Chapter 36. Sepsis and Septic Shock

Of the 120 million patients presenting to the emergency department (ED) in the United States per year, 2.9% or over 600,000 carry a diagnosis of severe sepsis and septic shock in the United States. The overall hospital mortality for sepsis, severe sepsis, and septic shock is 15%, 20%, and 45%, respectively. Sepsis is responsible for 9% of the deaths or 210,000 deaths per year in the United States. By comparison, 180,000 persons die of acute myocardial infarction and 200,000 die of lung or breast cancer annually. Many patients with severe sepsis and septic shock present to the ED where there are often long delays before transfer to an intensive care unit (ICU) bed. Sepsis is the most expensive diagnosis admitted to hospitals accounting for over $50 billion in health care costs each year. It is because of these aforementioned factors that the ED has become a logical focal point for sepsis diagnosis and treatment. It is during the first 6 hours that sepsis management can improve outcomes in one of every six patients who present with the disease.1

Risk factors and comorbidities that increase the incidence and mortality include age, gender, race, multidrug-resistant organisms, severity of chronic illnesses, and overcrowding of the EDs.2 Because ED visits increase with age, a higher proportion of patients will come from this population. The incidence of sepsis in adult patients over age 85 is 26.2/1,000 versus 0.2/1,000 in children.3 Over 115 million patients present to the ED per year and 2.9% of these patients have sepsis that accounts for over 600,000 annually, with a mean stay of 4.7 hours. Approximately 20.4% of the patients stay longer than 6 hours.4 The ED accounts for 50% of all hospital admissions for sepsis. Because of ED overcrowding and long lengths of stay, the ED becomes a significant portal of entry to realize improved outcomes.5

A series of pathogenic events are responsible for the transition from simple infection or sepsis to severe sepsis and septic shock. When an organism enters the body, the reaction is a systemic response that creates a systemic inflammatory response syndrome (SIRS). This reaction may be self-limited or may create a generalized systemic response. SIRS is the result of a release of proinflammatory and anti-inflammatory mediators. There is also a release of apoptotic proteins and an activation of the coagulation cascade. These processes can lead to malignant microvascular injury, thrombosis, and diffuse endothelial disruption, resulting in impaired tissue oxygenation. Furthermore, there is an imbalance between oxygen delivery and oxygen consumption. When organ dysfunction accompanies this response, this marks the onset of severe sepsis. Global tissue hypoxia and cytopathic (cellular) hypoxia develops, leading to multiple organ dysfunction and irreversible shock.

The American College of Chest Physicians (ACCP), the Society of Critical Care Medicine (SCCM), and International Sepsis Definitions Conference (ISDC) provided the definition of SIRS. SIRS is the systemic inflammatory response that is seen in a variety of infectious or noninfectious insults. The conference saw SIRS as a continuum, worsening and evolving to sepsis, severe sepsis, and septic shock (Table 36-1).6 The definitions were still vague and not specific; for the 2001 International Sepsis Definition Conference definitions, the SIRS definition was expanded in the attempt to increase early diagnosis (Table 36-2).7

Biomarkers as diagnostic, therapeutic, and prognostic tools in sepsis continue to evolve. C-reactive protein (CRP) and lactate were the first to be used in risk stratification. A more sensitive and specific marker is procalcitonin (PCT); it has higher sensitivity and specificity.8,9 PCT is currently being used in European countries to risk-stratify septic patients. In addition, PCT has been incorporated in recent guidelines as a first choice in management of sepsis. Recently, other markers have shown similar sensitivity and specificity as PCT. These markers include interleukin (IL)-1 receptor antagonist, protein C, and neutrophil gelatinase–associated lipocalin, which are key components of sepsis pathophysiology including inflammation, activation of coagulation, and renal/organ dysfunction, respectively.10 This multimarker approach may offer an advantage for identifying patients likely to develop severe sepsis and earlier implementation of sepsis bundles.

