Systemic Inflammatory Response Syndrome (SIRS)
• Diagnosis is based on the presence of two or more of the following:
• Temperature >38°C or <36°C
• Pulse >90 beats/min
• Respirations >20 breaths/min or arterial partial pressure of carbon dioxide (Paco2) <32 mm Hg
• White blood cell count >12,000 or <4000 cells/mm3 or >10% bands
• Diagnosis is based on a finding of SIRS plus proven or suspected infection as the cause.
• Diagnosis is based on a finding of sepsis plus organ dysfunction of one or more major systems (typically kidney, lung, or heart; less often, central nervous system).
• Diagnosis is based on a finding of severe sepsis plus persistent hypotension despite aggressive fluid resuscitation (i.e., vasopressors are required to maintain mean arterial pressure >65 mm Hg).
Supportive Laboratory Findings
• Lactic acidosis
• Acute renal failure
• Leukocytosis or leukopenia
• Elevated erythrocyte sedimentation rate or C-reactive protein
Supportive Physical Findings
• Decreased capillary refill or mottling of skin
• Mental status changes or obtundation
• Tachypnea or respiratory failure
• Anuria or oliguria
Initial Six Hours
• Initiate fluid resuscitation with crystalloid or colloid to achieve central venous pressure of 12 mm Hg (or 15 mm Hg if intubated).
• Add dopamine (Intropin) or norepinephrine (Levophed) for persistent hypotension (mean arterial pressure <65 mm Hg).
• Obtain blood, urine, and other appropriate cultures (cerebrospinal fluid, abscess drainage, catheter tip, tissue, sputum).
• Administer empiric antimicrobial therapy.
• Perform appropriate imaging studies with urgent source control as indicated and allowed by clinical status; remove potentially infected foreign bodies.
• All interventions should be undertaken simultaneously and initiated within 1 hour after making a presumptive diagnosis of sepsis.
• Maintain glycemic control with a target blood glucose level of less than 150 mg/dL.
• Use unfractionated or low-molecular-weight heparin for prophylaxis against deep venous thrombosis.
• Use a histamine 2 (H2) blocker or proton pump inhibitor for gastric ulcer prophylaxis.
• Initiate therapy with dobutamine (Dobutrex) for low cardiac output in the face of adequate filling pressures.
• Consider therapy with activated protein C (drotrecogin alfa [Xigris]) for patients with an Acute Physiology and Chronic Health Evaluation (APACHE) II score of 25 or greater.
• Consider therapy with hydrocortisone (Solu-Cortef)1 for patients with continued hypotension despite adequate fluid resuscitation and vasopressors.
• Achieve adequate sedation.
1 Not FDA approved for this indication.
The true prevalence and incidence of sepsis remain unknown. A study by Martin and colleagues in 2003 analyzed data from the National Hospital Discharge Survey from 1979 to 2000. Although some limitations apply, particularly shifts in the understanding of and use of coding, their findings have the advantage of assessing a large sample size over a prolonged period of observation. The most striking finding was a rise in the incidence of sepsis in the United States over this period, from an annual occurrence of 164,000 cases in 1979 to 660,000 cases in 2000. This reflects an average annual increase of 8.7% for more than 20 years. Coincident with this increase was a rise in the percentage of patients with a diagnosis of sepsis due to fungal organisms (4.6% in 2000). Gram-positive organisms supplanted gram- negative organisms as the largest group of pathogens after 1987 (52% versus 37%).
Mortality was highest among African American men and was associated with failure of three or more organs. Men were more likely to become septic than women (relative risk, 1.90). In-hospital mortality fell from 27.8% to 17.9% between 1995 and 2000. Total mortality increased, most likely as a result of the increase in the total burden of disease. This increase, although perhaps in part attributable to greater use of the diagnostic code for sepsis, is thought to have resulted from a combination of aging of the population, greater numbers of immunosuppressed individuals surviving longer, greater use of prosthetic devices, and greater numbers of and more invasive medical interventions. Diabetes, hypertension, chronic obstructive pulmonary disease, congestive heart failure, and HIV infection all increased as a proportion of medical conditions contributing to sepsis, whereas the proportion of patients with cancer and that of pregnant patients declined. The percentage of patients with one or more organs failing (i.e., severe sepsis) increased from 17% to 35% of all patients with a diagnosis of sepsis.
