Fluid resuscitation for people with sepsis

In The BMJ this week, Patel and colleagues (doi:10.1136/bmj.g4561) report a timely meta-analysis of clinical trials of human albumin for volume expansion and resuscitation in adults with sepsis.1 After carefully analysing data from 16 randomised trials including more than 4000 patients, they found no statistically significant survival benefit from using albumin overall or in any predefined subgroup. These new results call into question the Surviving Sepsis Campaign’s recommendation to consider giving albumin to patients who “require substantial amounts of crystalloid.”2

Most (84.3%) of the data in the analyses by Patel et al came from three large multicentre trials that compared albumin solutions with crystalloid fluids.3 4 5 Participants in these trials were enrolled either on admission to intensive care with severe sepsis, or 6-24 hours after the onset of severe sepsis while in the intensive care unit. Early resuscitation before enrolment was not controlled: in the ALBIOS study, half the participants had received either albumin or another colloid as part of early goal-directed treatment before enrolment.5 Study …