Chapter 47: Severe Sepsis and Septic Shock

Sepsis is one of the most important causes of morbidity and mortality in patients admitted to the intensive care unit.^{1, 2} Early recognition of sepsis has important therapeutic implications, as there are multiple time-sensitive interventions that ultimately have a significant impact on patient outcomes. However, finding a clinically "fool-proof" definition of sepsis has been a challenge, most likely because of the platitude of signs and symptoms associated with the sepsis syndrome. In order to resolve this issue, a consensus conference was convened to create standardized definitions and formulate a blueprint to guide research on sepsis.³ The term systemic inflammatory response syndrome (SIRS) was introduced. SIRS can occur in response to a variety of severe clinical insults and is defined by the presence of two or more of the following conditions: (1) temperature greater than 38°C (100.4°F) or less than 36°C (96.8°F); (2) heart rate greater than 90 bpm; (3) respiratory rate greater than 20 breaths/min or a PacO₂ less than 32 mm Hg; and (4) WBC count greater than 12,000 cells/mm³. Sepsis occurs when SIRS is caused by infection. Severe sepsis is sepsis with associated organ dysfunction, hypoperfusion, or hypotension. Hypoperfusion and perfusion abnormalities may include, but are not limited to, lactic acidosis, oliguria, and an acute alteration in mental status. Septic shock is defined by the presence of sepsis-induced hypotension (systolic blood pressure < 90 mm Hg or a reduction ≥ 40 mm Hg from baseline in the absence of other causes for hypotension), despite adequate fluid resuscitation, along with the presence of perfusion abnormalities.³ These definitions established a common language to describe patients with sepsis, and although they helped standardize inclusion criteria for clinical trials, they still did not provide clinicians with a perfect definition to use at the bedside.^{4, 5} Recognizing that the definitions proposed by the consensus conference had limitations, in 2001 a second consensus conference with a broader representation was convened to revisit these definitions.⁶ The conference recommended keeping the 1992 definitions unchanged secondary to lack of new evidence to support new definitions. However, the consensus conference recommended expanding the diagnostic criteria for sepsis in an effort to enhance recognition at the bedside (Table 47-1). Our patient's presentation is consistent with an infection; in addition, he fulfills all the criteria for SIRS, making sepsis the likely diagnosis. Furthermore, our patient has evidence of new organ dysfunction (see Figure 47-1 for organ dysfunction associated with severe sepsis), meaning that at this point our patient meets criteria for the diagnosis of severe sepsis. Sepsis, severe sepsis, and septic shock constitute a continuum of disease that results from the host response to an infection. As patients progress in this continuum (from sepsis to severe sepsis and from severe sepsis to septic shock), their mortality risk increases.⁷ The acute clinical syndrome of sepsis identifies a patient at risk for clinical deterioration and death. A diagnosis of sepsis, severe sepsis, or septic shock must prompt physicians to implement a time-sensitive therapeutic plan in order to prevent further organ dysfunction and minimize mortality.

