Seventy-nine RCTs with a total of 4,755 participants were included in the review.
In 14 of the 79 included trials there were potential confounding factors, such as the administration of concomitant albumin to all groups, or significantly higher baseline serum albumin levels in the control groups. The presence of these factors might have reduced the true treatment effects.
In 58 trials there was some form of clinical benefit in the albumin group compared with the control group, in 20 trials there was no significant difference between groups, and in one trial there was a negative effect in the albumin group compared with the control group.
Cardiac surgery (31 trials, 1,559 participants): albumin administration resulted in lower fluid requirements, higher colloid oncotic pressure, reduced pulmonary oedema with respiratory impairment and greater haemodilution compared with crystalloids; albumin reduced post-operative bleeding compared with hydroxyethylstarch.
Non-cardiac surgery (17 trials, 999 participants): albumin administration resulted in lower fluid requirements and reduced pulmonary and intestinal oedema compared with crystalloids.
Hypoalbuminaemia (9 trials, 536 participants): higher doses of albumin administration resulted in reduced morbidity.
Ascites (10 trials, 942 participants): albumin administration resulted in reduced haemodynamic derangements, morbidity and length of stay, and improved survival after spontaneous bacterial peritonitis.
Sepsis (4 trials, 104 participants): albumin administration resulted in reduced pulmonary oedema and pulmonary dysfunction compared with crystalloids; hydroxyethylstarch induced abnormalities of haemostasis.
Burns (4 trials, 197 participants): albumin administration resulted in decreased complications compared with crystalloids.
Brain injury (4 trials, 418 participants): albumin administration resulted in reduced mortality, disability and neurological deficits.