A study by Dombrovskiy and associates described an even greater increase in the proportion of severe sepsis, from 25.6% in 1993 to 43.8% in 2003. They also found an absolute increase in the number of cases of sepsis, as well as an increase in total mortality, again likely due to increased incidence. Case-fatality rates fell from 45.8% to 37.8%, but mortality was substantially higher at both the beginning and the end of their study period than was described by Martin. These findings suggest acceleration in both severity and incidence of severe sepsis and septic shock.
Sepsis is the tenth most common cause of death in the United States.
The incidence is highest during the winter, most likely reflecting increased respiratory viral infections that precede the development of community-acquired pneumonia, which is itself a medical condition with increased risk of developing severe sepsis.
A remarkable aspect of sepsis, severe sepsis, and septic shock is the relatively small number of associated pathogens out of the more than 1000 microorganisms known to cause human disease. Staphylococcus aureus, streptococci, enterococci, and gram-negative rods are the most commonly isolated species. Escherichia coli, Klebsiella, Pseudomonas, Serratia, Acinetobacter, Enterobacter, Citrobacter, and Neisseria meningitidis (a gram-negative coccus) constitute the most common gram-negative pathogens. This relative paucity of etiologic bacterial pathogens has implications for empiric antimicrobial therapy. The emergence of resistant organisms, particularly methicillin-resistant S. aureus (MRSA) and gram-negative bacteria harboring extended- spectrum β-lactamases, increases the risk of antibiotic failure and attendant mortality. Fungal pathogens, particularly non-candidal species, continue to increase in incidence. This is most likely the result of better empiric management of gram-negative and candidal sepsis in patients with hematologic malignancies.
Although sepsis is most typically a disease of the elderly, the immune suppressed, and those with chronic medical problems, otherwise young and healthy individuals can also be stricken with no clear cause or known risk factor, such as in meningococcemia, toxic shock syndrome, and necrotizing fasciitis.
The term sepsis derives from a Greek word that generally implies putrefaction. It also has a colloquial meaning, understood by most laypeople to mean a serious, potentially overwhelming infection. Historically, it has been intuitively understood by physicians to mean an infection, once localized, that has now disseminated and is life threatening. Bacteremia is implied if not always proven. Sepsis was usually fatal in the preantibiotic era, and its morbidity and mortality remain substantial.
In 2001, a consensus conference, sponsored by the American College of Chest Physicians and the Society of Critical Care Medicine and involving the European Society of Intensive Care Medicine, the Surgical Infection Society, and the American Thoracic Society, was convened to arrive at a specific definition of sepsis. Achieving such a definition is critically important to ongoing research on interventions aimed at improving mortality. Their deliberations, building on the work of a previous conference and published in 2003, elaborated several key concepts. They defined systemic inflammatory response syndrome (SIRS) as a state of immune activation characterized by the findings of fever with tachycardia, tachypnea, and/or leukocytosis or leukopenia. Although sensitive, this definition is so overly broad as to be rather unhelpful to an experienced clinician who would not need to resort to such terms to understand that a patient is seriously ill.
Nonetheless, it provides a useful construct in beginning to arrive at a specific definition of sepsis. Both infectious and noninfectious processes—severe burn and pancreatitis being the most notable examples of the latter—can cause SIRS.
Sepsis is defined as the finding of SIRS in the presence of known or suspected invasion by a microbe into a normally sterile body site.
