Purpose: We set out to assess the resuscitation fluid requirements and physiological and clinical responses of intensive care unit (ICU) patients resuscitated with 20% albumin versus 4–5% albumin. Methods: We performed a randomised controlled trial in 321 adult patients requiring fluid resuscitation within 48 h of admission to three ICUs in Australia and the UK. Results: The cumulative volume of resuscitation fluid at 48 h (primary outcome) was lower in the 20% albumin group than in the 4–5% albumin group [median difference − 600 ml, 95% confidence interval (CI) − 800 to − 400; P < 0.001]. The 20% albumin group had lower cumulative fluid balance at 48 h (mean difference − 576 ml, 95% CI − 1033 to − 119; P = 0.01). Peak albumin levels were higher but sodium and chloride levels lower in the 20% albumin group. Median (interquartile range) duration of mechanical ventilation was 12.0 h (7.6, 33.1) in the 20% albumin group and 15.3 h (7.7, 58.1) in the 4–5% albumin group (P = 0.13); the proportion of patients commenced on renal replacement therapy after randomization was 3.3% and 4.2% (P = 0.67), respectively, and the proportion discharged alive from ICU was 97.4% and 91.1% (P = 0.02). Conclusions: Resuscitation with 20% albumin decreased resuscitation fluid requirements, minimized positive early fluid balance and was not associated with any evidence of harm compared with 4–5% albumin. These findings support the safety of further exploration of resuscitation with 20% albumin in larger randomised trials. Trial registration: http://www.anzctr.org.au. Identifier ACTRN12615000349549.
- Critical care
- Fluid therapy