The Surviving Sepsis Campaign (SSC)11 was initiated in 2002 to improve the diagnosis, survival, and management of patients with sepsis by addressing the challenges associated with it. It developed three phases to carry its mission statement. Phase I started in October 2002 at the Barcelona Declaration at the 15th Annual Congress of the European Society for Intensive Care Medicine. It called for a 25% global reduction of mortality by 2009. The first-phase agenda includes the following points: public awareness, provider education, international standardization of care, improving early diagnosis and treatment, and providing access to post-ICU patient care.12

The second phase of the campaign was released as a set of guidelines for sepsis management in March 2004 and April 2004. The biggest achievement of this phase was the availability of unrestricted educational grants. The updated sepsis guidelines were released in 2008 and used the GRADE system. The GRADE system is a system developed to evaluate literature that includes grade of recommendation, assessment, development, and evaluation.

The early detection of high-risk patients facilitates the rapid implementation of the sepsis bundles. High-risk patients are defined as patients who remain hypotensive (systolic blood pressure <90 mm Hg) after a 20–40 mL/kg fluid challenge or who have a lactate greater than 4 mmol/L.13 Base deficit may be useful when present. However, up to 20% of patients may have a normal base deficit but elevated lactate.14 While single lactate levels are helpful, serial lactates (every 3–6 hours) carry significant prognostic implications for morbidity and mortality.15

In multivariate studies, inadequate antibiotic therapy was shown to be a significant risk factor for sepsis. There is a 10% increase in hospital mortality among patients who receive inadequate therapy.16 In addition, the failure to initiate antibiotics prior to or within the first hour of onset of hypotension is associated with an increased mortality of 7.6% for each hour antibiotic was delayed.17 The recommendation for early administration of intravenous (IV) broad-spectrum antibiotics should be within 3 hours of arrival to the ED and 1 hour for non-ED ICU admissions.18

Prior to initiating any antibiotic treatment, appropriate cultures should be obtained, and, if needed, lumbar puncture or stool cultures should be collected. The goal is to start the appropriate antibiotics that will most likely attack the offending organism. Another consideration is the local antibiotic resistance patterns. Drug cost, pharmacokinetics, and interaction should also be considered when selecting the appropriate antibiotic.19 The selection of the appropriate antibiotics also has significant outcome implications (Table 36-3).

The most common cause of sepsis is pneumonia (47%), intra-abdominal processes (18%), urinary tract infection (UTI) (18%), and bloodstream infection (12%). Other sites of infection include skin infections (rarely), nosocomial meningitis, and indwelling catheters.16,18

There are two different divisions of pneumonia: either nosocomial pneumonia or pneumonia without risk factors of multidrug-resistant pathogens. In nosocomial pneumonia, the most common organism is Enterobacteriaceae (30–40%), Pseudomonas aeruginosa (17–30%), and Staphylococcus aureus (7–15%). Treat this type of infection with a β-lactam active against P. aeruginosa ± aminoglycoside ± glycopeptide or linezolid if methicillin-resistant S. aureus (MRSA) is suspected. For pneumonia without risk factors for multidrug-resistant pathogens, the most common organism is S. aureus (45%), Haemophilus influenzae (20%), gram-negative bacilli (20%), and Streptococcus pneumoniae (9%). The recommended empirical antibiotic for this group is third-generation cephalosporin ± macrolide if intracellular bacteria are suspected.19

Another leading cause of sepsis includes UTI (60–70%), most commonly from Enterobacteriaceae including Escherichia coli (40%). This infection is usually treated with ciprofloxacin or if P. aeruginosa is suspected, ceftriaxone or ceftazidime ± aminoglycoside.19 A bloodstream infection related to a catheter caused by Staphylococcus sp. (50%) and Enterobacteriaceae (30%) should be treated with a glycopeptide or linezolid with β-lactam that has activity against P. aeruginosa.21

Other causes of sepsis include intra-abdominal sepsis (usually gram-negative bacilli [60%] that include E. coli [40%], P. aeruginosa [30%]); gram-positive cocci (30%) include Enterococcus species.22 Anaerobes and fungi can also cause sepsis. Treatment for gram-negative includes piperacillin–tazobactam or third/fourth-generation cephalosporin + metronidazole, or impenem ± fluconazole ± aminoglycoside.