Severe sepsis is sepsis that has become more generalized. The cardinal finding is organ dysfunction that is unrelated to the primary site of infection. Other typical findings include hyperglycemia, thrombocytopenia, hyperbilirubinemia, acidosis, coagulopathy, edema, oliguria, hypotension, ileus, hypoxia, and poor perfusion. Heart, kidney, and respiratory failure are the most common forms of organ dysfunction. Altered sensorium is also common. Septic shock refers to the presence of hypotension with systolic blood pressure lower than 90 mm Hg or mean arterial pressure lower than 60 mm Hg despite adequate fluid resuscitation. These terms (sepsis, severe sepsis, and septic shock) are all part of a continuum, implying a progressively graver degree of illness with associated increasing mortality. No single symptom, physical finding, organ dysfunction, or laboratory abnormality serves to make or exclude the diagnosis, although isolation of a microorganism in the setting of such findings is highly suggestive. Increasing numbers of physical findings and other abnormalities correlate with an increasing likelihood of the diagnosis of sepsis.
The consensus conference put forward a staging system for sepsis based on host predisposition, nature of infection, host response, and organ dysfunction (mnemonic: PIRO). By this, it is understood that a given patient may be predisposed to infection based on medical conditions such as diabetes, vascular disease, sickle cell anemia, or inherited or acquired immunodeficiencies. Additionally, there are likely to be more subtle genetic predispositions to the development of sepsis, as well as age-related factors. The nature of infection in terms of the site, inoculum, and virulence of the infecting organism clearly plays a role in determining the development of sepsis. The host response—ranging from localization and clearing of the infection without deleterious effect on the host to a state of immunologic dissonance whereby the inflammatory response is itself the driver of pathology—is signaled by the development of organ dysfunction. This staging system allows patients to be stratified at different points and serves as a template for evaluating the efficacy of various interventions as the characterization of sepsis and research into novel therapies progresses.
Infection of the immunocompetent host by a microorganism typically leads to immune activation. This serves to isolate the site and source of infection. Local tissue is often damaged, but eventually the infection is cleared, and repair and regeneration occur. This process is highly regulated, with a number of different cell types and mediators involved, all in delicate balance between proinflammatory and antiinflammatory effects. The patient may have few or no symptoms, or there may be systemic evidence of infection (i.e., SIRS). Sepsis is the failure of localization such that the process becomes generalized and leads to tissue destruction remote from the site of infection. Why the immune system enters this state of dysregulation remains unknown, although an enormous amount of research over the last 4 decades has elucidated many of the pathways and mediators involved. Tumor necrosis factor, platelet-activating factor, interleukins, eicosanoids, interferons, and nitric oxide are among the biologically active molecules characterized to date. Particular microbes also contribute to this process through the elaboration of toxins (typically by gram- positive organisms) and endotoxins (gram-negative–derived lipopolysaccharide). These events lead to tissue destruction as a result of ischemic insult, direct cytotoxicity, and accelerated apoptosis. The characterization of inflammatory mediators has led to attempts to modify the immune response through the use of novel therapies such as monoclonal antibodies directed against tumor necrosis factor. To date such attempts have not met with success, and investigation continues.
The diagnosis of sepsis ultimately relies on the clinical suspicion of infection in the setting of SIRS. A constellation of other supportive evidence establishes a greater or lesser likelihood of the presence of sepsis (see the Current Diagnosis box). It is rare that specific microbiologic evidence for infection is available in a manner timely enough to determine that sepsis is present or to help guide the initial, typically urgent, therapy. When it is available, it usually is a Gram stain or other type of specialized microbiology stain that confirms the presence of a potential pathogen in a site where none should be (e.g., gram-positive cocci in cerebrospinal fluid). Early therapy therefore relies on aggressive resuscitative measures and the administration of empiric antibiotics.
