosts ("able %?G.%). "hese microbes probably e/ploit deficiencies in innate host defenses (e.g., phagocytes, complement, and natural antibodies) to survive ithin the body. (icrobial pathogens, in contrast, are able to circumvent innate defenses by elaborating to/ins or other virulence factors. In both cases, the body can fail to kill the invaders despite mounting a vigorous inflammatory reaction that can result in severe sepsis. "he septic response may also be induced by microbial e/oto/ins that act as superantigens (e.g., to/ic shock syndrome to/in #' =hap. #%$). )o&t Me$ a"!&m& for Se"&!"# M!$robe& Animals have e/4uisitely sensitive mechanisms for recogniLing and responding to conserved microbial molecules. Recognition of the lipid A moiety of lipopolysaccharide (A:S, also called endotoxin' =hap. ##+) is the best.studied e/ample. A host protein (A:S.binding protein, or A7:) binds lipid A and transfers the A:S to =5#+ on the surfaces of monocytes, macrophages, and neutrophils. A:S then is passed to (5.%, hich interacts ith "olllike receptor ("AR) + to form a molecular comple/ that transduces the A:S recognition signal to the interior of the cell. "his signal rapidly triggers the production and release of mediators, such as tumor necrosis factor ("6F' see belo ), that amplify the A:S signal and transmit it to other cells and tissues. 7acterial peptidoglycan and lipoteichoic acids elicit responses in animals that are generally similar to those induced by A:S' hereas these molecules also bind =5#+, they interact ith different "ARs. 8aving numerous "AR.based receptor comple/es (#& different "ARs have been identified so far in humans) allo s animals to recogniLe many conserved microbial molecules. "he ability of some of the "ARs to serve as receptors for host ligands (e.g., hyaluronans, heparan

sulfate, saturated fatty acids) raises the possibility that these molecules play a role in producing noninfectious sepsis.like states. Hther host pattern.recognition proteins that are important for sensing microbial invasion and initiating host inflammation include the intracellular 6H5# and 6H5% proteins, hich recogniLe discrete fragments of bacterial peptidoglycan' complement (principally the alternative path ay)' mannose.binding lectin' and =.reactive protein. "he ability to recogniLe certain microbial molecules may influence both the potency of the host defense and the pathogenesis of severe sepsis. For e/ample, (5.%*"AR+ best senses A:S that has a he/aacyl lipid A moiety (i.e., one ith si/ fatty acyl chains). (ost of the commensal aerobic and facultatively anaerobic gram.negative bacteria that trigger severe sepsis and shock (including E" coli, #lebsiella, and Enterobacter) make this lipid A structure. !hen they invade human hosts, often through breaks in an epithelial barrier, infection is typically localiLed to the subepithelial tissue. 7acteremia, if it occurs, is intermittent and lo .grade, as these bacteria are efficiently cleared from the bloodstream by "AR+.e/pressing Mupffer cells and splenic macrophages. "hese mucosal commensals seem to induce severe sepsis most often by triggering severe local tissue inflammation rather than by circulating ithin the bloodstream. In contrast, gram.negative bacteria that do not make he/aacyl lipid A ($ersinia pestis, %rancisella tularensis, &ibrio vulnificus, 'seudomonas aeruginosa, and (urkholderia pseudomallei, among others) are poorly recogniLed by (5.%*"AR+. "hese bacteria usually enter the body via nonmucosal routesDe.g., as a result of bites, cuts, or inhalationDand initially induce relatively little inflammation. !hen they do trigger severe sepsis, it is often in the setting of massive bacterial gro th throughout the body. Jngineering a virulent strain of $" pestis to produce he/aacyl lipid A has rendered it avirulent in mice' this result attests to the importance of "AR+.based bacterial recognition in host defense. For most gram.negative bacteria, the pathogenesis of sepsis thus depends, at least in part, on hether the bacteriumKs A:S is sensed by the host receptor (5.%*"AR+. Lo$al a"% S*&tem!$ )o&t +e&,o"&e& to I"-a%!"# M!$robe& Recognition of microbial molecules by tissue phagocytes triggers the production and@or release of numerous host molecules (cytokines, chemokines, prostanoids, leukotrienes, and others) that increase blood flo to the infected tissue, enhance the permeability of local blood vessels, recruit neutrophils to the site of infection, and elicit pain. "hese phenomena are familiar elements of local inflammation, the bodyKs frontline innate immune mechanism for eliminating microbial invaders. Systemic responses are activated by neural and@or humoral communication ith the hypothalamus and brainstem' these responses enhance local defenses by increasing blood flo to the infected area, augmenting the number of circulating neutrophils, and elevating blood levels of numerous molecules (such as the microbial recognition proteins discussed