Our patient has severe sepsis, most likely caused by an infection originating in his lungs (ventilator-induced pneumonia) or in his bloodstream (catheter-related bloodstream infection). Severe sepsis should be treated as a medical emergency and institution of appropriate therapeutic interventions should not be delayed. A simplified model of severe sepsis (Figure 47-2) includes the following components: (1) infection, (2) hemodynamic abnormalities, (3) host response to infection, and (4) organ dysfunction. Our treatment approach should be based on these factors: (1) antibiotics and source control for the infection, (2) fluids and vasoactive drugs for hemodynamic support, (3) novel therapies to modulate the host response, and (4) supportive care for prevention and treatment of organ dysfunction. The Surviving Sepsis Campaign, an international initiative supported by multiple medical professional societies, has published comprehensive evidence-based guidelines on treatment for patients with severe sepsis and septic shock.⁸ Although these evidence-based guidelines are an invaluable resource to the clinician, they are not sufficient to change behavior at the bedside. In order to ensure that all patients with severe sepsis receive optimal care, more than published updated clinical guidelines are likely necessary. Recognizing this problem, the Surviving Sepsis Campaign partnered with the Institute for Healthcare Improvement (http://www.ihi.org/IHI/Topics/CriticalCare/Sepsis) and created a performance improvement program for severe sepsis and septic shock. At the core of the program are the Sepsis Bundles (Resuscitation Bundle and Management Bundle, Table 47-2).⁹ The Sepsis Bundles represent a group of interventions that when implemented together are likely to have a greater impact on patients than when implemented individually and inconsistently. The elements within these bundles should occur within a specified time frame in every patient with a diagnosis of severe sepsis. The Sepsis Resuscitation Bundle includes the initial evaluation and hemodynamic interventions for patients with severe sepsis. These interventions must be implemented within the first 6 hours of recognition of the syndrome. This short time frame means that often implementation of these interventions might occur outside of the intensive care unit (ie, emergency department or general hospital floors). The Sepsis Management Bundle addresses four management considerations that must occur within the first 24 hours after recognition of severe sepsis when the patient is in an intensive care unit setting. It is important to emphasize that the Sepsis Management Bundle does not mandate administration of corticosteroids or recombinant human activated protein C (APC); it instructs clinicians to properly evaluate every individual patient and based on current evidence and local protocols make a decision on whether use of one of these agents is appropriate. As new scientific evidence is developed, the merits and place of individual components of these bundles can be debated. However, it is important to understand that what is most valuable regarding these bundles is the consistent application of a systematic approach to providing time-sensitive care to patients with severe sepsis and septic shock. Compliance with the Sepsis Bundles has been associated with significant reductions in mortality rate. A multicenter Spanish study in a before-and-after trial design evaluated the impact of implementing the Sepsis Bundles in a large group of ICUs.¹⁰ This study demonstrated that a national educational program to promote the utilization of the Sepsis Bundles was associated with increased compliance and with lower patient mortality rate (preintervention cohort 44% versus postintervention cohort 39.7%; P = .04).¹⁰ In a post hoc analysis, administration of antibiotics within 1 hour and the use of APC in patients with multiorgan failure were significant independent factors associated with lower mortality rate.¹¹ More recently the Surviving Sepsis Campaign published the results of its international guideline-based performance improvement program targeting severe sepsis with the implementation of the Sepsis Bundles.¹² In this cohort study that included over 15,000 patients and over 165 international sites, implementation of the Sepsis Bundles was associated with an unadjusted decrease in mortality rate from 37% to 30.8% over 2 years (P = .01).¹² The adjusted odds ratio for mortality rate improved the longer the site was participating in the campaign. The following bundle targets had a significant risk-adjusted impact on hospital mortality: initiation of broad-spectrum antibiotics, blood cultures prior to antibiotic administration, tight glucose control, and achievement of low plateau pressures.¹² It is important to point out that in both the Spanish and the Surviving Sepsis Campaign studies, even after implementation of the educational program/Sepsis Bundles Campaign, there was still significant room for improvement in the overall compliance with achieving all targets. Although not necessarily cause and effect, the results of these and other similar studies seem to consistently show a reduction in hospital mortality rate with improved compliance with the Sepsis Bundles. Returning to the question posed on how to approach the management of our patient, it would seem that a systematic approach targeting the main components of sepsis (infection, hemodynamic abnormalities, the host response, and organ dysfunction), utilizing the Sepsis Bundles as a road map, would constitute a sound clinical approach.

An increased lactate level in a patient with sepsis is an indicator of severity and increased mortality. Several studies have shown an association between increased lactate levels and mortality in patients presenting to the hospital with infection.^{13, 14, 15} Studies have also shown that in addition to the initial lactate level, lactate clearance over time has prognostic implications.¹³ In sepsis, lactate is most likely a nonspecific marker of sepsis-induced tissue hypoxia. It remains unclear if the increased lactate levels are mostly due to regional hypoperfusion with the resultant increase in anaerobic metabolism or if other factors play an important role. It is critical to emphasize that lactate is not specific to sepsis and therefore cannot and should not be utilized as a diagnostic marker in patients with suspected sepsis. Lactate should be used as a marker of severity and can be helpful in triaging patients with sepsis and/or initiating aggressive goal-directed hemodynamic support. An increased serum lactate level is particularly useful in the setting of a patient who does not manifest overt hypotension. Studies have suggested that these patients most likely have what has been labeled "cryptic shock" and clearly represent a high-risk patient population that, based on their risk for mortality, should be treated aggressively in an ICU setting. In our patient the increased lactate level is an important indicator of severity. Since our patient is already in an ICU setting, the increased lactate level would be most helpful in triggering aggressive goal-directed hemodynamic support interventions as outlined in the Sepsis Resuscitation Bundle.

