Explore this free Rotation Guide to see what Rotation Prep has to offer.
Unlock all of Rotation Prep by linking your NEJM Resident 360 account
to your personal or institutional subscription to NEJM Knowledge+.
Intravascular volume repletion is crucial to resuscitating critically ill adults. Fluids, electrolytes, and blood products have historically been given liberally because they were considered natural and essential elements of human physiology, but increasingly we are recognizing that, like other medications, they must be administered with careful consideration of indication, type, dose, frequency, and adverse effects. In this section, we review the following topics:
Curbside Consults: In this interview, Dr. Kathryn Hibbert reviews the biology underlying the effects of various fluids on the body and how sepsis disrupts the body’s natural response to corticosteroids. We also discuss the major points from these four new papers in the March 1, 2018 issue of NEJM and existing literature to get you up to speed on the evidence behind practice.
Dr. Hibbert is an Instructor in Medicine at Harvard Medical School, Director of the MGH Medical Intensive Care Unit, and Site Director for the Harvard Pulmonary and Critical Care fellowship.
An ideal intravenous (IV) fluid does not exist for all situations, but data gathered over the last 15 years have transformed our understanding of safe and appropriate fluid resuscitation. Keep the following evidence-based principles in mind as you select the best option for your patient:
1. Colloids are not superior to crystalloids in most cases and may be harmful in some situations.
Colloids, such as albumin and blood, exert higher oncotic pressure and theoretically are retained entirely in the vasculature. Consequently, to achieve the same hemodynamic change, a smaller amount of colloid than crystalloid should be required (often described in a 1:3 ratio). However, in practice, the observed difference is smaller, likely due to increased vascular permeability (see figure below).
Albumin also has theoretical benefits of being an antioxidant, an anti-inflammatory, a carrier for drugs, and an acid-base buffer. Derived from blood, it is expensive to produce and distribute.
Because data from a meta-analysis suggested that giving albumin-containing fluid increased mortality, investigators conducted the randomized SAFE trial that compared 4.0% albumin and 0.9% normal saline in approximately 7,000 adult ICU patients and found no difference in 28-day mortality.
A post-hoc analysis of the SAFE trial in patients with traumatic brain injury showed higher 24-month mortality with albumin.
Because results from the SAFE trial suggested a possible benefit from albumin in the subgroup of patients with severe sepsis, the ALBIOS trial compared 20% albumin plus crystalloid versus crystalloid alone in ICU patients with severe sepsis. Subgroup analysis suggested a benefit from albumin in patients with septic shock, compared with patients without septic shock.
Semisynthetic colloids were created to circumvent the availability and expense of albumin. However, one semisynthetic colloid, hydroxyethyl starch (HES), has been associated with increased mortality and adverse events. HES is not recommended for fluid resuscitation, and other semisynthetic colloids should be used with caution.
Albumin is beneficial in select situations, such as reducing the incidence of renal impairment and death in patients with cirrhosis and spontaneous bacterial peritonitis.
Role of the Endothelial Glycocalyx Layer in the Use of Resuscitation Fluids
Figure 1. The structure and function of the endothelial glycocalyx layer, a web of membrane-bound glycoproteins and proteoglycans on endothelial cells, are key determinants of membrane permeability in various vascular organ systems. Panel A shows a healthy endothelial glycocalyx layer, and Panel B shows a damaged endothelial glycocalyx layer and resultant effect on permeability, including the development of interstitial edema in some patients, particularly those with inflammatory conditions (e.g., sepsis). (Source: Resuscitation Fluids. N Engl J Med 2013.)
2. Among crystalloids, different solutions have different adverse event profiles.
Normal saline (0.9% sodium chloride [NaCl]) has higher concentrations of Na and Cl (154 mEq/L) than plasma. Large-volume infusion leads to hyperchloremic metabolic acidosis (owing to a compensatory decrease in bicarbonate concentration and acid-buffering capacity), renal vasoconstriction, acute kidney injury, and hyperkalemia.