above) that have anti.infective functions. CYTOKINES AND OTHER MEDIATORS =ytokines can e/ert endocrine, paracrine, and autocrine effects (=hap. 3&;). "6F.<Nstimulates leukocytes and vascular endothelial cells to release other cytokines (as ell as additional "6F.<), to e/press cell.surface molecules that enhance neutrophil. endothelial adhesion at sites of infection, and to increase prostaglandin and leukotriene production. !hereas blood levels of "6F.<Nare not elevated in individuals ith localiLed infections, they increase in most patients ith severe sepsis or septic shock. (oreover, I9 infusion of "6F.<Ncan elicit the characteristic abnormalities of SIRS. In animals, larger doses of "6F.<Ninduce shock, disseminated intravascular coagulation (5I=), and death. Although "6F.<Nis a central mediator, it is only one of many proinflammatory molecules that contribute to innate host defense. =hemokines, most prominently interleukin (IA) ;, attract circulating neutrophils to the infection site. IA.#<Ne/hibits many of the same activities as "6F.<. "6F.<, IA.#<, interferon (IF6) <, IA.#%, and other cytokines probably interact synergistically ith one another and ith additional mediators. 8igh.mobility group 7.#, a transcription factor, can also be released from cells and interact ith microbial products to induce host responses late in the course of the septic response. COAGULATION FACTORS Intravascular thrombosis, a hallmark of the local inflammatory response, may help all off invading microbes and prevent infection and inflammation from spreading to other tissues. Intravascular fibrin deposition, thrombosis, and 5I= can also be important features of the systemic response. IA.? and other mediators promote intravascular coagulation initially by inducing blood monocytes and vascular endothelial cells to e/press tissue factor (=hap. G$). !hen tissue factor is e/pressed on cell surfaces, it binds to factor 9IIa to form an active comple/ that can convert factors O and IO to their enLymatically active forms. "he result is activation of both e/trinsic and intrinsic clotting path ays, culminating in the generation of fibrin. =lotting is also favored by impaired function of the protein =*protein S inhibitory path ay and depletion of antithrombin and protein =, hile fibrinolysis is prevented by increased plasma levels of plasminogen activator inhibitor #. "hus, there may be a striking propensity to ard intravascular fibrin deposition, thrombosis, and bleeding' this propensity has been most apparent in patients ith intravascular endothelial infections such as meningococcemia (=hap. #3?). =ontact.system activation occurs during sepsis but contributes more to the development of hypotension than to 5I=. CONTROL MECHANISMS Jlaborate control mechanisms operate ithin both local sites of inflammation and the systemic circulation. Local Control Mechanisms 8ost recognition of invading microbes ithin subepithelial tissues typically ignites immune responses that rapidly kill the invader and then subside to allo tissue recovery. "he anti.inflammatory forces that put out the fire and clean up the battleground

include molecules that neutraliLe or inactivate microbial signals. Among these molecules are A:S' intracellular factors (e.g., suppressor of cytokine signaling 3) that diminish the production of proinflammatory mediators by neutrophils and macrophages' anti.inflammatory cytokines (IA.#&, IA.+)' and molecules derived from essential polyunsaturated fatty acids (lipo/ins, resolvins, and protectins) that promote tissue restoration. S stemic Control Mechanisms "he signaling apparatus that links microbial recognition to cellular responses in tissues is less active in the blood. For e/ample, hereas A7: plays a role in recogniLing the presence of A:S, in plasma it also prevents A:S signaling by transferring A:S molecules into plasma lipoprotein particles, hich se4uester the lipid A moiety so that it cannot interact ith cells. At the high concentrations found in blood, A7: also inhibits monocyte responses to A:S, and the soluble (circulating) form of =5#+ strips off A:S that has bound to monocyte surfaces. Systemic responses to infection also diminish cellular responses to microbial molecules. =irculating levels of anti.inflammatory cytokines (e.g., IA.? and IA.#&) increase even in patients ith mild infections. Plucocorticoids inhibit cytokine synthesis by monocytes in vitro' the increase in blood cortisol levels early in the systemic response presumably plays a similarly inhibitory role. Jpinephrine inhibits the "6F.<Nresponse to endoto/in infusion in humans hile augmenting and accelerating the release of IA.#&' prostaglandin J% has a similar BreprogrammingC effect on the responses of circulating monocytes to A:S and other bacterial agonists. =ortisol, epinephrine, IA.#&, and =.reactive protein reduce the ability of neutrophils to attach to vascular endothelium, favoring their demargination and thus contributing to leukocytosis hile preventing neutrophil.endothelial adhesion in uninflamed organs. "he available evidence thus suggests that the bodyKs systemic responses to in0ury and infection normally prevent inflammation ithin organs distant from a site of infection. "he acute.phase response increases the blood concentrations of numerous molecules that have anti.inflammatory actions. 7lood levels of IA.# receptor antagonist (IA.#Ra) often greatly e/ceed those of circulating IA.#<, for e/ample, and this e/cess may result in inhibition of the binding of IA.#<Nto its receptors. 8igh levels of soluble "6F receptors neutraliLe "6F.<Nthat enters the circulation. Hther acute.phase proteins are protease inhibitors' these may neutraliLe proteases released from neutrophils and other inflammatory cells. Or#a" D*&f."$t!o" a"% S o$/ As the bodyKs responses to infection intensify, the mi/ture of circulating cytokines and other molecules becomes very comple/E elevated blood levels of more than G& molecules have been found in patients ith septic shock. Although high concentrations of both pro. and anti.inflammatory molecules are found, the net mediator balance in the plasma of these e/tremely sick patients