Skin infection and nosocomial meningitides are rare causes of sepsis. Skin infection organisms include Streptococcus sp. and Staphylococcus sp.23 While nosocomial meningitides may be caused by gram-negative bacilli, more commonly, Acinetobacter sp., Staphylococcus sp., and Streptococcus sp. can be the causative organism.24

Source control is strategic in the early management of sepsis. It includes drainage of infected fluids, debridement of infected soft tissues, removal of infected devices or foreign bodies, and, finally, definite measures to correct anatomic derangement resulting in ongoing microbial contamination and to restore optimal function. A clear diagnostic approach to source identification and eradication should be determined within the first 6 hours of sepsis presentation.25

A goal-directed hemodynamic resuscitation within the first 6 hours of severe sepsis/septic shock includes a systematic approach to restoration of systemic oxygen delivery through a manipulation of preload (volume), afterload (blood pressure), and contractility (stroke volume) in order to preserve effective tissue perfusion while avoiding excessive increases in myocardial oxygen consumption (i.e., avoiding tachycardia and maintaining coronary perfusion pressure). There have been multiple studies that have shown when early goal-directed therapy (EGDT) is performed within the first 6 hours of disease presentation, outcomes are improved (Table 36-4).

Specifically, patients are managed by (Table 36-5): (1) fluid resuscitation with either crystalloid or colloid to achieve a central venous pressure (CVP) goal of 8–12 mm Hg, (2) vasoactive agents to achieve a mean arterial pressure (MAP) goal of 65–90 mm Hg, (3) blood transfusion to a hematocrit ≥30%, (4) inotrope therapy, and (5) intubation, sedation, and paralysis as necessary to achieve a central venous oxygen saturation (ScvO2) ≥70% as measured by intermittent or continuous central venous monitoring (Figure 36-1). EGDT should be employed as the first means of resuscitation with simultaneous prioritization of appropriate empirical antimicrobials and source control.19

Fluid Therapy

Decreased oral intake, increased insensible losses, arterial and venous dilation, and transudation of fluid into the extravascular fluid compartment lead to volume depletion. Because of these factors, initial fluid resuscitation should consist of rapid delivery of at least 20 mL/kg of crystalloid or colloid equivalent. If the patient needs further fluid resuscitation, it should be guided by CVP measurement. If CVP <8 mm Hg, the patient should receive 500-mL bolus every 30 minutes until the CVP >8 mm Hg.26,27 The type of fluid has not been shown to have outcome benefit. See Table 36-6 for fluid choices.

Early fluid administration should not be confused with the adverse effects of late or liberal fluid administration in acute lung injury. The Fluids and Catheters Treatment Trial (FACTT) followed patients for 43 hours after being admitted to the ICU and 24 hours after developing a lung injury. The study showed no difference in the 60-day mortality rate.33 Conservative fluid management showed significantly improved lung function, decreased need for mechanical ventilation, and improvement of central nervous system function secondary to decreased need for sedation.34 It must be remembered that this is after the initial resuscitation.

Vasopressor Use

The homeostatic balance between vasodilatation and vasoconstriction is altered in severe sepsis and septic shock.35 In sepsis, the predominant feature is systemic vasodilatation. Nitric oxide synthesis production increases in sepsis; the vascular smooth muscle relaxes. The endogenous vasoconstriction is less studied. In septic shock, smooth muscle is poorly responsive to norepinephrine. When the blood pressure decreases, endogenous arginine vasopressin (AVP) increases. AVP is released by the neurohypophysis to induce water conservation by the kidneys; thus, it helps to regulate osmotic pressures and cardiovascular homeostasis.36

Exogenous Vasopressors

If a patient continues to be hypotensive with vasopressor dependency after adequate volume resuscitation, central venous and arterial access should be obtained. In EGDT, if CVP <8 mm Hg, the patient should be given boluses of 500-mL aliquots every 30 minutes until CVP >8 mm Hg. In a severe hypotensive patient, vasopressors should be started sooner. Vasopressors should be started if MAP <65 mm Hg and CVP >8 mm Hg, or prior to fluid resuscitation in the presence of very low MAP.19,37Table 36-7 shows some exogenous vasopressors with their side effects.

Inotropic Therapy

Prior to EGDT, ICU-based studies using inotropic therapy targeting supranormal goals of oxygen delivery were associated with an increased mortality rate.53 In EGDT, the timing, patient selection, and physiologic goals were different. Dobutamine was started at a lower dose and titrated up to achieve ScvO2 ≥70% (Figure 36-1). For an initial resuscitation in a busy ED, ScvO2 can be used as a convenient surrogate for mixed venous oxygen saturation (SvO2).26,54 A resuscitation endpoint of SvO2 of 65% and ScvO2 of 70% is recommended by the SSC.

