Efficacy and Safety of Colloids Versus Crystalloids for Fluid Resuscitation in Critically Ill Patients

Background: Two recent systematic reviews of the literature and meta-analyses have suggested that colloids administration might be deleterious in critically ill patients. Objective: To compare the effects on hospital mortality of crystalloids and colloids when given for fluid resuscitation in critically ill patients. Setting: Adult intensive care units (ICUs) in several European countries. Study design: A multinational, randomised, controlled trial performed on two parallel groups. Intervention: Any type of crystalloids (control group) versus any type of colloids (including albumin). Patients: All patients above the legal age of consent and hospitalised in an intensive care unit, who need fluid resuscitation (according to the physician). Pregnant women, moribund patients, brain dead patients, and patients who have a known allergy to colloids or severe head injury or major burns (\> 20% of body surface) or dehydration will not be included. Primary endpoint: 28-day mortality. Hypothesis: Assuming a hospital mortality rate of 20% in the crystalloids group, a 0.05 type I error, 3010 patients are needed to show a difference between the 2 groups of 5% with a 90% probability (two-sided test).