[Excess mortality in critically ill patients after treatment with human albumin].

According to the results of a systematic review of randomized clinical studies administration of human albumin to critically ill patients is associated with excess mortality, compared with withholding albumin or administration of crystalloid fluids. The study appears to be well done. Also, there are various explanatory pathophysiological mechanisms supporting the association. However, a favourable effect of albumin in certain patient groups cannot be excluded. Alternatives to albumin are available in most clinical situations, but unfortunately, they are not completely without drawbacks. The use of albumin has to be limited; it might only be abolished when a better effect of other fluids, such as synthetic solutions, is demonstrated.