may actually be anti.inflammatory. For e/ample, blood leukocytes from patients ith severe sepsis are often hyporesponsive to agonists such as A:S. In patients ith severe sepsis, persistence of leukocyte hyporesponsiveness has been associated ith an increased risk of dying. Apoptotic death of 7 cells, follicular dendritic cells, and =5+Q " lymphocytes also may contribute significantly to the immunosuppressive state. ENDOTHELIAL IN!URY (ost investigators have favored idespread vascular endothelial in0ury as the ma0or mechanism for multiorgan dysfunction. In keeping ith this idea, one study found high numbers of vascular endothelial cells in the peripheral blood of septic patients. Aeukocyte.derived mediators and platelet.leukocyte.fibrin thrombi may contribute to vascular in0ury, but the vascular endothelium also seems to play an active role. Stimuli such as "6F.<Ninduce vascular endothelial cells to produce and release cytokines, procoagulant molecules, platelet.activating factor (:AF), nitric o/ide, and other mediators. In addition, regulated cell.adhesion molecules promote the adherence of neutrophils to endothelial cells. !hile these responses can attract phagocytes to infected sites and activate their antimicrobial arsenals, endothelial cell activation can also promote increased vascular permeability, microvascular thrombosis, 5I=, and hypotension. "issue o/ygenation may decrease as the number of functional capillaries is reduced by luminal obstruction due to s ollen endothelial cells, decreased deformability of circulating erythrocytes, leukocyteplatelet. fibrin thrombi, or compression by edema fluid. Hn the other hand, studies using orthogonal polariLation spectral imaging of the microcirculation in the tongue found that sepsis.associated derangements in capillary flo could be reversed by applying acetylcholine to the surface of the tongue or giving nitroprusside intravenously' these observations suggest a neuroendocrine basis for the loss of capillary filling. H/ygen utiliLation by tissues may also be impaired by a state of BhibernationC in hich A": production is diminished as o/idative phosphorylation decreases because of mitochondrial dysfunction' nitric o/ide or its metabolites may be responsible for inducing this response. Remarkably, poorly functioning BsepticC organs usually appear normal at autopsy. "here is typically very little necrosis or thrombosis, and apoptosis is largely confined to lymphoid organs and the gastrointestinal tract. (oreover, organ function usually returns to normal if patients recover. "hese points suggest that organ dysfunction during severe sepsis has a basis that is principally biochemical, not anatomic. . 7efore this vasodilatory phase, many patients e/perience a period during hich o/ygen delivery to tissues is compromised by myocardial depression, hypovolemia, and other factors. 5uring this BhypodynamicC period, the blood lactate concentration is elevated, and central venous o/ygen saturation is lo . Fluid administration is usually follo ed by the hyperdynamic, vasodilatory phase during

hich cardiac output is normal (or even high) and o/ygen consumption is independent of o/ygen delivery. "he blood lactate level may be normal or increased, and normaliLation of the central venous o/ygen saturation (SvH%) may reflect either improved o/ygen delivery or leftto. right shunting. :rominent hypotensive molecules include nitric o/ide, <.endorphin, bradykinin, :AF, and prostacyclin. Agents that inhibit the synthesis or action of each of these mediators can prevent or reverse endoto/ic shock in animals. 8o ever, in clinical trials, neither a :AF receptor antagonist nor a bradykinin antagonist improved survival rates among patients ith septic shock, and a nitric o/ide synthetase inhibitor, A.6Pmethylarginine 8=l, actually increased the mortality rate. Se-ere Se,&!&' ( S!"#le Pat o#e"e&!&0 In some cases, circulating bacteria and their products almost certainly elicit multiorgan dysfunction and hypotension by directly stimulating inflammatory responses ithin the vasculature. In patients ith fulminant meningococcemia, for e/ample, mortality rates have correlated ell ith blood endoto/in levels and ith the occurrence of 5I= (=hap. #3?). In most patients ith nosocomial infections, in contrast, circulating bacteria or bacterial molecules may reflect uncontrolled infection at a local tissue site and have little or no direct impact on distant organs' in these patients, inflammatory mediators or neural signals arising from the local site seem to be the key triggers for severe sepsis and septic shock. In a large series of patients ith positive blood cultures, the risk of developing severe sepsis as strongly related to the site of primary infectionE bacteremia arising from a pulmonary or abdominal source as eightfold more likely to be associated ith severe sepsis than as bacteremic urinary tract infection, even after the investigators controlled for age, the kind of bacteria isolated from the blood, and other factors. A third pathogenesis may be represented by severe sepsis due to superantigen.producing Staphylococcus aureus or Streptococcus pyogenes, since the " cell activation induced by these to/ins produces a cytokine profile that differs substantially from that elicited by gram.negative bacterial infection. In summary, the pathogenesis of severe sepsis may differ according to the infecting microbe, the ability of the hostKs innate defense mechanisms to sense it, the site of the primary infection, the presence or absence of immune defects, and the prior physiologic status of the host. Penetic factors may also be important. For e/ample, studies in different ethnic groups have identified associations bet een allelic polymorphisms in "AR+, caspase #%A, "6F.<, and IF6.<Ngenes and the risk of developing severe sepsis. Further studies in this area are needed. 1LINI1(L M(NIFEST(TIONS "he manifestations of the septic response are usually superimposed on the symptoms and signs of the patientKs underlying illness and primary infection. "he rate at hich signs and symptoms develop may differ from patient to patient, and there are striking individual variations in

