Fluid therapy in septic shock. BE, base excess; CI, cardiac index; CVP, central venous pressure; MAP, mean arterial pressure; NA, noradrenaline; ScvO2, central venous oxygen saturation; SVRI, systemic vascular resistance index.
In the course of sepsis, unnecessary fluid might aggravate degradation of the endothelial glycocalyx and thus contribute to edema due to capillary leakage and not enhance perfusion. A continuous positive fluid balance lasting for days is a significant predictor of mortality.[12] However, it is possible that a positive fluid balance is only a marker of the severity of illness.
In the “fluids and catheter treatment trial,” 1,000 patients with acute lung injury or acute respiratory distress syndrome were randomized to a conservative fluid administration strategy compared with a more liberal strategy.[13] The trial lasted for 7 days, and the liberal group was brought into a positive fluid balance of 1 liter/day. In the conservative group, the fluid balance was nearly 0. The conservative strategy improved oxygenation, increased the number of ventilator-free days, and reduced the length of stay in the ICU. Patients were relatively young (average about 50 years), and those with overt renal failure were excluded from the trial. However, fluid therapy in this trial was started on average 43 hours after admission to the ICU and 24 hours after the establishment of acute lung injury. These patients were, in other words, already optimized with early fluid administration.
This study underlines that what is beneficial in the early course of sepsis might not be beneficial later. In the later course of the disease, between 2 and 7 days, a more restricted fluid management strategy should be instituted. Not all critically ill patients who are fluid responders should automatically receive fluids. The intensivist should always consider whether the fluid bolus would be beneficial for the patients. Thus, it is prudent to treat patients with severe sepsis or septic shock with EGDT during the first 6–24 hours only.
Crystalloids versus colloids
The “Surviving Sepsis Campaign” recommends administering crystalloids as the initial fluid.[14] Subgroup analysis of meta-analyses has demonstrated that crystalloid resuscitation was associated with a lower mortality in trauma patients. In contrast, albumin resuscitation in some subgroup analysis has been associated with a better outcome in patients with septic shock. This tendency was observed in two large prospective randomized trials.[15,16]
Fluid resuscitation with albumin results in a greater and faster increase in cardiac filling and cardiac output than crystalloid resuscitation in septic hypovolemia. Albumin remains in circulation for a longer time than the crystalloids, and crystalloids require that more fluid is used in a patient to attain the same goals, whereby more edema might develop.[17] Hydroxy-ethyl starch and dextran are no longer used in critically ill septic patients, owing to increased morbidity and mortality.[18]
The Surviving Sepsis Campaign recommends initial fluid resuscitation with crystalloid and consideration of the addition of albumin in patients who continue to require substantial amount of crystalloids to maintain adequate mean arterial pressure.[14]
Blood transfusions
Anemia is a common problem in critically ill patients. Harmful effects of anemia include increased risk of cardiac-related morbidity and mortality as well as a general decrease in oxygen-carrying capacity.[19].
The consequences of anemia may be deleterious in this population because critical illness is often associated with increased metabolic demands. A surgical procedure will accentuate metabolic demands, and intraoperative blood loss reduces oxygen delivery. Therefore, an optimal hemoglobin level must be maintained. However, the criteria for an optimal hemoglobin level in critical illness are not clearly defined. A Canadian study indicated that a liberal use of transfusions (100–120 g/l) might result in increased hospital mortality rates compared with a more restrictive transfusion regime (70–90 g/l).[20] A later observational multicenter study confirmed that there is an association between transfusions and diminished organ function as well as between transfusions and mortality.[21] A recent study showed no effect on mortality and morbidity between patients receiving liberal (9 g/l) versus restricted transfusion regimen (70 g/l).[22] There seems to be a very delicate balance between the harmful effects of anemia on organ function and the harmful effects of transfusion. In the future, treatment with blood transfusion should probably be individualized.
Fluid responsiveness in septic shock
It is important that the critically ill patient is hemodynamically stable and normovolemic before being transported from the ICU or emergency room to the operation theater. This is not always possible in, for example, patients with ongoing and uncontrolled bleeding.
Patients with severe sepsis or septic shock are often resuscitated with fluid challenges. In this process, a large amount of fluid is administered under close monitoring to evaluate the hemodynamic response. The “Surviving Sepsis Campaign” recommends an initial fluid challenge of 20 ml/kg of crystalloids administered over 30 min. If the patient is in severe shock, a more rapid infusion might be necessary. Repeated fluid challenges are performed as long as the patient improves hemodynamically. Clinical signs of hemodynamic improvement might be increasing arterial pressure, decreasing heart rate, increasing urine output, or improvement in capillary refill time.
