Hydroxyethyl starch versus other fluids for non-septic patients in the intensive care unit: a meta-analysis of randomized controlled trials

INTRODUCTION: Use of hydroxyethyl starch (HES) in septic patients is reported to increase the mortality and incidence of renal replacement therapy (RRT). However, whether or not use of HES would induce the same result in non-septic patients in the intensive care unit (ICU) remains unclear. The objective of this meta-analysis was to evaluate 6% HES versus other fluids for non-septic ICU patients. METHODS: Randomized controlled trials (RCTs) were searched from Pubmed, OvidSP, Embase database and Cochrane Library, published before November, 2013. A meta-analysis was made on the effect of 6% HES versus other fluids for non-septic ICU patients, including mortality, RRT incidence, bleeding volume, red blood cell (RBC) transfusion and fluid application for non-septic patients in ICU. RESULTS: Twenty-two RCTs were included, involving 6,064 non-septic ICU patients. Compared with the other fluids, 6% HES was not associated with decreased overall mortality (RR = 1.03, 95%CI: 0.09 to 1.17; P = 0.67; I(2) = 0). There was no significant difference in RRT incidence, bleeding volume and red blood cell transfusion between 6% HES group and the other fluid groups. However, patients in HES group received less total intravenous fluids than those receiving crystalloids during the first day in ICU (SMD = -0.84; 95%CI: -1.39 to -0.30; P = 0.003, I(2) = 74%). CONCLUSIONS: This meta-analysis found no increased mortality, RRT incidence, bleeding volumes or RBC transfusion in non-septic ICU patients, but the sample sizes were small and the studies generally were of poor quality.