Balanced solutions (e.g., lactated Ringer's solution or Plasma-Lyte) have more physiologic concentrations of Na and Cl and contain lactate or acetate as anionic buffers (bicarbonate is unstable in plastic containers) as well as potassium (K) and other cations. Large-volume infusion may lead to hyponatremia and metabolic alkalosis.
Two pragmatic single-center clinical trials (SMART) and SALT-ED) that compared normal saline to balanced solutions found that balanced solutions were associated with a small but significant decrease in major adverse kidney events. Study clinicians could choose saline for patients with relative contraindications to balanced solutions (e.g., hyperkalemia and brain injury when hypernatremia may be preferred to reduce brain swelling).
In the open-label multicenter BICAR-ICU trial, treatment with 4.2% bicarbonate solution to increase pH >7.3 in patients with severe metabolic acidosis (pH <7.2) and SOFA score >3 or lactate >2mM did not improve mortality at 28 days or reduce organ failure at 1 week. However, patients with acute kidney injury (AKIN score, 2−3) showed improved mortality and a reduction in organ failure. Adverse effects of bicarbonate infusion included metabolic alkalosis, hypernatremia, and hypocalcemia.
3. Patients’ fluid requirements depend on the clinical circumstances, and conservative fluid administration may be beneficial.
Excessive fluid administration can lead to tissue edema and organ dysfunction, especially in situations with increased vascular permeability from inflammation (see figure above).
In the FACCT trial, a conservative fluid-management strategy (central venous pressure [CVP] goal <4 mm Hg) was associated with improved outcomes in patients with acute respiratory distress syndrome (ARDS). However, the mean time to intervention was 43 hours after ICU admission, when most patients are past the acute phase of sepsis.
In contrast, for the initial resuscitation of patients with sepsis-induced hypoperfusion, the Surviving Sepsis Campaign recommends ≥30 mL/kg of crystalloids within the first 3 hours, followed by fluids administered guided by frequent reassessments of how a patient might improve (see algorithm below).
Several metrics can be used at the bedside to determine if a patient is fluid responsive (e.g., passive leg raise, pulse pressure variation with ventilation, and change in CVP). Passive leg raise may be the most reliable test (positive likelihood ratio, 11) according to a recent review.
Blood transfusion is also often required in critically ill patients, but growing evidence supports the safety of conservative transfusion thresholds (hemoglobin concentration <7 g/dL), reducing the need for blood transfusion in critically ill patients and avoiding unnecessary risks.
The 1999 TRICC trial found no difference in mortality or severity of organ dysfunction between a restrictive (hemoglobin <7 g/dL) and liberal (hemoglobin <9 g/dL) transfusion threshold in a general ICU patient population.
The 2014 follow-up TRISS trial compared the same restrictive or liberal transfusion thresholds as the 1999 TRICC trial but in patients with septic shock. This trial also found no difference in mortality and ischemic events between the two transfusion thresholds.
Note: Patients with acute coronary syndrome were excluded from the TRISS trial, and evidence is lacking for this patient population and other populations, including patients with hematologic disorders, cancer, thrombocytopenia, or acute neurologic disorders.
The AABB (formerly the American Association of Blood Banks) recommends a restrictive hemoglobin threshold <7 g/dL for adults who are hemodynamically stable, including critically ill patients, and a threshold of <8 g/dL for those with underlying cardiovascular disease. The Society of Critical Care Medicine recommends a similar threshold of <7 g/dL for general critically ill patients and those with stable cardiac disease who are hemodynamically stable, and to reserve the <8 g/dL threshold for patients with acute coronary syndrome.
Ultimately, the hemoglobin threshold should not be the only trigger for transfusion. Blood transfusion is preferred for resuscitation of patients with hemorrhagic shock and indicated for anyone with hemodynamic instability or inadequate oxygen delivery.
The risks associated with blood transfusion are generally low. See Transfusion Reactions in the Hematology rotation guide for more information.