Adequate fluid resuscitation cannot, however, be based only on normalization of vital signs. Traditionally, physicians have used static hemodynamic values such as CVP or the pulmonary artery occlusion pressure (PAOP) to evaluate whether the patient would benefit from further fluid challenge. There is increasing evidence that estimates of intravascular volume based on CVP or PAOP do not reliably predict the patient’s response to a fluid challenge [12,23].
In addition to the Swan–Ganz (SG) catheter, cardiac output can be measured by pulse contour analysis. This method also estimates the global end-diastolic volume and intrathoracic blood volume. These new static preload parameters correlate better with cardiac index than the traditionally measured CVP. However, static preload measurements are inaccurate and must be supplemented with more dynamic measures.
Functional hemodynamic parameters, such as systolic pressure variation (SPV), pulse pressure variation (PPV), and passive leg raising are more sensitive indices of fluid responsiveness. SPV and PPV can be used only in sedated, mechanically ventilated patients with rather large tidal volumes. With more invasive monitoring, cardiac output can be measured and used as an adjunct when evaluating the response to a fluid challenge. Cardiac output measurement may also help to identify the minority of patients who have a low cardiac output despite adequate fluid resuscitation.
After fluid resuscitation, septic shock is often hyperdynamic with high cardiac output, low systemic vascular resistance (SVR), and reduced MAP. This hyperdynamic state is, however, often confined to the large vessels, whereas the regional microcirculation is compromised.
As there is no perfect hemodynamic parameter, the patient’s response to fluid administration must be evaluated together with other parameters, such as the ScvO2. A multimodal monitoring approach has to be instituted. The goal of fluid resuscitation in the study by Rivers et al. [1] was a ScvO2 greater than 70%. If a SG catheter was used, a mixed venous oxygen saturation (SvO2) greater than 65% could be the goal.
SvO2 has been considered to be the gold standard to monitor whole-body perfusion. This might be true in hemorrhagic shock, but in septic shock the SvO2 is often normal or even supernormal owing to reduced oxygen extraction at the microvascular level. In contrast, serum lactate is a useful measure of anaerobic metabolism, and base excess is often negative if the organs are not adequately perfused.[24] Monitoring serum lactate as well as the base excess improves the overall evaluation of the patient’s response to fluid challenge.
Treatment with vasopressors and inotropic agents
The “Surviving Sepsis Campaign” recommends that noradrenaline is used as the initial vasopressor agent.[14] A new, large multicenter study was not able to show any significant difference in mortality in septic patients treated with noradrenaline compared with dopamine, although those treated with dopamine had more arrhythmic events.[25] There has also been some concern that the use of noradrenaline in patients who are inadequately fluid resuscitated would increase blood pressure because of vasoconstriction, and thereby reduce the blood flow to the organs.
It might be necessary to use noradrenaline to restore MAP during the early course of septic shock, before the patient is adequately fluid resuscitated. Some caution is recommended as animal studies have shown that noradrenaline-masked hypovolemia is associated not only with renal failure but also with cardiomyocyte necrosis.[26]
Echocardiography cannot provide continuous hemodynamic data but can be used initially to determine the type of shock or cardiac function when septic shock is evident. When the patient with septic shock does not respond to initial therapy, they should be monitored with a SG catheter or pulse contour analysis. Then cardiac output can be measured and SVR calculated. When this monitoring is instituted, the vasoconstrictor noradrenaline should be used to increase MAP guided by SVR, whereas dobutamine can be used to increase cardiac output, if necessary. When adequately fluid-resuscitated, the septic patient most often has hyperdynamic shock with high cardiac output and low SVR. Only in a minority of fluid-resuscitated septic patients is it necessary to administer dobutamine to increase cardiac output. In the study by Rivers et al. 15.4% of the patients who received EGDT were treated with dobutamine.[1] By administration of noradrenaline, the low SVR can be increased and a MAP ≥65 mmHg obtained. The goal is, however, not to normalize SVR.
Adrenaline is not used very often in septic patients as it can impair the splanchnic circulation in septic shock. Compared with noradrenaline there was no difference in mortality, but adrenaline was associated with more adverse effects.[27]
Full access? Get Clinical Tree
