There is no specific diagnostic test for sepsis. Diagnostically sensitive findings in a patient with suspected or proven infection include fever or hypothermia, tachypnea, tachycardia, and leukocytosis or leukopenia (Table 19-1); acutely altered mental status, thrombocytopenia, an elevated blood lactate level, respiratory alkalosis, or hypotension also should suggest the diagnosis. The systemic response can be quite variable, however. In one study, 36% of patients with severe sepsis had a normal temperature, 40% had a normal respiratory rate, 10% had a normal pulse rate, and 33% had normal white blood cell counts. Moreover, the systemic responses of uninfected patients with other conditions may be similar to those characteristic of sepsis. Examples include pancreatitis, burns, trauma, adrenal insufficiency, pulmonary embolism, dissecting or ruptured aortic aneurysm, myocardial infarction, occult hemorrhage, cardiac tamponade, postcardiopulmonary bypass syndrome, anaphylaxis, tumor-associated lactic acidosis, and drug overdose.
TREATMENT Severe Sepsis and Septic Shock
Patients in whom sepsis is suspected must be managed expeditiously. This task is best accomplished by personnel who are experienced in the care of the critically ill. Successful management requires urgent measures to treat the infection, to provide hemodynamic and respiratory support, and to remove or drain infected tissues. These measures should be initiated within 1 h of the patient’s presentation with severe sepsis or septic shock. Rapid assessment and diagnosis are therefore essential.
ANTIMICROBIAL AGENTS Antimicrobial chemotherapy should be started as soon as samples of blood and other relevant sites have been obtained for culture. A large retrospective review of patients who developed septic shock found that the interval between the onset of hypotension and the administration of appropriate antimicrobial chemotherapy was the major determinant of outcome; a delay of as little as 1 h was associated with lower survival rates. Use of “inappropriate” antibiotics, defined on the basis of local microbial susceptibilities and published guidelines for empirical therapy (see below), was associated with fivefold lower survival rates, even among patients with negative cultures.
It is therefore very important to promptly initiate empirical antimicrobial therapy that is effective against both gram-positive and gram-negative bacteria (Table 19-3). Maximal recommended doses of antimicrobial drugs should be given intravenously, with adjustment for impaired renal function when necessary. Available information about patterns of antimicrobial susceptibility among bacterial isolates from the community, the hospital, and the patient should be taken into account. When culture results become available, the regimen can often be simplified because a single antimicrobial agent is usually adequate for the treatment of a known pathogen. Meta-analyses have concluded that, with one exception, combination antimicrobial therapy is not superior to monotherapy for treating gram-negative bacteremia; the exception is that aminoglycoside monotherapy for P. aeruginosa bacteremia is less effective than the combination of an aminoglycoside with an antipseudomonal β-lactam agent. Empirical antifungal therapy should be strongly considered if the septic patient is already receiving broad-spectrum antibiotics or parenteral nutrition, has been neutropenic for ≥5 days, has had a long-term central venous catheter in place, or has been hospitalized in an ICU for a prolonged period. The chosen antimicrobial regimen should be reconsidered daily in order to provide maximal efficacy with minimal resistance, toxicity, and cost.
Most patients require antimicrobial therapy for at least 1 week. The duration of treatment is typically influenced by factors such as the site of tissue infection, the adequacy of surgical drainage, the patient’s underlying disease, and the antimicrobial susceptibility of the microbial isolate(s). The absence of an identified microbial pathogen is not necessarily an indication for discontinuing antimicrobial therapy because “appropriate” antimicrobial regimens seem to be beneficial in both culture-negative and culture-positive cases.
REMOVAL OF THE SOURCE OF INFECTION Removal or drainage of a focal source of infection is essential. In one series, a focus of ongoing infection was found in ~80% of surgical ICU patients who died of severe sepsis or septic shock. Sites of occult infection should be sought carefully, particularly in the lungs, abdomen, and urinary tract. Indwelling IV or arterial catheters should be removed and the tip rolled over a blood agar plate for quantitative culture; after antibiotic therapy has been initiated, a new catheter should be inserted at a different site. Foley and drainage catheters should be replaced. The possibility of paranasal sinusitis (often caused by gram-negative bacteria) should be considered if the patient has undergone nasal intubation or has an indwelling nasogastric or feeding tube. Even in patients without abnormalities on chest radiographs, computed tomography (CT) of the chest may identify unsuspected parenchymal, mediastinal, or pleural disease. In the neutropenic patient, cutaneous sites of tenderness and erythema, particularly in the perianal region, must be carefully sought. In patients with sacral or ischial decubitus ulcers, it is important to exclude pelvic or other soft tissue pus collections with CT or magnetic resonance imaging (MRI). In patients with severe sepsis arising from the urinary tract, sonography or CT should be used to rule out ureteral obstruction, perinephric abscess, and renal abscess. Sonographic or CT imaging of the upper abdomen may disclose evidence of cholecystitis, bile duct dilation, and pus collections in the liver, subphrenic space, or spleen.