Early Goal-Directed Resuscitation: The First Six Hours
Initial treatment of sepsis should focus on correction of hemodynamic parameters, early administration of antibiotics, and source control of potential sites of infection. The 2008 guidelines from the Surviving Sepsis Campaign, an international initiative to improve sepsis outcomes, emphasized the importance of aggressive fluid resuscitation. Therapy should be implemented according to a protocol directed at achieving the following specific goals:
• Central venous pressure 8 to 12 mm Hg, or 12 to 15 mm Hg in those who are mechanically ventilated or have decreased left ventricular compliance
• Central venous or mixed venous oxygen saturation 70% or greater
• Mean arterial pressure (MAP) 65 mm Hg or greater
• Urine output 0.5 mL/kg/hour or greater
Administration of fluid boluses of 1000 mL or more of crystalloids or 300 to 500 mL of colloids over 30 minutes should begin as soon as hypoperfusion is recognized. There is no evidence that one type of fluid is superior to the other, although crystalloid is substantially cheaper. In cases of profound intravascular volume depletion, more rapid and more frequent fluid administration may be needed.
Hemodynamic improvement (decreased heart rate, increased blood pressure, increased urine output) and the goal of optimizing central venous pressure should direct the need for continued infusion of fluid while avoiding the development of volume overload and pulmonary edema. Transfusion of packed red blood cells should be considered if anemia is present, with a goal of achieving a hemoglobin level of 7.0 to 9.0 g/dL. If tissue hypoperfusion or hypoxia persists (central mixed venus oxygen saturation <70%) despite achieving a central venous pressure of 12 mm Hg, therapy with dopamine (Intropin) or norepinephrine (Levophed) should be initiated with a goal of achieving a MAP of 65 mm Hg. There is no role for the use of low- dose dopamine for renal protection. An arterial line for more precise and continuous measurement of blood pressure should be inserted as soon as possible after the initiation of vasopressor therapy. Ideally, vasopressors should not be introduced to increase MAP until after the fluid deficit has been corrected. However, in cases of severe shock, vasopressor therapy may be needed early in the resuscitation effort to improve perfusion to the peripheral vascular beds. If there is no response to dopamine and norepinephrine, the patient should be treated with epinephrine (Adrenalin).1
Appropriate antibiotics should be administered within 1 hour after diagnosis of severe sepsis or septic shock, because mortality increases in a linear fashion with each hour of delay. All efforts should be made to obtain appropriate cultures, in particular at least two sets of blood cultures. At least one of these should be peripheral, with the second from any long-term (>48 hours) vascular device. Cultures of urine, sputum, wounds, abscesses, and cerebrospinal fluid should also be obtained as appropriate and before the administration of antibiotics, assuming that such specimens can be obtained during the first hour.
Specific antibiotic choices are discussed later.
A survey for potential sources of infection should be performed, and early resuscitation efforts should occur concomitantly. Elimination of the source of infection is critical to reversing septic shock. Conditions that require emergent intervention, such as necrotizing fasciitis, cholangitis, and intestinal infarction, should be ruled out within the first 6 hours after presentation. Potentially infected indwelling devices should be removed as soon as possible. Necrotic tissue should be débrided and abscesses drained if either condition is detected.
Practitioners must consider the risks and benefits of the specific invasive procedures and the timing of such interventions for each patient individually. Every effort should be made to limit the invasiveness of necessary procedures, to avoid further stress in patients with an already hemodynamically fragile state. Imaging studies such as computed tomography of the head, chest, abdomen, and pelvis are necessary to identify or rule out potential sources of infection. An exception to the mandate to drain or débride infected collections is the presence of infected pancreatic necrosis, in which case surgical intervention should be delayed.
After hemodynamic parameters have been stabilized with fluid and vasopressors, cultures have been obtained, antibiotics have been administered, and initial source control of infected foci has been achieved, other interventions may be appropriate. Many of these are typical components of good critical care.
Patients who have adequate left ventricular filling pressures (as determined by a central venous pressure ≥12 mm Hg) but low cardiac output may benefit from therapy with dobutamine (Dobutrex) to increase cardiac output and improve tissue perfusion.