Blood Product Administration

The rationale for transfusing a severely septic or a septic shock patient due to impaired marrow response and erythropoietin levels may be to improve low hemoglobin levels in the presence of global tissue hypoxia.55 When septic shock or global tissue hypoxia is combined with anemia, transfusion is recommended. When resuscitating patients with large volumes of fluid, a dilutional anemia is common. In the presence of an elevated or low ScvO2, transfusion to a hematocrit of 30% is recommended.

When should packed red blood cells, fresh frozen plasma, antithrombin, and platelets be transfused? As per guidelines, red blood cells should be transfused when hemoglobin is <7 g/dL to a target of 7–9 g/dL except in patients with significant coronary artery disease or acute hemorrhage where hemoglobin >10 g/dL is maintained.19 Fresh frozen plasma should not be transfused to correct coagulopathy unless there is bleeding or a planned invasive procedure. Antithrombin therapy should not be used. Transfuse platelets when the count is <5,000/mm3 regardless of bleeding, or if count is 5,000–30,000/mm3 with significant bleeding risk.

Corticosteroids and the Septic Patient

In the stress of sepsis, the adrenal response may not be sufficient; thus, a relative adrenal insufficiency results. This has given rise to the existence of adrenal dysfunction. While early studies have shown that steroid replacement has no benefit and may cause harm,56–58 Annane et al in 2002 showed that replacement of hydrocortisone at 50 mg every 6 hours showed improved outcomes and decreased vasopressor use.59 More recently, the Corticosteroid Therapy of Septic Shock (CORTICUS) study enrolled a wider population range.60 Specifically, patients with systolic blood pressure less than 90 mm Hg for 1 hour regardless of vasopressor use, septic shock patients from the ICU up to the first 72 hours, and patients where etomidate was used were enrolled. By comparison, Annane et al only enrolled persistent hypotensive patients for 1 hour or greater and septic shock patients for the first 8 hours, and excluded any person where etomidate was used within 6 hours.59 CORTICUS showed no significant difference in 28-day mortality between corticosteroid and placebo.60 However, in a subset analysis of patients similar to the first trial, the mortality benefit was significant.

A clinical caveat is to withhold the use of steroid for 6–8 hours until the patient reaches the endpoints of EGDT. There is a 14% reduction in vasopressor use during the first 6 hours by giving fluids alone. By delaying steroid use in this setting, there will be a decreased use. In the latest 2008 guidelines of the SSC, the use of IV hydrocortisone is recommended for adult patients with persistent hypotensive septic shock where the blood pressure does not respond to fluid therapy and vasopressors. The guidelines went one step further by making the cosyntropin stimulation test elective and recommending against the use of dexamethasone as a substitute to hydrocortisone.19

In an ED rapid sequence intubation (RSI), etomidate is commonly used.61 In ICU studies, continuous infusion of etomidate in septic patients led to adrenal insufficiency; this in turn led to an increase in mortality.62 Another study in surgical patients showed that prolonged use of etomidate lowered plasma cortisol level.63 The use of a single dose of etomidate during RSI is currently under debate. Recently, ketamine is being used as an alternative RSI agent with a favorable hemodynamic profile. Further studies are needed in this area.

Activated Protein C

During sepsis, the inflammatory complexes (cytokines and coagulation mediators) in the body start to damage endothelial cells and coagulation in the microcirculation leading to end-organ dysfunction and finally death.64 The body inhibits coagulation by tissue factor pathway inhibitor, antithrombin, and the protein C system.65 Endogenous activated protein C inhibits thrombosis, inflammation, and apoptosis. Severe sepsis patients with low activated protein C are associated with increased mortality.66 The Recombinant Human Activated Protein C Worldwide Evaluation in Severe Sepsis (PROWESS) study demonstrated reduction of mortality rate from 30.9% in the placebo to 24.7% in the treatment group; the study was stopped early after these findings.64 The increased survival was confirmed only in higher severities of illness, as reflected by a higher APACHE II score ≥25 or multiple organ dysfunction. The Extended Evaluation of Recombinant Human Activated Protein C (ENHANCE) study showed similar mortality benefit as PROWESS with increase of bleeding; if recombinant human activated protein C (rhAPC) is started within the first 24 hours, then the mortality was 22.9% versus 27.4% if rhAPC is started >24 hours.67

In the ED, the use of rhAPC is limited only to instances where admission to the ICU is delayed.19 Pursuant to the current guidelines, rhAPC is recommended in the presence of sepsis-induced acute organ dysfunction associated with a clinical assessment of high risk of death (APACHE II scores ≥25) and with no absolute contraindications or relative contraindications that outweigh benefit (Table 36-8).

