Since its original description in 1832, fluid resuscitation has become the cornerstone of early and aggressive treatment of severe sepsis and septic shock. However, questions remain about optimal fluid composition, dose, and rate of administration for critically ill patients. This article reviews pertinent physiology of the circulatory system, pathogenesis of septic shock, and phases of sepsis resuscitation, and then focuses on the type, rate, and amount of fluid administration for severe sepsis and septic shock, so providers can choose the right fluid, for the right patient, at the right time.
Key points
- •
Fluid resuscitation is the cornerstone of resuscitation in patients with severe sepsis and septic shock.
- •
Fluids should be considered as medications; it is imperative to consider the type, dose, and duration of intravenous fluid therapy in sepsis.
- •
Crystalloids remain the intravenous fluid of choice in sepsis resuscitation. Balanced solutions may be preferred to normal saline and colloids.
- •
It is important to know the difference between empiric fluid loading and a fluid challenge in the assessment of fluid responsiveness.
- •
Excessive and indiscriminate fluid administration can increase organ dysfunction and mortality in patients with severe sepsis and septic shock.
Introduction
Intravenous fluid (IVF) therapy began in 1832, when Dr Thomas Latta administered “two drachms of muriate and two scruples of carbonate, of soda, to sixty ounces of water” to 6 patients with hypovolemic shock during the cholera epidemic in London. Dr Latta described an “immediate return of the pulse and improvement in respiration” with the administration of repeated small amounts of his specific fluid. Since the time of Dr Latta’s original description, fluid resuscitation has become the cornerstone of early and aggressive treatment of patients with severe sepsis and septic shock. However, fundamental questions still remain about the optimal fluid composition, dose, and rate of fluid administration for critically ill patients.
In 2001, Rivers and colleagues published the landmark Early Goal-directed Therapy (EGDT) trial, which showed a significant mortality benefit in septic patients who received aggressive IVF therapy as a component of resuscitation targeted to specific hemodynamic end points. In the EGDT trial, IVFs were administered to target a predefined value for central venous pressure (CVP). Since the publication of the EGDT trial, there have been significant changes in the approach to fluid resuscitation in patients with sepsis. Specifically, aggressive fluid resuscitation that results in volume overload and organ dysfunction has been associated with increased patient mortality. In addition, CVP has been shown to be an unreliable marker of intravascular volume status and fluid responsiveness. The use of CVP to guide IVF therapy did not improve mortality in 3 recent randomized trials designed to compare EGDT with current usual care. Furthermore, the administration of IVFs in pursuit of a select CVP goal has been implicated in the development of volume overload.
In addition to the potential harm of overzealous fluid administration, the composition of select fluids may affect patient-centered outcomes. Specifically, the supraphysiologic concentration of chloride in 0.9% normal saline has been associated with adverse effects on the pulmonary, circulatory, gastrointestinal, coagulation, and renal organ systems. The recognition of these adverse effects has led to a greater awareness of the phases of sepsis resuscitation and a focus on appropriate fluid administration. Physiology of the circulatory system, the pathogenesis of septic shock, and the phases of sepsis resuscitation is discussed elsewhere in this issue.
Physiology of the circulatory system
Frank-Starling Curve
Mean arterial blood pressure (MAP) is determined by cardiac output (CO) and systemic vascular resistance (SVR), and can be calculated according to the following equation:
The primary determinants of CO are heart rate (HR) and stroke volume (SV). In order to maintain CO, blood ejected from the left ventricle (LV) must traverse the circulatory system, return to the right atrium and right ventricle, and transit the pulmonary circulation. In this way, CO is coupled with venous return (VR). The Frank-Starling curve ( Fig. 1 ) illustrates the volume of the LV at the end of diastole (preload) and directly influences SV. Any increases in preload result in an increase in SV, until a plateau is reached. Beyond this point, any additional preload in the form of IVF fails to significantly increase SV and leads to fluid overload, impaired cardiac function, pulmonary edema, and interstitial edema.