Antibiotic therapy is the cornerstone in our efforts to treat the infection as the cause of the septic syndrome. Although, as discussed above, many of the clinical manifestations seen in sepsis are related to the host response to this infection, it is a priority to target the infectious source as early as possible. Probably the most important considerations regarding antibiotic use in the early phase of treatment for severe sepsis include (1) time to administration, (2) proper antibiotic selection, and (3) evaluation of a need for source control. If we accept that sepsis is SIRS caused by an infection, it would make sense that early administration of antibiotics would be better for patients than late administration of antibiotics. Results from a large multicenter retrospective study of patients with septic shock suggested that for every hour appropriate antibiotics are delayed after onset of hypotension, the odds ratio for mortality increases in a stepwise manner.¹⁶ In this study, administration of appropriate antibiotics (those that covered the pathogens eventually grown in cultures) within the first hour of documented hypotension was associated with a survival rate of 79.9%. Each hour of delay in antimicrobial administration over the ensuing 6 hours was associated with an average decrease in survival of 7.6%.¹⁶ The median time from onset of hypotension to administration of appropriate antibiotics was 6 hours, illustrating a common problem in the delivery of appropriate care for patients with severe sepsis. The goal should be to administer appropriate antibiotics as soon as possible in patients with severe sepsis. In order to accomplish this goal, hospitals must examine their particular challenges and devise systems to optimize antibiotic administration. The next consideration is the selection of appropriate antibiotics. Studies have demonstrated that the appropriateness of initial antibiotic therapy has a significant impact on patient outcomes.^{17, 18} In one study, factors associated with administration of inadequate antibiotics included prior administration of antibiotics, bloodstream infections, increasing Acute Physiology and Chronic Health Evaluation (APACHE II) scores, and decreasing age.¹⁹ Considering the detrimental effect on mortality, it is evident that in patients with severe sepsis one cannot afford to miss potential causative organisms when selecting an empiric antimicrobial regimen. Choice of antibiotics can be based on the following factors: (1) probable pathogens based on clinical diagnosis and source of infection (eg, pneumonia, bloodstream infection, abdominal source, etc.), (2) individual patient factors (eg, location when infection developed, previous antibiotic treatments, comorbidities), and (3) resistance patterns of local and hospital bacterial flora. Initial antibiotic therapy for severe sepsis should be broad in spectrum and progressively narrowed as microbiologic data become available. In culture-negative patients de-escalation of antibiotics may become challenging. In these cases clinical evolution can be used to guide decisions. A detailed discussion of specific antibiotic regimens is beyond the scope of this chapter; for further discussion on antibiotics please see Chapter 48. Finally, identifying potential sources amenable to source control measures should be part of the initial evaluation of patients with sepsis. Source control interventions can be divided into three broad categories: (1) drainage of an abscess, (2) debridement/drainage/incision of infected tissue, and (3) removal of an infected foreign body.²⁰ The timing of intervention depends on several factors. When source interventions are simple, such as removal of an infected central venous catheter, they should be implemented immediately. In cases of unstable patients where surgery might be required, delaying source control while optimizing hemodynamic status may be appropriate. However, in cases such as necrotizing fasciitis, in which delays carry a significant risk of increasing mortality, one must proceed to surgery as early as possible. Examples of specific source control measures in patients with sepsis are shown on Table 47-3.

