European Journal of Anaesthesiology



Original Article

Albumin administration – what is the evidence of clinical benefit? A systematic review of randomized controlled trials


G. R. Haynes a1, R. J. Navickis a2 and M. M. Wilkes a2c1
a1 Medical University of South Carolina, Department of Anesthesia and Perioperative Medicine, Charleston, South Carolina, USA
a2 Hygeia Associates, Grass Valley, California, USA

Article author query
haynes gr   [PubMed][Google Scholar] 
navickis rj   [PubMed][Google Scholar] 
wilkes mm   [PubMed][Google Scholar] 

Summary

Background and objective: The advantages of albumin over less costly alternative fluids continue to be debated. Meta-analyses focusing on survival have been inconclusive, and other clinically relevant end-points have not been systematically addressed. We sought to determine whether albumin confers significant clinical benefit in acute illness compared with other fluid regimens.

Methods: Database searches (MEDLINE, EMBASE, Cochrane Library) and other methods were used to identify randomized controlled trials comparing albumin with crystalloid, artificial colloid, no albumin or lower-dose albumin. Major findings for all end-points were extracted and summarized. A quantitative meta-analysis was not attempted.

Results: Seventy-nine randomized trials with a total of 4755 patients were included. No significant treatment effects were detectable in 20/79 (25%) trials. In cardiac surgery, albumin administration resulted in lower fluid requirements, higher colloid oncotic pressure, reduced pulmonary oedema with respiratory impairment and greater haemodilution compared with crystalloid and hydroxyethylstarch increased postoperative bleeding. In non-cardiac surgery, fluid requirements, and pulmonary and intestinal oedema were decreased by albumin compared with crystalloid. In hypoalbuminaemia, higher doses of albumin reduced morbidity. In ascites, albumin reduced haemodynamic derangements, morbidity and length of stay and improved survival after spontaneous bacterial peritonitis. In sepsis, albumin decreased pulmonary oedema and respiratory dysfunction compared with crystalloid, while hydroxyethylstarch induced abnormalities of haemostasis. Complications were lowered by albumin compared with crystalloid in burn patients. Albumin-containing therapeutic regimens improved outcomes after brain injury.

Conclusions: Albumin can bestow benefit in diverse clinical settings. Further trials are warranted to delineate optimal fluid regimens, in particular indications.

(Accepted January 2003)


Key Words: BRAIN INJURIES; CARDIAC SURGICAL PROCEDURES; INFECTION, sepsis; PATHOLOGICAL PROCESSES, ascites; PROGNOSIS, treatment outcome; RANDOMIZED CONTROLLED TRIALS; SERUM ALBUMIN; SURGERY; VITAL STATISTICS, morbidity; WOUNDS AND INJURIES, burns.

Correspondence:
c1 Correspondence to: Mahlon Wilkes, Hygeia Associates, 17988 Brewer Road, Grass Valley, California 95949, USA. E-mail: mwilkes@hygeiaassociates.com; Tel: +1 530 268 2300; Fax: +1 530 268 2301