presentation. 8yperventilation is often an early sign of the septic response. 5isorientation, confusion, and other manifestations of encephalopathy may also develop early on, particularly in the elderly and in individuals ith pree/isting neurologic impairment. Focal neurologic signs are uncommon, although pree/isting focal deficits may become more prominent. 8ypotension and 5I= predispose to acrocyanosis and ischemic necrosis of peripheral tissues, most commonly the digits. =ellulitis, pustules, bullae, or hemorrhagic lesions may develop hen hematogenous bacteria or fungi seed the skin or underlying soft tissue. 7acterial to/ins may also be distributed hematogenously and elicit diffuse cutaneous reactions. Hn occasion, skin lesions may suggest specific pathogens. !hen sepsis is accompanied by cutaneous petechiae or purpura, infection ith )eisseria meningitidis (or, less commonly, H" influenzae) should be suspected (Fig. #3?.#)' in a patient ho has been bitten by a tick hile in an endemic area, petechial lesions also suggest Rocky (ountain spotted fever (Fig. #?-.#). A cutaneous lesion seen almost e/clusively in neutropenic patients is ecthyma gangrenosum, usually caused by '" aeruginosa. It is a bullous lesion, surrounded by edema that undergoes central hemorrhage and necrosis (Fig. #+G.#). 8istopathologic e/amination sho s bacteria in and around the all of a small vessel, ith little or no neutrophilic response. 8emorrhagic or bullous lesions in a septic patient ho has recently eaten ra oysters suggest &" vulnificus bacteremia, hile such lesions in a patient ho has recently suffered a dog bite may indicate bloodstream infection due to Capnocytophaga canimorsus or C" cynodegmi. PeneraliLed erythroderma in a septic patient suggests the to/ic shock syndrome due to S" aureus or S" pyogenes. Pastrointestinal manifestations such as nausea, vomiting, diarrhea, and ileus may suggest acute gastroenteritis. Stress ulceration can lead to upper gastrointestinal bleeding. =holestatic 0aundice, ith elevated levels of serum bilirubin (mostly con0ugated) and alkaline phosphatase, may precede other signs of sepsis. 8epatocellular or canalicular dysfunction appears to underlie most cases, and the results of hepatic function tests return to normal ith resolution of the infection. :rolonged or severe hypotension may induce acute hepatic in0ury or ischemic bo el necrosis. (any tissues may be unable to e/tract o/ygen normally from the blood, so that anaerobic metabolism occurs despite near.normal mi/ed venous o/ygen saturation. 7lood lactate levels rise early because of increased glycolysis as ell as impaired clearance of the resulting lactate and pyruvate by the liver and kidneys. "he blood glucose concentration often increases, particularly in patients ith diabetes, although impaired gluconeogenesis and e/cessive insulin release on occasion produce hypoglycemia. "he cytokine.driven acute.phase response inhibits the synthesis of transthyretin hile enhancing the production of =.reactive

protein, fibrinogen, and complement components. :rotein catabolism is often markedly accelerated. Serum albumin levels decline as a result of decreased hepatic synthesis and the movement of albumin into interstitial spaces, hich is promoted by arterial vasodilation. M(2O+ 1OMPLI1(TIONS 1ar%!o,.lmo"ar* 1om,l!$at!o"& 9entilation.perfusion mismatching produces a fall in arterial :H% early in the course. Increasing alveolar capillary permeability results in an increased pulmonary ater content, hich decreases pulmonary compliance and interferes ith o/ygen e/change. :rogressive diffuse pulmonary infiltrates and arterial hypo/emia (:aH%@FIH%, >%&&) indicate the development of the acute respiratory distress syndrome (AR5S). AR5S develops in )G&, of patients ith severe sepsis or septic shock. Respiratory muscle fatigue can e/acerbate hypo/emia and hypercapnia. An elevated pulmonary capillary edge pressure (1#; mm8g) suggests fluid volume overload or cardiac failure rather than AR5S. :neumonia caused by viruses or by 'neumocystis may be clinically indistinguishable from AR5S. Sepsis.induced hypotension (see BSeptic Shock,C above) usually results initially from a generaliLed maldistribution of blood flo andblood volume and from hypovolemia that is due, at least in part, to diffuse capillary leakage of intravascular fluid. Hther factors that may decrease effective intravascular volume include dehydration from antecedent disease or insensible fluid losses, vomiting or diarrhea, and polyuria. 5uring early septic shock, systemic vascular resistance is usually elevated and cardiac output may be lo . After fluid repletion, in contrast, cardiac output typically increases and systemic vascular resistance falls. Indeed, normal or increased cardiac output and decreased systemic vascular resistance distinguish septic shock from cardiogenic, e/tracardiac obstructive, and hypovolemic shock' other processes that can produce this combination include anaphyla/is, beriberi, cirrhosis, and overdoses of nitroprusside or narcotics. 5epression of myocardial function, manifested as increased end.diastolic and systolic ventricular volumes ith a decreased e0ection fraction, develops ithin %+ h in most patients ith severe sepsis. =ardiac output is maintained despite the lo e0ection fraction because ventricular dilatation permits a normal stroke volume. In survivors, myocardial function returns to normal over several days. Although myocardial dysfunction may contribute to hypotension, refractory hypotension is usually due to a lo systemic vascular resistance, and death results from refractory shock or the failure of multiple organs rather than from cardiac dysfunction per se. +e"al 1om,l!$at!o"& Hliguria, aLotemia, proteinuria, and nonspecific urinary casts are fre4uently found. (any patients are inappropriately polyuric' hyperglycemia may e/acerbate this tendency. (ost renal failure is due to acute tubular necrosis induced by hypotension or capillary in0ury, although some patients also have glomerulonephritis,