Venous Return
Blood is returned to the right atrium via the superior and inferior vena cavae. VR is determined by the gradient between the mean systemic filling pressure (Pms) and the right atrial pressure, which can be approximated by the CVP. VR is also affected by any condition that affects the resistance to return (RVR), such as abdominal ascites. VR can be calculated according to the following equation:
The venous circulation contains approximately 70% of the total blood volume and can be divided into a stressed and unstressed volume. The stressed volume is the main determinant of Pms and directly affects both VR and CO. The unstressed volume can be converted to the stressed volume either through additional volume expansion (ie, fluid bolus) or venoconstriction with the use of vasopressor medications. As a result of these relationships, IVF administration only augments VR and CO if it increases Pms more than the CVP. In addition to VR and CO, organ blood flow depends on perfusion pressure. Perfusion pressure is determined by subtracting the CVP from the MAP. Under normal conditions, autoregulation maintains constant organ blood flow over a wide range of blood pressures. However, when the patient becomes hypotensive, autoregulation fails and organ blood flow becomes more dependent on CVP. These critical points emphasize why IVF administration can be difficult. For a fluid bolus to increase VR, the Pms must increase more than the CVP. If both Pms and CVP increase with IVF administration, VR and CO remain unchanged. If the CVP increases more than Pms, VR decreases, organ perfusion pressure worsens, and venous congestion and organ injury ensues.
Endothelial Glycocalyx
The vascular endothelium plays a critical role in fluid homeostasis and its function is important to understanding fluid administration in patients with severe sepsis and septic shock. In 1896, Ernest Starling described the traditional view that fluid was filtered at the arterial end of capillaries and reabsorbed by the venous system, and that both hydrostatic and colloid oncotic pressures governed this filtration and absorption. In recent years, the discovery of the endothelial glycocalyx later (EGL) has revised the traditional Starling description. The EGL is a complex web of hairlike glycoproteins and proteoglycans that lines the vascular endothelium. The EGL has numerous functions, including the regulation of vascular permeability by limiting tissue edema and modulation of inflammation through the prevention of leukocyte and cytokine adhesion. The EGL has also been shown to hold up to 25% of the intravascular fluid volume. Sepsis damages the EGL and results in the loss of vascular integrity, capillary leak, vasodilatation, and hypovolemia. In addition, volume overload results in the release of natriuretic peptides, which have been shown to cause shedding of the EGL and further contribute to interstitial edema and inhibit lymphatic drainage.
Hemodynamic instability in septic shock
In order to understand the goals of fluid resuscitation in sepsis, it is pertinent to review the pathophysiology of sepsis. Sepsis has recently been defined as “life-threatening organ dysfunction caused by a dysregulated host response to infection.” Patients with septic shock are defined as those with a lactate value greater than 2 mmol/L who require vasopressor medications to maintain an MAP greater than or equal to 65 mm Hg despite adequate fluid resuscitation. Although dehydration may accompany septic shock, it is not the primary factor that results in hemodynamic instability. The primary mechanisms that cause circulatory compromise in septic shock include systemic vasodilatation, increased vascular permeability, and myocardial dysfunction.
In septic shock, the EGL and vascular endothelium are damaged by inflammation and lose the ability to regulate microvascular blood flow. This loss of regulation results in both arterial and venous dilatation and causes maldistribution of blood flow. The release of nitric oxide and prostacyclin have been implicated in the pathogenesis of vasodilatation. Arterial vasodilatation results in systemic hypotension, whereas venous dilatation increases the unstressed volume and decreases VR. In addition, endothelial and EGL injury result in the activation of the coagulation cascade, leukocyte adhesion, platelet aggregation, reduced red blood cell deformability, disruption of the junctions between cells, alterations in cell-to-cell signaling, and the production of tissue edema. All of these processes can decrease the diffusion of oxygen to cellular mitochondria and increase the accumulation of metabolic waste products. These pathologic changes are summarized in Fig. 2 .
In addition to pathologic vasodilatation and damaged endothelium, myocardial dysfunction has been shown to be a common occurrence in patients with septic shock. Up to 50% of patients with septic shock have LV systolic dysfunction. Furthermore, up to 62% have LV diastolic dysfunction, whereas as many as 31% of patients have right ventricular dysfunction. Diastolic dysfunction is characterized by a stiff, poorly compliant LV. This condition impairs filling and SV and has been shown to be associated with increased mortality in septic shock. Excessive fluid administration in the presence of diastolic dysfunction inevitably leads to pulmonary edema, pulmonary hypertension, right ventricular dysfunction, and decreased CO. Adequate preload, maintenance of sinus rhythm, and avoidance of tachycardia are of paramount importance to maximize LV filling and augment SV in these patients.
