Colloids in the intensive care unit.

Link to article at PubMed

Colloids in the intensive care unit.

Am J Health Syst Pharm. 2012 Oct 1;69(19):1635-42

Authors: Kruer RM, Ensor CR

Abstract

Purpose The most recent published evidence on the use of colloids versus crystalloids in critical care is reviewed, with a focus on population-dependent differences in safety and efficacy. Summary Colloids offer a number of theoretical advantages over crystalloids for fluid resuscitation, but some colloids (e.g., hydroxyethyl starch solutions, dextrans) can have serious adverse effects, and albumin products entail higher costs. The results of the influential Saline Versus Albumin Fluid Evaluation (SAFE) trial and a subsequent SAFE subgroup analysis indicated that colloid therapy should not be used in patients with traumatic brain injury and other forms of trauma due to an increased mortality risk relative to crystalloid therapy. With regard to patients with severe sepsis, two meta-analyses published in 2011, which collectively evaluated 82 trials involving nearly 10,000 patients, indicated comparable outcomes with the use of either crystalloids or albumins. For patients requiring extracorporeal cardiopulmonary bypass (CPB) during heart surgery, the available evidence supports the use of a colloid, particularly albumin, for CPB circuit priming and postoperative volume expansion. In select patients with burn injury, the published evidence supports the use of supplemental colloids if adequate urine output cannot be maintained with a crystalloid-only rescue strategy. Conclusion The results of the SAFE trial and a subgroup analysis of SAFE data suggest that colloids should be avoided in patients with trauma and traumatic brain injury. There are minimal differences in outcome between crystalloids and hypo-oncotic or iso-oncotic albumin for fluid resuscitation in severe sepsis; in select populations, such as patients undergoing cardiac surgery, the use of iso-oncotic albumin may confer a survival advantage and should be considered a first-line alternative.

PMID: 22997116 [PubMed - in process]

Leave a Reply

Your email address will not be published. Required fields are marked *