Ventilation with Low Tidal Volume

As mentioned previously, the lung is the most common site of infection. The inflammatory cascade in the lung leads to pulmonary parenchymal damage. This damage causes dysfunction in the respiratory system that may lead to acute respiratory distress syndrome (ARDS). ARDS is defined as an acute onset of respiratory failure with a ratio of arterial partial pressure of oxygen and fractional inspired concentration of oxygen (Pao2/FiO2) <200 regardless of positive end-expiratory pressure (PEEP), bilateral infiltrates, and pulmonary capillary wedge pressure (PCWP).68 A study was performed using lung-protective ventilation (LPV) (6 mL/kg tidal volumes and a plateau pressure <30 cm H2O) versus conventional volumes (ARDSNET study).69 This study showed that in patients with lung injury, LPV was associated with 9% absolute reduction in mortality. The ARDSNET study also showed that permissive hypercapnea (hypoventilation and mild respiratory acidosis [pH 7.3–7.45]) was tolerated.

Guidelines recommend in severe sepsis or septic shock to avoid high tidal volume coupled with high plateau pressures by decreasing tidal volumes to 6 mL/kg (based on predicted body weight) with the goal of maintaining end-inspiratory plateau pressure <30 cm H2O. Set PEEP based on severity of oxygenation deficit. Consider intermittent prone positioning treatment for patients who require potentially injurious FiO2 levels or plateau pressure >30 cm H2O.19 To prevent ventilator-associated pneumonia, raise head of bed 45° unless contraindicated.

Glycemic Control

Hyperglycemia is noted to be prevalent in critically ill patients. Intensive glucose control between 80 and 110 mg/dL showed to be more beneficial in surgical intensive care unit (SICU) versus medical intensive care unit (MICU) patients.70,71 In the MICU population, there was not any benefit to intensive glucose control. In fact, there were patients who developed hypoglycemia.71

The 2008 SSC recommends that after the first 6 hours, blood glucose <150 mg/dL and greater than lower limit of laboratory normal be maintained.19 In addition, continuous insulin and glucose infusion should be used and the blood glucose levels should be monitored every 30–60 minutes until the levels are stabilized; once stabilized, the levels should be monitored every 4 hours. The Campaign also recommends a longer ICU stay when patients receive IV insulin. All patients on IV insulin therapy should receive a glucose calorie source; caution should be observed when using point-of-care glucose to make sure that the plasma values are not overestimated.

Prophylaxis

Prophylactic unfractionated heparin or low-molecular-weight heparin should be administered as soon as the patient is admitted. A mechanical compression device should be used in patients with a contraindication for heparin. Stress ulcer prophylaxis should be provided with histamine 2 (H2) receptor antagonists.19

The evidence for the 6- and 24-hour bundle has shown repeatedly to improve outcome in numerous follow-up studies in both adults and children.72,73 EGDT in particular has been shown to be cost-effective by decreasing sepsis-related hospital costs by 20% even after taking into account all the training and equipment needed for this initiative.74 If an institution averages >16 septic patients per year, then the institution will average a 32.6% reduction of cost. These cost analysis studies have been repeated and validated.75,76

There are multiple models an institution can adopt. A coordinated patient care model requires convergence of multiple expertise and resources to be initiated by the ED until a patient arrives to the ICU. A second model requires a rapid response team that is mobile and can move within the hospital.77 This model covers the entire hospital and transfers to the ICU would be initiated for the management part of the bundle. A third model requires early ICU admission, and allows the ICU intensivist to start EGDT.

Early sepsis management is associated with morbidity, mortality, and health care resource consumption. Similar to acute myocardial infarction, stroke, and trauma, the management of sepsis is time-sensitive and requires emergent expertise in the ED.