Phases of resuscitation
The complex mechanisms that produce hemodynamic alterations in severe sepsis and septic shock make it difficult to recommend a one-size-fits all approach to fluid resuscitation. Importantly, patients with severe sepsis and septic shock can present along a spectrum of illness and the need for fluid therapy may vary for each patient. A recent conceptual model of circulatory shock has been published that identifies 4 phases of resuscitation: rescue, optimization, stabilization, and de-escalation. These phases are depicted and described in Fig. 3 and Table 1 . The goals of fluid therapy depend on the patient’s phase of illness at presentation. In general, patients present to the emergency department in the rescue or optimization phase, whereas the stabilization and de-escalation phases typically occur during the inpatient setting.
| Rescue | Optimization | Stabilization | De-escalation | |
|---|---|---|---|---|
| Treatment goal | Shock reversal/life salvage | Adequate tissue perfusion | Zero-to-negative daily fluid balance | Fluid accumulation reversal/edema resolution |
| Time course | Minutes | Hours | Day | Up to weeks |
| Hemodynamic targets | Autoregulatory thresholds of perfusion pressure | Micro/macrocirculatory blood flow parameters | Weaning of vasopressors with stable hemodynamic conditions | Return to premorbid/chronic values of pressure and flow |
| Treatment options | Rapid fluid boluses + vasopressors | Repeated fluid challenges + vasopressors + inotropes | Maintenance fluids + decreasing/chronic vasoactive agents | Diuretics or other means of fluid removal |
Rescue Phase
The rescue phase represents the initial minutes to hours of resuscitation and is characterized by profound shock, hypotension, and impaired organ perfusion. Rapid fluid administration is needed to prevent cardiovascular collapse and death. The optimal MAP for sepsis resuscitation has not been established. Current guidelines recommend IVF administration to target a MAP of 65 mm Hg. A recent trial did not find a mortality difference when patients with septic shock were resuscitated to a MAP of 80 to 85 mm Hg compared with a MAP goal of 60 to 65 mm Hg. Of note, the rate of new-onset atrial fibrillation was higher in patients randomized to the higher MAP target. In patients with a history of chronic hypertension, the need for renal replacement therapy was lower in patients randomized to the higher MAP target. However, this finding was not associated with an improvement in mortality.
Because the microcirculation is the site where most derangements occur in sepsis, a targeted approach to microcirculatory resuscitation is appealing. Importantly, perfusion of the microcirculation is regulated by blood flow, not arterial blood pressure. Microcirculatory impairment may persist despite an MAP greater than 65 mm Hg. At present, microcirculatory-guided resuscitation remains primarily investigational.
Although most emergency department patients present in the rescue phase of resuscitation, it is important for emergency providers to be familiar with subsequent phases in the event that the patient has a prolonged length of stay.
Optimization Phase
The optimization phase is characterized by a careful assessment of intravascular volume status and a determination of the need for further fluid administration. For some patients, vasopressor medications are initiated and titrated during this phase. The optimal end points of resuscitation continue to be debated. Notwithstanding, additional IVF administration should be guided by the results of tests to determine fluid responsiveness and the performance of a fluid challenge. Fluid responsiveness and the fluid challenge are discussed in detail later in this article.
Stabilization
The goals of the stabilization phase of resuscitation are to maintain intravascular volume, replace ongoing fluid losses, support organ dysfunction, and avoid iatrogenic harm with unnecessary and indiscriminate IVF administration.
De-escalation
The de-escalation phase is characterized by organ recovery and weaning from mechanical ventilation and vasopressor support. Excess fluid that was accumulated during the previous phases of resuscitation is actively removed with the use of diuretic medications or renal replacement therapy (ie, ultrafiltration). The goal of this phase is to achieve an overall negative fluid balance.
Selection of fluid
In the past, the selection of which fluid to administer to patients with severe sepsis or septic shock has largely been based on geography, marketing, availability, cost, and even the type of provider training (medical vs surgical). Fluids can largely be separated into crystalloid and colloid solutions. Crystalloid solutions can be further divided into unbalanced and balanced solutions, whereas colloid solutions primarily include albumin, dextran, and hydroxyethyl starch solutions. The compositions of commonly used crystalloid solutions are listed in Table 2 .
