A 78-year-old nursing home resident is admitted for fever and altered mental status. She has a history of recurrent urinary tract infections, hypertension, diabetes mellitus and dementia. The nursing home staff noticed she was less responsive than usual. When they checked vital signs, her temperature was 38.4°C and her heart rate was 105 beats per minute. She was transferred to the emergency department, where her blood pressure was 100/65 mm Hg. Laboratory tests revealed a serum leukocyte count of 13 × 109/L, and urine tests were positive for leukocytes and nitrates. She was given one liter of fluids and admitted to your service. Soon after she arrives on the ward, a nurse calls to tell you the patient's blood pressure is now 85/60 mm Hg. How should you manage this patient?
As a hospitalist, you may be asked to intervene on a patient like this one, who is in sepsis or septic shock. Whether you manage these patients in the intensive care unit (ICU) or the hospital ward, early recognition and intervention can significantly improve outcomes.
What is sepsis?
Sepsis is a condition in which a patient has an infection and a resultant systemic inflammatory response is evident by two or more of the following clinical signs: temperature greater than 38°C or less than 36°C; heart rate greater than 90 beats/minute; respiratory rate greater than 20 breaths/minute; and white blood cell count greater than 12 × 109/L, less than 4 × 109/L, or greater than 10% immature forms. Severe sepsis is sepsis with evidence of end-organ dysfunction due to hypoperfusion. This can be manifested by findings such as lactic acidosis, oliguria, or altered mental status. The most severe form, septic shock, is defined as severe sepsis with persistent hypotension in spite of fluid resuscitation (Crit Care Med. 1992;20:864-874).
Distinction in severity of sepsis is important as it impacts survival. In one prospective study, mortality increased from 16% for sepsis to 46% for septic shock (JAMA. 1995;273:117-123). Other research has suggested mortality for severe sepsis to be 30% and as high as 57% for septic shock (N Engl J Med. 2001;345:1368-1377).
This clinical distinction is also important because it may dictate where the patient is treated. A patient with sepsis may be managed in the ward, but those with severe sepsis or septic shock should be transferred to an ICU where additional monitoring and aggressive therapies may be initiated.
What is the impact of sepsis?
The annual incidence of sepsis in the U.S. as of the year 2000 was 240 per 100,000 (N Engl J Med. 2003;348:1546-1554). For comparison, the annual incidence of lung cancer in the U.S. during 2000 was 92 per 100,000 (MMWR Morb Mortal Wkly Rep. 2011;60:1243-1247). During that year, the estimated incidence of ST-elevation myocardial infarction was 80 per 100,000 (Am J Cardiol. 2009;104:5-8).
How do we manage sepsis?
Like most disorders, management of sepsis begins with the ABCs (i.e., airway, breathing, and circulation). Due to metabolic acidosis, respiratory alkalosis will often occur. This leads to an increased load on the respiratory system. At a minimum, due to increased myocardial demand and lactic acidosis, supplemental oxygen may be required to help improve oxygen delivery to vital tissues. Those with preexisting lung disease, such as chronic obstructive pulmonary disease (COPD), may require respiratory assistance.
Patients who experience sepsis encephalopathy may require airway protection via intubation and mechanical ventilation. Furthermore, sepsis is the most common cause of acute lung injury or acute respiratory distress syndrome (ARDS). Up to 43% of those with sepsis may develop ARDS (Am J Respir Crit Care Med. 1995;151:293-301), often necessitating lung-protective mechanical ventilation strategies to minimize further injury. For these reasons alone, some patients with sepsis and most patients with severe sepsis will need to be transferred to the ICU for optimum management of their care.
Fluid resuscitation is the foundation on which circulation and perfusion is supported. Due to the release of various cytokines and vasoactive mediators, vasodilation and increased endothelial permeability lead to the reduction of effective circulating volume. Early in the course of sepsis, patients may not be hypotensive but may show signs of hypoperfusion, such as lactic acidosis, oliguria or altered mental status. Early, aggressive fluid administration is essential to mitigate the irreversible effects of tissue hypoperfusion. Even before vasopressors are used, enough fluid must be administered (often 5 to 7 liters) to allow some stability in hemodynamics.
Patients in sepsis may respond to fluid administration alone, but those in severe sepsis and septic shock may need more aggressive monitoring (e.g., central venous pressures and mixed venous oxygen saturation) and vasopressors to ensure adequate hemodynamic resuscitation. For initial resuscitative efforts, early goal-directed therapy has been validated and shown to significantly improve outcomes in those with severe sepsis and septic shock (N Engl J Med. 2001;345:1368-1377). During acute resuscitation, fluids should be administered in boluses of 500 to 1000 mL as clinical assessments are done. Excessive fluid administrations have adverse consequences, particularly in those with ARDS. To optimize the quantity of fluids administered, early goal-directed therapy uses a central venous catheter. Fluids are given until the central venous pressure is at least 8 mm Hg (12 mm Hg if mechanically ventilated). If the patient is still hypotensive, vasopressors should then be added.
In conjunction with early fluid resuscitation, another basic tenet of sepsis management includes early, effective antibiotic administration. To treat any potential pathogens, it is recommended to keep antibiotic coverage broad initially, then de-escalate the regimen based on culture results. Risk of death increases by nearly 8% per hour during the first six hours of delay of appropriate antibiotics in those with septic shock (Crit Care Med. 2006;34:1589-1596). Furthermore, inappropriate antibiotics early in the course of septic shock significantly increases the risk of death, by fivefold and up to 17-fold if the patient has primary bacteremia (Chest. 2009;136:1237-1248).
Crystalloid or colloid?