renal cortical necrosis, or interstitial nephritis. 5rug.induced renal damage may complicate therapy, particularly hen hypotensive patients are given aminoglycoside antibiotics. 1oa#.lo,at * Although thrombocytopenia occurs in #&*3&, of patients, the underlying mechanisms are not understood. :latelet counts are usually very lo (>G&,&&&@<A) in patients ith 5I=' these lo counts may reflect diffuse endothelial in0ury or microvascular thrombosis. Ne.rolo#!$ 1om,l!$at!o"& !hen the septic illness lasts for eeks or months, Bcritical.illnessC polyneuropathy may prevent eaning from ventilatory support and produce distal motor eakness. Jlectrophysiologic studies are diagnostic. Puillain.7arre syndrome, metabolic disturbances, and to/in activity must be ruled out. L(3O+(TO+Y FINDINGS Abnormalities that occur early in the septic response may include leukocytosis ith a left shift, thrombocytopenia, hyperbilirubinemia, and proteinuria. Aeukopenia may develop. "he neutrophils may contain to/ic granulations, 5ohle bodies, or cytoplasmic vacuoles. As the septic response becomes more severe, thrombocytopenia orsens (often ith prolongation of the thrombin time, decreased fibrinogen, and the presence of 5.dimers, suggesting 5I=), aLotemia and hyperbilirubinemia become more prominent, and levels of aminotransferases rise. Active hemolysis suggests clostridial bacteremia, malaria, a drug reaction, or 5I=' in the case of 5I=, microangiopathic changes may be seen on a blood smear. 5uring early sepsis, hyperventilation induces respiratory alkalosis. !ith respiratory muscle fatigue and the accumulation of lactate, metabolic acidosis ( ith increased anion gap) typically supervenes. Jvaluation of arterial blood gases reveals hypo/emia, hich is initially correctable ith supplemental o/ygen but hose later refractoriness to #&&, o/ygen inhalation indicates right.to.left shunting. "he chest radiograph may be normal or may sho evidence of underlying pneumonia, volume overload, or the diffuse infiltrates of AR5S. "he electrocardiogram may sho only sinus tachycardia or nonspecific S"*". ave abnormalities. (ost diabetic patients ith sepsis develop hyperglycemia. Severe infection may precipitate diabetic ketoacidosis, hich may e/acerbate hypotension (=hap. 33;). 8ypoglycemia occurs rarely. "he serum albumin level, initially ithin the normal range, declines as sepsis continues. 8ypocalcemia is rare. DI(GNOSIS "here is no specific diagnostic test for the septic response. 5iagnostically sensitive findings in a patient ith suspected or proven infection include fever or hypothermia, tachypnea, tachycardia, and leukocytosis or leukopenia ("able %?G.#)' acutely altered mental status, thrombocytopenia, an elevated blood lactate level, or hypotension also should suggest the diagnosis. "he septic response can be 4uite variable,

ho ever. In one study, 3?, of patients ith severe sepsis had a normal temperature, +&, had a normal respiratory rate, #&, had a normal pulse rate, and 33, had normal hite blood cell counts. (oreover, the systemic responses of uninfected patients ith other conditions may be similar to those characteristic of sepsis. 6oninfectious etiologies of SIRS ("able %?G.#) include pancreatitis, burns, trauma, adrenal insufficiency, pulmonary embolism, dissecting or ruptured aortic aneurysm, myocardial infarction, occult hemorrhage, cardiac tamponade, post*cardiopulmonary bypass syndrome, anaphyla/is, and drug overdose. 5efinitive etiologic diagnosis re4uires isolation of the microorganism from blood or a local site of infection. At least t o blood samples (#& mA each) should be obtained (from different venipuncture sites) for culture. 7ecause gram.negative bacteremia is typically lo .grade (>#& organisms@mA of blood), prolonged incubation of cultures may be necessary' S" aureus gro s more readily and is detectable in blood cultures ithin +; h in most instances. In many cases, blood cultures are negative' this result can reflect prior antibiotic administration, the presence of slo .gro ing or fastidious organisms, or the absence of microbial invasion of the bloodstream. In these cases, PramKs staining and culture of material from the primary site of infection or of infected cutaneous lesions may help establish the microbial etiology. "he skin and mucosae should be e/amined carefully and repeatedly for lesions that might yield diagnostic information. !ith over helming bacteremia (e.g., pneumococcal sepsis in splenectomiLed individuals, fulminant meningococcemia, or infection ith &" vulnificus* (" pseudomallei, or $" pestis), microorganisms are sometimes visible on buffy coat smears of peripheral blood. SJ9JRJ SJ:SIS A65 SJ:"I= S8H=M :atients in hom sepsis is suspected must be managed e/peditiously. "his task is best accomplished by personnel ho are e/perienced in the care of the critically ill. Successful management re4uires urgent measures to treat the infection, to provide hemodynamic and respiratory support, and to eliminate the offending microorganism. (ost emergency centers no aim to initiate these measures ithin # h of the patientKs presentation ith severe sepsis or shock. Rapid assessment and diagnosis are therefore essential. (NTIMI1+O3I(L (GENTS Antimicrobial chemotherapy should be initiated as soon as samples of blood and other relevant sites have been cultured. A large retrospective revie of patients ho developed septic shock found that the interval bet een the onset of hypotension and the administration of appropriate antimicrobial chemotherapy as the ma0or determinant of outcome' a delay of as little as # h as associated ith lo er survival rates. It is important, pending culture results, to initiate empirical antimicrobial therapy that is effective against both gram.positive and gram.negative bacteria (Table 265-4). (a/imal recommended doses of antimicrobial