HEMODYNAMIC, RESPIRATORY, AND METABOLIC SUPPORT The primary goals are to restore adequate oxygen and substrate delivery to the tissues as quickly as possible and to improve tissue oxygen utilization and cellular metabolism. Adequate organ perfusion is thus essential. Circulatory adequacy is assessed by measurement of arterial blood pressure and monitoring of parameters such as mentation, urine output, and skin perfusion. Indirect indices of oxygen delivery and consumption, such as central venous oxygen saturation, may also be useful. Initial management of hypotension should include the administration of IV fluids, typically beginning with 1–2 L of normal saline over 1–2 h. To avoid pulmonary edema, the central venous pressure should be maintained at 8–12 cmH2O. The urine output rate should be kept at >0.5 mL/kg per hour by continuing fluid administration; a diuretic such as furosemide may be used if needed. In about one-third of patients, hypotension and organ hypoperfusion respond to fluid resuscitation; a reasonable goal is to maintain a mean arterial blood pressure of >65 mmHg (systolic pressure >90 mmHg). If these guidelines cannot be met by volume infusion, vasopressor therapy is indicated. Titrated doses of norepinephrine should be administered through a central catheter. If myocardial dysfunction produces elevated cardiac filling pressures and low cardiac output, inotropic therapy with dobutamine is recommended. Dopamine is rarely used.
In patients with septic shock, plasma vasopressin levels increase transiently but then decrease dramatically. Early studies found that vasopressin infusion can reverse septic shock in some patients, reducing or eliminating the need for catecholamine pressors. Although vasopressin may benefit patients who require less norepinephrine, its role in the treatment of septic shock seems to be a minor one overall.
CIRCI (see “Adrenal Insufficiency,” above) should be strongly considered in patients who develop hypotension that does not respond to fluid replacement therapy. Hydrocortisone (50 mg IV every 6 h) should be given; if clinical improvement occurs over 24–48 h, most experts would continue hydrocortisone therapy for 5–7 days before slowly tapering and discontinuing it. Meta-analyses of recent clinical trials have concluded that hydrocortisone therapy hastens recovery from sepsis-induced hypotension without increasing long-term survival.
Ventilator therapy is indicated for progressive hypoxemia, hypercapnia, neurologic deterioration, or respiratory muscle failure. Sustained tachypnea (respiratory rate, >30 breaths/min) is frequently a harbinger of impending respiratory collapse; mechanical ventilation is often initiated to ensure adequate oxygenation, to divert blood from the muscles of respiration, to prevent aspiration of oropharyngeal contents, and to reduce the cardiac afterload. The results of recent studies favor the use of low tidal volumes (6 mL/kg of ideal body weight, or as low as 4 mL/kg if the plateau pressure exceeds 30 cmH2O). Patients undergoing mechanical ventilation require careful sedation, with daily interruptions; elevation of the head of the bed helps to prevent nosocomial pneumonia. Stress-ulcer prophylaxis with a histamine H2-receptor antagonist may decrease the risk of gastrointestinal hemorrhage in ventilated patients.
Erythrocyte transfusion is generally recommended when the blood hemoglobin level decreases to ≤7 g/dL, with a target level of 9 g/dL in adults. Erythropoietin is not used to treat sepsis-related anemia. Bicarbonate is sometimes administered for severe metabolic acidosis (arterial pH <7.2), but there is little evidence that it improves either hemodynamics or the response to vasopressor hormones. DIC, if complicated by major bleeding, should be treated with transfusion of fresh-frozen plasma and platelets. Successful treatment of the underlying infection is essential to reverse both acidosis and DIC. Patients who are hypercatabolic and have acute renal failure may benefit greatly from intermittent hemodialysis or continuous veno-venous hemofiltration.
GENERAL SUPPORT In patients with prolonged severe sepsis (i.e., that lasting more than 2 or 3 days), nutritional supplementation may reduce the impact of protein hypercatabolism; the available evidence favors the enteral delivery route. Prophylactic heparinization to prevent deep venous thrombosis is indicated for patients who do not have active bleeding or coagulopathy; when heparin is contraindicated, compression stockings or an intermittent compression device should be used. Recovery is also assisted by prevention of skin breakdown, nosocomial infections, and stress ulcers.