The 2008 Surviving Sepsis guidelines recommend fluid resuscitation with either crystalloids or colloids, and there are no data to favor one type over another (Crit Care Med. 2008;36:296-327). Due to the lack of mortality benefit and increased cost of albumin, crystalloids are often preferred. Following are profiles of various types of fluid used for replacement therapy.
Normal saline contains 0.9% sodium chloride with an osmolarity of 308 mOsmol/L. This is similar to the osmolarity of serum. Thus, normal saline, rather than 0.45% or 3% saline, is typically used for initial volume resuscitation. Various trials have confirmed the efficacy of normal saline administration in sepsis management. However, due to its volume of distribution, saline volume of one to three times the colloid volume may be required. Depending on total volume infused, hyperchloridemic non-gap metabolic acidosis may ensue. A few trials have evaluated the efficacy and safety of hypertonic saline (7.5% saline) in sepsis to minimize the effect of the total volume of saline infused. While there are some theoretical advantages and possible temporary benefits, more clinical studies are required before regular use of hypertonic saline can be recommended. As of the last update, the Surviving Sepsis guidelines do not recommend using hypertonic solution for volume resuscitation.
Due to a lower content of sodium and chloride, lactated ringers may be a better option for hyperchloridemic non-gap metabolic acidosis. It is also isotonic to serum, but data on its use in sepsis management are limited. This solution should be used with caution as it contains 4 mEq of potassium in each liter of fluid. Thus, if the septic patient has acute renal failure and is anuric, the additional potassium load via this replacement fluid can potentially worsen electrolyte disturbances. The lactate in the fluid is converted to bicarbonate and thus this fluid is considered alkalinizing. But, in those with liver failure, this conversion may be limited, thus requiring that this fluid be used with caution in those with hepatic dysfunction.
Albumin may seem appropriate for initial fluid resuscitation, especially in those who may be hypoalbuminemic, such as patients with liver failure. However, numerous trials have failed to show the superiority of albumin. One meta-analysis has shown that even in those with hypoalbuminemia, there was no significant benefit with albumin (Ann Intern Med. 2001;135:149-164). In a prospective study, nearly 7,000 patients admitted to the intensive care unit were randomized to either saline or albumin as the initial resuscitation fluid (N Engl J Med. 2004;350:2247-2256). This trial failed to show any benefit of albumin in days spent in the intensive care unit, hospital stay, days on mechanical ventilation, days of renal-replacement therapy or mortality. In a subgroup analysis of those with severe sepsis, there was a trend toward decreased mortality but it did not reach statistical significance. A recent meta-analysis of albumin in sepsis suggested that albumin was associated with a decrease in mortality (Crit Care Med. 2011;39:386-391). But most of the trials in this meta-analysis used other colloids such as hetastarch (rather than saline) as the control fluid.
Various synthetic colloids (e.g., hydroxyethyl starch [HES], dextran, gelatin) have been tested in sepsis and are frequently used in Europe. However, the lack of any benefit over crystalloids and increased risks of anaphylactic reactions (incidence rate ratio compared to albumin of 2.3 for dextran, 4.5 for HES, and 12.4 for gelatin), coagulopathy and renal impairment prevent most clinicians from using synthetic colloids in critically ill patients (Anesth Analg. 2011;112:156-164). In fact, high doses of HES have been associated with higher mortality (N Engl J Med. 2008;358:125-139).
If there is obvious blood loss resulting in hemorrhagic shock, the best replacement therapy is blood since it can most effectively restore lost intravascular volume. As acute administration of a large volume of fluid can result in dilution of red blood cells, early goal-directed therapy guidelines recommend keeping the hematocrit above 30% during the first six hours of resuscitation to maintain adequate oxygen-carrying capacity. The more liberal transfusion guideline in the early management of sepsis is to mitigate the effect of tissue hypoxia by maintaining oxygen delivery to vital tissues. This recommendation is in contrast to a more conservative transfusion recommendation of withholding transfusion until hemoglobin is less than 7 g/dL once tissue hypoperfusion has resolved (N Engl J Med. 1999;340:409-417).
Early and goal-directed management of septic patients involves many decisions. One important aspect of early resuscitation is adequate volume replacement to optimize hemodynamics and tissue perfusion. Numerous trials have sought to identify the ideal volume expander. Based on cost and comparable efficacy with colloids, crystalloids are often favored. Early in the resuscitative efforts, a more liberal blood transfusion strategy may be beneficial in improving oxygen delivery. Regardless of the type of fluids used, excessive infusion may result in various complications. Examples include hyperchloridemic non-gap metabolic acidosis (with normal saline) or increased morbidity in those with ARDS. Thus, to guide adequacy of replacement, monitoring central venous pressure (to greater than 8 mm Hg but less than 12 mm Hg) and other signs of adequate perfusion, such as urine output, has been shown to improve outcomes.
Although early fluid resuscitation is critical in successful management of patients with sepsis, other important aspects of management include the appropriate use of antibiotics, vasopressors and lung-protective mechanical ventilation strategies. These other aspects of management are discussed in the Surviving Sepsis Campaign (Crit Care Med. 2008;36:296-327) and the early goal-directed therapy guidelines (N Engl J Med. 2001;345:1368-1377).
Back to our patient
Our 78-year-old patient should have received supplemental oxygen and early antibiotics, ideally before blood and urine cultures were obtained. She should now be given additional fluid boluses with close monitoring of her respiration, oxygen saturation, blood pressure, urine output and mental status. If she stabilizes, her care can continue in the ward. However, if there are any signs of deterioration, she should be transferred to the ICU for closer monitoring and additional interventions, such as respiratory assistance or central venous catheter placement.