drugs should be given intravenously, ith ad0ustment for impaired renal function hen necessary. Available information about patterns of antimicrobial susceptibility among bacterial isolates from the community, the hospital, and the patient should be taken into account. !hen culture results become available, the regimen can often be simplified, as a single antimicrobial agent is usually ade4uate for the treatment of a kno n pathogen. (eta.analyses have concluded that, ith one e/ception, combination antimicrobial therapy is not superior to monotherapy for treating gram.negative bacteremia' the e/ception is that aminoglycoside monotherapy for '" aeruginosa bacteremia is less effective than the combination of an aminoglycoside ith an antipseudomonal <.lactam agent. (ost pa INITI(L (NTIMI1+O3I(L T)E+(PY FO+ SE5E+E SEPSIS 6IT) NO O35IO7S SO7+1E IN (D7LTS 6IT) NO+M(L +EN(L F7N1TION 1l!"!$al 1o"%!t!o" ("t!m!$rob!al +e#!me"& (I"tra-e"o.& T era,*) Immunocompetent adult "he many acceptable regimens include (#) ceftria/one (% g 4%+h) or ticarcillinclavulanate (3.# g 4+*?h) or piperacillin.taLobactam (3.3-G g 4+*?h)' (%) imipenemcilastatin (&.G g 4?h) or meropenem (# g 4;h) or cefepime (% g 4#%h). Pentamicin or tobramycin (G*- mg@kg 4%+h) may be a""e" to either regimen. If the patient is allergic to <.lactam agents, use ciproflo/acin (+&& mg 4#%h) or levoflo/acin (G&&*-G& mg 4#%h) #l$s clindamycin (?&& mg 4;h). If the institution or the community has a high prevalence of (RSA isolates, add vancomycin (#G mg@kg 4#%h) to each of the above regimens. 6eutropeniaa (>G&& neutrophils@<A) Regimens include (#) imipenem.cilastatin (&.G g 4?h) or meropenem (# g 4;h) or cefepime (% g 4;h)' (%) ticarcillin.clavulanate (3.# g 4+h) or piperacillin.taLobactam (3.3-G g 4+h) #l$s tobramycin (G*- mg@kg 4%+h). 9ancomycin (#G mg@kg 4#%h) should be added if the patient has an infected vascular catheter, if staphylococci are suspected, if the patient has received 4uinolone prophyla/is, if the patient has received intensive chemotherapy that produces mucosal damage, if the institution has a high incidence of (RSA infections, or if there is a high prevalence of (RSA isolates in the community. Splenectomy =efota/ime (% g 4?*;h) or ceftria/one (% g 4#%h) should be used. If the local prevalence of cephalosporin.resistant pneumococci is high, a"" vancomycin. If the patient is allergic to <.lactam drugs, vancomycin (#G mg@kg 4#%h) #l$s ciproflo/acin (+&& mg 4#%h) or levoflo/acin (-G& mg 4#%h) or aLtreonam (% g 4;h) should be used. I9 drug user 6afcillin or o/acillin (% g 4;h) #l$s gentamicin (G*- mg@kg 4%+h). If the local prevalence of (RSA is high or if the patient is allergic to <.lactam drugs, vancomycin (#G mg@kg 4#%h) ith gentamicin should be used. AI5S =efepime (% g 4;h), ticarcillin.clavulanate (3.# g 4+h), or piperacillin.taLobactam (3.3-G g 4+h) #l$s tobramycin (G*- mg@kg 4%+h) should be used. If the patient is allergic to <.lactam drugs, ciproflo/acin (+&& mg 4#%h) or levoflo/acin (-G& mg

4#%h) #l$s vancomycin (#G mg@kg 4#%h) #l$s tobramycin should be used. tients re4uire antimicrobial therapy for at least # eek' the duration of treatment is typically influenced by factors such as the site of tissue infection, the ade4uacy of surgical drainage, the patientKs underlying disease, and the antimicrobial susceptibility of the bacterial isolate(s). +EMO5(L OF T)E SO7+1E OF INFE1TION Removal or drainage of a focal source of infection is essential. Sites of occult infection should be sought carefully. Ind elling I9 catheters should be removed and the tip rolled over a blood agar plate for 4uantitative culture' after antibiotic therapy has been initiated, a ne catheter should be inserted at a different site. Foley and drainage catheters should be replaced. "he possibility of paranasal sinusitis (often caused by gram.negative bacteria) should be considered if the patient has undergone nasal intubation. In patients ith abnormalities on chest radiographs, =" of the chest may identify unsuspected parenchymal, mediastinal, or pleural disease. In the neutropenic patient, cutaneous sites of tenderness and erythema, particularly in the perianal region, must be carefully sought. In patients ith sacral or ischial decubitus ulcers, it is important to e/clude pelvic or other soft tissue pus collections ith =" or (RI. In patients ith severe sepsis arising from the urinary tract, sonography or =" should be used to rule out ureteral obstruction, perinephric abscess, and renal abscess. )EMODYN(MI18 +ESPI+(TO+Y8 (ND MET(3OLI1 S7PPO+T "he primary goals are to restore ade4uate o/ygen and substrate delivery to the tissues as 4uickly as possible and to improve tissue o/ygen utiliLation and cellular metabolism. Ade4uate organ perfusion is thus essential. Initial management of hypotension should include the administration of I9 fluids, typically beginning ith #*% A of normal saline over #*% h. "o avoid pulmonary edema, the pulmonary capillary edge pressure should be maintained at #%*#? mm8g or the central venous pressure at ;*#% cm 8%H. "he urine output rate should be kept at 1&.G mA@kg per hour by continuing fluid administration' a diuretic such as furosemide may be used if needed. In about one.third of patients, hypotension and organ hypoperfusion respond to fluid resuscitation' a reasonable goal is to maintain a mean arterial blood pressure of 1?G mm8g (systolic pressure, 1$& mm8g) and a cardiac inde/ of I+ A@min per m%. If these guidelines cannot be met by volume infusion, vasopressor therapy is indicated. =irculatory ade4uacy is also assessed by clinical parameters (mentation, urine output, skin perfusion) and, hen possible, by measurements of o/ygen delivery and consumption. A study of Bearly goal.directed therapyC (JP5") found that prompt resuscitation based on maintenance of the SvH% at 1-&, as associated ith significantly improved survival of patients ho ere admitted to an emergency department ith severe sepsis. "he treatment algorithm

