|
|
A Comparison of Albumin and Saline for Fluid Resuscitation in the Intensive Care Unit.The SAFE Study Investigators. N Engl J Med 2004; 350: 2247-56.Reviewer: Michael H. Wall, MD
Abstract Excerpt: This multi-centered, randomized, double-blinded trial evaluated 28-day mortality of a heterogenous population of ICU patients resuscitated with 4% albumin versus normal saline. Patients admitted to 16 closed ICU's in Australia and New Zealand between November 2001 and June 2003 who needed fluid resuscitation to maintain or increase intravascular volume were eligible for inclusion in this study. (Patients following cardiac surgery, liver transplantation or burns were excluded). Six thousand, nine hundred, ninety-seven patients were randomized (according to institution and presence or absence of trauma) then received specially manufactured "masked" 500 cc bottles and infusion sets. The amount and rate of administration was determined by each clinician. Maintenance fluids, replacement fluids, nutrition, blood products, monitoring (CVP vs. PAC) and all other aspects of patient care were not controlled. The primary outcome variable was 28-day mortality. There were many secondary outcome variables including: new organ failures, duration of ventilation, etc. There were no differences in baseline characteristics between groups except the CVP was higher in the albumin group (9.0 ± 4.7 mmHg vs. 8.6 ± 4.6 mmHg). About 44% of the patients in both groups were admitted to the ICU from the emergency room or hospital floor and 42% from the operating room. Patients in the albumin group received less fluid on day one and two (1183 ± 973 cc vs 1565 ± 1536 and 602 ± 892 cc vs 954 ± 11484 respectively). The albumin group had higher serum albumin levels on day 1-4 and received more blood (about ¬ unit over days 1-4) than the saline group. On days 1-4 there was no difference in MAP between groups but the CVP was higher (by about 1 mmHg) in the albumin group. HR was lower only on day one in the albumin group (88 ± 20 bpm vs. 90 ± 21 bpm). There were no significant differences in 28-day mortality, mechanical ventilation, ICU or hospital stay, new single or multiple organ failures between groups. Subgroup analysis showed a trend toward improved outcome in trauma patients who were resuscitated with saline, and patients with sepsis who were resuscitated with albumin. However, the authors commented that further large prospective studies of these subgroups need to be done. The authors concluded "albumin and saline should be considered clinically equivalent treatments..." and the choices of resuscitation fluid include clinician preferences, safety and cost. Comments: The crystalloid vs colloid debate surrounding fluid resuscitation rages on. Two published meta-analyses came to different conclusions regarding albumin or fluid resuscitation in the ICU - one showed a 6% increase in the risk of death in ICU patients given albumin, the other showed no significant difference in death rate. However, both meta-analyses were limited by small inadequately powered studies. This important trial was the first specifically designed and adequately powered to prospectively evaluate mortality in ICU patients, and overall they found no difference in 28-day mortality. The trends toward improved outcomes in subgroups with trauma (particularly trauma without head injury) resuscitated with saline and sepsis resuscitated with albumin will have to be prospectively evaluated. A similar trial needs to be done in patients undergoing elective cardiac surgery. Until these additional trials are done the debate will continue! Table of Contents:
© Society of Cardiovascular Anesthesiologists Questions or comments? Please send email to webmaster@scahq.org |