Background

A meta-analysis of randomized controlled trials was recently published in BMC Surgery that compared the use of human albumin with 6% hydroxyethyl starches 130/0.4 for cardiopulmonary bypass prime and perioperative fluid management in pediatric and adult cardiac surgery patients. The two plasma expanding solutions are described as equivalent for efficacy and safety outcomes, and, on that basis, the preferential use of hydroxyethyl starches 130/0.4 was recommended for economic reasons because of the higher unit costs of human albumin solutions.

Results

In addition to the fact that trials were mostly small, single-center studies and the number of total participants was low, making the meta-analysis underpowered for several outcomes, selective reporting of data for ICU length of stay was identified. Re-calculation of statistics at higher precision showed that ICU length of stay of patients in the human albumin group was significantly shorter than that of patients in the 6% hydroxyethyl starches 130/0.4 group (standard mean difference − 0.181, 95% confidence interval − 0.361 to − 0.001, P = 0.049), which may offset any proposed economic advantage of using 6% hydroxyethyl starches 130/0.4. At the same time, the renal safety of 6% hydroxyethyl starches 130/0.4 in surgical patients is under regulatory review.

Conclusions

Underpowered trials and selective reporting may impair the validity of the meta-analysis. A more cautious conclusion about the interchangeability between human albumin and 6% hydroxyethyl starches 130/0.4 in cardiac surgery should have been reached.