Case Vignette
A Woman with Septic Shock
Rebecca E. Berger, M.D.
Ms. Jones is a 65-year-old woman with a history of hypertension who presents to the emergency department with a 3-day history of chills and dysuria. The only medication she is taking is amlodipine, at a dose of 10 mg daily; she had had normal electrolyte levels and renal function at a routine visit 6 weeks earlier. On arrival at the emergency department, she reports feeling dizzy.
She is 165 cm (65 in.) tall and weighs 70 kg (154 lb). Her temperature is 38.6°C (101.5°F), heart rate 125 beats per minute, blood pressure 85/55 mm Hg (mean arterial pressure, 65 mm Hg), respiratory rate 28 breaths per minute, and oxygen saturation as measured by pulse oximetry 94% while she is breathing ambient air. A physical examination reveals dry mucous membranes; undetectable jugular venous pulsation; tachycardia without gallops, rubs, or murmurs; clear lungs; and warm extremities. She has tenderness on palpation of her suprapubic region. You begin intravenous administration of a bolus of crystalloid solution.
Laboratory testing shows a creatinine level of 1.8 mg per deciliter (159 μmol per liter) (normal range, 0.5 to 1.1 mg per deciliter [44 to 97 μmol per liter]), blood urea nitrogen 76 mg per deciliter (27 mmol per liter) (normal range, 7 to 20 mg per deciliter [2 to 7 mmol per liter]), lactate 5.0 mmol per liter (normal value, <2.0), anion gap 25 mmol per liter (normal range, 8 to 15), white-cell count 20,000 per cubic millimeter (normal range, 4500 to 11,000), and hemoglobin 9.0 g per deciliter (normal range, 12.0 to 15.5). Urinalysis shows 3+ leukocyte esterase, more than 100 white cells per high-power field, and abundant bacteria.
You make a presumptive diagnosis of sepsis from a urinary source and begin treatment with intravenous antibiotics to target likely urinary pathogens. Ultrasonography of the kidneys and bladder reveals no hydronephrosis or evidence of obstruction.
After administration of 2100 ml of crystalloid fluid (30 ml per kilogram of body weight), the patient’s jugular venous pressure is 8 cm of water, but her systemic arterial pressure has decreased to 80/50 mm Hg (mean arterial pressure, 60 mm Hg). During the 3 hours that she has been in the emergency department, she has produced 20 ml of urine, as measured through a Foley catheter that was placed on her arrival.
You place a central venous catheter and initiate a norepinephrine infusion, which you adjust with a goal of raising her mean arterial pressure to 65 to 70 mm Hg. She is transferred to the intensive care unit (ICU); on arrival in the ICU, her mean arterial pressure is 65 mm Hg while she is receiving 40 μg of norepinephrine per minute, and her heart rate is 100 beats per minute. A chest radiograph shows early evidence of acute lung injury and good central catheter placement. Her arterial oxygen saturation is 100% while she is receiving 4 liters of oxygen through a nasal cannula.
You are aware that there are two main approaches to the management of septic shock in a patient such as Ms. Jones. One approach involves serial measurement of central venous pressure, central venous oxygen saturation (Scvo2), and hemoglobin, and following the early, goal-directed therapy (EGDT) protocol, in which specified targets are used for the initiation of inotropic agents or transfusion of red cells.1 For example, if the central venous pressure is less than 8 mm Hg, additional fluid resuscitation is administered; if the Scvo2 is less than 70%, the patient receives a transfusion of red cells until a hematocrit goal of at least 30% is reached, and if the Scvo2 remains less than 70%, inotropic support is initiated.
The second approach involves continuing intravenous administration of antibiotics and vasopressors, guided by clinical signs including blood pressure and urine output, without serial central venous pressure monitoring, serial Scvo2 monitoring, transfusion of red cells, or administration of inotropic agents. You are undecided about which of these approaches would maximize the chance of survival for your patient with septic shock.
Treatment Options
Which of the following treatment strategies should you pursue for this patient?
1. Follow the EGDT protocol.
2. Monitor the patient and administer treatment on the basis of clinical signs.
To aid in your decision making, each of these approaches is defended in a short essay by an expert in the field. Given your knowledge of the patient and the points made by the experts, which option would you choose? Make your choice, vote, and offer your comments at NEJM.org.