Activation of the hypothalamic-pituitary axis and the consequent increase in serum cortisol levels are vital aspects of the body’s acute stress response to shock. Recent data suggest that critical illness– related corticosteroid insufficiency is more prevalent in septic shock than previously thought, with rates as high as 60%. Therapy with corticosteroids is indicated only for those patients who have continued hypotension in the face of adequate fluid resuscitation and vasopressor support. Hydrocortisone (Solu-Cortef)1 should be administered intravenously 200–300 mg/day for seven days, either divided every 6 hours or as a continuous infusion. Dexamethasone (Decadron)1 should not be used unless hydrocortisone is not available. Because of the unclear long-term benefits and the known immunosuppressive side effects of corticosteroids, patients should be weaned from hydrocortisone as soon as vasopressors are no longer necessary. If another form of corticosteroid other than hydrocortisone is used, then fludrocortisone (Florinef)1 at a dose of 50 mcg/day should be added for mineralocorticoid effect.
Activated Protein C
Patients who are at increased risk of death with Acute Physiology and Chronic Health Evaluation (APACHE) II scores of 25 or higher and those with multiple organ dysfunction may benefit from infusion of activated protein C (drotrecogin alfa [Xigris]). This drug has numerous contraindications, including current active bleeding, recent (within 3 months) hemorrhagic stroke, recent (within 2 months) severe head trauma or intracranial or intraspinal surgery, trauma with a risk of life-threatening bleeding, presence of an epidural catheter, and intracranial neoplasm or mass lesion or evidence of herniation. It is not recommended for use in children. It is given as a 96-hour continuous infusion.
Maintenance of the blood glucose concentration lower than 150 mg/dL is associated with decreased mortality and length of stay in the intensive care unit. Control should be achieved with intravenous insulin, paying close attention to serum glucose levels every 1 to 2 hours until stable, with adjustments made on the basis of a validated protocol. Patients receiving intravenous insulin should simultaneously receive some form of glucose as a calorie source to minimize the risk of hypoglycemia.
Sedation and Paralytics
Sedation and treatment of pain should be aggressively managed according to validated protocols. Daily interruption of sedation allows for more accurate titration of drug and decreases the total time of mechanical ventilation. Paralytics should be avoided or used only briefly if required.
Patients should receive prophylaxis for deep venous thrombosis with either low-molecular-weight heparin or unfractionated heparin unless contraindicated by severe thrombocytopenia, recent intracranial bleeding, or coagulopathy. Those patients who cannot receive heparin should receive prophylaxis with graduated compression stockings or intermittent compression devices. Patients who are at especially high risk for deep venous thrombosis (e.g., prior history of clot, orthopaedic surgery, trauma) should receive both pharmacologic and mechanical prophylaxis. Low-molecular-weight heparin is preferred to unfractionated heparin in high-risk patients.
Patients should receive prophylaxis with a proton pump inhibitor or a histamine 2 (H2) blocker to prevent upper gastrointestinal bleeding.
There is no role for the administration of bicarbonate to correct acidosis or improve hemodynamic status.
Antibiotic choices should take into account the most likely pathogens for the suspected site or process. In general, initial empiric therapy (Table 1) should be broad, with an intention to narrow therapy once a microorganism has been isolated or a more precise clinical diagnosis has been made. Such a reevaluation should take place approximately 72 hours after the initiation of therapy. Numerous studies have documented the mortality associated with initial therapy that did not include agents active against the pathogen eventually isolated. In general, drugs from the β-lactam and related classes of antibiotics should be preferred for at least a part of most empiric regimens.