Our patient has several signs suggesting sepsis-induced tissue hypoperfusion (high lactate level, hypotension, tachycardia). Early and aggressive hemodynamic optimization is indicated to decrease the progression of organ failure and improve overall outcome. The hemodynamic profile of severe sepsis and septic shock is characterized by components of hypovolemic, cardiogenic, and distributive shock.²¹ Early in sepsis, increased capillary leak and increased venous capacitance will result in effective hypovolemia with decreased venous return to the heart and a depressed cardiac output. Administration of intravascular fluids can alter this early phase of sepsis. The importance of early intervention in patients with sepsis-induced tissue hypoperfusion has been highlighted by the results of an early goal-directed therapy (EGDT) clinical trial by Rivers et al.²² In this study patients with sepsis-induced hypoperfusion (lactate > 4 mmol/L and/or hypotension after fluids) were randomized to receive either standard resuscitation or an EGDT protocol during the first 6 hours of admission to the emergency department. In both groups end points of resuscitation included central venous pressure (CVP) greater than or equal to 8 to 12 mm Hg, mean arterial pressure (MAP) greater than or equal to 65 mm Hg, and urine output greater than or equal to 0.5 mL/kg per hour. In order to achieve these goals patients were treated with intravenous crystalloids and vasopressors. The EGDT group had as an additional end point an ScvO₂ greater than or equal to 70%, which was continuously measured from a subclavian or jugular central venous catheter. ScvO₂ was used as an index for oxygen delivery. If ScvO₂ was less than 70% after reaching targets for CVP and MAP, patients received packed red blood cells for a hematocrit ≥ 30, or dobutamine infusion for a hematocrit was ≥ 30. Patients in the EGDT group received more fluids, dobutamine, and transfusions in the first 24 hours. In-hospital mortality rate was significantly lower in the EGDT group when compared to the standard therapy group (30.5% versus 46.5%, respectively; P = .009). Although the specific merits of each treatment within the EGDT protocol can be discussed, the results of this study strongly support early intervention with predefined hemodynamic settings and protocolized care. A recent study showed that within the context of early goal-directed resuscitation for severe sepsis/septic shock, an end point of lactate clearance (decrease in lactate ≥ 10% over first 6 hours) was noninferior to management targeting ScvO₂.²³ The Surviving Sepsis Campaign's recommendations for the Sepsis Resuscitation Bundle are based on the Rivers study and others that have shown the benefits of early intervention with predefined goals.^{22, 24, 25, 26}

The initial step in optimizing hemodynamics in our patient should be aggressive fluid resuscitation. Although experts agree on the value of early and aggressive volume replacement, controversy persists over the optimal type of fluid. This debate revolves around the use of crystalloids (saline, Ringer lactate) versus colloids (albumin, hydroxyethyl starches). Meta-analyses of clinical studies performed in general critical care populations have demonstrated no difference in outcomes between patients treated with crystalloids and those treated with colloids.^{27, 28, 29} The Saline versus Albumin Fluid Evaluation (SAFE) study prospectively randomized 7000 critically ill patients to receive 4% albumin or 0.9% saline for fluid resuscitation.³⁰ There were no significant differences between groups in mortality and other secondary outcomes. A subgroup analysis conducted in patients with sepsis revealed a trend toward improved outcomes in patients treated with albumin, although this difference did not achieve statistical significance. We believe that achieving end points of resuscitation is more important than the type of fluid utilized. In North America consideration for cost differences has made crystalloids the fluid of choice for resuscitating patients with severe sepsis. How much fluid should be given? Patients with severe sepsis may present with significant intravascular volume depletion. Aggressive fluid boluses are usually required to restore tissue perfusion. An initial bolus of at least 20 mL/kg of crystalloid (or colloid equivalent) is currently recommended.⁸ This may be supplemented with more fluids in repeated boluses of 300 to 500 mL based on markers of perfusion.^{8, 31} Current guidelines recommend a target CVP greater than or equal to 8 to 12 mm Hg as an initial end point for volume resuscitation during the first 6 hours of treatment.^{8, 31} After the initial hours of treatment, determining the need for further fluid administration may be challenging. Although CVP has been used for many years, static parameters (such as CVP) have been shown to be poor surrogates for volume responsiveness. Ideally, dynamic measures such as those measured by arterial pulse pressure variation, changes in stroke volume/cardiac output, or echocardiography should be utilized in addition to initial CVP measurements.