included rapid administration of fluids, antibiotics, and vasopressor support' erythrocyte transfusion (to maintain the hematocrit above 3&,)' and administration of dobutamine if fluids, erythrocytes, and pressors did not result in an SvH% of 1-&,. "he e/tent to hich the different components of the JP5" algorithm contribute to the overall effect has not been e/amined in controlled trials. In particular, neither the use of SvH% to manage therapy nor the need for continuous SvH% monitoring ith a pulmonary artery catheter has been formally confirmed. A multicenter study (sponsored by the 6ational Institutes of 8ealth) of the efficacy of the JP5" approach is in progress. In patients ith septic shock, plasma vasopressin levels increase transiently but then decrease dramatically. Studies have found that vasopressin infusion can reverse septic shock in some patients, reducing or eliminating the need for catecholamine pressors. An ade4uately po ered and randomiLed trial of vasopressin infusion has not been performed. 9asopressin is a potent vasoconstrictor that may be most useful in patients ho have vasodilatory shock and relative resistance to other pressor hormones. Adrenal insufficiency is very likely hen the plasma cortisol level is >#G <g@dA in a patient ith severe sepsis. Penerally accepted criteria for partial adrenal insufficiency have not been devised' ma0or problems have been the inability to raise cortisol levels in e/tremely stressed individuals above high baseline values in response to cosyntropin (<#*%+*A="8) (=hap. 33?) and the high fre4uency of hypoalbuminemia, hich decreases total but not free (active) plasma cortisol levels. Adrenal insufficiency should be strongly considered in septic patients ith refractory hypotension, fulminant meningococcal bacteremia, disseminated tuberculosis, AI5S, or prior use of glucocorticoids, megestrol, etomidate, or ketoconaLole. 8ydrocortisone (G& mg I9 every ? h) may be given as a trial therapeutic intervention. If clinical improvement occurs over %+*+; h, most e/perts ould continue hydrocortisone therapy, tapering and discontinuing it after G*- days. Improved recommendations regarding hydrocortisone therapy may come from the Juropean =HR"I=2S trial. 9entilator therapy is indicated for progressive hypo/emia, hypercapnia, neurologic deterioration, or respiratory muscle failure. Sustained tachypnea (respiratory rate, 13& breaths@min) is fre4uently a harbinger of impending respiratory

collapse' mechanical ventilation is often initiated to ensure ade4uate o/ygenation, to divert blood from the muscles of respiration, to prevent aspiration of oropharyngeal contents, and to reduce the cardiac afterload. "he results of recent studies favor the use of lo tidal volumes (? mA@kg of ideal body eight, or as lo as + mA@kg if the plateau pressure e/ceeds 3& cm8%H). :atients undergoing mechanical ventilation re4uire careful sedation, ith daily interruptions' elevation of the head of the bed helps to prevent nosocomial pneumonia. Stress.ulcer prophyla/is ith a histamine 8%.receptor antagonist may decrease the risk of gastrointestinal hemorrhage in ventilated patients. "he use of erythrocyte transfusion continues to be debated. In the study of JP5", packed erythrocytes ere given to raise the hematocrit to 3&, if the patientKs SvH% as >-&,. "he e/tent to hich this intervention contributed to the improvement reported in patients ho received the JP5" regimen is uncertain. 7icarbonate is sometimes administered for severe metabolic acidosis (arterial p8 >-.%), but there is little evidence that it improves either hemo. dynamics or the response to vasopressor hormones. 5I=, if complicated by ma0or bleeding, should be treated ith transfusion of fresh.froLen plasma and platelets. Successful treatment of the underlying infection is essential to reverse both acidosis and 5I=. :atients ho are hypercatabolic and have acute renal failure may benefit greatly from hemodialysis or hemofiltration. GENE+(L S7PPO+T In patients ith prolonged severe sepsis (i.e., that lasting more than % or 3 days), nutritional supplementation may reduce the impact of protein hypercatabolism' the available evidence, hich is not strong, favors the enteral delivery route. :rophylactic hepariniLation to prevent deep venous thrombosis is indicated for patients ho do not have active bleeding or coagulopathy. Recovery is also assisted by preventing skin breakdo n, nosocomial infections, and stress ulcers. Investigators in 7elgium reported in %&&# that maintaining blood glucose levels in the normal range (;&*##& mg@dA) greatly improved survival rates among patients ho had 0ust undergone ma0or surgery and had received I9 glucose feeding for the previous %+ h. "he same group then studied intensive glucose control in critically ill medical patients and found a survival benefit only for patients ho remained in the intensive care unit for I3 days. 8ypoglycemia as much more common in the intensive.insulin group. 2ntil more e/perience ith intensive glucose control is reported, it seems reasonable to maintain glucose levels of >#G& mg@dA during the first 3 days of severe sepsis and then to target the normoglycemic range if the patient remains in the intensive care unit for a longer period. Fre4uent monitoring of blood glucose levels is indicated to avoid hypoglycemia during intensive insulin therapy. OT)E+ ME(S7+ES 5espite aggressive management, many patients ith severe sepsis or septic shock die. 6umerous interventions have been tested for their ability to improve survival in patients ith severe sepsis. "he list includes endoto/in.neutraliLing proteins, inhibitors of cycloo/ygenase or nitric o/ide synthase, anticoagulants, polyclonal immunoglobulins,