Empiric Antibiotic Choices for Severe Sepsis*
|Source||Antibiotic and Dose||Comments|
|Community- acquired pneumonia||Ceftriaxone (Rocephin) 2 g q24h plus azithromycin (Zithromax) 500 mg q24h||Should include atypical coverage; alternative is moxifloxacin (Avelox)|
|Health care– associated pneumonia||Piperacillin/tazobactam (Zosyn) 4.5 g q6h or meropenem (Merrem)1 2 g q8h3plus vancomycin (Vancocin)1 1 g q12h||Should cover for Pseudomonas and other resistant gram-negative rods|
|Neutropenia/fever||Piperacillin/tazobactam1 4.5 g q6h or meropenem1 2 g q8h3||Consider empiric fungal coverage for prolonged neutropenia|
|Abdominal sepsis||Ampicillin/sulbactam (Unasyn) 3 g q6h or piperacillin/tazobactam 4.5 g q6h||Consider coverage for yeast, MRSA|
|Urosepsis||Ampicillin/sulbactam1 3 g q6h or piperacillin/tazobactam1 4.5 g q6h||Obtain imaging and decompression as appropriate|
|Foreign body/vascular catheter–related sepsis||Piperacillin/tazobactam1 4.5 g q6h or meropenem1 2 g q8h3plus vancomycin 1 g q12h||Vascular catheters or other foreign bodies should be removed urgently|
|Meningitis||Ceftriaxone 2 g q12h plus vancomycin1 750 mg q8h plus rifampin (Rifadin)1 600 mg q24h||Consider steroid therapy before or simultaneously with administration of antibiotics|
|Soft-tissue infection||Cefazolin (Ancef) 2 g q8h plus vancomycin 1 g q12h||Image for abscess with débridement as appropriate|
|Necrotizing fasciitis||Ampicillin/sulbactam 3 g q6h plus vancomycin 1 g q12h plus clindamycin (Cleocin) 900 mg q8h||Obtain urgent surgical consultation|
|Unknown||Piperacillin/tazobactam 4.5 g q6h or meropenem 2 g q8h3plus vancomycin 1 g q12h plus tobramycin (Tobrex) 7 mg/kg q24h||Obtain appropriate imaging studies, especially of abdomen, pelvis, central nervous system|
1 Not FDA approved for this indication.
3 Exceeds dosage recommended by the manufacturer.
* In all cases, prior antimicrobial therapy, kidney and liver dysfunction, and the probable source of sepsis should be carefully considered. Always consider coverage for methicillin- resistant Staphylococcus aureus (MRSA) in areas where incidence in bloodstream isolates is >10%.
Special considerations include patients with neutropenia and fever, who should always be treated with at least one agent active against Pseudomonas. Some debate continues regarding the use of two anti- pseudomonal drugs as part of the initial antibiotic regimen. Given the possibility of resistance on the part of this pathogen, it would seem reasonable to use two drugs initially, until Pseudomonas has been isolated (if present) and its susceptibilities are known, allowing coverage to be narrowed. The Surviving Sepsis Campaign guidelines advocate this approach. There is no benefit in treating with two drugs known to be active in an attempt to achieve a supposed synergy.
Patients with hematologic malignancies are at increased risk for sepsis from fungal organisms. Severe sepsis or septic shock in such patients warrants empiric treatment with an echinocandin, a broad-spectrum azole such as voriconazole (Vfend) or posaconazole (Noxafil), or amphotericin (Fungizone).
MRSA continues to increase in incidence nationally and is now common as a community-acquired pathogen. It is also to be suspected as a cause of postinfluenza bacterial pneumonia. Empiric treatment with an antibiotic active against this bacterium, such as vancomycin (Vancocin), linezolid (Zyvox), or daptomycin (Cubicin), should be strongly considered in septic patients in communities where the rate of MRSA in bloodstream infections exceeds 10%. This pathogen should always be considered in a patient with a long-term intravenous catheter, prosthetic device, or other indwelling foreign body. Although vancomycin-resistant S. aureus is extremely rare, caution should be taken when using daptomycin and linezolid as empiric therapy, because resistance has been reported.
Prior administration of antibiotics and the attendant risk of infection by a pathogen resistant to the previous therapy should be considered in arriving at a course of empiric therapy. A typical scenario is a patient who presents with a catheter-related bloodstream infection while taking vancomycin. One would expect a gram-negative bacterium, a fungal organism, or, potentially, a vancomycin-resistant enterococcus as the pathogen. Recent hospitalization or residence in a nursing home places patients at risk for colonization and subsequent infection with resistant gram-negative rods.