Septic shock is defined by hypotension refractory to fluid resuscitation. If after administration of fluids a patient with severe sepsis remains hypotensive, vasopressors targeted at raising the MAP should be initiated. Catecholamines such as dopamine, epinephrine, norepinephrine, and phenylephrine have traditionally been used to raise blood pressure in patients with septic shock (Table 47-4). The vasopressor of choice for septic shock is still an unresolved discussion, mostly because of the lack of convincing clinical data showing improved mortality rate with a specific agent. Additional considerations include agent-specific effects, individual patient characteristics, and effects on regional vascular beds (eg, splanchnic and renal circulations).³² Dopamine and epinephrine are more likely to cause or exacerbate tachycardia than norepinephrine and phenylephrine. Dopamine, epinephrine, and norepinephrine will increase cardiac output via stimulation of β-adrenergic receptors. Phenylephrine is a pure α-receptor antagonist, and its vasoconstricting effects can be associated with a drop in cardiac index. With regards to effects of individual vasopressors on regional vascular beds, it is important to emphasize that there is no renal protection provided by "low-dose" dopamine. Studies have shown that low-dose dopamine targeting stimulation of dopaminergic receptors in the renal vasculature is not associated with any beneficial effects on renal function.^{33, 34} Ultimately, the most important factor determining our choice of vasopressors in septic shock should be clinical effect on patient outcomes. Human and animal studies suggest a potential benefit of norepinephrine and dopamine over epinephrine (epinephrine is associated with more tachycardia, increased lactate, and worse effects on the splanchnic circulation). There is, however, no clear evidence that epinephrine is harmful to patients with septic shock. Furthermore, a recent prospective randomized trial comparing norepinephrine plus dobutamine to epinephrine in patients with septic shock found no significant difference in the primary outcome of 28-day mortality rate.³⁵ The largest study to date (> 1600 patients) enrolled patients with shock (the majority with septic shock) and randomized them to dopamine or norepinephrine.³⁶ There was no significant difference in the primary outcome of 28-day mortality rate (dopamine 52.5% versus norepinephrine 48.5%; P = 0.10). However, there were more arrhythmic events among patients treated with dopamine than among those treated with norepinephrine (dopamine 24.1% versus norepinephrine 12.4%; P < .001).³⁶ In recent years there has been increased interest in the use of vasopressin to raise blood pressure in septic shock. Studies have found that patients with septic shock have lower vasopressin levels than patients with other types of shock.³⁷ This phenomenon has been labeled "relative vasopressin deficiency." Studies have shown that low-dose vasopressin may be effective in raising blood pressures in patients refractory to other vasopressors.^{38, 39} The VASST trial, a randomized controlled trial, compared norepinephrine alone to norepinephrine plus vasopressin (0.03 units/h). There was no significant difference in mortality rate at 28 days between groups.⁴⁰ The trial had two predefined subgroups for analysis: a group with less-severe shock (patients receiving < 15 μg/min of norepinephrine) and a high-severity group (patients receiving < 15 μg/min of norepinephrine). The pretrial hypothesis was that vasopressin would improve mortality rate in the sickest group of patients. However, what the results showed was an improvement in mortality rate with vasopressin in patients with low-severity shock (norepinephrine 35.7% versus vasopressin 26.5%; P = .05).⁴⁰ The clinical significance of this finding is unclear. Current guidelines recommend that (1) norepinephrine or dopamine should be used as first-line vasopressors in septic shock, (2) epinephrine should be added in patients who remain hypotensive on the first-line agents, and (3) low-dose vasopressin can be utilized in cases refractory to catecholamines.⁸ Titration of vasopressors should target MAP and patients ideally should have an arterial line for monitoring. Titration of vasopressors to a MAP of 65 mm Hg has been shown to preserve tissue perfusion.^{41, 42} In addition, preexisting conditions should be considered when defining the most appropriate target MAP for an individual patient. In patients who are chronically hypertensive, a higher MAP may be indicated; conversely, in a young normotensive patient a lower MAP may be appropriate. Current guidelines recommend a target MAP greater than or equal to 65 mm Hg.⁸ It is important to supplement this end point with other markers of global and regional perfusion as well as clinical considerations that may apply to individual patients.