The role of tight control of the blood glucose concentration in recovery from critical illness has been addressed in numerous controlled trials. Meta-analyses of these trials have concluded that use of insulin to lower blood glucose levels to 100–120 mg/dL is potentially harmful and does not improve survival rates. Most experts now recommend using insulin only if it is needed to maintain the blood glucose concentration below ~180 mg/dL. Patients receiving intravenous insulin must be monitored frequently (every 1–2 h) for hypoglycemia.
OTHER MEASURES Despite aggressive management, many patients with severe sepsis or septic shock die. Numerous interventions have been tested for their ability to improve survival rates among patients with severe sepsis. The list includes endotoxin-neutralizing proteins, inhibitors of cyclooxygenase or nitric oxide synthase, anticoagulants, polyclonal immunoglobulins, glucocorticoids, a phospholipid emulsion, and antagonists to TNF-α, IL-1, platelet-activating factor, and bradykinin. Unfortunately, none of these agents has improved rates of survival among patients with severe sepsis/septic shock in more than one large-scale, randomized, placebo-controlled clinical trial. Many factors have contributed to this lack of reproducibility, including (1) heterogeneity of the patient populations studied, the primary infection sites, the preexisting illnesses, and the inciting microbes; and (2) the nature of the “standard” therapy also used. A dramatic example of this problem was seen in a trial of tissue factor pathway inhibitor. Whereas the drug appeared to improve survival rates after 722 patients had been studied (p = .006), it did not do so in the next 1032 patients, and the overall result was negative. This inconsistency argues that the results of a clinical trial may not apply to individual patients, even within a carefully selected patient population. It also suggests that, at a minimum, a sepsis intervention should show a significant survival benefit in more than one placebo-controlled, randomized clinical trial before it is accepted as routine clinical practice. In one attempt to reduce patient heterogeneity in clinical trials, experts have called for changes that would restrict these trials to patients who have similar underlying diseases (e.g., major trauma) and inciting infections (e.g., pneumonia). Other investigators have proposed using specific biomarkers, such as IL-6 levels in blood or the expression of HLA-DR on peripheral-blood monocytes, to identify the patients most likely to benefit from certain interventions.
Recombinant activated protein C (aPC) was the first immunomodulatory drug to be approved by the U.S. Food and Drug Administration (FDA) for the treatment of patients with severe sepsis or septic shock. Approval was based on the results of a single randomized controlled trial in which the drug was given within 24 h of the patient’s first sepsis-related organ dysfunction; the 28-day survival rate was significantly higher among aPC recipients who were very sick (APACHE II score, ≥25) before infusion of the protein than among placebo-treated controls. Subsequent trials failed to show a benefit of aPC treatment in patients who were less sick (APACHE II score, <25) or in children, and, a decade after its licensure by the FDA, the drug was withdrawn from the market when a European trial failed to confirm its efficacy in adults with sepsis. Agents in ongoing or planned clinical trials include intravenous immunoglobulin, a polymyxin B hemofiltration column, and granulocyte-macrophage colony-stimulating factor, which has been reported to restore monocyte immunocompetence in patients with sepsis-associated immunosuppression.
A careful retrospective analysis found that the apparent efficacy of all sepsis therapeutics studied to date has been greatest among the patients at greatest risk of dying before treatment; conversely, use of many of these drugs has been associated with increased mortality rates among patients who are less ill. It is possible that neutralizing one of many different mediators may help patients who are very sick, whereas disrupting the mediator balance may be harmful to patients whose adaptive defense mechanisms are working well. This analysis suggests that if more aggressive early resuscitation improves survival rates among sicker patients, it will become more difficult to obtain additional benefit from other therapies; that is, if an intervention improves patients’ risk status, moving them into a “less severe illness” category, it will be harder to show that adding another agent to the therapeutic regimen is beneficial.
THE SURVIVING SEPSIS CAMPAIGN An international consortium has advocated “bundling” of multiple therapeutic maneuvers into a unified algorithmic approach that will become the standard of care for severe sepsis. In theory, such a strategy would improve care by mandating measures that seem to bring maximal benefit, such as the rapid administration of appropriate antimicrobial therapy, fluids, and blood pressure support. Caution may be engendered by the fact that three of the key elements of the initial algorithm were eventually withdrawn for lack of evidence; moreover, the benefit of the current sepsis bundles has not been established in randomized controlled clinical trials.