Option 1: Follow the EGDT Protocol
Option 2: Monitor the Patient and Administer Treatment on the Basis of Clinical Signs
Option 1 (61)
Option 2 (61)
Option 1
Follow the EGDT Protocol
Emanuel Rivers, M.D.
Ms. Jones has been admitted to the ICU with septic shock and is receiving vasopressors to elevate her mean arterial pressure. Shortly after her arrival, her condition deteriorates, and intubation and mechanical ventilation are initiated because of acute lung injury. The increased lactate level and low Scvo2 indicate inadequacy of systemic oxygen delivery (hypoxia, anemia, or decreased cardiac output) to meet demands (increased work of breathing).
The EGDT protocol expanded the landscape of sepsis management outside the ICU with a series of steps.1 Step one is early detection of patients at high risk for infection according to the criteria for diagnosis of the systemic inflammatory response syndrome, followed by culturing of appropriate specimens and initiation of antibiotics. Step two is risk stratification on the basis of serum lactate levels, response to fluid challenge if the patient has hypotension, or both, for appropriate disposition. Patients who are stratified for risk on the basis of lactate level and a fluid challenge of 30 ml per kilogram have more than 19% lower mortality than patients who are not stratified in this way.2 Early risk stratification also reduces mortality from acute cardiopulmonary deterioration, which may occur in up to 20% of patients in the early course of septic shock.2 These initial steps alone or in combination significantly affect mortality.2
The remaining steps of the EGDT protocol include effective hemodynamic management of preload, afterload, and cardiac contractility and assessment of perfusion to balance systemic oxygen delivery with demands by measurement of Scvo2 and central venous pressure. Early placement of a central venous catheter has been associated with improved outcomes.2 A low Scvo2 on admission to the ICU is associated with mortality that is at least 10% higher than that with a normal Scvo2.3 Normalization of Scvo2 in acute lung injury is associated with decreased duration of mechanical ventilation and 15% lower mortality.4 If the Scvo2 is low and the partial pressure of arterial oxygen (Pao2) is normal, effective hemodynamic support begins with transfusion of one unit of packed red cells to attain a hemoglobin level above 10 g per deciliter. Although the hemoglobin target in this hemodynamic phenotype (low Scvo2 and increased lactate level) is not known, transfusion has not been associated with increased complications and may decrease the risk of death.5,6 After correction of arterial oxygen content with transfusion, the remaining variable that has to be addressed to correct oxygen delivery is decreased cardiac output (myocardial suppression, which can occur in up to 15% of patients). Inotropic agents, such as dobutamine, are included in the EGDT algorithm to increase cardiac output.2
After the original EGDT trial was performed and the Surviving Sepsis Campaign was initiated, the standard of care changed, and mortality from sepsis has decreased over the past decade.2 A recent meta-analysis of three trials (Protocolized Resuscitation in Sepsis Meta-Analysis [PRISM]) concluded that there was no mortality benefit of protocolized care for sepsis.7 However, the PRISM trials provided steps one and two of the EGDT protocol as usual care for all treatment groups before randomization, and the care was unblinded. As a result, many of the patients had reached normal Scvo2 and central venous pressure values by the time of randomization and also had a lower baseline illness severity, as evidenced by the fact that mechanical ventilation rates were lower than those in the original EGDT trial. These patients had little or no chance to benefit from the later steps in the EGDT algorithm that targeted Scvo2-guided effective hemodynamic management. In addition, ICU admission in these three trials occurred within 2 to 3 hours after presentation, as compared with 6 to 8 hours in the original EGDT study. Although early admission to the ICU is a worthwhile goal, it is not a universal reality; thus, the results are not generalizable.
The results of the PRISM trials confirm that early intervention strategies, including early detection of sepsis, risk stratification, early administration of antibiotics, and appropriate fluid resuscitation, improve the outcomes in patients with severe sepsis and septic shock. All these steps were components of the original EGDT protocol and led to historically low mortality rates in both the control groups and the intervention groups in the PRISM studies.2 However, because of the limitations of the PRISM trials with respect to the patient populations and trial methods, the potential benefit of the EGDT steps that involved effective hemodynamic management was diminished, which increased the probability of equivalency among the treatment groups.2
In the case of Ms. Jones, I would continue monitoring her condition by serial measurements of central venous pressure, Scvo2, and lactate levels and following the EGDT protocol. EGDT is more effective than usual care across a broader range of hemodynamic phenotypes, including in patients receiving mechanical ventilation. This strategy maximizes her chances of survival from septic shock.