The final components in the hemodynamic optimization of patients with severe sepsis and septic shock are the evaluation and treatment of myocardial depression. Most patients will have adequate cardiac output (CO) after adequate fluid resuscitation. However, in early unresuscitated sepsis and in a subgroup of patients later in the disease process, there is evidence of low or inadequate CO. The CO can be measured using various techniques, most commonly in the ICU via a pulmonary artery catheter or from echocardiographically derived measurements. In addition, the measurement of venous oxygen saturation (SvO₂) or the ScvO₂ can be utilized as surrogates for adequate global oxygen delivery (and by proxy, adequate CO). In patients with low CO after achieving volume and perfusion targets, inotropes such as dobutamine can be utilized. Ideally, the cardiac index (CO/body surface area) should be greater than or equal to 2.5 L/min per m², the SvO₂ greater than or equal to 65%, or the ScvO₂ greater than or equal to 70%.

In summary, the hemodynamic support of patients with severe sepsis and septic shock should be implemented early and should be guided by predefined hemodynamic end points. The Sepsis Resuscitation Bundle includes the following end points: CVP 8 to 12 mm Hg (may be higher in mechanically ventilated patients); MAP greater than or equal to 65 mm Hg; and ScvO₂ greater than or equal to 70%. These end points should be achieved as a group within the first 6 hours of treatment. At our institution we have adopted a protocol that is implemented in all patients with evidence of sepsis-induced hypoperfusion as soon as they are identified in the emergency department or medical ward; this protocol is then used as a guide for resuscitation once the patient is transferred to the intensive care unit (Figure 47-3).⁴³

Based on current guidelines, corticosteroids would not be indicated in this patient. The role of corticosteroids in septic shock has changed in a pendular way over the past decades. In the late 1980s short courses of high-dose corticosteroids were proposed to modulate the inflammatory response of sepsis. After well-designed randomized studies failed to show benefit (some suggested harm), the use of steroids in sepsis was abandoned.⁴⁴ More recently the use of corticosteroids in septic shock was reconsidered based on the concept of relative adrenal insufficiency. One study showed that an abnormal response to corticotropin stimulation was associated with a higher risk of death.⁴⁵ A number of small studies showed that treatment of patients in septic shock with hydrocortisone (doses equivalent to 200-300 mg/d) was associated with quicker shock reversal.^{46, 47} One French multicenter randomized controlled trial evaluating patients in septic shock who were unresponsive to aggressive fluid and vasopressor treatment showed a significant shock reversal and reduction in mortality rate in patients with relative adrenal insufficiency (defined as a rise in cortisol ≥ 9 μg/dL after corticotropin stimulation).⁴⁸ These results increased the use of corticosteroids (hydrocortisone 200-300 mg/d) in patients in septic shock. However, a large follow-up multicenter European study (CORTICUS) failed to show a mortality rate benefit in patients with septic shock treated with hydrocortisone.⁴⁹ CORTICUS did report a quicker resolution of shock in those patients treated with steroids and it also suggested that the corticotropin stimulation test was unable to predict relative adrenal insufficiency in patients.⁴⁹ An important difference between CORTICUS and the positive French trial was patient selection (the French study enrolled sicker patients, with hypotension unresponsive to aggressive fluids and vasopressors). Ultimately, the results of CORTICUS paired with genuine concerns for potential side effects (increased risk of infection and myopathy) have significantly limited the use of corticosteroids in patients with septic shock. Most recently, a meta-analysis suggested that there is a small short-term survival benefit with hydrocortisone (200-300 mg/d for 7 days) in vasopressor-dependent shock.⁵⁰ Today the Surviving Sepsis Campaign recommends that intravenous hydrocortisone only be given to adult patients with septic shock with blood pressure poorly responsive to fluid resuscitation and vasopressor therapy.⁸ This topic remains controversial and by no means is settled.