glucocorticoids, and antagonists to "6F.<, IA.#, :AF, and bradykinin. 2nfortunately, none of these agents has improved rates of survival among patients ith severe sepsis@septic shock in more than one large, randomiLed, placebo.controlled clinical trial. "his lack of reproducibility has had many contributing factors, including (#) heterogeneity in the patient populations studied and the inciting microbes and (%) the nature of the BstandardC therapy also used. A dramatic e/ample of this problem as seen in a trial of tissue factor path ay inhibitor (F!#9 265-1). !hereas the drug appeared to improve survival rates after -%% patients had been studied (p R .&&?), it did not do so in the ne/t #&3% patients, and the overall result as negative. "his inconsistency, even ithin a carefully selected patient population, argues strongly that a sepsis intervention should sho significant survival benefit in more than one placebo.controlled clinical trial before it is accepted as part of routine clinical practice. Recombinant activated protein = (a:=) as the first drug to be approved by the 2.S. Food and 5rug Administration (F5A) for the treatment of patients ith severe sepsis or septic shock. In a single randomiLed controlled trial in hich drug or placebo as given ithin %+ h of the patientKs first sepsis.related organ dysfunction, %;.day mortality as significantly lo er among recipients of a:= than among patients ho received placebo (%+.-, vs. 3&.;,' p >.&&G). In addition, a:= recipients ere more likely than placebo recipients to have severe bleeding (3.G, vs. %,). Survival improved only for patients ho had an A:A=8J II score of I%G during the %+ h before initiation of a:= infusion. (idtrial changes in the protocol and drug ere follo ed by improvement in the apparent efficacy of a:=. "he F5A approved a:= for use in adults (1#; years of age) ho meet the A:A=8J II criterion and have a lo risk of hemorrhage.related side effects. a:= is administered as a constant I9 infusion of %+ <g@kg per hour for $? h. Jach patientKs clotting parameters must be monitored carefully. a:= should not be given to patients ho have platelet counts of >3&,&&&@<A or to patients ho have dysfunction of one organ system and have had surgery during the previous 3& days. "reatment ith a:= should not be started 1%+ h after the onset of severe sepsis, nor should it be used in the patient subsetsDe.g., patients ith pancreatitis or AI5SDthat ere e/cluded from the clinical trial. Although the theoretical rationale for treating septic patients ith anticoagulants is strong and studies have found that a:= may have anti.inflammatory and anti.apoptotic properties in vitro, t o additional randomiLed, placebo.controlled trials of a:= ere stopped hen interim analyses sho ed lack of efficacy. Hne trial as in children, and the other as in adults ith A:A=8J II scores of F%G. a:= has not been tested again in the patient population for hich it as approved by the F5AE adults ith high A:A=8J II scores. Some e/perts have advocated BbundlingC of multiple therapeutic maneuvers into a unified, algorithmic approach to management that ould become the standard of care for severe sepsis. "he proposed resuscitation

(?.h) bundle incorporates most of the elements discussed above for acute (JP5") resuscitation. "he management (%+.h) bundle includes three measures of uncertain or marginal benefitE tight control of blood glucose, administration of lo .dose hydrocortisone, and treatment ith a:=. 7undling of therapies obscures the efficacy and to/icity of the individual interventions and allo s little room for individualiLing therapy. "he use of bundling in an industry.sponsored marketing program for a:= (the Surviving Sepsis =ampaign) has also been criticiLed. A careful retrospective analysis found that the apparent efficacy of all sepsis therapeutics studied to date has been greatest among the patients at greatest risk of dying before treatment' conversely, use of many of these drugs has been associated ith increased mortality rates among patients ho are less ill. "he authors proposed that neutraliLing one of many different mediators may help patients ho are very sick, hereas disrupting the mediator balance may be harmful to those hose adaptive defense mechanisms are still orking. "his analysis suggests that if more aggressive early resuscitation improves survival rates among sicker patients, it should become more difficult to sho additional benefit from other therapies' that is, if early resuscitation improves patientsK status, moving them into a Bless severe illnessC category, the addition of another agent is less likely to be beneficial. P+OGNOSIS Appro/imately %&*3G, of patients ith severe sepsis and +&*?&, of patients ith septic shock die ithin 3& days. Hthers die ithin the ensuing ? months. Aate deaths often result from poorly controlled infection, immunosuppression, complications of intensive care, failure of multiple organs, or the patientKs underlying disease. :rognostic stratification systems such as A:A=8J II indicate that factoring in the patientKs age, underlying condition, and various physiologic variables can yield estimates of the risk of dying of severe sepsis. Hf the individual covariates, the severity of underlying disease most strongly influences the risk of dying. Septic shock is also a strong predictor of short. and long.term mortality. =ase.fatality rates are similar for culture.positive and culture.negative severe sepsis. P+E5ENTION :revention offers the best opportunity to reduce morbidity and mortality. In developed countries, most episodes of severe sepsis and septic shock are complications of nosocomial infections. "hese cases might be prevented by reducing the number of invasive procedures undertaken, by limiting the use (and duration of use) of ind elling vascular and bladder catheters, by reducing the incidence and duration of profound neutropenia (>G&& neutrophils@<A), and by more aggressively treating localiLed nosocomial infections. Indiscriminate use of antimicrobial agents and glucocorticoids should be avoided, and optimal infectioncontrol measures (=hap. #%G) should be used. Several studies point to associations bet een allelic polymorphisms in specific genes and risk of

severe sepsis' if these associations prove to be broadly applicable, such polymorphisms can be used prospectively to identify high.risk patients and to target preventive and@or therapeutic measures to them. Studies indicate that G&*-&, of patients ho develop nosocomial severe sepsis or septic shock have e/perienced a less severe stage of the septic response (e.g., SIRS, sepsis) on at least one previous day in the hospital. Research is needed to develop ad0unctive agents that can damp the septic response before organ dysfunction or hypotension occurs.