Crystalloids

Crystalloids are salt solutions used for intravenous infusion. Presently, none of the commercially available solutions contains electrolyte and buffer concentrations comparable to those of plasma; in particular, bicarbonate is very rarely found in these solutions.

Isotonic saline (0.9% sodium chloride) is still frequently used despite concentrations of sodium and chloride (154 mmol/L) well above those found in plasma. Alternatively, so-called balanced salt solutions include Ringer’s lactate (or Hartmann solution) and Ringer’s acetate. The concentrations of electrolytes in these solutions better resemble those of plasma, and most preparations include sodium, chloride, potassium, and calcium. However, bicarbonate is normally not included because bicarbonate-containing fluids, particularly those in plastic containers, have a shorter shelf-life. Instead, the balanced solutions are buffered with lactate or different combinations of acetate, gluconate, and malate.

Colloids

Colloid solutions contain large molecules that prolong the time the fluid remains in the circulation. The molecules used to obtain the colloidal and thereby volume expansion are human albumin or synthetically modified sugars or collagens. The most frequently used synthetic colloid solutions are hydroxyethyl starch (HES), gelatin, and dextrans.

Colloid solutions have been extensively administered for volume expansion in critically ill patients, but clinical practice has varied, mainly because of regional differences in the types of colloid solutions available. Moreover, in recent years, results of large trials have examined the effects and side effects of colloids in critically ill patients. These trials raise questions regarding the clinical efficacy of colloids and have clearly demonstrated some harmful effects in groups of critically ill patients.

In the following section, I describe what fluids to give to critically ill patients and what fluids to avoid. My recommendations are based on the results of recently updated high-quality systematic reviews and recently conducted high-quality randomized trials. These findings allow evidence-based choices to be made and are likely to improve care and outcome and reduce costs.

In General ICU Patients, What Fluids Should I Give and What Should I Avoid?

The short answer is to give crystalloids and avoid the synthetic colloids (HES, gelatins, and dextrans). The systematic review of the Cochrane Collaboration comparing crystalloid with colloid solutions showed that crystalloids were associated with improved mortality compared with HES solutions, whereas there were no differences in mortality between crystalloids and the other colloid solutions analyzed (albumin, gelatins, and dextrans). The increased mortality with the use of HES is most likely mediated through impaired kidney function and hemostasis, resulting in an increased use of renal replacement therapy and increased bleeding ( Table 20-1 ). A recently published large randomized trial, the CRISTAL (Colloids Versus Crystalloids for the Resuscitation of the Critically Ill) trial, was not included in the Cochrane analysis. The CRISTAL trial compared any crystalloid to any colloid solution in ICU patients with shock. The results indicated that colloids (mainly HES) versus crystalloids (mainly saline) improved 90-day mortality, which was a secondary outcome measure. The primary outcome measure, 28-day mortality, did not differ between the groups, and the trial had high risk of bias in several domains (unblinded, uncertain allocation concealment, and baseline imbalance). Moreover, the results differed from those of the high-quality trials, and the editor argued in an accompanying editorial for cautious interpretation of these findings and advocated that crystalloids should be the first-line fluid in patients with shock.

The interpretation of the Cochrane meta-analyses on crystalloid versus gelatin and dextran solutions was hampered by low-quality trial data and few patients and events and therefore wide confidence intervals on the point estimates. Because gelatins and dextrans have registered side effects that are similar to those observed with HES, these synthetic colloids should be avoided in critically ill patients. This recommendation is substantiated by the fact that no patient group in any high-quality trial has been shown to benefit from any colloid solution, including the trial using human albumin ( Table 20-1 ), although the latter showed that albumin is safe to administer to critically ill patients, excluding those with trauma. As a blood product fractionated from human plasma, albumin is an expensive and limited resource; thus it is of limited value in many health-care systems. The same may be said of red blood cells, plasma, and platelets, but these products have a larger potential for harm than albumin, and their safety has not been ensured in large trials in the critically ill. Liberal transfusion of red blood cells may even increase the mortality of general ICU patients, as observed in the Transfusion Requirement in Critical Care (TRICC) trial. Thus blood products should only be used for patients with severe anemia or bleeding or as a prophylactic in high-risk patients with severely impaired hemostasis.

Overall, the recommendations of the European Society of Intensive Care Medicine Task Force on colloids support the notion that crystalloid solutions should be preferred over colloids in the vast majority of critically ill patients.

The choice between the different crystalloid solutions is more difficult because there are no data from high-quality trials supporting this decision. Saline may be preferred in patients at risk of brain edema because the high plasma concentrations of sodium may reduce brain swelling and intracranial pressure. Conversely, severe acidosis may be worsened by saline-induced hyperchloremia. A balanced crystalloid solution may be the better choice in these patients. Observations in cohort studies have suggested that the use of balanced crystalloids reduces the risk of acute kidney injury and even mortality as compared with the use of saline. These results need to be confirmed in randomized trials before implementing them into clinical practice because of the high risk of bias in observational studies of critically ill patients.