Research in patients at high risk of complications is now needed

Intravenous albumin has been widely used in the treatment of shock since the early 1940s. In a recent large European study on the occurrence of sepsis in acutely ill patients, 11% of patients received albumin at some point during their stay in intensive care.1 Despite such widespread use, there are few randomised controlled trials on albumin administration in critically ill patients, and even fewer data on which subgroups of patients are likely to benefit most. Because low baseline serum albumin concentrations are associated with worse outcomes,2 3 data on the use of albumin in patients with low albumin are of considerable interest. In this week's BMJ a subgroup analysis of the saline versus albumin fluid evaluation (SAFE) study4 assesses the effect of baseline serum albumin concentration on outcomes of resuscitation.5

A meta-analysis published in the BMJ in 1998 suggested that resuscitation with albumin solutions rather than crystalloids increased the absolute risk of death by 6%.6 In 2004, prompted in part by the meta-analysis, …