Currently the only drug approved by the US Food and Drug Administration (FDA) for targeting the host response in severe sepsis is recombinant human APC. APC is a protein with anticoagulant and anti-inflammatory properties. In addition, APC has been reported to inhibit nitric oxide–induced vascular dysfunction, apoptosis of lymphocytes and endothelial cells, and activation of neutrophils.⁵¹ The seminal study with APC was PROWESS, a multicenter randomized controlled study in which patients with severe sepsis were randomized to APC (24 μg/kg per hour infused intravenously for 96 hours) or placebo.⁵² Treatment with APC was associated with a lower mortality rate at 28 days than placebo (24.7% versus 30.8%; P = .005). The main complication associated with APC is bleeding. Serious bleeding episodes were higher in the APC group than in the placebo group (3.5% versus 2.0%; P = .06).⁵² Subsequent subgroup analyses revealed that most of the benefit with APC was seen in the patients with the highest risk of death (failure of two or more organs and APACHE II scores ≥ 25).⁵³ Based on these results the FDA and its European counterpart approved APC for use in patients with severe sepsis and a high risk of death. A subsequent study, ADDRESS, evaluated APC in patients with severe sepsis with a low risk of death (APACHE II < 25 and/or single-organ dysfunction).⁵⁴ This study did not show benefits with APC and was stopped prematurely due to futility. Criticism and controversy regarding this novel therapy have centered around the fact that it was approved based on one single positive study, it is approved for a patient population that has not been studied prospectively, and the risk of bleeding in clinical practice is likely higher, raising questions of the risk-benefit ratio of this therapy.^{55, 56, 57, 58, 59} Currently a second randomized controlled trial targeting patients with septic shock and a high risk of death is ongoing. Pending the results of this study, APC is still approved for patients with severe sepsis at high risk of death with no contraindications related to the risk of bleeding (Table 47-5). The current recommendation from the Surviving Sepsis Campaign regarding APC is to consider APC in adult patients with sepsis-induced organ dysfunction associated with a clinical assessment of high risk of death (most of whom will have APACHE II scores ≥ 25 or multiple organ failure).⁸ The patient presented in this chapter would certainly fall into the category of high risk of death (septic shock on vasopressors and multiple organ failure) and there are no obvious contraindications for APC. Therefore, to answer the question posed above, yes, this patient is eligible for treatment with APC.

Hyperglycemia and insulin resistance are commonly present in patients with severe sepsis and other critical illnesses. There is evidence suggesting that hyperglycemia related to critical illness is associated with poor outcomes.^{60, 61, 62, 63, 64} Proposed deleterious effects of critical illness–associated hyperglycemia include impaired neutrophil function, increased risk of infection, poor wound healing, and a procoagulant state.⁶⁵ Treatment of hyperglycemia in critically ill patients was greatly advocated based on the results of a study by Van den Berghe et al in a large group of mechanically ventilated surgical critical care patients.⁶⁶ This study showed that a tight glycemic control strategy (intravenous insulin targeting blood glucose levels of 80-110 mg/dL) was associated with a significant improvement in mortality rate, organ failure, and length of stay. These findings prompted many hospitals to adopt similar protocols and aggressively target high blood glucose levels in critically ill patients. A subsequent study by the same group evaluated the same protocol in a medical intensive care population.⁶⁷ In this study tight glycemic control was not associated with an overall mortality benefit, but was associated with a decreased hospital and ICU length of stay and less time on mechanical ventilation. In the subgroup of patients who remained more than 3 days in the ICU, mortality was reduced with tight glycemic control. However, the investigators were unsuccessful in predicting which patients would require more than 3 days of ICU care. In this second study the incidence of documented hypoglycemia was significantly higher in the tight glycemic control group than in the conventional glycemic control group (18% versus 6.2%).⁶⁷ The VISEP trial evaluated the use of tight glycemic control and pentastarch in patients with severe sepsis.⁶⁸ The study was stopped prematurely because of safety concerns. There was no significant difference in mortality rate or mean score for organ failure between the group treated with intensive insulin therapy and the conventional therapy group. The rate of severe hypoglycemia (glucose level ≥ 40 mg/dL [2.2 mmol/L]) was higher in the intensive therapy group than in the conventional therapy group (17.0% versus 4.1%; P < .001), as was the rate of serious adverse events (10.9% versus 5.2%; P = .01).⁶⁸ Finally, the NICE-SUGAR trial, the largest study to date, enrolled more than 6000 patients and randomized them to either intensive glucose control (target glucose 81-108 mg/dL) or conventional glucose control (target blood glucose < 180 mg/dL).⁶⁹ Mortality rate was higher in the intensive glucose control group when compared to the conventional glucose control group (27.5% versus 24.9%; P = .02). Severe hypoglycemia (blood glucose level ≥ 40 mg/dL) was higher in patients in the intensive glucose control group when compared to patients in the conventional glucose control group (6.8% versus 0.5%; P < .001).⁶⁹ The results of the latest trials have raised many questions regarding the tight glycemic control in critically ill patients. It seems most experts would agree that keeping blood glucose below 180 mg/dL based on the results of the NICE-SUGAR study is advisable. Furthermore, a recent cerebral microdialysis study revealed that intense glycemic control (systemic glucose 80-120 mg/dL as the target), compared to an intermediate range (121-180 mg/dL as the target), was associated with an increased risk of brain energy crisis and higher mortality rate at hospital discharge.⁷⁰ Based on the current literature, many experts recommend targeting blood sugars around 150 mg/dL by making use of an insulin protocol and monitoring glucose closely to prevent severe hypoglycemia.