Prognosis and Limits of Care
Patients who present with severe sepsis or septic shock often have substantial prior medical morbidity, decreasing their chance of survival. The overall mortality rate remains between 20% and 40%. Patients often have expressed wishes regarding limits of care to family members or others close to them before becoming ill. It is always appropriate to discuss goals of therapy, possible and probable outcomes, and plans for further evaluation and treatment with patients (if possible) and their proxies in all instances. Decisions to proceed with or limit care should be made within the context of a patient’s expressed or expected wishes and should take into account unfolding clinical data and circumstances. Time spent in this endeavor can substantially decrease the amount of futile care rendered to a patient and lead to care that more truly reflects the patient’s wishes regarding life-prolonging measures. The stress and anxiety experienced by family members may also be reduced.
Substantial progress has been made in the last 3 decades in decreasing the mortality associated with severe sepsis and septic shock. Nevertheless, the mortality rate remains unacceptably high, and the overall incidence and severity of disease appear to be increasing, by as much as 1.5% annually by some estimates. The risk of death for an individual patient appears to stabilize approximately 6 months after the original illness. Many patients who do survive remain with the same risk factors (e.g., diabetes, vascular disease, prosthetic devices, immunosuppression) that contributed to their infection and therefore are at risk for recurrence. Moreover, certain organisms, such as MRSA, resistant gram-negative rods, and fungi, remain difficult to treat, and success rates are relatively low despite aggressive, timely, and prolonged therapy.
There is some cause for optimism. The Surviving Sepsis Campaign has now entered phase III. This is a program in which a core set of recommendations, described in the guidelines, is being implemented with opportunities to measure outcomes, assess physician behavior, and provide feedback to improve survival through evidence-based interventions. This effort involves more than 12,000 patients in 239 hospitals in 17 countries and will undoubtedly change the future course of this lethal disease.
1. Bone R.C. Immunologic dissonance: A continuing evolution in our understanding of the systemic inflammatory response syndrome (SIRS) and the multiple organ dysfunction syndrome (MODS). Ann Intern Med. 1996;125:680–687.
2. Delinger R.P., Levy M.M., Carlet J.M., et al. Surviving Sepsis Campaign: Internatonal guidelines for management of severe sepsis and septic shock—2008. Crit Care Med. 2008;36:296–327.
3. Dombrovskiy V.Y., Martin A.A., Sunderram J., Paz H.L. Rapid increase in hospitalization and mortality rates for severe sepsis in the United States: A trend analysis from 1993 to 2003. Crit Care Med. 2007;35:1244–1250.
4. Ibrahim E.H., Sherman G., Ward S., et al. The influence of inadequate antimicrobial treatment of bloodstream infections on patient outcomes in the ICU setting. Chest. 2000;118:146– 155.
5. Jimenez M.F., Marshall J.C. Source control in the management of sepsis. Intensive Care Med. 2001;27(Suppl. 1):S49–62.
6. Kumar A., Roberts D., Wood K.E., et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med. 2006;34:1589–1596.
7. Leibovici L., Shraga I., Drucker M., et al. The benefit of appropriate empirical antibiotic treatment in patients with bloodstream infection. J Intern Med. 1998;244:379–386.
8. Levy M.M., Fink M.P., Marshall J.C., et al. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Intensive Care Med. 2003;29:530–538.
9. Martin G.S., Mannino D.M., Easton S., et al. The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med. 2003;348:1546–1554.
10. McDonald J.R., Friedman N.D., Stout J.E., et al. Risk factors for ineffective therapy in patients with bloodstream infection. Arch Intern Med. 2005;165:308–313.
11. Miller P.J., Wenzel R.P. Etiologic organisms as independent predictors of death and morbidity associated with bloodstream infection. J Infect Dis. 1987;156:471–477.
12. Sasse K.C., Nauenberg E., Long A., et al. Long-term survival after intensive care unit admission with sepsis. Crit Care Med. 1995;23:1040–1047.
1 Not FDA approved for this indication.