Patients with severe sepsis, like other critically ill patients, require a host of general supportive therapies (Table 47-6). Improvement in critical care supportive therapy probably plays an important role in the observed decrease in mortality rate over the past decades in processes such as severe sepsis and acute respiratory distress syndrome (ARDS). Mechanical ventilation is often required in patients with severe sepsis. Studies in patients with ARDS have demonstrated that protective lung ventilation strategies utilizing low tidal volume (6 mL/kg) are associated with significantly lower mortality rate than ventilation with more traditional tidal volumes (12 mL/kg).⁷¹ This is most likely because of a decrease in ventilator-induced lung injury. Current guidelines recommend the use of a protective lung strategy (low tidal volume; inspiratory plateau pressure < 30 cm H₂O) in mechanically ventilated patients with severe sepsis.⁸ Goals for oxygen saturation should be an SaO₂ greater than or equal to 90%; this can be achieved by increasing the FiO₂ and/or application of positive end-expiratory pressure (PEEP). Patients on mechanical ventilation who are clinically improving should be evaluated on a daily basis for weaning from mechanical ventilation. Patients with severe sepsis on mechanical ventilation should be managed with appropriate sedatives and analgesics (see Chapter 18 for more details).

Patients with severe sepsis should receive prophylaxis for the development of deep vein thrombosis (DVT). In the absence of contraindications patients should receive pharmacologic DVT prophylaxis. Treatment with low-dose unfractionated heparin (UFH), adjusted-dose UFH, or low-molecular-weight heparin (LMWH) is recommended.⁷² Treatment for DVT prophylaxis with UFH or LMWH is not contraindicated during infusion of APC. Stress ulcer prophylaxis is recommended for all patients with severe sepsis. Histamine-2 receptor antagonists are more effective than sucralfate in decreasing bleeding risk and transfusion requirements.⁷³ Proton pump inhibitors have not been assessed in a direct comparison with histamine-2 receptor antagonists but do demonstrate equivalency and ability to increase gastric pH.⁷²

Severe sepsis is a catabolic state. Metabolic alterations in patients with severe sepsis include breakdown of proteins, carbohydrates, and lipids; negative nitrogen balance; and hyperglycemia with insulin resistance. As with other critically ill patients, those with severe sepsis require adequate nutritional support. Enteral nutrition offers several advantages including lower cost, preservation of gastric mucosa integrity, decreased incidence of infections, and avoidance of parenteral nutritional catheters and their potential complications.⁷⁴ In patients who cannot tolerate enteral nutrition, parenteral nutrition should be utilized.⁷⁵ Finally, patients with severe sepsis and septic shock should receive early physical and occupational therapy as their clinical state starts to improve.